Citation
Prosocial messages in children Saturday morning television programs

Material Information

Title:
Prosocial messages in children Saturday morning television programs
Creator:
Hoover, Ferris Edward
Publication Date:
Language:
English
Physical Description:
vii, 68 leaves : forms ; 29 cm

Subjects

Subjects / Keywords:
Television and children ( lcsh )
Television and children ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 63-68).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Department of Communication.
Statement of Responsibility:
by Ferris Edward Hoover Jr.

Record Information

Source Institution:
|University of Colorado Denver
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
22839477 ( OCLC )
ocm22839477
Classification:
LD1190.L48 1990 .H66 ( lcc )

Downloads

This item has the following downloads:


Full Text
PROSOCIAL MESSAGES IN CHILDREN'S SATURDAY
MORNING TELEVISION PROGRAMS
by
Ferris Edward Hoover Jr.
B.A., University of Colorado, 1988
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Arts
Department of Communication
1990


This thesis for the Master of Arts
degree by
Ferris Edward Hoover Jr.
has been approved for the
Communication Department
by
Richard L. Dukes
Michael Z. Hackman
l-h-qo
Date


Hoover, Ferris Edward, Jr. (M.A., Communication)
Prosocial Messages in Children's Saturday Morning
Television Programs
Thesis directed by Associate Professor Donald D. Morley
This study was done to test a methodology for
measuring the existence and recognition of prosocial
messages in children's Saturday morning television
programs. A panel of four adults was trained to count
the occurrence of prosocial behaviors in sample episodes
of the five top-rated Saturday morning children's
programs. Per hour occurrence ranged from 22 per hour to
48.5 per hour. Fourth grade children were surveyed to
determine if some degree of recognition of prosocial
messages can be measured. The method used was to ask the
respondent's opinion of whether or not the major
character in a program would exhibit prosocial traits.
The results of the study were somewhat inconsistent. The
rank ordering of prosocial content of the programs by the
children was sharply different from that of the adults
with the children's most prosocial program being the
least prosocial according to the adult panel count. The
surprising results could have been due to one of several
factors. Among them might be that the intended


connotations of the words used in the questions might not
be the same as those assigned by the respondents. Or,
the one-time count of a single sample episode of a
program by an adult unfamiliar with the usual plot
structure of that program would not necessarily coincide
with the overall view taken by a child who chooses to
watch the program regularly. The trained adult panel
method seems to be a valid measurement of the occurrence
of prosocial messages, but the method for determining the
recognition of such messages by children needs
improvement.
The form and content of this abstract are approved,
recommend its publication.
Signed,
iv


CONTENTS
Tables...........................................vl
Acknowledgments ............................... vii
CHAPTER
1. INTRODUCTION ................................... 1
2. SOME HISTORICAL ASPECTS OF
MEDIA EFFECTS RESEARCH ......................... 4
Social Learning Theory ...................... 11
The 1982 NIMH Report............................15
Prosocial Research ............................ 18
Other Research..................................23
3. RATIONALE...........,.........................31
4. METHODS....................................... 36
Procedures......................................36
Instruments.....................................40
Data Analysis.................................. 41
5. RESULTS........................................ 45
6. CONCLUSIONS.....................................51
APPENDIX
A. ADULT RATING PANEL CODING SHEET .............. 61
B. FOURTH GRADE STUDENT SURVEY INSTRUMENT .... 62
BIBLIOGRAPHY ........................................ 63


TABLES
Table
5.1. Interrater Reliability ........................ 45
5.2. Adult Rating Panel Observation of Prosocial
Behaviors.......................................46
5.3. Fourth Grader Ranking of Positiveness Ratios . 47
vi


ACKNOWLEDGMENTS
I would like to express special appreciation to Don
Morley, my thesis director, and Kim Walker, who acted as
second reader, for the extra effort they put into guiding
me in the development and execution of the research and
the literature search.
Thanks go to the Pikes Peak Broadcasting Corporation
for providing me with the Nielsen ratings.
Thanks to all the members of the Communication
Department at UCCS for the support they have given during
the process of getting this thesis done. The other
Graduate Teaching Assistants get special thanks for their
participation in the rating panel and the suggestions
they had for overcoming the innumerable obstacles that
seemed to appear from nowhere. Without the support
freely given by the people around me, this thesis would
have never been completed.
vii


CHAPTER 1
INTRODUCTION
Mass media effects have been the subject of numerous
research efforts among the disciplines of psychology,
sociology, political science and communication. Early
research was generally done by scientists from
disciplines other than communication, and a great deal of
what is being done today is done by other than
communication scholars. Actual internal effects are
probably best studied by behavioral scientists or
cognitive psychologists while the mechanisms and
processes involved in the formulation and the reception
of messages through a mass medium are more appropriate
areas for a communication scholar.
Other social science disciplines generally utilize
controlled exposure and observation of behavior to
determine whether or not a linkage between media
consumption and behavioral change exists. This type of
methodology is best suited to examining the internal
processes that may be occurring in the subjects being
studied. If the intent is to examine what sort of
messages are discernable in a particular form of program


and whether or not the messages are recognizable, a
different methodology is needed.
Although examining the internal processes which lead
to behavioral changes is important, this does not lead to
a single generalizable conclusion which explains all
possible effects that media consumption might cause. An
easily accepted idea, that might cause many people some
concern about possible effects of media consumption, is
that whatever effects occur are the same for everyone.
This is a misconception according to Becker (1987) :
Too often, when people who do not understand
mass communication processes or mass communication
research think about the effects of the mass media,
they think in terms of all or none. That is, they
think if mass communication has a particular effect
on one person it must have it on everyone who uses
the mass media or who was exposed to that content.
Or if they do not see an effect on everyone, they
believe it must be due to something other than mass
communication, (p. 457)
The idea that mass media have different effects on
different people provides many paths for a communication
researcher to follow. The goal of this thesis was to
find a method to explore one of the factors that might
contribute to whatever effects might occur. The factor
to be examined was whether or not prosocial messages can
be reliably identified by adults and recognized by
children in children's television programming. First, it
was necessary to become familiar with what research has
2


been done in the area of mass media effects in order to
choose a useful and logical method of research. To this
end, a review was done of some of the research from all
the disciplines involved that has dealt with what effects
the medium of television might have on children. The
next step was the development of a valid measure of the
prosocial message content of the programs in question and
a valid method of indirectly measuring the comprehension
of the messages by the intended audience. This involved
finding methods that have been used in past research and
adapting them to this project. The measures which were
developed were then executed as a trial of their
viability. The methodologies which have been used
primarily to search for the adverse effects of mass media
were adaptable to the search for possible prosocial
effects. The criteria which were used in the examination
of whether or not prosocial messages are present and
recognizable in Saturday morning children's television
constituted the major difference between this research
and other explorations of media effects.
3


CHAPTER 2
SOME HISTORICAL ASPECTS OF MEDIA EFFECTS RESEARCH
Since the 1920s popular media have been under attack
by parents, educators and others for causing such
antisocial outcomes as an erosion of moral standards and
aggression. The fears that are the basis for such
feelings are rooted in the theory of uniform influences
of mass media (Lowery & DeFleur, 1988). That is, the
mass media have powerful effects on their audiences and
these effects are nearly the same for all audience
members. The Payne Fund Studies done in 1929 and 1930
were the first attempt to assess the effects of motion
pictures on children and adolescents. Some of their
findings indicated that adolescents in particular used
the actions, styles, and attitudes depicted in the movies
as models for their own behavior and style of dress
(Lowery & DeFleur, 1988).
The next modern mass medium to come under fire was
the comic book. Dr. Frederick Wertham, a noted New York
psychiatrist published Seduction of the Innocent in 1954
after having published several articles in popular family
magazines of the time. In Seduction of the Innocent.


Wertham attacked comic books for having a "bad influence"
on their young audience. Although Wertham's methods were
scientifically discredited according to Lowery and
DeFleur, he did arouse national attention and contributed
to the decline of comic book sales. Wertham was
generally ignored by social scientists, but his work is
an illustration of uniform influence theory in that he
described his research subjects as "normal" even though
virtually all of them were referred to him because of
some social problem such as delinquency. Because the
content analyses conducted by Wertham showed that a great
deal of socially unacceptable behavior was depicted, he
concluded that so called "crime comics" were severely
affecting children (Lowery & DeFleur, 1988).
The increasing popularity of television as the
medium viewed by children again aroused concern over the
effects of a mass medium during the 1960s. The first
major effort to determine what effects, if any,
television might have on children was published by
Professors Wilbur Schramm, Jack Lyle and Edwin Parker
(1961). Their series of reports investigated why
children watch television and what they learn from it.
The summation of the findings of this study is the often
quoted:
5


For some children, under some conditions, some
television is harmful. For other children under the
same conditions, or for the same children under
other conditions, it may be beneficial. For most
children, under most conditions, most television is
probably neither harmful nor particularly
beneficial. (Schramm, Lyle, & Parker, 1961, p. 13)
This conclusion did not satisfactorily answer the
question of whether or not television was causing adverse
effects on children.
The debate continues today as indicated by stories
of consumers pressuring advertisers to stop supporting
programs that are "objectionable'? ("idea of boycotting,"
1989). Nor has the concept of direct and uniform
influences disappeared, as illustrated by retiring
Surgeon General C. Everett Koop's 1989 call for
restricting and modifying alcohol-related advertising to
reduce the number of teenage drunken driving accidents
("Koop takes aim," 1989). In short, the public continues
to believe that the media serves as a powerful cause of
socially unacceptable behavior.
The National Commission on the Causes and Prevention
of Violence studied violence on television during the
late 1960s. The resultant manuscript, Violence and the
Media, edited by Robert K. Baker and Sandra J. Ball
(1969), among its intended research purposes, analyzed
the portrayals of violence in television programming,
measured the violent experiences of Americans, and
6


compared television portrayals of violence with the
perception of the audience as to the existence of
violence in society. This was done in order to determine
the existence of long- and/or short-term effects on
accepted norms pertaining to violence (Lowery & DeFleur
1988). Their conclusions confirmed the findings of
short-term effects that had been measured in previous
experimental studies. The report then suggested that
long-term studies should be conducted to determine
whether or not long-term effects existed.
The call for further research was answered when the
Surgeon General appointed an advisory committee to
supervise a massive research effort in 1969. The report
that resulted included more than 40 scientific papers and
an overview written by The Surgeon General's Scientific
Advisory Committee in 1969. The summation of the
research effort, Television and Growing Up (The Surgeon
General's Scientific Advisory Committee on Television and
Social Behavior, 1971), was that viewing television
violence could be causally linked to aggressive behavior,
but only in children who were inclined in that direction.
It was further argued that the context in which the
violence was viewed had a large influence on the actual
effects. In the words of the committee:
7


Thus, the two sets of findings converge in three
respects: a preliminary and tentative indication of
a causal relation between viewing violence on
television and aggressive behavior, an indication
that any such causal relation operates only on some
children (who are predisposed to be aggressive); and
an indication that it operates only in some
environmental contexts. Such tentative and limited
conclusions are not very satisfying. They represent
substantially more knowledge than we had two years
ago, but they leave many questions unanswered.
(Surgeon General's Scientific Advisory Committee,
1971, pp. 18-19)
Instead of focusing attention on which children in
which contexts were most affected by viewing violence on
television, the question that the public, as well as the
scientific community, most wanted to answer was that of
how much violence was portrayed on television. A
comprehensive study conducted by George Gerbner (1971)
of the Annenberg School of Communications at the
University of Pennsylvania measured the number and types
of violent acts, who the perpetrators were, and who the
victims were in both prime time and Saturday morning
programming. From this study emerged the use of the
violence index that has been cited by both the popular
press and the scientific community to argue that almost
all of the programs viewed by children contain violence.
Of note is the fact that Gerbner1s study found Saturday
morning cartoons to have the highest level of violent
content with a violence index of 251.1 compared to the
8


next highest index of 241.9 for crime, western and action
adventure programs (Gerbner, 1971, p. 67).
However, these findings must be tempered by the fact
that the study counted violent acts without regard to
context or intent. Gerbner was investigating content as
being symbolic material without regard to artistic merit
and without distinguishing between the relative impact of
any individual action and any other action occurring
within the same program (Gerbner, 1971). For example, a
pie in the face was weighted equally with the gunning
down of an unarmed person. Gerbner's measurement of
violent content would seem to be the basis for many
proposals which periodically are reported in the popular
press for control of the content of children's programs
by government agencies. The recurrence of these
proposals indicates that popular wisdom still supports
the idea of uniform and universal effects.
Since most of the questions raised by Baker and Ball
in 1969 about what might cause long-term effects were not
answered despite the massive effort, the committee
continued with a call for more research and specified
some areas to be covered. Among them were such areas as
television's effects in the context of other mass media,
individual developmental history, and other environmental
influences including the home environment. They also
9


called for research on the relationship between televised
violence and aggression, specifically in the areas of
predispositional characteristics of individuals, age
differences, effects of labeling, contextual cues and
other program factors, and longitudinal influences of
television. The modeling and imitation of prosocial
behavior, the role of environmental factors, including
the mass media, in the teaching and learning of values
about violence and the effects of such learning on social
development were also included as areas to be researched
(Surgeon General's Scientific Advisory Committee, 1971,
p. 19).
The scientific community responded by multiplying
their efforts during the decade that followed. According
to Lowery and DeFleur (1988) "approximately 90. percent of
all research publications on television's influence on
behavior appearedl" and more than 3000 titles are
included in the information that was published on
television violence between 1971 and 1982 (p. 353). This
accumulation led to the publication in 1982 of Television
and Behavior! Ten Years of Scientific Progress and
Implications for the Eighties by the National Institute
of Mental Health (NIMH) at the direction of the Surgeon
General. This report is a synthesis of the knowledge
gained from the effort instigated by the 1971 Television
10


and Growing Up and included not only research done on the
effects of television violence but many other areas. The
areas germane to this thesis will be discussed.
Social Learning Theory
The studies done to determine whether or not
television violence could be causally linked to
aggressive behavior were generally based on social
learning theories which posit that children learn by
observation or incidentally. One of the foremost
researchers and theoreticians in this area is Albert
Bandura, who began working in the area during the early
1960s. His work is well represented in his 1977 book
Social Learning Theory. It provides a framework for
describing the processes experienced by children in their
social development and the continuing effects of media
throughout most people's lives.
Bandura (1977) posits that observational learning,
which is a primary factor in social learning, has four
components: attention, retention, reproduction, and
motivation. Social learning according to Bandura is a
three-way interaction between the person, the behavior,
and the situation. The person part of the interaction
includes the developmental stage of the individual which
affects the ability to model a behavior on a delayed
11


basis. Another factor is the self-regulatory capacity of
an individual. This falls into the motivational
component of observational learning in that it affects
the determination of whether or not the outcome of a
behavior is considered to be valuable to an individual.
The situational aspect of the interaction is influenced
by reinforcement of the learning process, which Bandura
describes as "facilitative" rather than necessary for
learning because other influences can determine what is
attended to and learning can occur without reinforcement.
That is, even if the situation surrounding the
observation of a behavior provides no reinforcement, the
behavior may be successfully performed at a later time.
Behaviors may also be duplicated in the short-term
because of a situation like a laboratory experiment, but
not retained in an individual's behavioral repertoire for
later use. Social learning theory does recognize that
the actual learning of a behavior is an individual
process, but Bandura indicates that the media have broad-
ranging effects in the population when he states: "It has
been shown that both children and adults acquire
attitudes, emotional responses, and new styles of conduct
through filmed and televised modeling" (Bandura, 1977,
p. 39).
12


Social learning theory has been used as a model for
explaining many findings of media effects research.
Other research using the theory was being done
concurrently with the work that Bandura was doing that
resulted in the publication of Social Learning Theory as
well as after its appearance. Schramm, Lyle, and Parker
(1961) theorize that most of what children learn from
television falls into the category of incidental
learning. As children mature, it has been demonstrated
that they acquire the ability to distinguish between
reality and fantasy in what they view (Corder-Bolz,
1982). During the period between three and eight years,
however, they identify with television characters who
they think are "real" and probably learn more behaviors
from them than they do as they mature and differentiate
between reality and fantasy.
Social learning theory as a model would explain
these findings in terms of increased ability to perform a
behavior after a lapse of time due to maturation and the
individual characteristics (i.e., the "person part" of
the interaction) being a determining factor in what
behaviors would be perceived as having valued outcomes.
Situational factors would include the perception of what
was desired of the subjects by significant social others
such as the researchers in an experimental setting and
13


what previous learning had been acquired. As long as it
is not assumed that a conscious effort is required to
learn a behavior, social learning theory would seem to be
a convenient and sensible model for explaining incidental
learning from the media, particularly that which takes
place in experimental settings.
Fishbein and Ajzen (1975) theorized that attitudes
and behaviors are learned on the basis of rewards and
punishments. That is, what is perceived to have a valued
outcome will be adopted. Whether or not the outcome of a
behavior will be valued will be largely dependent on what
attribution of responsibility for a behavior is assigned
to the model. Five levels of attribution are described
by Fishbein and Ajzen. They are: association with the
behavior, commission of the behavior, foreseeability of
the result, the intentionality of the model and
intentionality with justification. Fishbein and Ajzen
also describe three factors which bear on what
attribution will be assigned to a model. These are: the
consistency of the behavior, the distinctiveness of the
behavior and consensus with other behaviors of the model.
Rosenthal and Zimmerman (1978) include cognition, defined
as the covert processing of information, and abstraction,
defined as going beyond discrete stimulus-response
associations, in their expansion of social learning
14


theory. They explain four specific behaviors as a
refinement of social learning theory. The first is
inhibition which is the result of learning what behaviors
result in punishment. The second is disinhibition which
results from learning what behaviors will not result in
immediate punishment. The third is facilitation which
results when an observed behavior serves as a reminder of
things learned at an earlier time. The fourth is novel
behavior which is something not previously modeled.
Rosenthal and Zimmerman (1978) posit that
enculturation is the result of modeling parental
behaviors and learning group standards and mores. This
process could include behaviors learned from modeling on
television. They also state that "the social context
factors in implementing learning have received limited
attention" (p. 267). This relates to the "situation
part" of the social learning interaction as described by
Bandura (1977) and reiterates a call for research in an
area that was specifically mentioned by the Surgeon
General's Scientific Advisory Committee in 1971.
The 1982 NIMH Report
Among the areas explored in the 1982 National
Institute of Mental Health Report were factors believed
to determine children's attention to television, how much
15


and what type of content was retained, and what processes
might explain the relationship between violence viewing
and behavior.
Collins (1982) postulated three factors for
determining children's attention to and comprehension of
television content. The factors relating to attention
were: viewer characteristics (e.g., mostly age-related),
content attributes (e.g., auditory and visual cues), and
comprehensibility (which again is a function of the
child's age and development). Research reviewed by
Collins suggested that older children have a more
extensive background making more material understandable
and have a longer attention span. Factors relating to
processing content were also found to be age related.
They were: knowledge of common formats such as narrative
stories and commercials, knowledge of commonly portrayed
situations and event sequences, and knowledge of the
forms and conventions such as the minor climax just
before a commercial break. Collins also posited that
younger children would be less likely to have the ability
to link actions with their consequences.
Huesmann (1982), in his contribution to the NIMH
report, describes five processes that might be used to
explain the relationship between violence viewing and
behavior. The first is observational learning that he
16


states has been verified in laboratory settings. The
persistence, or retention of observational learning is
dependent on whether or not it is reinforced in some
manner. The second is attitude change which can alter
the acceptance of violence by an individual. This
process of attitude change can be modified by prosocial
reinforcement or outside attitude training. The third is
the arousal process which is somewhat of a dichotomy.
That is, viewing violence could cause an "overload"
leading to hyperactivity or could have an anesthetizing
effect due to sensory "overload" and thereby lead to the
need for increased aggressive behaviors to achieve normal
arousal levels. The fourth is catharsis, that is,
watching violence reduces the need for aggressive
behavior because normal arousal levels are reached by the
act of viewing. Huesmann states that the catharsis
postulation has been rejected by research. The fifth is
justification. Aggressive individuals watch violent
programming in order to justify their own behavior as
"normal" and reduce guilt levels.
Freedman (1984, 1986) challenged the conclusions of
several articles in the 1982 NIMH report that causally
linked viewing television violence with aggressive
behavior. He found in reviewing the research for the
NIMH report that "The bulk of the correlations fall
17


between .10 and .20." (1984, p. 237). He described these
correlations which were found in both laboratory studies
and field research as "mild" and judged them insufficient
to extrapolate to a cause and effect determination. He
concluded that: "(a) exposure to and preference for
violent programming correlates with aggressive behavior
and (b) there is little convincing evidence that viewing
violence on television in natural settings causes an
increase in subsequent aggressiveness" (1984, p. 244).
This position would also dispute any argument for the
possibility of television viewing having beneficial
effects on the audience.
Cook, Kendzierski and Thomas (1983) also argued that
the 1982 NIMH report failed to establish the causal link
between viewing television violence and aggressive
behavior. Additionally, they criticized the report for
focusing on behavioral studies only and its failure to
discuss aspects of television programming other than its
possible influence on children.
Prosoclal Research
Bryan and Walbek (1970a, 1970b) in studies with
second through fifth grade children found that prosocial
behavior modeled by both live and videotaped models
affected the propensity to share game winnings from a
18


subsequent session. The actions of the model seemed to
have more effect than what was said while the game to be
played was demonstrated. Inconsistency on the part of
the model was found to have no significant effect. Stein
and Bryan (1972), however, found that preaching about
following rules while practicing violation of those rules
led to an increase in cheating by their third and fourth
grade female subjects. When the model preached and
practiced rule compliance, cheating was nearly
eliminated.
In a study of three- to five-year-old children who
were tested before and after being exposed to violent,
neutral or prosocial programming, Stein, Friedrich and
Vondracek (1972) found that those who were most
aggressive according to the baseline test exhibited the
greatest increase in aggressive behaviors after exposure
to the violent programming. They also found that the
group receiving the prosocial treatment exhibited a
higher level of self controlling behaviors than the other
groups.
In an experimental study on three- to six-year-old
children, Friedrich and Stein (1975) found that prosocial
behavior could be learned from television and that the
subjects applied the behaviors. They also found that
verbal labeling training and role playing exercises
19


enhanced the learning experience. Gorn, Goldberg, and
Kanungo (1976) also found that prosocial television could
have clearcut short-term effects on the attitude of
three- to five-year-old children toward members of other
racial and ethnic groups.
Singer and Singer (1976) in a study of three- and
four-year-old children in a day care center over a two-
week period used a prosocial program along with
structured play sessions to determine the effect of adult
mediation on the level of imaginativeness showed by the
children. The findings indicated that imaginativeness of
play could be increased by the presence of an adult
mediator.
Poulos, Harvey and Liebert (1976) in their content
analysis of Saturday morning television found that
prosocial acts were being depicted on children's
programs. In the fifty programs which made up their
primary sample the combined network number of prosocial
acts per half-hour was higher than the number of
aggressive acts (a mean of 10.7/half-hour versus
6.10/half-hour). They also found that one quarter of the
sample contained no examples of aggression. The one
shortcoming noted in Saturday morning television was the
lack of examples of preventing or eliminating violence.
20


In their 1979 study, Friedrich-Cofer, Huston-Stein,
Kipnis, Susman, and Clewett, also found that
reinforcement of the model provided by prosocial
programming can increase its effect. Their study of two-
to five-year-old children in a Head Start program done
over an eight-week exposure period found increased levels
of prosocial behavior in children who were provided
reinforcement through associated play materials and a
greater increase when special teacher training was added
to the reinforcement.
Some findings in the 1982 NIMH report also indicated
that viewing television need not necessarily affect
behavior in an adverse way. Singer (1982) reported
finding that when children's viewing was monitored and
mediated by an adult, it helped to improve the children's
imagination. Corder-Bolz (1982) called for the teaching
of television literacy and parental participation in
children's viewing in order for children to better
discriminate between reality and fiction and to improve
the learning process when educational programs were
viewed. Rushton (1982) reviewed numerous experiments
that were designed to determine if television could have
beneficial as well as detrimental effects on the
audience. He defines prosocial material as "that which
specifies things that are socially desirable and in some
21


way benefit another person or society at large" (p. 249).
He describes four types of prosocial influences. First
are those which promote altruism, that is, generosity,
helping behavior, and cooperating. The second is
friendly behavior; the third is self-control, which is
resisting temptation and delaying gratification; and the
fourth is coping with fears. Rushton's review of both
laboratory and naturalistic research designs indicated
that viewing prosocial material correlated with an
improvement in prosocial behavior in the short-term and
on delayed measures.
However, Rushton (1982) drew a conclusion that
supported not only the argument for beneficial effects
from television viewing, but also supported the widely
held contention that viewing television has detrimental
effects on behavior. Specifically, "Television is much
more than mere entertainment; it is also a major source
of observational learning experiences, a setter of norms.
It determines what people judge to be appropriate
behavior in a variety of situations. Indeed it might be
that television has become one of the most important
agencies for socialization that our society possesses"
(p. 255).
Finally, Forge and Phemister (1987) found that
preschool children demonstrated more behaviors of
22


sharing, cooperation, delay of gratification, and
positive social contacts after exposure to prosocial
programming. The direction of their research was to
determine whether or not the mode of presentation,
animated versus non-animated, made a difference. The
mode of presentation was not found to be significant but
prosocial versus neutral program topics did seem to
affect the subjects behavior. This would lead one to
believe that the effects of television viewing would be a
function of what sorts of messages were being presented.
Other Research
Given that the studies reviewed in the 1982 NIMH
report did not fully establish the causal link between
viewing violence on television and aggressive behavior,
it could be suggested that establishing a causal link
between viewing prosocial television and increases in
prosocial behavior would be equally difficult. Although
the social learning theory work of Bandura and others
provides an explanation of how observed effects could
have occurred, other works published since 1980 provide
additional information about how and why television might
influence audience members.
Corder-Bolz (1980) also found that intervention by
an adult mediator can affect the learning process and the
23


formation of attitudes about sex roles and violence. The
findings were that primary social agents, those that can
demand compliance with their expectations such as
parents, teachers, neighbors, the clergy, institutions,
organizations, and peer groups, have the most effect on
attitudes and beliefs. If secondary social agents, those
that cannot demand compliance, such as mass media,
present information contradictory to that presented by
the primary social agents, it is likely to be
discredited.
Huesmann, Eron, Klein, Brice, and Fischer (1983)
approach the question from the standpoint of cognition.
They suggest intervention in the form of teaching that
television violence is an unrealistic portrayal, that
aggressive behaviors depicted on television are not as
acceptable in the real world as they are in televised
stories, and that one should not behave in the same ways
as the aggressive characters on television. This
intervention would reduce the effect of violent
portrayals by altering the cognitive process. They also
posit that children who are normally less aggressive will
perceive television violence as less realistic and will
be less likely to model it. However, a didactic course
of treatments designed to produce attitude changes
produced no significant results in their second and
24


fourth grade student subjects. The researchers
considered it possible that their didactic treatment was
not directed at exactly the right variables and this may
have affected the outcome. A second treatment that
involved having the subjects produce their own arguments
about why television violence is not realistic and should
not be modeled did produce some measurable results.
However, their primary finding was that identification
with a character was the best predictor of whether or not
a child was peer nominated as an aggressive individual.
They also noted that the subjects who initially had the
lowest level of identification showed the most effect.
Of special interest is the fact that the amount of
television viewed was found to have no significant effect
on the other findings.
Berkowitz (1984) discussed some factors that extend
the "person part" of the social learning interaction
beyond age and development. He posited that reaction to
a message was dependent on interpretation by the
individual as well as the ideas previously held and the
thoughts activated by the message. He called this a
"priming" effect. That is, a media event leads the
viewer to thoughts about similar things without conscious
effort. This effect could be caused by both pro- and
antisocial situations depicted in media. However,
25


Berkowitz states that a semantically related concept,
such as a previously acquired set of words or images,
must be available to the individual for the effect to
occur and that the effect of "priming" decreases over
time. Berkowitz also found that viewing fictional
material and focusing on the aesthetic aspects of a
program reduces the impact of aggression facilitating
ideas. Tamborini, Zillmann, and Bryant (1985) expanded
this concept with their finding that short-term "priming"
occurs after a single exposure to a program but long-
term effects result from "repriming" which occurs with
multiple exposures.
Christenson (1986) determined that the perception of
moral lessons improves with age. In a study of
kindergartners, first, third, fourth, and sixth graders,
the highest level moral lesson as determined by an adult
panel was more likely to be perceived by older age
groups. The kindergarten and first grade group had no
members perceive the highest level lesson while the sixth
grade group had 39 percent of its members perceive the
lesson.
Tan (1986) describes some possible models that can
be used to track how children might acquire the cultural
values that determine their individual attitudes toward
the viewing of violent television programming. The major
26


structure is the coorientation model, that is,
communication requires the participation of at least two
persons. According to Tan, three criteria must be met
for coorientation to exist. The first is congruency,
which is similarity between one person's cognitions and
the perceptions of another person's cognitions. The
second is agreement, which is the extent to which two
people have the same salience evaluations. The third is
accuracy, which is the extent to which the estimate of
another's cognitions matches what the other actually
thinks (Tan, 1986).
Tan (1986) places coorientation within families into
two patterns and describes four variations of family
types based on whether one or the other or both patterns
exist. The first pattern is socioorientation in which
parents stress the child's relationships with others with
the desired outcome of conflict avoidance. The second
pattern is concept orientation in which parents stress
seeking new ideas, looking at all sides of issues and
forming one's own opinions.
The first family type is a laissez faire attitude in
which neither socioorientation nor concept orientation
are stressed and the children are "undirected." The
second family type is the protective family in which
socioorientation is stressed, social harmony is valued
27


and the children are "obedient." The third family type
is pluralistic in which concept orientation is stressed.
In this family type there is no constraint in
interpersonal relationships and open discussion is
encouraged. The fourth family type is consensual in
nature with both orientations being stressed and
discussion but not debate is encouraged (Tan, 1986).
Some behavioral patterns can be associated with
family types. In pluralistic families the children are
generally more competent, active in public affairs, more
receptive to contradictory ideas and less persuasible.
Junior and senior high school students from pluralistic
type families also use media more extensively to learn
about public affairs and are more positive about the
political system. Same age students from protective
families generally rank lower in use of media for gaining
information, are less politically active, and less
positive about the political system (Tan, 1986).
In addition to discussing what contributions are
made by immediate family members to the acquisition and
development of attitudes, Tan (1986) describes some
possible effects of violent television and some possible
prosocial effects of television. The effects of violent
television are: learning new acts, disinhibition and
facilitation of aggression, reinforcement of attitudes
28


held prior to observation, vicarious reinforcement of
behavior patterns, and postobservation reinforcement of
modeling behavior observed on television. The prosocial
effects are divided into cognitive effects and behavioral
effects. The cognitive effects are learning useful
information and development of cognitive skills, that is,
perceptual discrimination, reasoning, and problem
solving. The behavioral effects are performance of
socially desirable acts, these are: helping others,
altruism, controlling aggressive impulses, delaying
gratification, persistence, explaining feelings,
resisting temptation, adhering to rules, and expressing
sympathy to others. Tan states that most of the research
done in the area of prosocial effects has been done with
preschool children but the results have been positive.
In an investigation involving preschoolers and first
and second graders, Nikken and Peeters (1988) found that
by age seven children realize that what is on television
is not necessarily real. Liebert and Sprafkin (1988)
found that by age eight, a majority of children
understand the purpose of commercial messages and that
among fourth grade children, 65 percent of the central
content of a program should be remembered. As a means of
reducing possible detrimental effects, they suggest that
teaching younger children that most programs do not
29


reflect reality could reduce aggressive reactions.
Liebert and Sprafkin also found that adult co-viewers
providing commentary on programs improves children's
comprehension of the content. These findings indicate
that perhaps the best way to limit possible undesirable
effects of television on children is to teach them how to
discriminate truth from fiction and provide commentary to
improve their comprehension.
Walker and Morley (1988) found that a liking for
violence and a generally aggressive attitude was related
to watching violent television and that different types
of violence are perceived in differing ways. They also
found that adolescents with aggressive behavioral
intentions found violent television to be more
aesthetically acceptable. These findings could be
explained by the proposition that the intended or
unintended intervention in the early learning processes
by primary social agents helped form the beliefs and
attitudes which made the watching of violent television
programs more desirable.
30


CHAPTER 3
RATIONALE
Review of research on the effects of mass media
reveals that the topic can cause the researcher a great
many problems. Finding causal links between viewing any
genre of television and subsequent behavior can become a
circular process that the researcher may not see. Are
aggressive tendencies acquired from primary social agents
with adolescents consequently watching violent
programming or, does watching violent programming lead to
aggressive tendencies in behavior? Cleary, the processes
that occur are difficult to isolate and study and the
artifacts that result from the research programs are
difficult to sort out. If children are being observed
outside the home and are being exposed to researcher
selected programming, is such a study externally valid?
The bulk of the studies relating to media effects
deal with the possible adverse effects. Exploring the
possibility that prosocial messages exist in Saturday
morning children's programs has not received a great deal
of attention. Measuring the existence and recognition of
these messages in the programs that the children chose to
view would begin to balance previous research.


If social learning theory is valid and children
learn from watching others' behavior either in person, or
through some mediated means, it can be posited that the
existence of prosocial messages in Saturday morning
television could provide the prosocial models for
children to emulate. Furthermore, if it can be
demonstrated that some recognition of the message occurs
in young viewers, then from a social learning theory
perspective it would follow that prosocial behavioral
changes may be occurring.
Existing prosocial research can be described as
highly restricted in that it occurs in laboratory
settings or is done with groups of subjects in other than
an at home setting (e.g., pre-school, lower grade
elementary school, or head start). Another problem is
that the researcher typically selects the material to be
watched rather than the subjects watching what they
normally select (Bryan and Walbek, 1970a, 1970b; Stein,
Friedrich, & Vondracek, 1972; Friedrich & Stein, 1975;
Gorn, Goldberg, & Kanungo, 1976; Friedrich-Cofer, Huston-
Stein, Kipnis, Susman, & Clewett, 1979; Singer & Singer,
1976). The ages of the subjects could also be questioned
in that normally all are six years old or younger and
consequently not of an age where they could be expected
to have a full comprehension of reality (Christenson,
32


1986? Nikken & Peeters, 1988). The present investigation
proposes a more naturalistic method that would not select
the programs viewed and use measures other than
behavioral observation in a laboratory setting to
determine what messages are recognized by the children.
In order to determine whether or not Saturday
morning programming contains prosocial messages, and
whether or not those messages are being recognized, two
things should be accomplished. First, the existence of
prosocial messages needs to be reliably determined based
on at least face valid criteria. This requires the use
of trained adult rating panels. Second, the recognition
of the messages should be measured using an instrument
based on the same criteria. In order to avoid the
artifacts that might occur in a laboratory setting, the
recognition measurement has to be indirect. This
requirement somewhat restricts the amount of information
that can be gathered in a small scale study. Another
consideration that arises from conducting research on a
small scale is the limited number of adult participants.
In order to fully examine whether or not prosocial
messages exist in Saturday morning programming, several
panels should be used and interrater reliability be
established not only within each panel but across panels.
33


Within the limitations of scales that exist, the
methodology should yield results that are indicative of
the efficacy of the training procedures used and the
reliability and validity of the survey instruments. A
shortcoming of controlled studies that do not attempt to
determine whether or not recognition of a particular type
of message has occurred in the subjects from a
communication standpoint is that the subjective processes
within the receivers is not examined. That is, the
simple observation of behavior before and after a
controlled exposure to some media event has value but
fails to take into account what the subjects of the study
think about the experience. By developing methods that
allow the subjects to express what message is being
recognized, it may be possible to describe more
completely the processes that occur during individually
selected media consumption.
Therefore, the purpose of the present study was to
develop and use methods for examining the possibility
that not all media effects are adverse. The first step
was to develop a methodology for determining whether or
not prosocial messages exist in Saturday morning
children's programs. The second step was to develop an
instrument which would indirectly measure children's
recognition of prosocial messages in those programs. The
34


third step was the execution of the methodology on a
trial basis. Lastly the results of the trial were
analyzed to determine whether or not the methodology was
viable and what improvements need to be made.
35


CHAPTER 4
METHODS
Procedures
The research methodology was built on three
elements. These were the Saturday morning television
programs, a trained adult rating panel, and fourth grade
school children. The only experimenter manipulation of
any sort was the training of adult rating panel to rate
selected programs on prosocial message content.
In order to assess the presence of prosocial
messages, a sample of Saturday morning programs was drawn
for a trained panel of adult observers to evaluate. The
programs used were selected based on Nielsen ratings.
The programs that held the top five positions for
Saturday morning ratings among six to eleven year old
children during the November 1989 rating period were
selected to maximize the likelihood of the survey
respondents having viewed them. Since there was a tie
for fifth place in the ratings, a disinterested party
blindly drew the name of the program to be used. The
five programs selected for the study and their rank
according to the ratings were: (1) "Slimer and the Real


Ghostbusters" (2) "Beetlejuice" (3) "Alvin and the
Chipmunks" (4) "Denver, the Last Dinosaur" and
(5) "Garfield and Friends." The actual episode of each
program to be recorded and shown to the adult panel was
selected by numbering the Saturdays during the period
30 December 1989 through 27 January 1990 and rolling a
die.
A panel consisting of four female upper division and
graduate students was recruited at the University of
Colorado at Colorado Springs. All of the panel members
were communication students but only one undergraduate
had some background in media effects research. The panel
had two training sessions during which interrater
reliability was established using examples of programs
not included in the study. The definitions of prosocial
behaviors given by Liebert and Sprafkin (1988) were used
in training the raters. They are:
Altruism consists of sharing, helping and
cooperation involving humans or animals.
Control of aggressive impulses involves
nonaggressive acts or statements that serve to
eliminate or prevent aggression by self or
others toward humans or animals.
Delay of gratification/task persistence consists
of the related acts of delay of gratification
and task persistence, expressed either
nonverbally or verbally.
Explaining feelings of self or others consists of
statements to another person(s) explaining
the feelings, thinking, or action of self or
others with the intent of effective positive
outcome,including increasing the understanding
37


of others, resolving strife, smoothing out
difficulties, or reassuring someone.
Reparation for bad behavior refers to behavior
that is clearly intended as reparation for an
act seen as a wrongdoing committed by the
person himself/herself.
Resistance to temptation refers to withstanding
the temptation to engage in behaviors generally
prohibited by society (e.g., stealing), which
may be prohibited in the program explicitly or
implicitly.
Sympathy is a verbal or behavioral expression
of concern for others and their problems.
(Liebert & Sprafkin, 1988, Table 10.3, p. 230)
These definitions provided the criteria to determine
the existence of prosocial messages and were explained to
the panel during the first training session. An example
of a program not broadcast on Saturday mornings was used
to illustrate behaviors that fit each of the definitions.
Raters were simply asked to count the occurrence of each
form of prosocial behavior during a practice rating done
on a different program to establish a preliminary level
of interrater reliability. The second training session
consisted of further discussion and group analysis of a
program to establish shared definitions for the panel
members. Again an example of a program not broadcast on
Saturday mornings was rated by the panel and interrater
reliability was checked. The final session for the panel
consisted of rating the selected episodes with the
commercials being skipped by means of fast forwarding the
38


tape. The programs rated aired as nominally three and one
half hours of programming.
A sample of 371 fourth graders from a local school
district was surveyed to ascertain whether or not
prosocial messages were being recognized by viewers of
the programs. There were 96 valid returns for a 25.9%
return rate. The responses came from 38 males and 57
females and one response with the gender left blank.
Demographics for the schools was not collected at the
time the surveys were distributed. When a comparison of
the respondents gender distribution to the gender
distribution of the school populations seemed to be
necessary, the school district was queried for the
information. As the term had been over for several
weeks, the data files had been purged and the information
was not available in a usable form. One survey was
returned without being completed and one respondent
indicated that neither of the programs had been watched.
Neither of these was counted as a valid response.
The distribution of the number of responses for each
program was not related to the Nielsen ratings. "Alvin
and the Chipmunks" which was rated number three had 25
responses, "Slimer and the Real Ghostbusters" rated
number one had 22 responses, "Beetlejuice" and "Garfield
and Friends" rated number two and five respectively had
39


18 responses each, and "Denver, the Last Dinosaur" rated
number four had 13 responses.
Fourth graders were determined to be capable of
reading the simple questions used on the survey
instrument. Further, only three simple answer choices of
yes, no, and don't know were provided (See Appendix B).
Two program surveys were given to each student in order
to increase the probability of getting one valid program
evaluation from each respondent. For those indicating the
same frequency of viewing for both programs, a coin flip
was used to determine which one would be used in the data
analysis. For differing frequencies of viewing, the
program watched most was used.
Instruments
The panel used a tally sheet with each of the
individual prosocial behaviors listed and briefly
defined. While viewing the program episodes selected,
they noted the occurrence of each behavior (See
Appendix A).
The survey forms for the fourth graders were
tailored to the programs. That is, each form had the name
of the program at the top and the questions pertained to
that program's title character (See Appendix B). Prior
to distribution, the questionnaires were paired equally
40


and shuffled so that the programs any particular child
was asked about occurred randomly. The two
questionnaires were stapled to a parental permission form
prior to being distributed. The sets of questionnaires
were broken down prior to being delivered to the
participating schools and were picked up after being
returned to the schools by those who chose to
participate.
The seven questions about the title characters
paralleled the Liebert and Sprafkin prosocial behaviors
listed on the trained rating panel's sheets and the
respondents were asked whether or not the main character
of the program would choose to behave in that manner.
For example, on the altruism behavior the respondents
were asked to reply yes, no, or don't know to the
statement, (Character Name) likes to share and help
others. Additionally, the respondents were asked if
they would like to have the title character as a friend,
how often they viewed the program in question, and their
gender and age (See Appendix B).
Data Analysis
The rating panel tally sheets were compiled to
determine a rate per program of each prosocial behavior
and a total rate. These results were put into a common
41


measurement on a per hour basis for ranking the programs
according to the rate of occurrence of prosocial
behaviors. Interrater reliability was established by
intercorrelating the four raters evaluations of each
program. Specifically, each rater's count of each type
of prosocial behavior were intercorrelated with the other
raters' counts of each type of prosocial behavior. This
yielded six correlations per program which were converted
to z scores and averaged to arrive at interrater
reliability for each program and across all programs.
The data from the children's survey instruments were
used to calculate a total positive rating score. This
score was the total of positive answers given by the
respondents for each program. A prosocial ratio score
was also determined for each program. This was done by
dividing the number of positive answers by the total of
the positive and negative answers. This ratio was used
to determine the prosocial message recognition rank of
the programs.
A one between analysis of variance was done to
determine if significant relationships existed among the
program rankings as determined by the adult panel and the
survey respondents' evaluations. The dependent variable
in this procedure was the student derived positiveness
ratio for each program. The independent variable was the
42


prosocial ranking of the program as determined by the
adult panel. Thus, if the children were recognizing
prosocial messages in the same way as the adults, the
ordering of the children's positiveness ratio means
should be the same as the prosocial ordering of the
adults.
Analyses of variance were also done using the
respondent's gender and the answer to the question about
whether or not they would like to have the main character
of the program as a friend as independent variables. The
positive answer totals for the various behaviors were
used as dependent variables in order to determine whether
there was more recognition of any of the behaviors based
on gender or the desirability of the character as a
friend.
In order to explore the relationship between whether
or not the various characters would be desirable as
friends and the number of positive opinions assigned to
the programs by the children, the correlation between
these items was checked. The number of positive opinions
was counted and used as a score in this case and the
friend dimension was left on its one to three scale.
Finally, a rank order correlation was done between
the prosocial rankings assigned by the adult panel and
children's prosocial rankings of the programs. A
43


positive correlation would indicate similarities in
adult-child recognition of prosocial messages while a
negative correlation would indicate dissimilarity.
44


CHAPTER 5
RESULTS
The first item that needed to be established was the
interrater reliability of the adult panel. The reliable
recognition of the occurrence of prosocial behaviors in
the programs was the basis for determining whether or not
the methodology used in this project was viable. The
statistical operation used for establishing reliability
in this case was the average intercorrelation of the
scores given by the four panel members. The resultant
average intercorrelations among raters are listed in
Table 5.1. The overall good intercorrelation among adult
raters indicates that the training sessions established a
shared set of definitions.
Table 5.1 Interrater Reliability
Program Title
Garfield and Friends
Slimer and the Real Ghostbusters
Denver, the Last Dinosaur
Alvin and the Chipmunks
Beetlejuice
Across all programs
Correlation
0.935
0.915
0.910
0.890
0.775
0.895


The question of the occurrence of prosocial
behaviors in Saturday morning children's programs that
were evaluated was the next consideration. The rate of
occurrence differed widely across the set of programs but
all were determined to portray at least some prosocial
behaviors. The adult panel means for the number of
prosocial occurrences on a per hour basis is given in
Table 5.2.
Table 5.2 Adult Rating Panel Observation of Prosocial
Behaviors
(Panel mean of occurrences per hour)
Rank Procrram Title Score
1. Alvin and the Chipmunks 48.5/hour
2. Beetlejuice 48.0/hour
3. Garfield and Friends 26.75/hour
4. Slimer and the Real Ghostbusters 23.7 5/hour
5. Denver, the Last Dinosaur 2 2.0/hour
In order to establish a basis for comparing what the
opinions of the fourth graders about the probability of
the main characters of the programs behaving in a
prosocial manner with the results of the adult panel's
evaluation, an index of -some sort was necessary. The
children's total positive answers about the main
characters for each program were divided by the total of
46


the positive and negative answers for each program to
determine a "positiveness ratio" for each program. The
"Don't Know" answers were considered to be a "no opinion"
and omitted from the calculation. The resultant ratios
established the ranking given in Table 5.3.
Table 5.3 Fourth Grader Ranking of Positiveness Ratios*
Rank Proaram Title Ratio
1. Denver, the Last Dinosaur 0.98
2. Slimer and the Real Ghostbusters 0.74
3. Alvin and the Chipmunks 0.42
4. Beetlejuice 0.34
5. Garfield and Friends 0.24
*(Ratio=Positive Answers/Positive+Negative Answers)
In a one between analysis of variance the number of
positive answers given for the program rated by each
respondent was entered as a separate case. This was used
as the dependent variable with the adult ranking of the
programs as the independent variable. The results
indicated that significant differences, F(4,92) = 20.86
p<.001, existed in the number of positive answers given
for each program. Specifically the results in Table 5.2
indicate that adults ranked the prosocialness of programs
from most to least as: (1) "Alvin and the Chipmunks"
47


(2) "Beetlejuice" (3) "Garfield and Friends" (4) "Slimer
and the Real Ghostbusters" (5) "Denver, the Last
Dinosaur." In contrast, the results in Table 5.3
revealed that children ranked the prosocialness of the
lead characters in these programs from most to least as:
(1) "Denver, the Last Dinosaur" (2) "Slimer and the Real
Ghostbusters" (3) "Alvin and the Chipmunks"
(4) "Beetlejuice" (5) "Garfield and Friends." The Tukey
HSD procedure indicated that the differences at the p<.05
level existed between both "Beetlejuice" and "Garfield
and Friends" and the other three programs. This
procedure was indicated by the sharp differences in the
ordering of the programs according to prosocial content
between the adult panel and the children. The results
indicate that the children rated the programs as being
significantly different in content.
Rank order correlation between the adult and
children's prosocial rankings of the five programs
indicated that the adults and children have different
opinions about the amount of prosocial behavior that
occurred in the programs (Spearman's rho = 0.60,
p0.164)* The correlation did not approach acceptable
significance due to a small sample of only five programs.
Analysis variance using gender as the independent
variable revealed significant differences in the
48


children's answers on two of the behavior descriptions.
The first was reparation for bad behavior which on the
survey constituted apologizing for wrongdoing,
F(l,91) = 6.45 p<.013. The second was sympathy which on
the survey was helping others to feel better,
F(1,91) = 8.89 p<.004. Males rated the characters as
exhibiting both of these traits more than did the
females.
For analysis of variance with the answer to whether
or not the respondent would like to have the main
character as a friend, there were significant differences
in the answers to the same two behavior descriptions.
That is, reparation for bad behavior, F(2,91) = 4.56
p<.013, and sympathy, F(2,91) = 6.76 p<.002, again were
the dimensions that were rated differently. Those who
would like the character as a friend rated them as
exhibiting both the traits more than did those who did
not desire the character as a friend.
The correlation between the total of positive
opinion answers assigned to the programs by children and
their desire to have the character as a friend was
contrary to expectations. The negative correlation of
-0.444 p<.001 across all programs, between these items
indicates that the less perceived prosocial behavior by a
character predicts greater desire by children to want the
49


character as a friend. This is well illustrated by the
fact that the program "Garfield and Friends" ranked
lowest on the prosocial scale derived from the children's
opinions and yet 13 of 18 total respondents indicated
they would like Garfield as a friend.
50


CHAPTER 6
CONCLUSIONS
The goal of this research was to develop a reliable
and valid methodology in order to examine the existence
and recognition of prosocial messages in Saturday morning
children's television programs. The bulk of previous
investigation of the effects of television viewing on
children has been focused on the possible detrimental
effects. The research that has explored the possibility
of prosocial effects occurring has been done largely in
laboratory settings and has used preschool age children
as subjects. In order to determine whether or not
prosocial messages exist in Saturday morning programs, it
was necessary to use adults as raters in order to
increase the objectivity of the observations. Fourth
grade students, who are of an age to understand that the
programs are fiction, were surveyed about their opinions
of how the characters in top rated Saturday morning
programs behave in order to establish whether or not the
messages were being recognized by the audience members.
The investigator did not control the programs students
viewed thereby minimizing experimental effects. Using a
national rating service's ratings to determine what


programs were to be investigated also eliminated
researcher bias from that part of the process. In effect
two parts of a communication process were investigated.
First, the content of a message, that is, the occurrence
of prosocial behaviors within the programs, was examined
as objectively as possible. Second, the receivers'
recognition of the content was indirectly measured by
soliciting their opinions about how the main characters
behave.
The question of whether or not the Liebert and
Sprafkin typology provided the basis for a reliable
method of measuring the prosocial message content of
Saturday morning children's programs can be at least
partially resolved. The high degree of interrater
reliability among the members of the adult rating panel
indicates that agreement can be reached about the meaning
of a set of criteria for such measurement. Although
unknown factors may have contributed to the ease of
establishing the common definitions that the panel used
in their rating, it is probable that a high degree of
interrater reliability could be established with
appropriate training of any similar rating panel.
The success with establishing a high degree of
interrater reliability among the adult panel members
implies that the occurrence of prosocial behaviors in the
52


program episodes that were rated is a valid finding.
This finding can be used to support the hypothesis that
prosocial models exist for the audience members to
emulate if other individual psychographic factors provide
the initial influence for such behavior. An important
fact to keep in mind is that the presentation of a
prosocial model for a child to emulate is not likely in
itself to cause an alteration of behavior patterns. The
influence of primary social agents would in large part
determine what behaviors would be acceptable for
emulation and this would provide the criteria that a
child used for categorizing the modeled behaviors as to
acceptability.
The survey instrument used in assessing the
recognition of prosocial messages by children yielded
somewhat ambiguous results. Although the respondents
gave positive answers about at least some of the
behaviors in question for all the programs, their ranking
of which programs contained the most prosocial content
varied widely from the rankings assigned by the adult
raters. It can be concluded from this that the
subjective recognition of the prosocial messages by the
children differs from the trained objective observation
of the programs by the adult raters. This may be a
function of identity with the characters in question.
53


Adult raters would have little or no identification with
the characters during a one-time observation of an
episode of a program while the regular child viewer would
develop identification with the main character over time.
The children's opinion of how a character behaves would
evolve from seeing how the character acted in several
episodes of a program. If the episode viewed by the
adult rating panel happened to be one in which the
prosocial content was low, differences would occur in the
rankings assigned.
Other possible processes which might account for the
fact that the children's opinions differed from the adult
counting of behaviors are recall and context. The
children's opinions would be based on the recall of final
outcomes of stories rather than concentration on
detecting and counting occurrences in a single episode.
This would tend to blend the memory of individual
antisocial acts that might have occurred into the context
of a prosocial outcome such as ultimately helping another
to solve a problem or escape from a perilous situation.
The differences among the children's opinions of
whether or not the main characters exhibit various
prosocial behaviors occurred in two areas. Reparation
for bad behavior and sympathy were rated differently
according to the gender of the child and whether or not
54


the character was desirable as a friend. These
differences could be attributed to the socialization
process. Nine and ten year old fourth graders have
probably started to develop gender related social
characteristics. Desirable characteristics in a friend
could be expected to be different for girls and boys
depending on what sort of role modeling had been done by
primary social agents.
The negative correlation between the overall
positive answer score for the programs and the
desirability of the characters as friends must also be
attributed to identification with the characters. Over a
period of time, it is reasonable that children would not
repeatedly choose to watch a particular program unless
they felt some sort of friendship based on identification
with the main character.
The method used in this project to determine whether
or not prosocial messages are included in the content of
Saturday morning children's television seems to be
reliable and valid. High interrater reliability was
established and the programs that were rated were found
to have some prosocial message content. The method used
to determine whether or not these messages are received
by the audience appears to have worked to a degree but
did not yield sufficiently clear results to suggest that
55


the prosocial messages are being recognized, if the
adult rankings are used as a standard of measurement, it
could be argued that the differences between the adult
panel and the children indicate that the children do not
recognize the prosocial messages in Saturday morning
programming. This is most likely a reflection of the
different ways in which the measurement was done. The
adult rating panel was counting overall occurrences to
include the actions of supporting characters in the
programs. This would mask somewhat the fact that the
main character was not behaving in what would be
considered a prosocial manner. The children were
restricted to evaluating the main character and this
would mask the prosocial actions of supporting
characters. In order to resolve this problem of
measurement, it would be necessary to word the survey
questions to cover the entire cast of characters in a
program or to have the adult rating panel count only
prosocial behaviors exhibited by the main character.
However, the fact that the children's opinions of the
prosocial content of the various programs were
significantly different indicates that their responses
were nonrandom. This finding suggests some sort of
systematic recognition of a perceived prosocial pattern
of behavior on the part of the characters by the
56


children. The finding of an inverse correspondence
between the adults' and children's prosocial rankings
muddies the finding that the children recognized some
prosocial messages and requires further investigation
with more comprehensive research tools.
The questions which arose during the analysis of the
results of the survey administered to the children
concern the validity of the method. The first problem
area which must be noted is the fact that the surveys
were accompanied by a parental consent form and were
completed by the children at home. This may have led to
consulting with the parents about the meanings of the
questions and what the "appropriate" answer might be.
This would introduce some parental bias into the answers
given and would not be detectable in the course of
analyzing the results. In order to avoid this, it would
be better to obtain parental permission separately from
the administration of the survey instrument.
The necessarily simple wording of the questions on
the survey was another problem area. Asking the children
their opinion of overall behavior patterns makes it easy
for them to answer but does not elicit a truly thoughtful
response. An instrument consisting of various scenarios
with the character name inserted would provide a method
for more precisely measuring children's opinions of a
57


character's behavior patterns. Additionally, it is
difficult to determine whether or not the behavioral
descriptions used carry the same connotation in the
children's minds as in the mind of an adult. Resolving
this problem would require interviewing a large sample of
children in order to determine what best describes to
them the various behavior patterns being investigated.
Once the most common connotations used by children in a
particular age group were determined, the questions used
in soliciting their opinions could be worded in a way
that would make them a more valid measurement.
Another extension of the opinion measurement would
be to recruit the parents of the children to also view
the programs and evaluate the main characters on the
Liebert and Sprafkin typology. This Would allow for
cross referencing the opinions of the parents and their
children to determine what similarities exist between the
opinions of primary social agents and the children using
a common basis of measurement. The results obtained
would provide indications of how much influence is
exerted by primary social agents on the attitudes of
children. This two pronged approach could also provide
indications as to how much the training of a rating panel
contributes to the reliability of its measurements.
Comparing the results obtained from two different adult
58


groups with those of a group of children would yield more
comprehensive findings on which to base conclusions and
determine what the next level of refinement should be in
order to more firmly establish the existence and
recognition of prosocial messages in Saturday morning
programming.
The results of this research indicate that the
prosocial messages do exist in Saturday morning
children's television programs and that it can be
reliably measured. There were also indications that
children recognize these messages. The reliability and
validity of the method used to measure the children's
recognition of prosocial messages needs to be greatly
improved over what was used in this project. This
improvement would require more resources to be available
to conduct the survey of children's opinions. The
possible influence of parental bias needs to be
eliminated by administering the survey outside the home.
Using a national rating service's ratings to select
subject programs and not controlling the children's
viewing should be retained in order to prevent researcher
bias from influencing the results.
It is evident that examining the possible effects of
television viewing on children in a naturalistic manner
is achievable. Taking such research outside the
59


laboratory and controlled field experiment setting
provides a more realistic picture of what children are
recognizing as the message of what they view. The
communication concept of the meaning of a message
residing in the receiver can be applied only by
determining what meaning is given to the message content
of television programs by the receiver, in this case the
children who are the intended audience of those who
produce what is broadcast on Saturday mornings.
60


APPENDIX A Adult Rating Panel Coding Sheet
CODING SHEET
NAME OF PROGRAM________________________________________
DATE TAPED_____________________________'____
As you note the occurrence of one of the listed
behaviors, mark its occurrence in the count column for
that behavior.
MESSAGE TYPE COUNT
ALTRUISM sharing, helping & cooperation involving
humans or animals.
CONTROL OF AGGRESSIVE IMPULSES nonaggressive acts
or statements that prevent aggression.
DELAY OF GRATIFICATION/PERSISTENCE acts of
delaying gratification & task persistence.
EXPLAINING FEELINGS statements explaining the
feelings, thinking or action of self or others.
REPARATION FOR BAD BEHAVIOR behavior intended as
reparation for wrongdoing by person.
RESISTANCE TO TEMPTATION ^ withstanding the
temptation to engage in prohibited behaviors.
SYMPATHY verbal or behavioral expression of
concern for others.
61


APPENDIX B Fourth Grade Student Survey Instrument
I watch (name of program) ALMOST ALWAYS SOMETIMES
ALMOST NEVER HAVEN'T SEEN IT
(Character name) likes to share
and help others. YES NO DON'T KNOW
(Character name) likes to fight. (Character name) keeps working YES NO DON'T KNOW
until the job is done before playing. YES NO DON'T KNOW
(Character name) helps settle arguments by explaining why they started. YES NO DON'T KNOW
(Character name) apologizes when he/she does something wrong. YES NO DON'T KNOW
(Character name) takes things that belong to someone else. YES NO DON'T KNOW
(Character name) would help someone who was feeling bad to feel better. YES NO DON'T KNOW
I would like (Character name) to be my friend. YES NO DON'T KNOW
I AM A BOY GIRL WHO IS YEARS OLD.
62


BIBLIOGRAPHY
Anderson, J. A., & Meyer, T. P. (1988). Mediated
communication a social action perspective. Newbury
Park, CA: Sage Publications.
Bandura, A. (1977). Social learning theory. Englewood
Cliffs, NJ: Prentice-Hall, Incorporated.
Becker, S. L. (1987). Discovering mass communication
(2nd ed.). Glenview, IL: Scott, Foresman and Company.
Berkowitz, L. (1984). Some effects of thoughts on anti-
and prosocial influences of media events: A cognitive-
neoassociation analysis. Psychological Bulletin. 95, 410-
427.
Bryan, J. H., & Walbek, N. H. (1970a). Preaching and
practicing generosity: Children's actions and reactions.
Child Development. 41. 329-353.
Bryan, J. H., & Walbek, N. H. (1970b). The impact of
words and deeds concerning altruism upon children. Child
Development. 41, 747-757.
Christenson, P. G. (1986). Children's perceptions of
moral themes in television drama. In M. L. McLaughlin
(Ed.), Communication Yearbook 9. Beverly Hills, CA: Sage
Publications, Incorporated.
Collins, W. (1982). Cognitive processing in television
viewing. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.),
Television and behavior: Ten years of scientific progress
and implications for the eighties (vol. 2) (pp. 9-23).
Rockville, MD: National Institute of Mental Health.
Comstock, G. (1982). Violence in television content: An
overview. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.),
Television and behavior: Ten years of scientific progress
and implications for the eighties (vol. 2) (pp. 108-
125). Rockville, MD: National Institute of Mental
Health.


Cook, T. D., Kendzierski, D. A., & Thomas, S. V. (1983).
The implicit assumptions of television research: An
analysis of the 1982 NIMH report on television and
behavior. Public Opinion Quarterly. 47, 161-201.
Corder-Bolz, C. (1980). Mediation: The role of
significant others. Journal of Communication. 30(3), 106-
118.
Corder-Bolz, C. (1982). Television literacy and critical
viewing skills. In D. Pearl, L. Bouthilet, & J. Lazar
(Eds.), Television and behavior: Ten years of scientific
progress and implications for the eighties (vol. 2) (pp.
91-103). Rockville, MD: National Institute of Mental
Health.
Dorr, A. (1982). Television and affective development
and functioning. In D. Pearl, L. Bouthilet, & J. Lazar
(Eds.), Television and behavior: Ten years of scientific
progress and implications for the eighties (vol. 2) (pp.
68-77). Rockville, MD: National Institute of Mental
Health.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude.
intention and behavior: An introduction to theory and
research. Reading, PA: Addison-Wesley Publishing
Company.
Forge, K. L. S., & Phemister, S. (1987). The effect of
prosocial cartoons oh preschool children. Child Study
Journal. 17(2). 83-88.
Freedman, J. L. (1984). Effect of television violence on
aggressiveness. Psychological Bulletin. 92., 227-246.
Freedman, J. L. (1986). Television violence and
aggression: A rejoinder. Psychological Bulletin. 100.
372-378.
Friedrich, L. K., & Stein, A. H. (1975). Prosocial
television and young children: The effects of verbal
labeling and role playing on learning and behavior.
Child Development. 46. 27-38.
Friedrich-Cofer, L., & Huston, A. C. (1986). Television
violence and aggression: The debate continues.
Psychological Bulletin. 100. 364-371.
64


Friedrich-Cofer, L. K., Huston-Stein, A. H., Kipnis, D.
M., Susman, E. J., & Clewett, A. S., (1979).
Environmental enhancement of prosocial television
content: Effects on interpersonal behavior, imaginative
play, and self-regulation in a natural setting.
Developmental Psychology. 15(6), 637-646.
Fry, D. L., & Fry, V. H. (1986). A semiotic model for
the study of mass communication. In M. L. McLaughlin
(Ed.), Communication Yearbook 9. Beverly Hills, CA:
Sage Publications, Incorporated.
Gerbner, G. (1971). Violence in television drama:
Trends and symbolic functions. In G. A. Comstock, &
E. A. Rubinstein (Eds.), Television and social behavior
reports and papers volume I: Media content and control.
Rockville, MD: National Institute of Mental Health.
Gorn, G. J., Goldberg, M. E., & Kanungo, R. N. (1976).
The role of educational television in changing intergroup
attitudes of children. Child Development. 47, 277-280.
Hirsch, P. M. (1981). On not learning from one's own
mistakes, a reanalysis of Gerbner et al.'s findings on
cultivation analysis part II. Communication Research.
8(1), 3-37.
Huesmann, L. R. (1982). Television violence and
aggressive behavior. In D. Pearl, L. Bouthilet, &
J. Lazar (Eds.), Television and behavior: Ten years of
scientific progress and implications for the eighties
(vol. 2) (pp. 126-137). Rockville, MD: National Institute
of Mental Health.
Huesmann, L. R., Eron, L. D., Klein, R., Brice, P., &
Fischer, P. (1983). Mitigating the imitation of
aggressive behaviors by changing children's attitudes
about media violence. Journal of Personality and Social
Psychology. 44. 899-910.
Huesmann, L. R., Lagerspetz, K., & Eron, L. D. (1984).
Intervening variables in the TV violence-aggression
relation: Evidence from two countries. Developmental
Psychology. 20. 746-775.
Idea of boycotting TV advertisers gaining. (1989, June.
3). Colorado Springs Gazette Telegraph, pp. El, E6.
65


Koop takes aim at alcohol. (1989, June 1). Colorado
Springs Gazette Telegraph, p. A6.
Liebert, R. M., & Sprafkin, J. (1988). The early window:
Effects of television on children and youth (3rd ed.).
New York: Pergamon Press.
Lowery, S., & DeFleur, M. L. (1988). Milestones in mass
communication research: Media effects (2nd ed.). New
York: Longman Inc.
Loye, D., Gorney, R., & Steele, G. (1977). Effects of
television: An experimental field study. Journal of
Communication. 27(3), 206-216.
Murray, J. P., Rubinstein, E. A., & Comstock, G. A.
(Eds.), (1972). Television and social behavior. Reports
and papers vol II: Television and social learning.
Rockville, MD: National Institute of Mental Health.
National Institute of Mental Health. (1982). Television
and behavior: Ten years of scientific progress and
implications for the eighties. (DHHS Publication No. ADM
82-1195). Washington, DC: U. S. Government Printing
Office.
Nikken, P., & Peeters, A. L. (1988). Children's
perceptions of television reality. Journal of
Broadcasting and Electronic Media. 32. 441-452.
Norusis, M. J./SPSS Inc. (1988). SPSS-X(tm) Advanced
statistics guide. (2nd ed.). Chicago: SPSS Inc.
Poulos, R. W., Harvey, S. E., & Liebert, R. M. (1976).
Saturday morning television: A profile of the 1974-75
children's season. Psychological Reports, 39, 1047-1057.
Rosenthal, T. L., & Zimmerman, B. J. (1978). Social
learning and cognition. New York: Academic Press, Inc.
Rushton, J. P. (1979). Effects of prosocial television
and film material on the behavior of viewers. In L.
Berkowitz (Ed.), Advances in experimental social
psychology. Vol. 12. New York: Academic Press, pp.
321-351.
66


Rushton, J. P. (1982). Television and prosocial
behavior. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.),
Television and behavior: Ten years of scientific progress
and implications for the eighties (vol. 2) (pp. 248-
257). Rockville, MD: National Institute of Mental
Health.
Schramm, W., Lyle, J., & Parker, E. B. (1961).
Television in the lives of our children. Stanford, CA:
Stanford University Press.
Singer, D. G. (1982). Television and the developing
Imagination of the child. In D. Pearl, L. Bouthilet, &
J. Lazar (Eds.), Television and behavior: Ten Vears of
scientific progress and implications for the eighties
(vol 2) (pp. 39-52). Rockville, MD: National Institute
of Mental Health.
Singer, J. (1980). The power and limitations of
television: A cognitive-affective analysis. In P. H.
Tannenbaum (Ed.), The entertainment functions of
television. (pp. 31-65). Hillsdale, NJ: Lawrence
Erlbaum Associates, Inc. Publishers.
Singer, J. L. and Singer, D. G. (1976). Can TV stimulate
imaginative play? Journal of Communication. 46(3).
74-80.
Stein, A. H., Friedrich, L. K., & Vondracek, F. (1972).
Television content and young children's behavior. In J.
P. Murray, E. A. Rubinstein, & G. A. Comstock (Eds.),
Television and social behavior: Reports and papers. vol
II: Television and social learning. Rockville, MD:
National Institute of Mental Health.
Stein, G. M., & Bryan, J. H. (1972). The effects of a
television model upon rule adoption behavior of children.
Child Development. 43, 268-273.
Surgeon General's Scientific Advisory Committee on
Television and Social Behavior. (1971). Television and
growing up: The impact of televised violence. Report to
the Surgeon General. Washington, DC: United States
Public Health Service.
67


Tamborini, R., Zillmann, D., & Bryant, J. (1985). Fear
and Victimization: Exposure to television and
perceptions of crime and fear. In R. N. Bostrora, fit B. H.
Westley (Eds.), Communication Yearbook 8. Beverly Hills,
CA: Sage Publications, Incorporated.
Tan, A. S. (1986). Mass communication theories and
research (2nd ed.). New York: MacMillan Publishing
Company.
Tannenbaum, P. H. (Ed.), (1980). The entertainment
functions of television. Hillsdale, NJ: Lawrence
Erlbaum Associates, Inc. Publishers.
Walker, K. B., & Morley, D. D. (1988). Attitudes and
parental factors as intervening variables in the
television violence-aggression relation. Unpublished
manuscript, University of Colorado at Colorado Springs,
Colorado Springs, CO.
Walker, K. B., St Morley, D. D. (in press). Adolescent
perceptions of context in violent media portrayals.
Speech Communication Annual.
Williams, F. (1986). Reasoning with statistics. (3rd
ed.). New York: Holt, Rinehart and Winston.
Wood, D. N. (1983). Mass media and the Individual. St.
Paul, MN: West Publishing Company.
68


both patient and physician, di
manding or controlling patient, patient had
special problems (alcohol, drugs, chronic pain).
Another communication problem identif|ed by physicians in frustrating visits was a
feeling of distress on their part during or ifter seeing the patient.
For each of the seven communicai ion problems identified by the physicians,
Levinson et al. (1993) discussed how each could be addressed by developing
communication skills that are oriented toward a more patient-centered, collaborative,
mutual participation model of clinical care. Characteristics of this model included: use
of open-ended questions, allowing the patient to talk uninterrupted at the outset,
active listening, repeating and summarizing the patients problems, avoiding the use of
medical jargon, and involving the patient in decision-making about treatment options.
They particularly noted the danger of the tendency for physicians to blame the patient;
and, concurrently emphasized the importance of discussion and education for
physicians in enhancing physician-patient clmmunications.
White, Levinson and Roter (1994) provided insight into one of the most
common challenges encountered by practicing physicians and one of the frustrations
identified by physicians in the Levinson et al (1993) study: the oh by the way...
phenomenon which occurs at the end of the medical visit. The objectives of the White
et al. (1994) study were to define and describe the physician-patient communications
during the closing phase of the medical visit
and to identify the types of
63


communication during the visit that might have contributed to the introduction of a
new problem at the close of the visit.
Audiotapes of 88 outpatient visits
with 20 primary care physicians were
analyzed for communication patterns using the Roter Interactional Analysis System.
The patients were predominantly white, highly educated (80% had some college
education), females (61%).
Closure was initiated by physicians
whether or not there were any more questions 25% of the time. In 21% of the
86% of the time and physicians asked
closures, patients presented new problems
not previously discussed. Both physician
and patient behaviors throughout the visit were associated with the presenting of new
problems during closure. The specific phj
sician behaviors associated with the
reduction in the occurrence of new problems at closure were: orienting the patient to
the flow of the visit, giving patients more information about the treatment plan, and
asking about patient beliefs. The specific patient behaviors associated with reduced
occurrence of new problems at closure wers: talking more about their therapy and
sharing emotions and feelings. Interestingly, the findings of this study are closely
related to the observational studies by Beckman and Frankel (1984) demonstrating
that patients are interrupted on average 18 seconds into their opening narrative; and,
the first problem presented may not be the primary problem.


Similar to other suggestions and
models for a more participatory, collaborative,
patient-centered approach to medical care, White et al. (1994) proposed a set of
communication skills to use before and during closure which may be helpful to
physicians in decreasing the frequency of new problems presented at closure. Before
closure: orient patient to the flow of the visit, address emotions and psychological
issues, explore patient beliefs, check for understanding, ask if the patient has additional
questions. During closure: orient the patient to the closure, provide a summary of the
visit, complete a final check for understanding, clarify the plan, provide encouragement
and reassurance, and consolidate rapport/caring.
Finally, a noteworthy practical observation was made regarding the concern
with time pressures in the everyday practice of medicine. These researchers pointed
out that instead of asking open-ended questions and whether or not the patient had
additional questions during closure, it may be more efficient to ask these questions
during the visit and before closure begins. Implicitly embedded in this observation is
the possibility that by using the aforementioned communication skills, overall clinical
efficiencies in terms of total encounter time may be achieved as well.
Doctor-Patient Communications: Effects on Outcomes
We have entered into an era of accountability in health care where the ultimate
purchasers of care (government and employers) are placing emphasis on value
65


enhancement. The notion of value enhancement recognizes that the cost of delivering
medical care and the associated resource
jtilization can be reduced even as quality is
improved. Because quality and value enhancement may become central to provider
differentiation and competitive advantage
in the market place, the importance of
providers understanding the impact of physician-patient communications on quality,
value enhancement, and outcomes is paramount.
Different dimensions have been used to define patient outcomes: satisfaction,
clinical indicators, compliance/adherence, quality of life, health status, patient-derived
measures of functional status, etc. The purpose of this section is to review some of
the impacts of doctor-patient communications on various outcome dimensions.
In their review of the literature, Onj
salient findings regarding the relationship o
et al. (1995) noted some of the more
f patient satisfaction, compliance, and
health outcomes to physician-patient communications:
Patients are frequently dissatisfied vrith the information they receive and
physicians tend to underestimate paiient desire for information.
Information-giving by providers to patients was significantly related to patient
satisfaction.
Time spent in discussing preventive care was positively associated with patient
satisfaction.
More time spent in the medical interview was positively associated with patient
satisfaction.
A more dominant and controlling physician communication style was
associated with less patient satisfaction.
66


Non-verbal behaviors such as frequent eye contact, appearing interested etc.
were positively associated with patient satisfaction.
Positive associations have been noted between patient-centered care behaviors
and patient satisfaction, compliance, resolution of patient concerns.
Compliance has been associated with more information-giving and positive
talk.
Patient-centered behaviors were related to higher self-reported compliance.
More patient controlling behaviors, less physician controlling behaviors and
more information-giving by physi patients were related to improved
and subjectively.
patient health status clinically, functionally,
In cancer patients, lack of information seems to be related to increased coping
problems, depression, anxiety, and uncertainty.
In her commentary on the effects of the physician-patient relationship on the quality of
health care, DiMatteo (1994) denoted similar findings:
Patient satisfaction with their physicians behavior has been positively
associated with the continuity of tile relationship.
More adequate histories have been| associated with higher patient satisfaction
and less delay in reporting symptoms.
Enhanced perceived personal control over health with collaboration and
TBDS.
Increased patient responsibility for jmedical decisions have been related to
lower likelihood of blaming the physician if things go wrong.
Increased adherence by patients when physicians provide more information.
Physician interpersonal characteristics and non-verbal communications have
been associated with patient health behaviors.
67


Patient outcomes can be effected by positive expectations about efficacy by
both the physician and the patient.
Improved recovery from surgery, decreased use of pain medications, shortened
hospital stays, improved clinical si:atus, improved symptoms, and enhanced
management of chronic conditions have all been associated with effective
physician-patient communication.
Roberts, Cox, Reintgen, Baile and Gibertini (1994) studied the influence of
physician communication on the psychological adjustment and decision-making of 100
post-surgical breast cancer patients. The Cancer Diagnostic Interview Scale (CDIS)
was used to assess the patients perceptions of their surgeons behavior during
the discussion in which they were informed of their cancer diagnosis. The SCL-90-R
Scale was used to assess psychological status.
Results from the CDIS assessment
revealed that the physicians caring attitude
was perceived as the most important factor related to psychological adjustment while
information giving was a much weaker factor. Those physician behaviors most closely
associated with a caring attitude were: I felt my doctor cared about me as a person,
my doctor understood my fears and concerns and my doctor is a warm and caring
person. The authors concluded that the use of psychotherapeutic techniques by
physicians can have a significant positive effect on a patients psychological well-
being. Based on this research and that from the social sciences literature, they
provided guidelines for what the surgeon should do in conducting the cancer
diagnostic interview: select an appropriate tune and place with the ideal being a
68


prearranged return office visit, verbalize empathy with direct eye contact, allow the
patient sufficient time to absorb the news and react emotionally, provide information
about the cancer and the treatment options, engage the patient in decision-making
about treatment.
In a very comprehensive study, Fox and Stein (1991) utilized random digit
dialed telephone interviews to investigate the effect of physician-patient
communication on mammography utilization in a random sample of 1,057 black,
Hispanic, and white women (mean age o:
53 years). Seven questions from the
interview were used to secure information about the quality of physician-patient
communication regarding mammography
Results revealed that 71% ofHisp anic women had never had a mammogram
and that more than half (55.3%) of the Hispanic women over age 50 years had never
had a mammogram. The most important predictor of whether women of all racial
groups had a mammogram, at any time or within the last year, was whether their
doctors had discussed mammography with them.
The data also suggested that within the physician-patient communication
process, physicians were not effectively communicating information about screening
and were discussing screening less frequently with Hispanic women than with white or
black women. The authors offered several possible explanations for this. First, fewer
than half of the Hispanic women had annu
d clinical breast exams that reduced the
69


opportunity for mammography referrals. Second, physicians were not using pediatric
visits for their children to offer family care. Third, Hispanic women tended not to
view the doctor as the most believable source of health information. Fourth, Hispanic
women cited more barriers to and anxieties about the use of mammography than did
black or white women. Finally, language differences may have created communication
difficulties. With the rapid population growth and purchasing power of the Hispanic
population in many concentrated areas ofjthe U.S. (Los Angeles, Houston, Phoenix,
Chicago, New York and others), the implications for provider organizations are
apparent.
In summary, the research reviewed here has provided evidence that patients,
and to some extent physicians, have been
patient communications and the physician
dissatisfactions have been borne primarily
frustrated with several aspects of doctor-
patient relationship. These frustrations and
out of traditional doctor-patient
relationships characterized by a Parsonian, physician-centered, paternalistic model
whereby the patient has played a predomii
antly passive, dependent role and the
physician has assumed an autonomous, professionally dominant and directive role in
the health and medical care decision-making process.
It is clear from the evidence presen
and non-physician researchers continue to
ted here that patients as well as physician
advocate a physician-patient relationship
and communication style characterized by informed collaboration, mutual
70


participation, and patient-centeredness. Most importantly, it is also clear from the
evidence presented here that such approaches can have profound effects on societal
outcomes of interest: enhanced patient satisfaction, improved clinical and functional
health status for patients, cost efficiencies, and possibly, enhanced job satisfaction for
physicians.
Given these implications, Ong et
al. (1995) noted that an important goal would
a theoretical framework would be useful
approach for improving physician-patien
be to establish a systematic theory and model of doctor-patient communication. Such
for: 1) providing a systematic, organized
communications and relationships; and 2)
generating hypotheses regarding such relationships. Thus, the purpose of the
following section is to review some models of doctor-patient communications and
relationships which have been championed by researchers and practitioners.
Collectively, these models may point us toward the theoretical framework and
practical models needed for advancing knowledge in this area; and subsequently, for
the successful reengineering of doctor-patient communications and relationships.
Models for Physician-Patient Communication and
Relationships |
In creating interpersonal relationsnips, there have been different opinions on
how to define and develop an optimal interpersonal relationship in the physician-
patient dyad. As early as 1956, Szasz and
Hollender described three types of
71


physician-patient relationships based largely on the degree of control required by each
participant for certain medical circumstances: activity-passivity, guidance-cooperation,
and mutual participation. According to Ong et al. (1995), some believe it is primarily
a social relationship; some claim that the relationship is determined by its therapeutic
qualities or client-centeredness; some describe the ideal as one which integrates
both patient-centered and physician-centered approaches; while others advocate a
eful models have been advocated for
itionship.
991) noted that the term patient-centered
Hunt, Joyce, Marinker, and Woodcock in
mutuality in the relationship. Several us
structuring the optimal doctor-patient rek
von Friederichs-Fitzwater et al. (1
medicine was first introduced by Balint,
1970. This was a contrasting model to the traditional illness-centered/disease-
centered/doctor-centered model. Patient-
follows:
centered medicine has been characterized as
In the patient-centered method, the physician attempts to
Ascertain the patients agenda... the essence of this
method is that the physiciaji tries to enter the patients
world, to see the illness through the patients eyes. This
is done by behavior that in\|ites and facilitates openness
by the patient. The doctorjs goal is to understand the
patients expectations, feelings, and fears.... bringing the
patients problems into the diagnosis, (von Friederichs-
Fitzwater et al.,1991;p,18)
von Friederichs-Fitzwater et al. (1991) pointed out that there has been no operational
definition of the model.
72


In 1977, Engel advocated a biopsychosocial model of disease and healing
which viewed disease as an open system process that incorporated mind, body, and
environment. This model was also in contrast to the traditional mind-body dualism
prevalent in the traditional paternalistic niomedical model. Keller and Carroll
(1994;p.l31) asserted that, While Engils biopsychosocial model of disease and
healing has received significant attention in primary care as an idea, it has received less
attention as a guide for physician behavior.
Emanuel and Dubler (1995) have characterized the ideal physician-patient
relationship as one that encompasses: choice, competence, communication,
compassion, continuity, and no conflict of interest. Their ideal tilted in favor of a
more patient-centered, collaborative, mutual participation approach: in terms of
choice, they have advocated for an informed patient who is provided choice among
treatment alternatives; in terms of communication, they have advocated for listening,
understanding preferences for information and involvement in decision-making,
assisting patients in dealing with the illness experience as well as the disease, and a
facilitative role for physicians in guiding sind directing patients through the illness
experience by recognizing values and feelings as well as life, family, and job
circumstances. From a legal perspective, :hey also pointed out that this type of
communication style can allay adverse events which may eventually lead to malpractice
litigation. In terms of compassion, they emphasized the importance of empathy; in
73


terms of continuity, they acknowledged the investment of time necessary for
prevention and chronic disease management; in terms of no conflicts of interest, it was
expected that the patients well-being take precedence over the physicians own
personal interests.
Keller and Carroll (1992) have id
influence the physician-patient relationshi
lentified three perspectives which they believe
iip: the physicians philosophy of disease and
health, the role which the physician assumes in the relationship, and the
communication strategies employed by the physician. They have advocated the E4"
model for physician-patient communications. This model goes beyond the traditional
biomedical tasks of find it and fix it to one which unites the communication tasks of
engage, empathize, educate and enlist wit
the model of Emanuel and Dubler (1995),
r the biomedical tasks of find and fix. Like
this model has favored a non-patemalistic
approach to physician-patient communications that emphasizes physician sensitivity to
patients preferences for information and involvement in decision-making.
Similar to Emanuel and Dubler (1995), Delbanco (1992) has advocated for a
more patient-centered view of doctor-patient interpersonal relationships. His
approach includes seven dimensions of care which should be considered when dealing
with each patient as a unique individual: respect for patients values, preferences and
expressed needs; communication and education; coordination and integration of care;
physical discomfort; emotional support ana alleviation of fears and anxieties;
74


involvement of family and friends; continuity and transition. Delbanco has asserted
that this approach should be: woven into the clinical encounter in bits and pieces until
practitioners become more comfortable with it; incorporated into the early years of
training in medical school; adopted by patients to guide encounters with physicians;
used to guide quality of care feedback from patients.
DiMatteo (1994) proposed a model based on collaborative TBDS. This model
was characterized as one which includes: mutual participation; explicit sharing of goals
and expected outcomes by both patient and physician; information exchange about
various treatment alternatives and their p otential effects; and negotiated treatment
decisions. To effect this model, she proceeded to identify the required physician and
patient behaviors: establishment of a con ducive atmosphere; allow patients to tell their
stories; physicians respond to patients illness experience with concern, empathy,
immediacy, and interest; physician encouragemen t of patients to ask questions;
physician exploration of patient goals, outcome expectations, values and quality of life
components; patient and doctor flexibility in considering various health management
strategies; physicians should assist patterns in organizing
information given to them; and patients should be allowed sufficient time to make
decisions.
Ballard-Reisch (1990) advanced a
making for physician-patient interactions.
theoretical model of participative decision
This model has three phases that guide the
75


doctor-patient interaction: the diagnostic phase; the exploration of treatment
alternatives phase; and the treatment decision, implementation, and evaluation phase.
In turn, each of these phases includes several stages.
The diagnostic phase includes information gathering and information
interpretation by both patient and physician. During this phase, the nature of the
relationship between the doctor and patient is negotiated. Within this negotiation,
there are four relational options: patient abdication, patient autonomy, collaboration,
or termination. Ballard-Reisch emphasized that collaboration is the ideal model for
participative decision making. The exploration of treatment alternatives phase
included the following stages: exploration of alternatives and criteria, establishment of
alternatives, and weighing of alternatives against criteria. The treatment decision,
implementation, and evaluation of implemented treatment (phase III) included the
stages of: alternative selection, decision implementation, and evaluation of
implemented treatment. Each of the three phases incorporates outside input from
other professionals, family members, and support groups.
Ballard and Reisch (1990) emphasized that the physician behaviors most
crucial to this model include: information sharing; gathering information about the
patients goals and expectations, values, beliefs, lifestyles, and sociocultural
relationships; active listening and providing feedback about understandings and
interpretations; involving family members and friends where appropriate; and
76


collaborative decision-making. The authors contended that viewing the doctor-patient
relationship as a three-phase process wnich progresses through eight distinct stages of
decision-making: 1) provides a structured framework upon which physician-patient
communications can be improved; and 2) tenders decision analysis researchers with a
framework for analyzing the quality of medical decisions.
Fredrikson (1995) has proposed an input-process-outcomes information-
exchange model of medical consultation. Central to the model and to the core of the
process component is information exchange which, like Ballard Reischs model
(1990), emphasizes a very flexible, parti Jipative flow and exchange of information
between doctor and patient in regards to
diagnostic and treatment information and
involvement in decision-making. Both mpdels have embodied input, process, and
outcome components.
The input component incorporates both patient and physician input. From the
patients perspective, this includes: frame of reference; motivations, goals, needs,
expectations, and personal information. Ijrom the physicians perspective, this
includes: motivations, goals, needs, expectations, and medical information. The
outcomes component is linked to the process component and encompasses specific
patient satisfaction outcomes related to factors of: patient understanding,
appropriateness of treatment, and physician relationship.
77


Kreps, OHair, and Clowers (1994) provided a conceptual model of
communications and how communications influences outcomes. They have referred to
this model as the Transformation Model
of Communications and Health Outcomes.
Theyve pointed out that this model is based on the systems theory transformation
model of input-process-output. In systems theory, inputs are the antecedents which
provide the raw materials for the processes necessary to accomplish the outputs. In
their model, antecedent conditions included health problem or risk and consumer
attitudes, beliefs, and expectations. The process component encompassed the
communication activities engaged in by tie patient and the physician:
provider/consumer message strategies, language used, non-verbal cues, channels, and
media. The outcomes of the communication and medical care activities comprised the
output component of the model: cognitive, behavioral, and physiological outcomes.
Finally, Ong et al. (1995) provided a theoretical framework which is very
analogous to the inputs-processes-outputs models described previously. Their model
includes three components: a) background variables, b) actual content of
Included in the background variables
ationship, types of patients and doctors,
dual content of communications component
communication, and c) patient outcomes,
component are: culture, doctor-patient rel
disease characteristics. Contained in the
were the actual communicative behaviors including both instrumental and affective
behaviors. Entailed in the patient outcomes component were short-term,
78


intermediate, and long-term outcomes in cluding: satisfaction, compliance, recall and
understanding of information, health status, and psychiatric morbidity.
Diffusion of Innovations: Constructs. Components.
and Determinants
Two of the primary aims of this study were to: 1) advance our understanding
about how physicians view enhanced participation of patients in medical decision
making through the integration of TBDS services, and 2) glean key insights on the
implications for working with physician organizations in optimizing the integration of
such services into medical care delivery systems. Since TBDS and the involvement of
patients in playing a more informed and active role in the medical decision-making
process can be considered a new innovation in the delivery of health care services,
the framework for studying the diffusion of innovations may assist in the fulfillment of
these aims. Therefore, the purposes of this section are to: review the basic com-
ponents and premises underlying the difiJsion of innovations, identify benchmark rates
for the adoption of new innovations and, where possible, delineate
factors/determinants influencing these rates.
Components of Diffusion
What is the diffusion of innovation and what are the major premises
underlying this concept? Diffusion has been defined as:
79


The process by which an innovation is communicated
through certain channels over time among the members
of a social system. The four main elements are the
innovation, communication channels, time, and the social
system. ... An innovation is an idea, practice, or object
that is perceived as new by an individual or an organization.
(Rogers, 1995, p. 10-11).
Innovations in health care include such t lings such as new drugs, treatment guidelines,
preventive services, procedures, diagnostics, therapeutics, and technologies. The rate
of adoption of an innovation over time hps been consistently demonstrated to follow
an S-shaped curve among various audiences including physicians, public health
administrators, teachers, and farmers (Rogers, 1995a).
This epidemiological S-shaped adoption curve characteristically has an early
slow phase involving a few innovators and early adopters. This is followed by an
accelerated middle phase whereby the rate of adoption usually takes off following
adoption of the idea by the early adopters. The rate of adoption in the final and third
phase is slow ending in incomplete penetration. In association with this adoption
curve, Ryan and Gross (1943) have described the different populations of adopters
within a given social system. They are characterized in five different categories.
The innovators generally comprise approximately 2-3% of the group. They
have been characterized as: well endowed with resources, venturesome, risk taking,
fascinated with novelty, willing to experiment and transcend geographical and social
boundaries to learn. Rogers (1995a) refers to them as cosmopolite. Locally, they


are generally somewhat disconnected socially, and are generally not the opinion
leaders. They play an important role by serving as the gatekeeper for the flow of new
ideas into the system.
The early adopters generally comprise around 12-15% of a group. They do
talk with innovators and selectively identify new ideas they would like to try out.
Locally, they are highly regarded for making prudent decisions in the adoption of new
ideas and they are socially well connected. They are regarded as significantly
influential opinion leaders with the early majority. For these reasons, they are often
utilized by change agents as the local evangelist for accelerating change.
The early majority comprise approximately one third of the group and are
characterized as very local in their perspectives, more focused on current needs and
local practical problems, and do not hold in high esteem lofty conceptual ideas,
science, or theory. Their innovation-decision making process has been characterized
as very deliberate and therefore slower than that of the innovators and early adopters.
They generally have a high state of readiness to hear about new innovations that are
relevant and practical for solving local problems from peers they know well and trust.
They look to the early adopters for indications of what is prudent to try in terms of
new innovations and serve as an important interpersonal network link between the
early adopters and late majority.
81


The next group, referred to as tne late majority, comprise approximately one
third of the group and have been described as skeptical about new innovations. They
distrust cosmopolite sources of knowledge,
for their learning channels, local proof, and
are more conservative in their views and
They look primarily to the early majority
cues in adopting new innovations. The late majority do not adopt new innovations
until they become the status quo and peep pressure eventually forces their adoption.
Finally, Ryan and Gross (1943) define the last 16% as the laggards. They
are the ultra-conservative traditionalists who are the last to change and to adopt new
innovations. They are isolated in their so
cial networks and are generally suspicious of
change, change agents, and innovations. The label given to this group may be an
injustice. Even though slow to adopt new innovations, this group is often useful and
valuable in making wise, systematic, calculated choices based on a rational view of
their own limited economic resources.
Adoption Rate Determinants
Rate of adoption has been defined as, The relative speed with which an
innovation is adopted by members of a social system and is measured by the number of
individuals who adopt a new idea in a specified period Rogers (1995a, p. 250). He
has identified at least five variables effecting the adoption rate of innovations.
82


First, the rate is influenced by how the potential adopter groups perceive the
attributes of the new idea. Rogers (1995a, pp. 250-251) emphasized that those
innovations that have the following attributes are more likely to see rapid adoption
when compared to those innovations that do not have these attributes.
1) Relative advantage the degree to which a new idea is perceived as better than
the idea that it supersedes.
2) Compatibility the degree to which an innovation is perceived a consistent with
the existing values, past experiences, and needs of potential adopters.
3) Complexity the degree to which a pew idea is perceived as relatively difficult to
understand and to use.
4) Trialability the degree to which an innovation can be experimented with on a
limited basis.
5) Observability the degree to which the results of an innovation are visible to
others.
Rogers (1995a) pointed out that 50-90% of the variance in adoption rates can be
explained by these five attributes. As we will see later, these attributes have indeed
been demonstrated to be influencing factors on the diffusion of innovations in health
and medical care.
83


A second variable influencing the rate of adoption is the decision-making process
itself. In the process of adopting a new innovation, individuals go through a decision-
making process that includes the following stages (Rogers, 1995a): 1) knowledge -
becoming aware that a new idea exists; 2) persuasion developing a favorable attitude
about the new idea; 3) decision detemiining whether to adopt the innovation; 4)
implementation putting the new idea into use; and 5) confirmation evaluating the
innovation for the purpose of continued
adoption, modification, or discontinuance.
Interestingly, this typology is conceptually congruent with the transtheoretical model
of behavior change that includes pre-con
emplation, contemplation, preparation,
action, and maintenance stages (Prochaska and DiClemente, 1983).
According to Rogers (1995a), three additional variables influencing adoption
rates include: 1) the nature of the communication channels within a social network,
2) the nature of the social system through which the innovation is diffusing, and 3) the
degree to which change agents are invol jed in promotional activities. He speculates
that when considering all five of these variables, the diffusion process is driven
predominantly by how the attributes are perceived and how this perception is
conveyed through the adopter groups interpersonal networks.
In addition to Rogers (1995a) classic social system theory of diffusion, Green
(1980) has provided another typology that holds significant potential for assessing
factors which influence the diffusion of innovations. It also provides a framework for
84


targeting interventions for enhancing the diffusion of a given innovation. In his
PRECEDE model, Green has classified these factors into three categories:
predisposing factors (physician knowleJge, attitudes, beliefs,), enabling factors
(skills/competence, resources, practice characteristics), and reinforcing factors
(reimbursement, peer support, positive feedback, enhanced self-esteem). The
application of this typology will be explored in a later section examining the diffusion
of preventive services.
Diffusion of Innovation in Health Care
The gap between medical science and medical practice (Greer, 1994) as well as
the gap between best achievable practices and actual practice (Berwick, 1998) is well-
established. Many medical innovations (new drugs, treatment guidelines, preventive
services, procedures, diagnostics, therapeutics, and technologies) may take years or
even decades to reach widespread adoption. Ford, Kaluzny, and Sondik (1990) have
poignantly noted that the time involved to achieve the full transfer of new treatments
continues to be problematic in medicine.
Even though an enormous body of scholarly research exists on the diffusion of
innovations, it is still not a well-understooL process. A practical standardized process
for success has not yet been codified. Therefore, it is not surprising to find wide
variation in the rates of adoption for new ideas in health and medical care services. The
85


purpose of this section is twofold. The arst is to review, where available, rates of
adoption related to new drugs, practice guidelines, and preventive services. The
second is to distill those determinants/factors that have been demonstrated to
predominantly influence diffusion and rates of adoption. Well begin by first
examining one of the earliest and most c
medical innovations.
assic sociometric studies on the diffusion of
Using pharmacy prescription records with information about the month in
which each doctor adopted the new drug. Coleman, Menzel, and Katz, (1966) studied
the spread of an antibiotic (tetracycline) among physicians in four Illinois communities.
They also collected data from physician respondents regarding other physicians in their
community with whom they had discussed the new drug. This information allowed the
researchers to trace the social networks through which the innovation was diffused.
The commercial sponsor of the study, Pfizer, provided academic detailers to
extensively promote the new drug.
This innovation spread rapidly reaching almost complete adoption in 17
months. Physicians reported that the academic detailers had little direct influence on
their decision to adopt. Instead, the researchers discovered that information about the
new drug spread primarily by word of mouth from physician to physician and,
generally, a physician did not adopt tetracycline until a respected peer had shared
his/her experiences with the new innovation. Greer (1994, p. 434) insightfully noted
86


that what is usually forgotten in this study is that the best predictor of general adoption
was not adoption by the physicians who were recognized as opinion leaders, but
adoption by those physicians who communicated with the most other physicians the
most often. She emphasized that, the more people talk and stimulate further talk, the
more there is opportunity for consensus, practical problem solving, and social pressure
to change.
In another prescription drug study, Peay and Peay (1988) examined the
adoption process for temazepam through interviews with 124 specialists and general
practitioners. They found that within 13 months after its release, 71% of the
physicians were familiar with the drug (average time between release and awareness
was about 17 weeks), 48% had prescribed it (average time between its introduction
and first prescription was 5.5 months), ana 27% preferred it to the alternatives.
In contrast to the Coleman, Menzel, and Katz (1966) study, they reported that
contact with academic detailers was the nlost consistent predictor of favorable
reception by physicians. Sixty-eight percent reported that their first news came from
commercial sources, 40% citing the detailers in particular. Only 28% reported
discussing this new drug with other doctors before deciding to prescribe. Eighteen
percent reported colleagues as the most important information sources in their first
decision to prescribe. Doctors who had more frequent contact with detailers (at least
once per week) were aware of the drug earlier, prescribed it earlier, and were more
87


Full Text

PAGE 1

PROSOCIAL MESSAGES IN CHILDREN'S SATURDAY MORNING TELEVISION PROGRAMS by Ferris Edward Hoover Jr. B.A., University of Colorado, 1988 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Department of communication 1990

PAGE 2

This thesis for the Master of Arts degree by Ferris Edward Hoover Jr. has been approved for the Communication Department by D. MorlelJ V Kim B. Walker Richard L. Dukes Pamela s. J= Michael z. Hackman l-l9-9D Date

PAGE 3

Hoover, Ferris Edward, Jr. (M.A., Communication) Prosocial Messages in Children's Saturday Morning Television Programs Thesis directed by Associate Professor Donald D. Morley This study was done to test a methodology for measuring the existence and recognition of prosocial messages in children's saturday morning television programs. A panel of four adults was trained to count the occurrence of prosocial behaviors in sample episodes of the five top-rated Saturday morning children's programs. Per hour occurrence ranged from 22 per hour to 48.5 per hour. Fourth grade children were surveyed to determine if some degree of recognition of prosocial messages can be measured. The method used was to ask the respondent's opinion of whether or not the major character in a program would exhibit prosocial traits. The results of the study were somewhat inconsistent. The rank ordering of prosocial content of the programs by the children was sharply different from that of the adults with the children's most prosocial program being the least prosocial according to the adult panel count. The surprising results could have been due to one of several factors. Among them might be that the intended

PAGE 4

connotations of the words used in the questions might not be the same as those assigned by the respondents. or, the one-time count of a single sample episode of a program by an adult unfamiliar with the usual plot structure of that program would not necessarily coincide with the overall view taken by a child who chooses to watch the program regularly. The trained adult panel method seems to be a valid measurement of the occurrence of prosocial messages, but the method for determining the recognition of such messages by children needs improvement. The form and content of this abstract are approved. I recommend its publication. Signed Donald D. Morley iv

PAGE 5

CONTENTS Tables . . . . . . . vi Acknowledgments . . CHAPTER 1. INTRODUCTION 2. SOME HISTORICAL ASPECTS OF MEDIA EFFECTS RESEARCH . . . . . . . . . . . Social Learning Theory . . . . . The 1982 NIMH Report . . . . . Prosocial Research . . . . . Other Research . . . . . 3 RATIONALE . . . . . . 4. METHODS . . . . . . . . Procedures . . . . . . . . 1 4 11 15 18 23 31 36 36 Instruments 4 o Data Analysis . . . . . . 41 5. RESULTS . . . . . . . . 45 6. CONCLUSIONS . . . . . . . 51 APPENDIX A. ADULT RATING PANEL CODING SHEET 61 B. FOURTH GRADE STUDENT SURVEY INSTRUMENT . 62 BIBLIOGRAPHY . . . . . . . . 63

PAGE 6

TABLES Table 5.1. Interrater Reliability . . . . 45 5.2. Adult Rating Panel Observation of Prosocial Behaviors 4 6 5.3. Fourth Grader Ranking of Positiveness Ratios 47 vi

PAGE 7

ACKNOWLEDGMENTS I would like to express special appreciation to Don Morley, my thesis director, and Kim Walker, who acted as second reader, for the extra effort they put into guiding me in the development and execution of the research and the literature search. Thanks go to the Pikes Peak Broadcasting Corporation for providing me with the Nielsen ratings. Thanks to all the members of the Communication Department at uccs for the support they have given during the process of getting this thesis done. The other Graduate Teaching Assistants get special thanks for their participation in the rating panel and the suggestions they had for overcoming the innumerable obstacles that seemed to appear from nowhere. Without the support freely given by the people around me, this thesis would have never been completed. vii

PAGE 8

CHAPTER 1 INTRODUCTION Mass media effects have been the subject of numerous research efforts among the disciplines of psychology, sociology, political science and communication. Early research was generally done by scientists from disciplines other than communication, and a great deal of what is being done today is done by other than communication scholars. Actual internal effects are probably best studied by behavioral scientists or cognitive psychologists while the mechanisms and processes involved in the formulation and the reception of messages through a mass medium are more appropriate areas for a communication scholar. Other social science disciplines generally utilize controlled exposure and observation of behavior to determine whether or not a linkage between media consumption and behavioral change exists. This type of methodology is best suited to examining the internal processes that may be occurring in the subjects being studied. If the intent is to examine what sort of messages are discernable in a particular form of program

PAGE 9

and whether or not the messages are recognizable, a different methodology is needed. Although examining the internal processes which lead to behavioral changes is important, this does not lead to a single generalizable conclusion which explains all possible effects that media consumption might cause. An easily accepted idea, that might cause many people some concern about possible effects of media consumption, is that whatever effects occur are the same for everyone. This is a misconception according to Becker (1987): Too often, when people who do not understand mass communication or mass communication research think about the effects of the mass media, they think in terms of all or none. That is, they -.think if mass communication has a particular effect on one person it must have it on everyone who uses the mass media or who was exposed to that content. Or if they do not see an effect on everyone, they believe it must be due to something other than mass communication. (p. 457) The idea that mass media have different effects on different people provides many paths for a communication researcher to follow. The goal of this thesis was to find a method to explore one of the factors that might contribute to whatever effects might occur. The factor to be examined was whether or not prosocial messages can be reliably identified by adults and recognized by children in children's television programming. First, it was necessary to become familiar with what research has 2

PAGE 10

been done in the area of mass media effects in order to choose a useful and logical method of research. To this end, a review was done of some of the research from all the disciplines involved that has dealt with what effects the medium of television might have on children. The next step was the development of a valid measure of the prosocial message content of the programs in question and a valid method of indirectly measuring the comprehension of the messages by the intended audience. This involved finding methods that have been used in past research and adapting them to this project. The measures which were developed were then executed as a trial of their viability. The methodologies which have been used primarily to search for the adverse effects of mass media were adaptable to the search for possible prosocial effects. The criteria which were used in the examination of whether or not prosocial messages are present and recognizable in Saturday morning children's television constituted the major difference between this research and other explorations of media effects. 3

PAGE 11

CHAPTER 2 SOME HISTORICAL ASPECTS OF MEDIA EFFECTS RESEARCH since the 1920s popular media have been under attack by parents, educators and others for causing such antisocial outcomes as an erosion of moral standards and aggression. The fears that are the basis for such feelings are rooted in the theory of uniform influences of mass media (Lowery & DeFleur, 1988). That is, the mass media have powerful effects on their audiences and these effects are nearly the same for all audience members. The Payne Fund Studies done in 1929 and 1930 were the first attempt to assess the effects of motion pictures on children and adolescents. Some of their findings indicated that adolescents in particular used the actions, styles, and attitudes depicted in the movies as models for their own behavior and style of dress (Lowery & DeFleur, 1988). The next modern mass medium to come under fire was the comic book. Dr. Frederick Wertham, a noted New York psychiatrist published Seduction of the Innocent in 1954 after having published several articles in popular family magazines of the time. In Seduction of the Innocent,

PAGE 12

wertham attacked comic books for having a "bad influence" on their young audience. Although Wertham's methods were scientifically discredited according to Lowery and DeFleur, he did arouse national attention and contributed to the decline of comic book sales. Wertham was generally ignored by social scientists, but his work is an illustration of uniform influence theory in that he described his research subjects as "normal" even though virtually all of them were referred to him because of some social problem such as delinquency. Because the content analyses conducted by Wertham showed that a great deal of socially unacceptable behavior was depicted, he concluded that so called "crime comics" were severely affecting children (Lowery & DeFleur, 1988). The increasing popularity of television as the medium viewed by children again aroused concern over the effects of a mass medium during the 1960s. The first major effort to determine what effects, if any, television might have on children was published by Professors Wilbur Schramm, Jack Lyle and Edwin Parker (1961). Their series of reports investigated why children watch television and what they learn from it. The summation of the findings of this study is the often quoted: 5

PAGE 13

For some children, under some conditions, television is harmful. For other children under the same conditions. or for the same children under other conditions, it may be beneficial. For most children, under most conditions, most television is probably neither harmful nor particularly beneficial. (Schramm, Lyle, & Parker, 1961, p. 13) This conclusion did not satisfactorily answer the question of whether or not television was causing adverse effects on children. The debate continues today as indicated by stories of consumers pressuring advertisers to stop supporting programs that are 11obj ("Idea of boycotting," 1989). Nor has the concept of direct and uniform influences disappeared, as illustrated by retiring Surgeon General c. Everett Keep's 1989 call for restricting and modifying alcohol-related-advertising to reduce the number of teenage drunken driving accidents ( "Koop takes aim," 1989). In short, the public continues to believe that the media serves as a powerful cause of socially unacceptable behavior. The National Commission on the Causes and Prevention of Violence studied violence on television during the late 1960s. The resultant manuscript, Violence and the Media, edited by Robert K. Baker and Sandra J. Ball (1969), among its intended research purposes, analyzed the portrayals of violence in television programming, measured the violent experiences of Americans, and 6

PAGE 14

compared television portrayals of violence with the perception of the audience as to the existence of violence in society. This was done in order to determine the existence of longand/or short-term effects on accepted norms pertaining to violence (Lowery & DeFleur 1988). Their conclusions confirmed the findings of short-term effects that had baen measured in previous experimental studies. The report then suggested that long-term studies should be conducted to determine whether or not long-term effects existed. The call for further research was answered when the surgeon General appointed an advisory committee to supervise a massive research effort in 1969. The report that resulted included more than 40 scientific papers and an overview written by The Surgeon General's Scientific Advisory Committee in 1969. The summation of the research effort, Television and Growing Ye (The Surgeon General's Scientific Advisory Committee on Television and Social Behavior, 1971), was that viewinq television violence could be causally linked to aggressive behavior, but only in children who were inclined in that direction. It was further arqued that the context in which the violence was viewed had a larqe influence on the actual effects. In the words of the committee: 7

PAGE 15

Thus, the two sets of findings converge in three respects: a preliminary and tentative indication of a causal relation between viewing violence on television and aggressive behavior, an indication that any such causal relation operates only on some children (who are predisposed to be aggressive); and an indication that it operates only in some environmental contexts. such tentative and limited conclusions are not very satisfying. They represent substantially more knowledge than we had two years ago, but they leave many questions unanswered. (Surgeon General's Scientific Advisory Committee, 1971, pp. 18-19) Instead of focusing attention on which children in which contexts were most affected by viewing violence on television, the question that the public, as well as the scientific community, most wanted to answer was that of how much violence was portrayed on television. A comprehensive study conducted by George Gerbner (1971) of the Annenberg School of Communications at the University of Pennsylvania measured the number and types of violent acts, who the perpetrators were, and who the victims were in both prime time and Saturday morning programming. From this study emerged the use of the violence index that has been cited by both the popular press and the scientific community to argue that almost all of the programs viewed by children contain violence. Of note is the fact that Gerbner's study found Saturday morning cartoons to have the highest level of violent content with a violence index of 251.1 compared to the 8

PAGE 16

next highest index of 241.9 for crime, western and action adventure programs (Gerbner, 1971, p. 67). However, these findings must be tempered by the fact that the study counted violent acts without regard to context or intent. Gerbner was investigating content as being symbolic material without regard to artistic merit and without distinguishing between the relative impact of any individual action and any other action occurring within the same program (Gerbner, 1971). For example, a pie in the face was weighted equally with the gunning down of an unarmed person. Gerbner's measurement of violent content would seem to be the basis for many proposals which periodically are reported in the popular press for control of the content of children's programs by government agencies. The recurrence of these proposals indicates that popular wisdom still supports the idea of uniform and universal effects. Since most of the questions raised by Baker and Ball in 1969 about what might cause long-term effects were not answered despite the massive effort, the committee continued with a call for more research and specified some areas to be covered. Among them were such areas as television's effects in the context of other mass media, individual developmental history, and other environmental influences including the home environment. They also 9

PAGE 17

called for research on the relationship between televised violence and aggression, specifically in the areas of predispositional characteristics of individuals, age differences, effects of labeling, contextual cues and other program factors, and longitudinal influences of television. The modeling and imitation of prosocial behavior, the role of environmental factors, including the mass media, in the teaching and learning of values about violence and the effects of such learning on social development were also included as areas to be researched (Surgeon General's Scientific Advisory Committee, 1971, p. 19) The scientific community responded by multiplying their efforts during the decade that followed. According to Lowery and DeFleur (1988) "approximately 90 percent of all research publications gn television's influence on behavior appeared!" and more than 3000 titles. are included in the information that was published on television violence between 1971 an"d 1982 (p. 353). This accumulation led to the publication in 1982 of Television and Behavior: Ten Years of Scientific Progress and Implications for the Eighties by the National Institute of Mental Health (NIMH) at the direction of the Surgeon General. This report is a synthesis of the knowledge gained from the effort instigated by the 1971 Television 10

PAGE 18

and Growing Ye and included not only research done on the effects of television violence but many other areas. The areas germane to this thesis will be discussed. Social Learning Theory The studies done to determine whether or not television violence could be causally linked to aggressive behavior were generally based on social learning theories which posit that children learn by observation or incidentally. One of the foremost researchers and theoreticians in this area is Albert Bandura, who began working in the area during the early 1960s. His work is well represented in his 1977 book Social Learning Theory. It provides a framework for describing the processes experienced by children in their social development and the continuing effects of media throughout most people's lives. Bandura (1977) posits that observational learning, which is a primary factor in social learning, has four components: attention, retention, reproduction, and motivation. Social learning according to Bandura is a three-way interaction between the person, the behavior, and the situation. The person part of the interaction includes the developmental stage of the individual which affects the ability to model a behavior on a delayed 11

PAGE 19

basis. Another factor is the self-regulatory capacity of an individual. This falls into the motivational component of observational learning in that it affects the determination of whether or not the outcome of a behavior is considered to be valuable to an individual. The situational aspect of the interaction is influenced by reinforcement of the learning process, which Bandura describes as "facilitative" rather than necessary for learning because other influences can determine what is attended to and learning can occur without reinforcement. That is, even if the situation surrounding the observation of a behavior provides no reinforcement, the behavior may be successfully performed at a later time. Behaviors may also be duplicated in the short-term because of a situation like a laboratory experiment, but not retained in an individual's behavioral repertoire for later use. Social learning theory does recognize that the actual learning of a behavior is an individual process, but Bandura indicates that the media have broadranging effects in the population when he states: "It has been shown that both children and adults acquire attitudes, emotional responses, and new styles of conduct through filmed and televised modeling" (Bandura, 1977, p. 39). 12

PAGE 20

Social learninq theory has been used as a model for explaininq many findinqs of media effects research. Other research usinq the theory was beinq done concurrently with the work that Bandura was doinq that resulted in the publication of Social Learning Theory as well as after its appearance. Schramm, Lyle, and Parker (1961) theorize that most of what children learn from television falls into the category of incidental learning. As children mature, it has been demonstrated that they acquire the ability to distinquish between reality and fantasy in what they view (Corder-Bolz, 1982). Durinq the period between three and eight years, however, they identify with television characters who they think are "real" and probably learn more behaviors from them than they do as they mature and differentiate between reality and fantasy. Social learninq theory as a model would explain these findinqs in terms of increased ability to perform a behavior after a lapse of time due to maturation and the individual characteristics (i.e., the "person part" of the interaction) beinq a determininq factor in what behaviors would be perceived as havinq valued outcomes. Situational factors would include the perception of what was desired of the subjects by siqnificant social others such as the researchers in an experimental settinq and 13

PAGE 21

what previous learning had been acquired. As long as it is not assumed that a conscious effort is required to learn a behavior, social learning theory would seem to be a convenient and sensible model for explaining incidental learning from the media, particularly that which takes place in experimental settings. Fishbein and Ajzen (1975) theorized that attitudes and behaviors are learned on the basis of rewards and punishments. That is, what is perceived to have a valued outcome will be adopted. Whether or not the outcome of a behavior will be valued will be largely dependent on what attribution of responsibility for a behavior is assigned to the model. Five levels of attribution are described by Fishbein and Ajzen. They are: association with the behavior, commission of the behavior, foreseeability of the result, the intentionality of the model and intentionality with justification. Fishbein and Ajzen also describe three factors which bear on what attribution will be assigned to a model. These are: the consistency of the behavior, the distinctiveness of the behavior and consensus with other behaviors of the model. Rosenthal and Zimmerman (1978) include cognition, defined as the covert processing of information, and abstraction, defined as going beyond discrete stimulus-response associations, in their expansion of social learning 14

PAGE 22

theory. They explain four specific behaviors as a refinement of social learning theory. The first is inhibition which is the result of learning what behaviors result in punishment. The second is disinhibition which results from learning what behaviors will not result in immediate punishment. The third is facilitation which results when an observed behavior serves as a reminder of things learned at an earlier time. The fourth is novel behavior which is something not previously modeled. Rosenthal and Zimmerman (1978) posit that enculturation is the result of modeling parental behaviors and learning group standards and mores. This process could include behaviors learned from modeling on television. They also state that "the social context factors in implementing learning have received limited attention" (p. 267). This relates to the "situation part" of the social learning interaction as described by Bandura (1977) and reiterates a call for research in an area that was specifically mentioned by the: Surgeon General's Scientific Advisory Committee in 1971. The 1982 H1!1H Report Among the areas explored in the 1982 National Institute of Mental Health Report were factors believed to determine children's attention to television, how much 15

PAGE 23

and what type of content was retained, and what processes might explain the relationship between violence viewing and behavior. Collins (1982) postulated three factors for determining children's attention to and comprehension of television content. The factors relating to attention were: viewer characteristics (e.g., mostly age-related), content attributes (e.g., auditory and visual cues), and comprehensibility (which again is a function of the child's age and development). Research reviewed by Collins suggested that older children have a more extensive background making more material understandable and have a longer attention span. Factors relating to processing content were also found to be age related. They were: knowledge of common formats such as narrative stories and commercials, knowledge of commonly portrayed situations and event sequences, and knowledge of the forms and conventions such as the minor climax just before a commercial break. Collins also posited that younger children would be less likely to have the ability to link actions with their consequences. Huesmann-(1982), in his contribution to the NIMH report, describes five processes that might be used to explain the relationship between violence viewing and behavior. The first is observational learning that he 16

PAGE 24

states has been verified in laboratory settings. The persistence, or retention of observational learning is dependent on whether or not it is reinforced in some manner. The second is attitude change which can alter the acceptance of violence by an individual. This process of attitude change can be modified by prosocial reinforcement or outside attitude training. The third is the arousal process which is somewhat of a dichotomy. That is, viewing violence could cause an "overload" leading to hyperactivity or could have an anesthetizing effect due to sensory "overload" and thereby lead to the need for increased aggressive behaviors to achieve normal arousal levels. The fourth is catharsis, that is, watching violence reduces the need for aggressive behavior because normal arousal levels are reached by the act of viewing. Huesmann states that the catharsis postulation has been rejected by research. The fifth is justification. Aggressive individuals watch violent programming in order to justify their own behavior as "normal" and reduce guilt levels. Freedman (1984, 1986) challenged the conclusions of several articles in the 1982 NIMH report that causally linked viewing television violence with aggressive behavior. He found in reviewing the research for the NIMH report that "The bulk of the correlations fall 17

PAGE 25

between .10 and .20." (1984, p. 237). He described these correlations which were found in both laboratory studies and field research as "mild" and judged them insufficient to extrapolate to a cause and effect determination. He concluded that: "(a) exposure to and preference for violent programming correlates with aggressive behavior and (b) there is little convincing evidence that viewing violence on television in natural settings causes an increase in subsequent aggressiveness" (1984, p. 244). This position would also dispute any argument for the possibility of television viewing having beneficial effects on the audience. Cook, Kendzierski and Thomas (1983) also argued that the 1982 NIMH report failed to establish the causal link between viewing television violence and aggressive behavior. Additionally, they criticized the report for focusing on behavioral studies only and its failure to discuss aspects of television programming other than its possible influence on children. Prosocial Research Bryan and Walbek (1970a, 1970b) in studies with second through fifth grade children found that prosocial behavior modeled by both live and videotaped models affected the propensity to share game winnings from a 18

PAGE 26

subsequent session. The actions of the model seemed to have more effect than what was said while the game to be played was demonstrated. Inconsistency on the part of the model was found to have no significant effect. Stein and Bryan (1972), however, found that preaching about following rules while practicing violation of those rules led to an increase in cheating by their third and fourth grade female subjects. When the model preached and practiced rule compliance, cheating was nearly eliminated. In a study of three-to five-year-old children who were tested before and after being exposed to violent, neutral or prosocial programming, Stein, Friedrich and Vondracek (1972) found that those who were most aggressive according to the baseline test exhibited the greatest increase in aggressive behaviors after exposure to the violent programming. They also found that the group receiving the prosocial treatment exhibited a higher level of self controlling behaviors than the other groups. In an experimental study on three-to six-year-old children, Friedrich and stein (1975) found that prosocial behavior could be learned from television and that the subjects applied the behaviors. They also found that verbal labeling training and role playing exercises 19

PAGE 27

enhanced the learning experience. Gorn, Goldberg, and Kanungo (1976) also found that prosocial television could have clearcut short-term effects on the attitude of three-to five-year-old children toward members of other racial and ethnic groups. Singer and Singer (1976) in a study of three-and four-year-old children in a day care center over a twoweek period used a prosocial program along with structured play sessions to determine the effect of adult mediation on the level of imaginativeness showed by the children. The findings indicated that imaginativeness of play could be increased by the presence of an adult mediator. Poulos, Harvey and Liebert (1976) in their c'ontent analysis of Saturday morning television found that prosocial acts were being depicted on children's programs. In the fifty programs which made up their primary sample the combined network number of prosocial acts per half-hour was higher than the number of aggressive acts (a mean of 10.7/half-hour versus 6.10/half-hour). They also found that one quarter of the sample contained no examples of aggression. The one shortcominq noted in Saturday morning television was the lack of examples of preventinq or eliminatinq violence. 20

PAGE 28

In their 1979 study, Friedrich-Cofer, Huston-stein, Kipnis, Susman, and Clewett, also found that reinforcement of the model provided by prosocial programming-can increase its effect. Their study of twoto five-year-old children in a Head start program done over an eight-week exposure period found increased levels of prosocial behavior in children who were provided reinforcement through associated play materials and a greater increase when special teacher training was added to the reinforcement. Some findings in the 1982 NIMH report also indicated that viewing television need not necessarily affect behavior in an adverse way. singer (1982) reported finding that when children's viewing was monitored and mediated by an adult, it helped to improve the children's imagination. Corder-Bolz (1982) called for the teaching of television literacy and parental participation in children's viewing in order for children to better discriminate between reality and fiction and to improve the learning process when educational programs were viewed. Rushton (1982) reviewed numerous experiments that were designed to determine if television could have beneficial as well as detrimental effects on the audience. He defines prosocial material as "that which specifies things that are socially desirable and in some 21

PAGE 29

way benefit another person or society at larqe" (p. 249). He describes four types of prosocial influences. First are those which promote altruism, that is, qenerosity, helpinq behavior, and cooperatinq. The second is friendly behavior; the third is self-control, which is resisting temptation and delaying qratificationr and the fourth is copinq with fears. Rushton's review of both laboratory and naturalistic research desiqns indicated that viewing prosocial material correlated with an improvement in prosocial behavior in the short-term and on delayed measures. However, Rushton (1982) drew a conclusion that supported not only the arqument for beneficial effects from television viewing, but also supported the widely held contention that viewinq television has detrimental effects on behavior. Specifically, "Television is much more than mere entertainment; it is also a major source of observational learninq experiences, a setter of norms. It determines what people judqe to be appropriate behavior in a variety of situations. Indeed it might be that television has become one of the most important aqencies for socialization that our society possesses" (p. 255). Finally, Forqe and Phemister (1987) found that preschool children demonstrated more behaviors of 22

PAGE 30

sharing, cooperation, delay of gratification, and positive social contacts after exposure to prosocial programming. The direction of their research was to determine whether or not the mode of presentation, animated versus non-animated, made a difference. The mode of presentation was not found to be significant but prosocial versus neutral program topics did seem to affect the subjects behavior. This would lead one to believe that the effects of television viewing would be a function of what sorts of messages were being presented. Other Research Given that the studies reviewed in the 1982 NIMH report did not fully establish the causal link between viewing violence on television and aggressive behavior, it could be suggested that establishing a causal link between viewing prosocial television and increases in prosocial behavior would be equally difficult. Although the social learning theory work of Bandura and others provides an explanation of how observed effects could have occurred, other works published since 1980 provide additional information about how and why television might influence audience members. Corder-Bolz (1980) also found that intervention by an adult mediator can affect the learning process and the 23

PAGE 31

formation of attitudes about sex roles and violence. The findings were that primary social agents, those that can demand compliance with their expectations such as parents, teachers, neighbors, the clergy, institutions, organizations, and peer groups, have the most effect on attitudes and beliefs. If secondary social agents, those that cannot demand compliance, such as mass media, present information contradictory to that presented by the primary social aqents, it is likely to be discredited. Huesmann, Eron, Klein, Brice, and Fischer (1983) approach the question from the standpoint of cognition. They suqgest intervention in the form of teachinq that television violence is an unrealistic portrayal, that aqqressive behaviors depicted on television are not as acceptable in the real world as they are in televised stories, and that one should not behave in the same ways as the aggressive characters on television. This intervention would reduce the effect of violent portrayals by altering the cognitive process. They also posit that children who are normally less aggressive will perceive television violence as less realistic and will be less likely to model it. However, a didactic course of treatments designed to produce attitude changes produced no significant results in their second and 24

PAGE 32

fourth grade student subjects. The researchers considered it possible that their didactic treatment was not directed at exactly the right variables and this may have affected the outcome. A second treatment that involved having the subjects produce their own arguments about why television violence is not realistic and should not be modeled did produce some measurable results. However, their primary finding was that identification with a character was the best predictor of whether or not a child was peer nominated as an aggressive individual. They also noted that the subjects who initially had the lowest level of identification showed the most effect. Of special interest is the fact that the amount of television viewed was found to have no significant effect on the other findings. Berkowitz (1984) discussed some factors that extend the "person part" of the social learning interaction beyond age and development. He posited that reaction to a message was dependent on interpretation by the individual as well as the ideas previously held and the thoughts activated by the message. He called this a "priming" effect. That is, a media event leads the viewer to thoughts about similar things without conscious effort. This_effect could be caused by both pro-and antisocial situations depicted in media. However, 25

PAGE 33

Berkowitz states that a semantically related concept, such as a previously acquired set of words or images, must be available to the individual for the effect to occur and that the effect of "priming" decreases over time. Berkowitz also found that viewing fictional material and focusing on the aesthetic aspects of a program reduces the impact of aggression facilitating ideas. Tamborini, Zillmann, and Bryant (1985) expanded this concept with their finding that short-term "priming" occurs after a single exposure to a program but longterm effects result from "repriming" which occurs with multiple exposures. Christenson (1986) determined that the perception of moral lessons improves with age. In a study of kindergartners, first, third, fourth, and sixth graders, the highest level moral lesson as determined by an adult panel was more likely to be perceived by older age groups. The kindergarten and first grade group had no members perceive the highest level lesson while the sixth grade group had 39 percent of its members perceive the lesson. Tan (1986) describes some possible models that can be used to track how children might acquire the cultural values that determine their individual attitudes toward the viewing of violent television programming. The major 26

PAGE 34

structure is the coorientation model, that is, communication requires the participation of at least two persons. According to Tan, three criteria must be met for coorientation to exist. The first is congruency, which is similarity between one person's cognitions and the perceptions of another person's cognitions. The second is agreement, which is the extent to which two people have the same salience evaluations. The third is accuracy, which is the extent to which the estimate of another's cognitions matches what the other actually thinks (Tan, 1986). Tan (1986) places coorientation within families into two patterns and describes four variations of family types based on whether one or the other or both patterns exist. The first pattern is socioorientation in which parents stress the child's relationships with others with the desired outcome of conflict avoidance. The second pattern is concept orientation in which parents stress seeking new ideas, looking at all sides of issues and forming one's own opinions. The first family type is a laissez faire attitude in which neither socioorientation nor concept orientation are stressed and the children are "undirected." The second family type is the protective family in which socioorientation is stressed, social harmony is valued 27

PAGE 35

and the children are "obedient." The third family type is pluralistic in which concept orientation is stressed. In this family type there is no constraint in interpersonal relationships and open discussion is encouraged. The fourth family type is consensual in nature with both orientations being stressed and discussion but not debate is encouraged (Tan, 1986). Some behavioral patterns can be associated with family types. In pluralistic families the children are generally more competent, active in public affairs, more receptive to contradictory ideas and less persuasible. Junior and senior high school students from pluralistic type families also use media more extensively to learn about public affairs and are more positive about the political system. Same age students from protective families generally rank lower in use of media for gaining information, are less politically active, and less positive about the political system (Tan, 1986). In addition to discussing what contributions are made by immediate family members to the acquisition and development of attitudes, Tan (1986) describes some possible effects of violent television and some possible prosocial effects of television. The effects of violent television are: learning new acts, disinhibition and facilitation of aggression, reinforcement of attitudes 28

PAGE 36

held prior to observation, vicarious reinforcement of behavior patterns, and postobservation reinforcement of modeling behavior observed on television. The prosocial effects are divided into cognitive effects and behavioral effects. The cognitive effects are learning useful information and development of cognitive skills, that is, perceptual discrimination, reasoning, and problem solving. The behavioral effects are performance of socially desirable acts, these are: helping others, altruism, controlling aggressive impulses, delaying gratification, persistence, explaining feelings, resisting temptation, adhering to rules; and expressing sympathy to others. Tan states that most of the research done in the area of prosocial effects has been done with preschool children but the results have been positive. In an investigation involving preschoolers and first and second graders, Nikken and Peeters (1988) found that by age seven children realize that what is on television is not necessarily real. Liebert and Sprafkin (1988) found that by age eight, a majority of children understand the purpose of commercial messages and that among fourth grade children, 65 percent of the central content of a program should be remembered. As a means of reducing possible detrimental effects, they suggest that teaching younger children that most programs do not 29

PAGE 37

reflect reality could reduce aggressive reactions. Liebert and Sprafkin also found that adult co-viewers providing commentary on programs improves children's comprehension of the content. These findings indicate that perhaps the best way to limit possible undesirable effects of television on children is to teach them how to discriminate truth from fiction and provide commentary to improve their comprehension. Walker and Morley (1988) found that a liking for violence and a generally aggressive attitude was related to watching violent television and that different types of violence are perceived in differing ways. They also found that adolescents with aggressive behavioral intentions found violent television to be more aesthetically acceptable. These findings could be explained by the proposition that the intended or unintended intervention in the early learning processes by primary social agents helped form the beliefs and attitudes which made the watching of violent television programs more desirable. 30

PAGE 38

CHAPTER 3 RATIONALE Review of research on the effects of mass.media reveals that the topic can cause the researcher a great many problems. Finding causal links between viewing any genre of television and subsequent behavior can become a circular process that the researcher may not see. Are aggressive tendencies acquired from primary social agents with adolescents consequently watching violent programming or, does watching violent programming lead to aggressive tendencies in behavior? Cleary, the processes that occur are difficult to isolate and study and the artifacts that result from the research programs are difficult to sort out. If children are being observed outside the home and are being exposed to researcher selected programming, is such a study externally valid? The bulk of the studies relating to media effects deal with the possible adverse effects. Exploring the possibility that prosocial messages exist in saturday morning children's programs has not received a great deal of attention. Measuring the existence and recognition of these messages in the programs that the children chose to view would begin to balance previous research.

PAGE 39

If social learninq theory is valid and children learn from watchinq others' behavior either in person, or throuqh some mediated means, it can be posited that the existence of prosocial messaqes in Saturday morninq television could provide the prosocial models for children to emulate. Furthermore, if it can be demonstrated that some recognition of the message occurs in younq viewers, then from a social learning theory perspective it would follow that prosocial behavioral changes may be occurring. Existinq prosocial research can be described as hiqhly restricted in that it occurs in laboratory settinqs or is done with groups of subjects in other than an at home setting (e.q., pre-school, lower grade elementary school, or head start). Another problem is that the researcher typically selects the material to be watched rather than the subjects watchinq what they normally select (Bryan and Walbek, 1970a, 1970b;. Stein, Friedrich, & Vondracek, 1972; Friedrich-& Stein, 1975; Gorn, Goldberq, & Kanungo, 1976; Friedrich-Cofer, Huston Stein, Kipnis, Susman, & Clewett, 1979; Sinqer & Singer, 1976). The aqes of the subjects could also be questioned in that normally all are six years old or younger and consequently not of an aqe where they could be expected to have a full comprehension of reality (Christenson, 32

PAGE 40

19867 Nikken & Peeters, 1988). The present investigation proposes a more naturalistic method that would not select the programs viewed and use measures other than behavioral observation in a laboratory setting to determine what messages are recognized by the children. In order to determine whether or not Saturday morning programming contains prosocial messages, and whether or not those messages are being recognized, two things should be accomplished. First, the existence of prosocial messages needs to be reliably determined based on at least face valid criteria. This requires the use of trained adult rating panels. Second, the recognition of the messages should be measured using an instrument based on the same criteria. In order to avoid the artifacts that might occur in a laboratory setting, the recognition measurement has to be indirect. This requirement somewhat restricts the amount of information that can be gathered in a small scale study. Another consideration that arises from conducting research on a small scale is the limited number of adult participants. In order to fully examine whether or not prosocial messages exist in Saturday morning programming, several panels should be used and interrater reliability be established not only within each panel but across panels. 33

PAGE 41

Within the limitations of scales that exist, the methodology should yield results that are indicative of the efficacy of the training procedures used and the reliability and validity of the survey instruments. A shortcoming of controlled studies that do not attempt to determine whether or not recognition of a particular type of message has occurred in the subjects from a communication standpoint is that the subjective processes within the receivers is not examined. That is, the simple observation of behavior before and after a controlled exposure to some media event has value but fails to take into account what the subjects of the study think about the experience. By developing methods that allow the subjects to express what message is being recognized, it may be possible to describe more completely the that occur during individually selected media consumption. Therefore, the purpose of the present study was to develop and use-methods for examining the possibility that not all media effects are adverse. The first step was to develop a methodology for determining whether or not prosocial messages exist in Saturday morning children's programs. The second step was to develop an instrument which would indirectly measure children's recognition of prosocial messages in those programs. The 34

PAGE 42

third step was the execution of the methodology on a trial basis. Lastly the results of the trial were analyzed to determine whether or not the methodology was viable and what improvements need to be made. 35

PAGE 43

CHAPTER 4 METHODS Procedures The research methodology built on three elements. These were the Saturday morning television programs, a trained adult rating panel, and fourth grade school children. The only experimenter manipulation of any sort was the training of adult rating panel to rate selected programs on prosocial message content. In order to assess the presence of prosocial messages, a sample of Saturday morning programs was drawn for a trained panel of adult observers to evaluate. The programs used were selected based on Nielsen ratings. The programs that held the top five positions for Saturday morning ratings among six to eleven year old children during the November 1989 ratinq period were selected to maximize the likelihood of the survey respondents having viewed them. Since there was a tie for fifth place in the ratings, a disinterested party blindly drew the name of the program to be used. The five programs selected for the study and their rank according to the ratings were: (1) "Slimer and the Real

PAGE 44

Ghostbusters11 (2) "Beetlejuice" (3) "Alvin and the Chipmunks11 (4) "Denver, the Last Dinosaur11 and (5) "Garfield and Friends." The actual episode of each program to be recorded and shown to the adult panel was selected by numbering the Saturdays during the period 30 December 1989 through 27 January 1990 and rolling a die. A panel consisting of four female upper division and graduate students was recruited at the University of Colorado at Colorado Springs. All of the panel members were communication students but only one undergraduate had some background in media effects research. The panel had two training sessions during which interrater reliability was established using examples of programs not included in the study. The definitions of prosocial behaviors given by Liebert and Sprafkin (1988) were used in training the raters. They are: Altruism -consists of sharing, helping and cooperation involving humans or animals. Control of aggressive impulses -involves nonaggressive acts or statements that serve to eliminate or preventaggression by self or others toward humans or animals. Delay of gratification/task persistence -consists of the related acts of delay of gratification and task persistence, expressed either nonverbally or verbally. Explaining feelings .of self or others -consists of statements to another person(s) explaining the feelings, thinking, or action of self or others with the intent of effective positive outcome,including increasing the understanding 37

PAGE 45

of others, resolving strife, smoothing out difficulties, or reassuring someone. Reparation for bad behavior -refers to behavior that is clearly intended as reparation for an act seen as a wrongdoing committed by the person himself/herself. Resistance to temptation -refers to withstanding the temptation to engage in behaviors generally prohibited by society (e.g., stealing), which may be prohibited in the program explicitly or implicitly. sympathy -is a verbal or behavioral expression of concern for others and their problems. (Liebert & Sprafkin, 1988, Table 10.3, p. 230) These definitions provided the criteria to determine the existence of prosocial messages and were explained to the panel during the first training session. An example of a program not broadcast on Saturday mornings was used to illustrate behaviors that fit each of the definitions. Raters were simply asked to count the occurrence of each form of prosocial behavior during a practice rating done on a different program to establish a preliminary level of interrater reliability. The second training session consisted of further discussion and group analysis of a program to establish shared definitions for the panel members. Again an example of a program not broadcast on Saturday mornings was rated by the panel and interrater reliability was checked. The final session for the panel consisted of rating the selected episodes with the commercials being skipped by means of fast forwarding the 38

PAGE 46

tape. The programs rated aired as nominally three and one half hours of programming. A sample of 371 fourth graders from a local school district was surveyed to ascertain whether or not prosocial messages were being recognized by viewers of the programs. There were 96 valid returns for a 25.9% return rate. The responses came from 38 males and 57 females and one response with the gender left blank. Demographics for the schools was not collected at the time the surveys were distributed. When a comparison of the respondents gender distribution to the gender distribution of the school populations seemed to be necessary, the school district was queried for the information. As the term had been over for several weeks, the data files had been purged and the information was not available in a usable form. One survey was returned without being completed and one respondent indicated that neither of the programs had been watched. Neither of these was counted as a valid response. The distribution of the number of responses for each program was not related to the Nielsen ratings. "Alvin and the Chipmunks" which was rated number three had 25 responses, 11Slimer and the Real Ghostbusters" rated number one had 22 responses, 11Beetlejuice" and "Garfield and Friends rated number two and five respectively had 39

PAGE 47

18 responses each, and "Denver, the Last Dinosaur" rated number four had 13 responses. Fourth graders were determined to be capable of reading the simple questions used on the survey instrument. Further, only three simple answer choices of yes, no, and don't know were provided (See Appendix B). Two program surveys were given to each student in order to increase the probability of getting one valid program evaluation from each respondent. For those indicating the same frequency of viewing for both programs, a coin flip was used to determine which one would be used in the data analysis. For differing frequencies of viewing, the program watched most was used. Instruments The panel used a tally sheet with each of the individual prosocial behaviors listed and briefly defined. While viewing the program episodes selected, they noted the occurrence of each behavior (See Appendix A). The survey forms for the fourth graders were tailored to the programs. That is, each form had the name of the program at the top and the questions pertained to that programs title character (See Appendix B). Prior to distribution, the questionnaires were paired equally 40

PAGE 48

and shuffled so that the programs any particular-child was asked about occurred randomly. The two questionnaires were stapled to a parental permission form prior to being distributed. The sets of questionnaires were broken down prior to being delivered to the participating schools and were picked up after being returned to the schools by those who chose to participate. The seven questions about the title characters paralleled the Liebert and Sprafkin prosocial behaviors listed on the trained rating panel's sheets and the respondents were asked whether or not the main character of the program would choose to behave in that manner. For example, on the altruism behavior the respondents were asked to reply yes, no, or dont know to the statement, (Character Name) likes to share and help others. Additionally, the respondents were asked if they would like to have the title character as a friend, how often they viewed the program in question, and their gender and age (See Appendix B). Data Analysis The rating panel tally sheets were compiled to determine a rate per program of each prosocial behavior and a total rate. These results were put into a common 41

PAGE 49

measurement on a per hour basis for rankinq the proqrams accordinq to the rate of occurrence of prosocial behaviors. Interrater reliability was established by intercorrelatinq the four raters evaluations of each program. Specifically, each rater's count of each type of prosocial behavior were intercorrelated with the other raters' counts of each type of prosocial behavior. This yielded six correlations per program which were converted to z scores and averaged to arrive at interrater reliability for each program and across all programs. The data from the children's survey instruments were used to calculate a total positive rating score. This score was the total of positive answers given by the respondents for each program. A prosocial ratio score was also determined for each program. This was done by dividing the number of positive answers by the total of the positive and neqative answers. This ratio was used to determine the prosocial message recognition rank of the programs. A one between analysis of variance was done to determine if siqnificant relationships existed amonq the program rankinqs as determined by the adult panel and the survey respondents' evaluations. The dependent variable in this procedure was the student derived positiveness ratio for each program. The independent variable was the 42

PAGE 50

prosocial rankinq of the proqram as determined by the adult panel. Thus, if the children were recoqnizinq prosocial messaqes in the same way as the adults, the orderinq of the children's positiveness ratio means should be the same as the prosocial orderinq of the adults. Analyses of variance were also done usinq the respondent's qender and the answer to the question about whether or not they would like to have the main character of the proqram as a friend as independent variables. The positive answer totals for the various behaviors were used as dependent variables in order to determine whether there was more recoqnition of any of the behaviors based on qender or the desirability of the character as a friend. In order to explore the relationship between whether or not the various characters would be desirable as friends and the number of positive opinions assigned to the programs by the children, the correlation between these items was checked. The number of positive opinions was counted and used as a score in this case and the friend dimension was left on its one to three scale. Finally, a rank order correlation was done between the prosocial rankings assigned by the adult panel and children's prosocial rankinqs of the programs. A 43

PAGE 51

positive correlation would indicate similarities in adult-child recognition of prosocial messages while a negative correlation would indicate dissimilarity. 44

PAGE 52

CHAPTER 5 RESULTS The first item that needed to be established was the interrater reliability of the adult panel. The reliable recognition of the occurrence of prosocial behaviors in the programs was the basis for determining whether or not the methodology used in this project was viable. The statistical operation used for establishing reliability in this case was the average intercorrelation of the scores given by the four panel members. The resultant average intercorrelations among raters are listed in Table 5.1. The overall good intercorrelation among adult raters indicates that the training sessions established a shared set of definitions. Table 5.1 Interrater Reliability Program Title Garfield and Friends Slimer and the Real Ghostbusters Denver, the Last Dinosaur Alvin and the Chipmunks Beetlejuice Across all programs Correlation 0.935 0.915 0.910 0.890 0.775 0.895

PAGE 53

The question of the occurrence of prosocial behaviors in saturday morning children's programs that were evaluated was the next consideration. The rate of occurrence differed widely across the set of programs but all were determined to portray at least some prosocial behaviors. The adult panel means for the number of prosocial occurrences on a per hour basis is given in Table 5.2. Table 5.2 Adult Rating Panel Observation of Prosocial Behaviors (Panel mean of occurrences per hour) Rank Program Title Score l. Alvin and the Chipmunks 48.5/hour 2. Beetlejuice 48.0/hour 3. Garfield and Friends 26.75/hour 4. Slimer and the Real Ghostbusters 23.75/hour 5. Denver, the Last Dinosaur 22.0/hour In order to establish a basis for comparing what the opinions of the fourth graders about the probability of the main characters of the programs behaving in a prosocial manner with the results of the. adult panel's evaluation, an index of-some sort was necessary. The children's total positive answers about the main characters for each program were divided by the total of 46

PAGE 54

the positive and negative answers for each program to determine a "positiveness ratio" for each program. The "Don't Know" were considered to be a "no opinion" and omitted from the calculation. The resultant ratios established the ranking given in Table 5.3. Table 5.3 Fourth Grader Ranking of Positiveness Ratios* Rank Program Title 1. Denver, the Last Dinosaur 2. Slimer and the Real Ghostbusters 3. Alvin and the Chipmunks 4. Beetlejuice 5. Garfield and friends Ratio 0.98 0.74 0.42 0.34 0.24 *(Ratio=Positive AnswersjPositive+Negative Answers) In a one between analysis of variance the number of positive answers given for the program rated by each respondent was entered as a separate case. This was used as the dependent variable with the adult ranking of the programs as the independent variable. The results indicated that significant differences, F(4,92) = 20.86 p<.OOl, existed in the number of positive answers given for each program. Specifically the results in Table 5.2 indicate that adults ranked the prosocialness of programs from most to least as: (l) "Alvin and the Chipmunks" 47

PAGE 55

(2) "Beetlejuice" (3) "Garfield and Friends" (4) "Slimer and the Real Ghostbusters" (5) "Denver, the Last Dinosaur." In contrast, the results in Table 5.3 revealed that children ranked the prosocialness of the lead characters in these proqrams from most to least as: (1) "Denver, the Last Dinosaur" (2) "Slimer and the Real Ghostbusters" (3) "Alvin and the Chipmunks" (4) "Beetlejuice" (5) "Garfield and Friends." The Tukey HSD procedure indicated that the differences at the p<.OS level existed between both "Beetlejuice" and "Garfield and Friends" and the other three programs. This procedurewas indicated by the sharp differences in the ordering of the programs according to prosocial content between the adult panel and the children. The results indicate that the children rated the proqrams as being siqnificantly different in content. Rank order correlation between the adult and children's prosocial rankinqs of the five proqrams indicated that the adults and children have different opinions about the amount of prosocial behavior that occurred in the proqrams (Spearman's rho = 0.60, p=O.l64). The correlation did not approach acceptable siqnificance due to a small sample of only five proqrams. Analysis variance usinq qender as the independent variable revealed siqnificant differences in the 48

PAGE 56

children's answers on two of the behavior descriptions. The first was reparation for bad behavior which on the survey constituted apologizing for wrongdoing, F(l,91) = 6.45 p<.Ol3. The second was sympathy which on the survey was helping others to feel better, F(l,91) = 8.89 p<.004. Males rated the characters as exhibiting both of these traits more than did the females. For analysis of variance with the answer to whether or not the respondent would like to have the main character as a friend, there were significant differences in the answers to the same two behavior descriptions. That is, reparation for bad behavior, F(2,9l) = 4.56 p<.Ol3, and sympathy, F(2,9l) = 6.76 p<.002, again were the dimensions that were rated differently. Those who would like the character as a friend rated them as exhibiting both the traits more than did those who did not desire the character as a friend. The correlation between the total of positive opinion answers assigned to the programs by children and their desire to have the character as a friend was contrary to eXpectations. The negative correlation of -0.444 p<.OOl across all programs, between these items indicates that the less perceived prosocial behavior by a character predicts greater desire by children to want the 49

PAGE 57

character as a friend. This is well illustrated by the fact that the program "Garfield and Friends" ranked lowest on the prosocial scale derived from the children's opinions and yet 13 of 18 total respondents indicated they would like Garfield as a friend. 50

PAGE 58

CHAPTER 6 CONCLUSIONS The goal of this research was to develop a reliable and valid methodology in order to examine the existence and recognition of prosocial messages in Saturday morning children's television programs. The bulk of previous investigation of the effects of television viewing on children has been focused on the possible detrimental effects. The research that has explored the possibility of prosocial effects occurring has been done largely in laboratory settings and has used preschool age children as subjects. In order to determine whether or not prosocial messages exist in Saturday morning programs, it was necessary to use adults as raters in order to increase the objectivity of the observations. Fourth grade students; who are of an age to understand that the programs are fiction, were surveyed about. their opinions of how the characters in top rated Saturday morning programs behave in order to establish whether or not the messages were being recognized by the audience members. The investigator did not control the students viewed thereby minimizing experimental effects. Using a national rating service's ratings to determine what

PAGE 59

programs were to be investigated also eliminated researcher bias from that part of the process. In effect two parts of a communication process were investigated. First, the content of a message, that is, the occurrence of prosocial behaviors within the programs, was examined as objectively as possible. Second, the receivers recognition of the content was indirectly measured by soliciting their opinions about how the main characters behave. The question of whether or not the Liebert and Sprafkin typology provided the basis for a reliable method of measuring the prosocial message content of saturday morning children's programs can be at least partially resolved. The high degree of interrater reliability among the members of the adult rating panel indicates that agreement can be reached-about the meaning of a set of criteria for such measurement. Although unknown factors may have contributed to the ease of establishing the common definitions that the panel used in their rating, it is probable that a high degree of interrater reliability could be established with appropriate training of any similar rating panel. The success with establishing a high degree of interrater reliability among the adult panel members implies that the occurrence of prosocial behaviors in the 52

PAGE 60

program episodes that were rated is a valid finding. This finding can be used to support the hypothesis that prosocial models exist for the audience members to emulate if other individual psychographic factors provide the initial influence for such behavior. An important fact to keep in mind is that the presentation of a prosocial model for a child to emulate is not likely in itself to cause an alteration of behavior patterns. The influence of primary social agents would in large part determine what behaviors would be acceptable for emulation and this would provide the criteria that a child used for categorizing the modeled behaviors as to acceptability. The survey instrument used in assessing the recognition of prosocial messages by children yielded somewhat ambiguous results. Although the respondents gave positive answers about at least some of the behaviors in question for all the programs, their ranking of which programs contained the most prosocial content varied widely from the rankings assigned by the adult raters. It can be concluded from this that the subjective recognition of the prosocial messages by the children differs from the trained objective observation of the programs by the adult raters. This may be a function of identity with the characters in question. 53

PAGE 61

Adult raters would have little or no identification with the characters during a one-time observation of an episode of a program while the regular child viewer would develop identification with the main character over time .. The children's opinion of how a character behaves would evolve from seeing how the character acted in several episodes of a program. If the episode viewed by the adult rating panel happened to be one in which the prosocial content was low, differences would occur in the rankinqs assiqned. Other possible processes which might account for the fact that the children's opinions differed from the adult countinq of behaviors are recall and context. The children's opinions would be based on the recall of final outcomes of stories rather than concentration on detectinq and counting occurrences in a single episode. This would tend to blend the memory of individual antisocial acts that might have occurred into the context . of a prosocial outcome such as ultimately helping another to solve a problem or escape from a perilous situation. The differences among the children's opinions of whether or not the main characters exhibit various prosocial behaviors occurred in two areas. Reparation for bad behavior and sympathy were rated differently according to the gender of the child and whether or not 54

PAGE 62

the character was desirable as a friend. These differences could be attributed to the socialization process. Nine and ten year old fourth graders have probably started to develop gender related social characteristics. Desirable characteristics in a friend could be expected to be different for girls and boys depending on what sort of role modeling had been done by primary social agents. The negative correlation between the overall positive answer score for the programs and the desirability of the characters as friends must also be attributed to identification with the characters. Over a period of time, it is reasonable that children would not repeatedly choose to watch a particular program unless they felt some sort of friendship based on identification with the main character. The method used in this project to determine whether or not prosocial messages are included in the content of Saturday morning children's television seems to be reliable and valid. High interrater reliability was established and the programs that were rated were found to have some prosocial message content. The method used to determine whether or not these messages are received by the audience appears to have worked to a degree but did not yield sufficiently clear results to suggest that 55

PAGE 63

the prosocial messages are being recognized. If the adult rankings are used as a standard of measurement, it could be arqued that the differences between the adult panel and the children indicate that the children do not recognize the prosocial messages in Saturday morning programming. This is most likely a reflection of the different ways in which the measurement was done. The adult rating panel was counting overall occurrences to include the actions of supporting characters in the programs. This would mask somewhat the fact that the main character was not behaving in what would be considered a prosocial manner. The children were restricted to evaluating the main character and this would mask the prosocial actions of supporting characters. In order to resolve this problem of measurement, it would be necessary to word the survey questions to cover the entire cast of characters in a program or to have the adult rating panel count only prosocial behaviors exhibited by the main character. However, the fact that the children's opinions of the prosocial content of the various programs were significantly different indicates that their responses were nonrandom. This finding suggests some sort of systematic recognition of a perceived prosocial pattern of behavior on the part of the characters by the 56

PAGE 64

children. The finding of an inverse correspondence between the adults' and children's prosocial rankings muddies the finding that the children recognized some prosocial messages and requires further investigation with more comprehensive research tools. The questions which arose during the analysis of the results of the survey administered to the children concern the validity of the method. The first problem area which must be noted is the fact that the surveys were accompanied by a parental consent form and were completed by the children at home. This may have led to consulting with the parents about the meanings of the questions and what the "appropriate" answer might be. This would introduce some parental bias into the answers given and would not be detectable in the course of analyzing the results. In order to avoid this, it would be better to obtain parental permission separately from the administration of the survey instrument. The necessarily simple wording of the questions_on the survey was another problem area. Asking the children their opinion of overall behavior patterns makes it easy for them to answer but does not elicit a truly thoughtful response. An instrument consisting of various scenarios with the character name inserted would provide a method for more precisely measuring children's opinions of a 57

PAGE 65

character's behavior patterns. Additionally, it is difficult to determine whether or not the behavioral descriptions used carry the same connotation in the children's minds as in the mind of an adult. Resolving this problem would require interviewing a large sample of children in order to determine what best describes to them the various behavior patterns being investigated. once the most common connotations used by children in a particular age group were determined, the questions used in soliciting their opinions could be worded in a way that would make them a more valid measurement. Another extension of the opinion measurement would be to recruit the parents of the children to also view the programs and evaluate the main characters on the Liebert and Sprafkin typology. This would allow for cross referencing the opinions of the parents and their children to determine what similarities exist between the opinions of primary social agents and the children using a common basis of measurement. The results -obtained would provide indications of how much influence is exerted by primary social agents on the attitudes of children. This two pronged approach could also provide indications as to how much the training of a rating panel contributes to the reliability of its measurements. Comparing the results obtained from two different adult 58

PAGE 66

groups with those of a group of children would yield more comprehensive findings on which to base conclusions and determine what the next level of refinement should be in order to more firmly establish the existence and recognition of prosocial messages in Saturday morning programming. The results of this research indicate that the prosocial messages do exist in Saturday morning children's television programs and that it can be reliably measured. There were also indications that children recognize these messages. The reliability and validity of the method used to measure the children's recognition of prosocial messages needs to be greatly improved over what was used in this project. This improvement would require more resources to be available to conduct the survey of children's opinions. The possible influence of parental bias needs to be eliminated by administering the survey outside the home. Using a national rating service s ratings to select subject programs and not controlling the children's viewing should be retained in order to prevent researcher bias from influencing the results. It is evident that examining the possible effects of television viewing on children in a naturalistic manner is. achievable. Taking such research outside the 59

PAGE 67

laboratory and controlled field experiment setting -provides a more realistic picture of what children are recognizing as the message of what they view. The communication concept of the meaning of a message residing in the receiver can be applied only by determining what meaning is given to the message content of television programs by the receiver, in this case the children who are the intended audience of those who produce what is broadcast on Saturday mornings. 60

PAGE 68

APPENDIX A -Adult Rating Panel Coding Sheet CODING SHEET NAME OF DATE TAPED---------------------------------As you note the occurrence of one of the listed behaviors, mark its occurrence in the count column for that behavior. MESSAGE TYPE COYNT ALTRUISM -sharinq, helpinq & cooperation involvinq humans or animals. CONTROL OF AGGRESSIVE IMPULSES -nonaqqressive acts or statements thatprevent aqqression. DELAY OF GRATIFICATION/PERSISTENCE -acts of delaying qratification & task persistence. EXPLAINING FEELINGS statements explaininq the feelinqs, thinkinq or action of self or others. REPARATION FOR BAD BEHAVIOR -behavior intended as reparation for wrongdoing by person. RESISTANCE TO TEMPTATION -withstandinq the temptation to engage in prohibited behaviors. SYMPATHY -verbal or behavioral expression of concern for others. 61

PAGE 69

APPENDIX B -Fourth Grade Student Survey Instrument I watch (name of program) ALMOST ALWAYS SOMETIMES ALMOST NEVER HAVEN'T SEEN IT (Character name) likes to share and help others. (Character name) likes to fight. (Character name) keeps working until the job is done before playing. (Character name) helps settle arguments by explaining why YES NO DON'T KNOW YES NO DON'T KNOW YES NO DON'T KNOW they started. YES NO DON'T KNOW (Character name) apologizes when he/she does something wrong. YES NO DON'T KNOW (Character name) takes things that belong to someone else. YES NO DON'T KNOW (Character name) would help someone who was feeling bad to feel better. YES NO DON'T KNOW I would like (Character name) to be my friend. YES NO DON'T KNOW I AM A BOY __ GIRL,_ __ WHO IS _____ YEARS OLD. 62

PAGE 70

BIBLIOGRAPHY Anderson, J. A., & Meyer, T. P. (1988). Mediated communication g social action perspective. Newbury Park, CA: Sage Publications. Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall, Incorporated. Becker, s. L. (1987). communication (2nd ed.). Glenview, IL: Scott, Foresman and Company. Berkowitz, L. (1984). Some effects of thoughts on antiand prosocial influences of media events: A cognitiveneoassociation analysis. Psychological Bulletin, 95, 410427. Bryan, J. H., & Walbek, N.H. (1970a). Preaching and practicing generosity: Children's actions and reactions. Child Development. 41, 329-353. Bryan, J. H., & Walbek, N.H. (1970b). The impact of words and deeds concerning altruism upon children. Child Development, 41, 747-757. Christenson, P. G. (1986). Children's perceptions of moral themes in television drama. In M. L. McLaughlin (Ed.), Communication Yearbook 9. Beverly Hills, CA: Sage Publications, Incorporated. Collins, w. (1982). Cognitive processing in television viewinq. In o. Pearl, L. Bouthilet, & J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (vol. 2) (pp. 9-23). Rockville, MD: National Institute of Mental Health. Comstock, G. (1982). Violence in television content: An overview. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.), Television and behavior: Ten years scientific progress gng implications for the eighties (vol. 2) (pp. 108125). Rockville, MD: National Institute of Mental Health.

PAGE 71

Cook, T. o., Kendzierski, o. A., & Thomas, s. v. (1983). The implicit assumptions of television research: An analysis of the 1982 NIMH report on television and behavior. Public Opinion Quarterly, 47, 161-201. corder-Bolz, c. (1980). Mediation: The role of significant others. Journal of Communication, 30(3), 106118. corder-Bolz, c. (1982). Television literacy and critical viewing skills. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (vol. 2) (pp. 91-103). Rockville, MD: National Institute of Mental Health. Dorr, A. (1982). Television and affective development and functioning. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (vol. 2) (pp. 68-77). Rockville, MD: National Institute of Mental Health. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to theory and research. Reading, PA: Addison-Wesley Publishing Company. Forge, K. L. s., & Phemister, s. (1987). prosocial cartoons on preschool children. Journal, 12(2), 83-88. The effect of Child Study Freedman, J. L. (1984). Effect of television violence on aggressiveness. Psychological Bulletin, 92, 227-246. Freedman, J. L. (1986). Television violence and aggression: A rejoinder. Psychological Bulletin, 100, 372-378.; Friedrich, L. K., & Stein, A. H. (1975). Prosocial television and young children: The effects of verbal labeling and role playing on learning and behavior. Child Development, 46, 27-38. Friedrich-Cofer, L., & Huston, A. c. (1986). Television violence and aggression: The debate continues. Psychological Bulletin, 100, 364-371. 64

PAGE 72

Friedrich-Cofer, L. K., Huston-stein, A. H., Kipnis, o. M., Susman, E. J., & Clewett, A. s., ( 1979). Environmental enhancement of prosocial television content: Effects on interpersonal behavior, imaginative play, and self-regulation in a natural setting. Developmental Psychology, 15(6), 637-646. Fry, D. L., & Fry, V. H. (1986). the study of mass communication. (Ed.), Communication Sage Publications, Incorporated. A semiotic model for In M. L. McLaughlin Beverly Hills, CA: Gerbner, G. (1971). Violence in television drama: Trends and symbolic functions. In G. A. Comstock, & E. A. Rubinstein (Eds.), Television and social behavior reports and papers volume I: Media content and control. Rockville, MD: National Institute of Mental Health. Gorn, G. J., Goldberg, M. E., & Kanungo, R. N. (1976). The role of educational television in changing intergroup attitudes of children. Child Development, 47, 277-280. Hirsch, P.M. (1981). On not learning from one's own mistakes, a reanalysis of Gerbner at al.'s findings on cultivation analysis part II. Communication Research, 3-37. Huesmann, L. R. (1982). Television violence and aggressive behavior. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (vol. 2) (pp. 126-137). Rockville, MD: National Institute of Mental Health. Huesmann, L. R., Eron, L. D., Klein, R., Brice, P., & Fischer, P. (1983). Mitigating the imitation of aggressive behaviors by changing children's attitudes about media violence. Journal of Personality and Social Psychology, !,!, 899-910 Huesmann, L. R., Lagerspetz, K., & Eron, L. D. (1984). variables in the TV violence-aggression relation: Evidence from two countries. Developmental Psychology, 746-775. Idea of boycotting TV advertisers gaininq. (1989, June. 3). Colorado Springs Gazette Telegraph. pp. El, E6. 65

PAGE 73

Koop takes aim at alcohol. Springs Gazette Telegraph. (1989, June 1). p. A6. Colorado Liebert, R. M., & Sprafkin, J. (1988). The early window: Effects of television on children and youth (3rd ed.). New York: Pergamon Press. Lowery, S., & DeFleur, M. L. (1988). Milestones in mass communication research: Media effects (2nd ed.). New York: Lonqman Inc. Loye, D., Gorney, R., & Steele, G. (1977). Effects of television: An experimental field study. Journal of Communication, 27(3), 206-216. Murray, J. P., Rubinstein, E. A., & Comstock, G. A. (Eds.), (1972). Television and social behavior. Reports and papers vol II: Television and social learning. Rockville, MD: National Institute of Mental Health. National Institute of Mental Health. (1982). Television and behavior: Ten years Qf scientific progress and implications for the eighties. (DHHS Publication No. ADM 82-1195). Washington, DC: u.s. Government Printinq Office. Nikken, P., & Peeters, A. L. (1988). Children's perceptions of television reality. Journal of Broadcasting and Electronic Media. 32, 441-452. Norusis, M. J./SPSS Inc. (1988). SPSS-X(tm) Advanced statistics guide. (2nd ed.). Chicago: SPSS Inc. 'Poulos, R. w., Harvey, s. E., & Liebert, R. M. (1976). Saturday morninq television: A profile of the 1974-75 children's season. Psychological Reports, 39, 1047-1057. Rosenthal, T. L., & Zimmerman, B. J. (1978). Social learning and cognition. New York: Academic Press, Inc. Rushton, J.P. (1979). Effects of prosocial television and film material on the behavior of viewers. In L. Berkowitz (Ed.), Advances !n experimental social psychology. Vol. 12. New York: Academic Press. pp. 321-351. 66

PAGE 74

Rushton, J.P. (1982). Television and prosocial behavior. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (vol. 2) (pp. 248-257). Rockville, MD: National Institute of Mental Health. Schramm, w., Lyle, J., & Parker, E. B. (1961). Television in the lives of children. Stanford, CA: Stanford University Press. Singer, D. G. (1982). Television and the developing Imagination of the child. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (vol 2) (pp. 39-52). Rockville, MD: National Institute of Mental Health. Singer, J. (1980). The power and limitations of television: A cognitive-affective analysis. In P. H. Tannenbaum (Ed.), The entertainment functions of television. (pp. 31-65). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Publishers. Singer, J. L. and Singer, D. G. (1976). Can TV stimulate imaginative play? Journal of Communication, 46(3), .74-80. Stein, A. H., Friedrich, L. K., & Vondracek, F. (1972). Television content and young children's behavior. In J. P. Murray, E. A. Rubinstein, & G. A. comstock (Eds.), Television and social behavior: Reports and papers, vol II: Television and social learning. Rockville, MD: National Institute of Mental Health. Stein, G. M., & Bryan, J. H. (1972). The effects of a television model upon rule adoption behavior of children. Child Development, 43, 268-273. Surgeon General's Scientific Advisory Committee on Television and Social Behavior. (1971). Television and growing YR: The impact .of televised violence. Report to the Surgeon General. Washington, DC: United States Public Health Service. 67

PAGE 75

Tamborini, R., Zillmann, D., & Bryant, J. (1985). Fear and Victimization: Exposure to television and perceptions of crime and fear. In R. N. Bostrom, & B. H. westley (Eds.), Communication Beverly Hills, CA: Sage Publications, Incorporated. Tan, A. s. (1986). research (2nd ed.). Company. Mass communication theories and New York: MacMillan Publishing Tannenbaum, P. H. (Ed.), (1980). The entertainment functions of television. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Publishers. Walker, K. B., & Morley, D. D. (1988). Attitudes and parental factors as intervening variables in the television violence-aggression relation. Unpublished manuscript, University of Colorado at Colorado Springs, Colorado springs, co. Walker, K. B., & Morley, D. D. (in press). Adolescent perceptions of context in violent media portrayals. Speech Communication Annual. Williams, F. (1986). Reasoning with statistics. (3rd ed.). New York: Holt, Rinehart and Winston. Wood, D. N. (1983). Mass media and the individual. st. Paul, MN: West Publishing Company. 68 . I



PAGE 1

PHYSICIANS' OPINIONS ABOUT PATIENT INVOLVEMENT IN I HEALTH AND MEDICI CARE DECISIONS AND TELEPHONE-BASED DECISION SUPPORT Jerry Wjyne Howell I B.A, University of Northern Colorado, 1977 MS., University of at Boulder, 1986 A thesis submitted to the I University of Golorado at Denver I in partial fulfillment of the requiremets for the degree of I Doctor of Philosophy Health and Behavioral Sciences 1J 9 .. B}.

PAGE 2

1999 by Jerry Wayne Howell All right reserved.

PAGE 3

This thesis for the Doctor of Philosophy degrjby Jerry Wayne Howell has been approved by Deborah S. Main Wendyhl.lynch Date

PAGE 4

Howell, Jerry Wayne (Ph.D., Health and Behavioral Sciences) Physicians' Opinions About Patient InJolvement in Health and Medical Care Decisions and Telephone-Based Decitn Support Thesis directed by Professor Deborah S. Main ABSTRACT Survey research (n = 914) and key interviews (n = 21) were used to assess physicians' opinions about: I) cobsumers/patients being more informed and taking a more active partnership role in Ihedical care decisions, 2) the incorporation of telephone-based decision support (TBDS) into care management, and 3) the effect a more active consumer/patient partnership and TBDS will have on health outcomes, patient satisfaction, doctor-patient clinical efPciencies, utilization, costs, and malpractice liability. The percentage of physicians agreeing that patients want more information about health and medibl issues and more active involvement in medical decision making was 900/o and s;s%, respectively. The percentage of physicians agreeing that physician and patient together should make decisions about minor and major medical conditions wasi56% and 68%, respectively, with 66% agreeing that they would like their patients to take a greater role in self-managing chronic conditions. Physicians believed that a more informed patient would result in better medical decisions leading to imprJved health outcomes, patient satisfaction, doctor-patient relationships, utilization efficiencies, and a potential reduction in medical liability. However, physicians in their preference for involving patients in decision making and acknowledged that patients have varying preferences for involvement. They believed that the knost prudent clinical approach is for the physician to be sensitive to each indivicWal patient's preference for information and involvement. Opinions varied regarding !the purposes, potential impacts, and reasons for why health plans had implemented TBDS services. Logistical regression analyses showed that primary care physicians werb more likely than referral specialists to: agree that their patients should take a grJater role in treating minor conditions; agree that their patients should take a greater rJte in self-managing chronic conditions; be knowledgeable about TBDS; have a positive opinion about TBDS; believe that iv

PAGE 5

the health plans implemented TBDS to respond to interests of the business community, to more efficiently care for capitated patients, improve quality of care, and member satisfaction; believe that would decrease the number of medically inappropriate visits and unnecessary office calls. Explanations for these findings, implications for working wit 1 physicians, and recommendations for future research are discussed. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. v

PAGE 6

DEDI
PAGE 7

Dr. Main for her unwavering '-'V1Ji1.1U'!;i11'-" in me, her boundless enthusiasm, her sage leadership, mentoring and in directing this thesis. Dr. Main, Dr. Iverson, Dr. Dr. Lynch, and Dr. Stulp for working together as a multidisciplinary in sharing their wealth of knowledge and experience to create and nurture 1 learning environment that should characterize all dissertation eXJ>eliJen<:es. Joe Mc<;Jloin for his patience and cal assistance in data management and statistical analysis. Lu Sandoval for her tenacity and in transcribing the key informant interviews. Sherry Holcomb for her and attention to detail in survey administration and data cotllectlon. David Nash, M.D., Zach M.D. and Mark Earnest, M.D., for their critical review of the qualitative Barb and Don Day for providing "home away from home." Pepper Hatch for his leadership "flying high cover" and providing consistent support and 1 to complete this work. All the physician subjects who l!fl!lCI
PAGE 8

co CHAPTER 1. IN'I'R.ODUCTION...... . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Specific Aims ......... l .............................................. 2 I Background and ..................................... 3 Overview ofResearcH Design and Setting ....................... 8 I Study Design ..................................................... 9 Study Site .......... j ............................................ 11 I Telephone-Based Decision Support Services .............. 12 Key Questions ..... j ............................................ 16 I Reconciling Quantitative and Qualitative Analysis ............ J ............................................ 18 Organization of the DJssertation ................................. 18 2. REVIEW OF LITERATURE ........................................... 20 Geographical Variatio sand Appropriate Use ofHealth Care .... 1 ............................................ 20 Demand-Side lnlerviions ...................................... 24 Perceived Need and Self-Care .............................. 24 Chronic Disease MLgement ............................... 27 Patient Preference id Shared Decisions ................... 29 V111

PAGE 9

Patient Preference for Information and Involvement in Decision Making ............................................... 32 Physician Perception lbout Patient Desire for Infonnation and Involvement .................................... 52 Doctor-Patient ComJunication and Relationships: Challenges/Barriers abd Effects on Outcomes . . . . . . . 54 Doctor-Patient coLunications: Challenges and Barriers ........ l ............................................ 55 ............. 65 Models for Physician-Patient Communication and Relationships. J ............................................ 71 Diffusion of Innovatijns: Constructs, Components, and Determinants ..... 11 ............................................ 79 Components ofDitsion ..................................... 79 Adoption Rate Determinants ................................. 82 Diffusion oflnnovltion in Health Care ..................... 85 Summary and ImplicJions for Research ..................... 117 3. QUANTITATIVE METHODS AND RESULTS ................... 119 Introduction ............ j . . . . . . . . . . . . . . . . . . . . 119 Methods ........................................................... 120 Subjects ........................................................ 120 I Physician Opinion Survey (POS) .......................... 120 lX

PAGE 10

POS Administration and Data Collection Protocol ............ l ........................................... 122 Analysis and Statistill Methods .............................. 123 Sample Descripti+ ......................................... 123 Description of Variables ..................................... 123 Selection of StatiJcal Tests ................................ 126 Results ............................................................. 129 Sample Description .......................................... 130 I Description of Variables ..................................... 132 Logistic Regressiot ......................................... 133 4. QUALITATIVE METHODS AND RESULTS ..................... 153 Introduction ............ l .......................................... 153 Methods ................ J .......................................... 154 Sample Selection .. 11 .......................................... 154 Physician Opinion Interview (POI) Guide ................ 156 Interview Structure and Administration ................... 157 Analyses ........................................................... 158 Transcription, Cod1g, Segmentation. . . . . . . . . . . 159 I Data Analysis and
PAGE 11

Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 163 Subjects ........................................................ 163 Opinions About Patient Involvement in Medical ............................................ 165 Opinions About Tilephone-Based Decision Support .l ........................................... 214 Closing Questions ........................................... 242 Synopsis: Overarching Themes ............................... 246 Theme 1: PhysiciaJs Support Idea of More Informed Adive Patients... . . . . . . . . . . . . 246 I Theme 2: Improved Health Outcomes: Chronic and Major Dlness j...... . . . . . . . . . . . . . . . . .. 247 Theme 3: Improved Patient Satisfaction (Patient Preference j. . . . . . . . . . . . . . . . . . . . 249 Theme 4: Doctor-Patient Relationship (Patient Preference)......................................... 250 Theme 5: Clinical Efficiencies for Minor and Chronic Illnesses....................................... 252 I Theme 6: Reduced Health Care Costs for Minor and Chronic Condi,.ms...... ... ......... ... ...... ... ...... 255 Theme 7: Reduced Potential for Medical Liability.. . . . . . . . . . . . . . . . . . . . . . . . . . . 256 Theme 8: Self-Efficacy Improved with More Informed Active Pahent .................................... 257 I Summary. Themes 1-8 ..................................... 257 X1

PAGE 12

Theme 9: Variance in Physician and Patient Preferences ...................................................... 258 Theme 10: TBDS -Uncertain or No Impact on Physicians' Practices and Patients .............................. 261 Theme 11: TBDS -Uack of Support by Practicing Physicians Will Diminish Its Value ............... 262 5. DISCUSSION ............................................................ 266 Key Quantitative and Qualitative Results .................... 266 Physicians' Opinions about Shared Decision Making ............. l ........................................... 267 Physicians' Opinijns about TelephoneBased Decision .................................... 277 Commentary on Jplications for Working with Physician Organizations ............................... 291 Limitations ........................................................ 304 Recommendations for Future Research ....................... 314 Conclusions ....................................................... 319 APPENDIX A PHYSICIAN OPINION SURVEY (POS) ........................... 321 I B. PHYSICIAN OPINION INTEVIEW (POI) ........................ 328 Xll

PAGE 13

C. QUALITATIVE ANAL VIIS CODEBOOK ........................ 334 D. VARIABLES WITH MISSING VALUES .......................... 340 REFERENCES ..................................................................... 343 xiii

PAGE 14

CHAPTER 1 I INTRODUCTION During the last decade ofU.S.1io-cultural history, the consumer has played an increasingly greater role in shaping the marketplace. The baby-boomer generation that has shaped virtually every other sectlr of the U.S. economy is emerging as the "new consumer'' for the healthcare markltplace as well. This new consumer is generally more informed; more derpandink about access, convenience, and service quality; generally more assertive in the d,ctor -patient relationship. In essence, they want to be more self-sufficient and independent. For this new consumer to be effective in shaping the healthcare they will need to technological tools and processes that can provide infoLtion and that can improve their ffi . h d I .lhi e ect1veness m t e octor-pat1ent re at10ns p. Consequently, this new era of conLerism in healthcare has stimulated an interest in the concept of demand manageLent and has spawned a rapidly growing increase in consumer health management id decision support services. Demand management has been defined as the use decision and self-management support systems to enable and encourage consumJs to make appropriate use of medical care (Vickery & Lynch, 1995). One of the conLmer health management services that has emerged has been telephone-based decisiol support (TBDS). Through a 1-800 1

PAGE 15

number, such services provide 24-hour per day access to specially trained nurse counselors. Using on-line decision supplrt protocols or algorithms, the nurse counselors provide decision support, selftgement support, psychosocial support, education and infonnation to consumers/patients on virtually any health care topic. The market potential for these types of sekces has been estimated to be somewhere I between 100-200 million lives (Equitable Securities Research, 1996). Even though the opinion of the may stands as a critical variable in determining the ultimate impact of more Jronned, actively involved patients and mos, reviews of published literature revLed no systematic assessment of physicians' viewpoints about these issues. This research will advance our understanding about how physicians view enhanced participation of patients in medical I decision-making and the integration of TBDS services into care management. This infonnation will provide key insights on thl implications for working with physicians in optimizing the integration of these comp\onents into the delivery of health care. Specific Aims A multi-method approach was used to adJess four central issues: The specific aims of this research were: 2

PAGE 16

Aim #1: To ascertain primary care and specialty physicians' opinions about more informed, actively involved patienls as partners in health and medical care decisions. Aim #2: To assess primary care and specialty physicians' opinions about the impact they believe consumers/patients bling more informed and taking a more active partnership role in health and medicali decisions will have on: the quality ofthe doctor-patient relationship, clinical efficiencies, health outcomes, patient satisfaction, health care utilization, health care costs, L malpractice liability. Aim# 3: To assess primary care Ld specialty physicians' opinions about the incorporation of TBDS services into care management. Aim# 4: To assess primary care and specialty physicians' opinions about the impact they believe a TBDS service will hlve on: the quality of the doctor-patient relationship, clinical efficiencies, health otmes, patient satisfuction, health care utilization, health care costs, and malpractice liability. Background and Significance Proposals and strategies fur reform (generally cost contairunent) have been derived chiefly from a supply-si(ie economic paradigm. Supply-side I strategies have historically emphasized global budgets, consumer tax incentives, cost\ shifting to consumers, capitated payments for physicians, technology management, 3

PAGE 17

hospital bed capacity control, use of gatekeepers, utilization review, and other forms of "managed care." In contrast, little eLphasis has been given to integrating the I biological sciences with demand-side social, cultural, psychological, and health risk factors which influence health care decisiln making, health care behaviors, and health I care utilization that ultimately impact health outcomes, patient satisfaction and health care costs. Recently, however, the concept oi "demand management" has received significant national interest with the incorporation of demand-side strategies into managed care delivery systems. Demand Lanagement has been defined as the use of decision and self-management support syslems to enable and encourage consumers to I make appropriate use of medical care (Vickery & Lynch, 1995). I The supposition behind demand-side strategies is that health care decisions and health care seeking behaviors are influenc1 by a complex set of psychological, social, and culturallilctors which are more relatJ to perceived need and patient preference rather than to morbidity (Vickery & LyncJ 1995). It appears that the access an individual has to infonnation and resourceJ influences perceived need and patient preference for services which in tum influelces behaviors associated with health and I medical decisions (Berkanovic Telesky, & Reeder, 1981; Bradley, Zia, & Hamilton, I 1996; Dentzer, 1991; Greenfield, Kaplan, W\ are, Yano & Frank, 1988; Levine & Britten, 1973; Lorig, Mazonson & Holm 1993; Mayo, Richman & Harris, 1990; 4

PAGE 18

Vickery, Kalmer, Lowry, Constantine, Wright & Loren, 1983). That is, a well-infonned patient will make ruJ1 nt decisions than a less-informed patient and will subsequently utilize health care resources differently. I This recent interest in demand-side strategies has subsequently spawned a rapidly growing increase in consumer he1th management and decision support I services. One such service has been TBJS. Through a 1-800 number, such services provide per day access to specially trained nurse counselors. With the aid of sophisticated on-line decision support pJtocols or algorithms, the nurse counselors provide decision support, self managemelt support, information, and psychosocial I fc ().I .. d.. d" support to consumers patients or: a tnflge or pre-VISit ects10ns regar mg acute minor illnesses and injuries (e.g., infectiol, headache, back pain, etc.); (b) shared d h h a1 h d I h d d. ectsions mt e t care proVl ers concermng maJor t erapeutic an Iagnosttc procedures (e.g., surgery, hospitalization, chemotherapy); c) chronic disease management (e.g., diabetes, asthma, hypertension); and d) lifestyle management (e.g. .gh lini. I a1 . ) exercise, stress, wei t management, c c preventive sernces There are over 60 million Amerijs enrolled in some form of managed care. Capitation has increased in prevalence as 1echanism for reimbursement. For managed I care to be ultimately successful, however, it must mature beyond strictly a mechanism for cost containment to one focused on mLng and managing value. Market dynamics suggest that employers and goveLent as the ultimate purchasers of care 5

PAGE 19

and patients as the ultimate consumers of care are looking beyond just cost as a basis by which they will select health care seJces. Even though cost will remain a primary consideration, they are looking to other ijdicators of value and quality such as patient satisfilction, access, convenience, and outcomes. With the advent of a more informed and assertive consumer, the health care industry is ripe for consumer health and dlcision support services. The market I potential for TBDS has been estimated to be at least I 00 million lives and as many as 200 million (Equitable Securities ResearcJ 1996). However, despite the recent growth of TBDS services, to date we knor very little about their impact on the quality of the doctor-patient relationship, clinical efficiencies, health outcomes, patient satisfaction, health care utilization, health jare costs, and malpractice liability. As these types of services gain visibility and bLme more widely incorporated into health care delivery systems as a "demand-side" stategy, an increasingly sophisticated American population of consumers, patientl, physicians, purchasers of medical care I (government and business), and insurers will pose questions about the medical efficacy and efficiency of such services. \ Physicians, in particular, are in clinically appropriate questions. Do consumers/patients want more information about health and medical care decisions? I Do consumers/patients want increased involvement in health and medical care decisions? Will TBDS services improve the\ quantity and quality of information 6

PAGE 20

desired by consumers/patients? If so, what impact will such services have on: the quality of the doctor-patient clinical efficiencies, health outcomes, patient satislilction, health care utilizatiot health care costs, and malpractice liability? Consumers and patients are desiring more infonnation related to health and medical care decisions. In 1988, GreenfiJld, Kaplan, Ware, Yano, and Frank prophetically stated, "We are entering an era of unprecedented sophistication and knowledge among American patients. . In the everyday practice of medicine, will 1 fh .\.aldi will" kind mvo vement o t e patient represent a cntlc sruptlon or 1t promote a new of relationship that enhances the roles of lhysician and patient in generating improved I outcomes" (p. 457)? Anecdotal reports TBDS decision support services have indicated that they support the doctor-patient relationship, enhance patient satisfaction and save money without sacrificing qualiJ of care (Barnett, 1995). Even though the opinion of the stands as a critical variable in detennining the ultimate impact such decilon support interventions will have on physician-patient relationships, clinical ,lencies, patient outcomes, medical care utilization, and costs, reviews of literature revealed no systematic assessment of physicians' viewpoints about these types of services. Nor has anyone tried to ascertain the implications of their opinions on the implementation, utilization, and ultimate impact of such services on health care quality, cost, and access factors. 7

PAGE 21

Consumer health management s pport services, including TBDS, have experienced rapid growth (Barnett, 199b; Equitable Securities Research, 1996). However, there is a dearth of informatiol regarding the opinion of physicians about such services and the effect physicians' might have on the ultimate effectiveness of these services. This reselch is one of the first to: (a) Advance our understanding about how physicians vi1 enhanced participation of patients in medical decision-making and the integration ofTBDS services into care management and (b) provide key insights on the implications fir working with physicians in optimizing the integration of such services in the of health care. Overview of Research Design and Setting' This research has been designed tJ be primarily descriptive and hypothesis generating. A multi-method approach uslg both quantitative (survey research) and qualitative (key infonnant discussions) melhods has been utilized to address the specific aims of this study. This section ptvides an overview of the study design and study site as well as a description of the telephone-based decision support services that were implemented within the community lhere the physician subjects for this study were practicing. Key questions explored l this study are presented. The reconciliation of quantitative and qualitatije data are discussed. 8

PAGE 22

Study Design A multi-method approach was chosen to enhance methodological rigor. Patton (1990) has emphasized the value lftriangulation or the use of multiple I methods in the research and evaluation of a given phenomena. Triangulation or multiple methods can involve the utilizatiln of different kinds of quantitative and qualitative data and methods. Therefore, both quantitative and qualitative approaches were employed concurrently to assess provider perceptions and opinions. Survey research and key infonnanl interviews were used to ascertain physicians' opinions about: 1. The increased involvement of as partners in health and medical care decisions. 2. The impact they believe consumers/patients being more infonned and taking a more active partnershib role in health and medical care decisions will have on: the quality of the relationship, clinical efficiencies, health outcomes, pltient satisfaction, health care utilization, health care costs, and liability 3. The integration ofiDDS servit into care management and the impact they believe a TBDS service will have on: the quality of the doctor-patient 9

PAGE 23

relationship, clinical efficiencies, health outcomes, patient satisfaction, health care utilization, health care costs, and malpractice liability. Table 1.1 denotes the specific questions from the Physician Opinion St (Appendix A) and the Physician Opinion Interview (Appendix B) used to address each of these issues. T bl 1 1 Ph . 0 . S 6 k nfl a e .. l}'Sictan pwon urvey an: ey_ 1 ormant mterview questiOns. I Issue Physician Opinion Survey Key Informant I Intenriew Patient involvement I B1, B2, B3, C2 1, 2, 4, 5, 6, 13 Doctor-patient D4b 4, 5,6,9 relationship Clinical efficiencies D4a \ 4, 5, 6, 9 Health outcomes D3, D4b \ 3, 4, 5, 6, 9 Patient satisfaction D3, D4b I 4, 5, 6, 9 Patient retention D4b I Continuity of care D4b I Health care costs B2,D3 \ 4, 5, 6, 9 Malpractice liability D3, D4a \ 4, 5, 6, 9 Physician D4a reimbursement Utilization and :82, D4a I 4, 5, 6, 9 appropriateness TBDS Service Dl, D2, D3, Elf E2 7, 8, 9, 10, 11, 12, 14 10

PAGE 24

Study Site The study site was the communi of Rochester, New York (NY). Located in upstate NY on Lake Ontario, Rochester las a largely stable population base ( + 3% from 1980-1996) of just over one millioJ with a demographic composition ( 16% minority) comparable to similar sized urbL areas in the Northeast United States. It has an export-oriented, high technology manjfacturing-based economy. Unemployment rate is usually a percentage point or more lower than the U.S. average with the cost of living slightly above the national average. Prior to this study, the Rochester system received recognition as a "model" for health care delivery. Rochester's success is attributed primarily to strong regional planning and cooperation that focused primarily on supply side management initiatives including: community rating and continuot open enrollment, near universal coverage, I technology assessment and control, limitei hospital bed capacity, inter-hospital cooperation, and a moderate managed care presence. Health insurance and medical d d aril b I d h d d care were provt e pnm y y two camls an t err respective m epen ent practice associations (IPAs): (a) Blue Cross and Blue Shield of the Rochester Area (BCBSRA) and the Rochester Community IPA (RCIPA) which owned 73% of market share, and (b) Preferred Care, an independent HMO, and the Rochester IP A (RIP A) which owned 13% of market share at the tile of this study. There were 11

PAGE 25

approximately 2100 total physicians com rising both RCIP A and RIP A with 95% of the total physicians providing services both BCBSRA and Preferred Care. Approximately 80,000 insured fi es from seven major employer groups and two major health plans in Rochester weJ offered a TBDS service. This service had just been implemented at the time of this ldy. Available 24 bows/day and toll-free, the telephone-based nurse counseling 1ce was staffed by nurse counselors trained in assisting, educating, and supporting peeple faced with choices about health and medical care. The counselors were requiJd to be registered nurses with a minimum of five years of current clinical experience jd a minimum of a bachelor of science degree . mnursmg. Nurses who staffed the TBDS service received 240 hours of structured education and training. The training focu{ed on imeractive learning, designed to familiarize them with concepts of commumcation, communication skills, and applications of these skills in interactions tth callers. Training emphasized nondirective, non-prescriptive techniques of communication as well as patient-centered rather than provider-centered communica+ns. Second, the nurse counselors were trained in psychosocial assessment and self+-care alternatives, and how to apply them to 12 I

PAGE 26

the phone counseling process. Third, they received training on how to use computer hardware and software which comprised re service's infunnation system. The training curriculum for the nurse counselors was comprised of three phases. Phase I consisted of 160 hours initial training in communications, infunnation systems, and protocols for tching medical infOrmation. Phase ll embodied an intensive 80-hour mentorship with a senior nurse counselor which included performance audits and feedbacJ. Phase Ill incorporated another 8-10 weeks of modified mentorsbip with a senior ni counselor, also with performance audits and feedback. Fallowing this formal, initi81 training, nurse counselors received 60 hours of continuing education per year on information systems, medical and health topic reviews, and research protocols; participated in weekly case review meetings; d k 1 . .. \ ed an too part m quarter y m-servtce trammg sesstons 10cus on customer servtce. The primary objectives of the nursl counselors were to provide non-directive, . d . If. I "al I non-prescnpttve ectston support, se -management support, soct support, stmp e information requests, and referrals to local agencies over the phone on virtually any health and medical issue: (a) triage or pre-provider decisions concerning acute minor illness or injuries (e.g., headache, back pJ sprained ankle); (b) major medical decisions shared with health care provi, (e.g. surgery, chemotherapy, advanced imaging, hospitalization); c) chronic disease management (e.g. diabetes, asthma, 13 I

PAGE 27

hypertension); (d) lifestyle changes to reduce risk of disease and injury (e.g. body fat loss, exercise, stress management). In addition to the TBDS service, hll recipients of the service received the fOllowing member conununication matJals: (a) an introductory packet (letter of I introduction, brochure, wallet cards, refrigerator magnet, initial quarterly newsletter) regarding the telephone counseling servJe; (b) a copy of the medical self-care reference guide, Take Care if Yourself; l) eight postcard reminders to use the service; and (d) a quarterly newsletter. In Februlry, May, August, and November, subjects received the quarterly newsletter. The quLerly newsletters emphasized specific, I usable information for personal health and medical care decisions regarding preprovider triage, major chronic diLe management, and healthy lifestyles. Each month in between newsletter issues, subjects received one of the reminder postcards promoting the benefits of calling the service for a wide range of health and medical issues. The self-care reference Jde featured clinical algorithms for over I 00 common medical problems for adults and ihildren. Promotional and educational materials were supplemented with comprlhensive and coordinated employer and insurer communications through e-mail, vJice-mail, employee letters, and newsletters explaining and promoting the service as pi of the benefit plan being offered to them.

PAGE 28

Physicians in the community were provided information about the TBDS service as one component of a comprehJsive and coordinated community-wide communications plan developed and impllmented by executive leaders from the . 'd 1 d I . F' h part1c1patmg proVI er, emp oyer, an msurer orgaruzat10ns. 1ve to SIX mont s pnor to the intervention, the governing boards ls well as the various standing committees of the two participating IP As were provided with information about the community-wide demand management initiative and the 1DS service. Three months prior to the intervention, every physician in each ofthe IPAs (RCIPA and RIPA) received a letter from their respective elected exJcutives explaining the service, asking for their support and inviting them to participlte in a conununity physician "town meeting" planned for three weeks prior tore implementation of' the TBDS. This letter was accompanied by a roster of the physician's participating patients. Two months prior to the intervention, each of4e lP As as weB as the county medical society included an article about the commbruty-wide demand management initiative I and the TBDS service in their respective prcian newsletters. During this same time frame, a press conference describing the oor ectives of the initiative, the community aspect of the project, and the TBDS servic was held. A subsequent press release was distributed to local and national media. Three weeks prior to the start date for the TBDS service, all physicians were provided the opportunity to attend one of community physicians' "town 15 I

PAGE 29

meetings" jointly sponsored by the co, medical society and the two IP As. During these town meetings, infurmation rel demand management, implementation of the mos service, evaluation of the servir and employer perspectives were shared by a panel of physician, employer, insurer, and vendor executive leaders. Physicians attending the meetings were afforded the \oppor1unity to express their views and to have questions or concerns addressed. Tio weeks prior to the intervention, a physician advisory panel comprised of phtsician executive and clinical leaders the IP As and the county mekcal society were provided with the opportunity to "test" the TBDS service. Key Questions Important questions explored in this study included: 1 Wh h b I ab at are p ys1c1ans oplDlons a out patients wantmg more uuormat10n out health and medical issues? 2. What are physicians' opinions about patients wanting more involvement and taking a more active role in the health and mebical care decision-making process? 2a. Who should primarily make decisiols about treatment for minor and major health issues? 2b. What are physicians' opinions about patients taking a greater role in the self-management of chronic disease? 16 I

PAGE 30

3. What are the opinions of physicians about the impact they believe a more I informed, active consumer/patient will have on: the quality of the doctor-patient relationship, clinical efficiencies, heallh outcomes, patient satisfaction, health care utilization, health care costs, and +ce liability? 4. What are the opinions of physicians ibout the incorporation of telephone based decision support (TBDS) for into the delivery of health care services? 5. What are the opinions of physicians about the impact they believe the I incorporation of TBDS into the health care delivery process will have on: the I quality of the doctor-patient relationship, clinical efficiencies, health outcomes, patient satisfaction, health care utilizJion, health care costs, and malpractice liability? 6. Do physicians' opinions differ by age, gender, specialty, practice structure, visit volume, and reimbursement structurej 7. What are the primary reasons physicians believed that the health plans implemented TBDS services? 8. Which independent variables (age, gender, specialty, practice structure, visit volume, and% of patients capitated) lethe most significant predictors of physicians' opinions?

PAGE 31

Reconciling Quantitative and Qualitative alysis The final step integrated the results of both quantitative and qualitative analyses to summarize physicians' opiniols about these key issues. Methods triangulation is often accompanied by the risk of divergent quantitative and qualitative data. The subsequent challenge is developing a single, well-integrated analysis and interpretation. For this study, steps wer1 taken to minimize this risk. The questions for the Physician Opinion Interviews (qualitative data) were derived directly from the Physician Opinion Survey (quantitative Ja). That is, the interview questions obtained physicians' opinions on the issues as the survey, but were formatted to derive more in-depth, information-rich understanding of that issue through the interview process. Nonetheless, where diJcordance between quantitative and qualitative data occurred, the quantitative were used as foundation for interpretation wbile the qualitative data Jvided additional explanatory power. Organization of le Dissertation This chapter introduces the signifiL of the study, states the specific aims, provides an overview of the research and outlines the organization of what is I to follow. Chapter 2 presents a review of iterature pertinent to health care utilization, demand-side strategies for improving health care, physician and patient 18 \

PAGE 32

perspectives about more informed and aiively involved patients as partners in health and medical issues, and a review of factors that influence the diffusion of new innovations in health care. Chapters 314 are dedicated, respectively, to quantitative and qualitative methods, analyses, and results. Chapter 5 presents a discussion of the integrated quantitative qualitative results including limitations, implications, recommendations for futuJ research, and conclusions. 19 I

PAGE 33

CJIMlTER2 I REVIEW Oi LITERATURE A review of literature pertinent t0 this study is presented in this chapter. The following topics are discussed: (a) health \care utilization, (b) demand-side strategies for improving health care, (c) physician r patient perspectives about more informed and actively involved patients as partners r health and medical issues, and (d) a review of factors that influence the diffusion of new innovations. Geographical Variations and Appropriate Use ofHealth Care In the current debate on health i reform, three very salient and inextricably linked questions central to health policy and strategy have been posited: (a) "Will we I need to ration effective care?'' ( Brook & Lohr, 1986); (b)" Do the geographical variations in the use of services have any jlinical significance?" (Brook, Lohr, Chassin, Kosecoff; Fink & Solomon, 198t; and c) "Is it prudent to focus exclusively on how to manage the supply of medical services, or does a new concept demand management -deseJVe serious considefati,n as well?" (Vickery & Lynch, 1995). Defining, measuring and classifying appropriateness has been recently delineated by Brook, Kamberg, Mayer-oles, Beers, et al. (1990): 20

PAGE 34

A service is provided when its benefits (expected positive exceed its costs (expected negative .... Based on literature review, we havb delineated three approaches by which to judge whethJr a service is appropriate. .. Benefit-risk approach .. .I a service is judged to be appropriate only if the heflth benefits to the patient exceed the negative health consequences to the patient by a wide enough margin so that procedure is worth doing .... Benefit-cost approach .. l extends the benefit-risk approach by considering the resourbe cost of providing the service .... Implicit approach ... for judging appropriateness may have been used, but clear, explicit definitions are specified .... Each of these approaches ... can be used to classifY a service as being misused, or overused. The underuse of a procedUr-e, office visit, or hospital stay is the failure to use these when they are indicated .... Misuse implies that are incorrectly used . . Overuse is the use of a office visit, or hospital Admission that is not indiJated (p 3). What have we learned from healtJ services research about the appropriateness I of medical care utilization? Vickery and Fries (1994) have pointed out that up to 700/o of all physician visits for new problems jy be medically unnecessary. Referencing a significant and growing body of evidence medical care utilization research, nationally prominent health services scientists, Brook and Lobr, estimated that, ... perhaps one-third of the resources devoted to health care today are being spent on ineffective or unproductive care" (1986, p. 6). After conducting a literature review regarding the appropriateness of acute medical care for the elderly, Brook et a1 (1990) stated that, "Vutually every included in this review fuund at least double-digit levels of inappropriate care.]. erhaps as much as one-fifth to one-quarter 21

PAGE 35

of acute hospital services or procedures felt to be used fur equivocal or inappropriate reasons and two-fifths to on.e-half of the medications studied were overused in outpatients" (p. 225). In addihon to these studies of overuse, they noted that the few studies that examined undenlse or misuse of services also documented the of these phenomena. Furthermore, they noted that for the services examined in these appropriateness studies, very little is knoi about the effects of such utilization on health outcomes. This can be viewed pessimistically as an indictment of the current U.S. health system or optimistically as I a window of opportunity. Even while awaiting more data on appropriateness, mhch work needs to be done to decrease the extent in the U.S. health care system. Depending on one's beliefs, that work could take the form ... at the clulicallevel ofbetter education or dissemination ofinformatit to patients ... (1990, p. 240). Eddy (1990, 1991, 1993) has tended to support this perspective by pointing out that resolving our societal conflict with health Le costs will require connecting value to cost and addressing unchecked demand 1 volume and intensity of services by involving patients as partners with physicians in clinical decision making. Vickery and Lynch (1995) have p+sed one such opportunity fur addressing volume and intensity of service utilization through the concept of "demand management." They have defined demand Lanagement as, "The use of decision and self-management support systems to enablj and encourage consumers to make 22

PAGE 36

appropriate use of medical care" (p. 4-5). Vickery and Lynch (1995) pointed out that the supposition behind demand manag4ent is that health care decisions are influenced by a complex set of psychological, soj and cultural variables, most of which are unrelated to morbidity (nature and of the illness). Similar to Brook & et al, (1990), I they have defined appropriate decision making as that which optimizes benefits rhile minimizing risk. However, Vickery and Lynch ( 1995) have provided an intriguing, expanded, and intuitively sound psychosocial-cultural dimension to apprJpriateness. They have indicated that two components of demand, perceived need ld patient preference, are central to demand management and hold considerable for enhancing more appropriate utilization of health and medical care ,ces. For example, in regards to perceived need, they have pointed to research by social scientists such as Berkanovic ( 1981) who has estimated that 42% of the variance in decisions to seek health care can be accounted for by psychosocial variables ( ersonal beliefs, social networks etc.) with morbidity explaining only 12%. Recent evidence has indicated tha an individual's access to information and resources is an important variable in healt care decisions and subsequent health care utilization. That is, a well-informed pati,t will make different decisions than a less informed patient. Since several of the moJ popular contemporary telephone-based decision support services generally includ pre-provider triage (self-care and perceived 23

PAGE 37

need), shared decisions for major medi issues (patient preference), and chronic disease management; the empirical basis for each of these decision-support components will be reviewed. "Demand-side Interventions This section will review the literatbe pertinent to various types of "demand side" interventions. Particular highlight 111 be given to the impact of these interventions on health outcomes, patient satisfaction, health care utilization, health care costs, cost-benefit, and clinical efficielcies. Perceived Need and Self-care Vickery and Lynch (1995) have hypothesized that self-care interventions directly influence an individual's perceptioJ of the need for health care by addressing the individual's capabilities to assess and m\anage medical problems outside of the formal medical care system. As noted by Jickery and Lynch (1995), there is a considerable amount of evidence supportink the effectiveness of self-care interventions in decreasing demand for medical care. Vickery, Kalmer, Lowry et al. (1983) conducted a prospective, randomized, controlled trial of self-care educational intelentions within a health maintenance organization to determine the effect of se}f.lare written communication materials on 24

PAGE 38

ambulatory health care utilization. Statistically significant decreases of 17% and 35% were found fur total medical visits and tor illness visits, respectively. They also estimated that these decreases in health ea.re utilization resulted in a savings of$2.50 to $3.50 for each dollar invested in the interventions. Lorig, Kraines, Brown and RicJ.dson (1985) conducted a study on the impact of a workplace health education progrJ on reducing medically unnecessary outpatient visits. The purpose of the was to evaluate a health education program's ability to help employees makl more informed decisions about when to seek professional medical care at the physicil' office or the emergency room. The main finding of this study was that the self-helJ intervention decreased self-reported physician visits (17% or 2.0 visits/househlld/year) in first-dollar-fee-for-service benefit plans. However, in contrast to the et al. (1983) study, no reduction was found among HMO members. The reselchers also examined whether or not these decreases in visits resUlted in poor medici\ outcomes. They tested this indirectly by examining increased hospitalization and o tpatient visit rates during treatment periods I of the study. They reported that, "Neither of these results was observed although the power for detecting such changes in this .Ldy is not great ... In future investigations of self-care educational programs, more alention must be given to the long-term health benefits ... (Lorig et al. 1985, p. 1053 ). 25

PAGE 39

Utilizing a case-control pre-post iudy design, Levine and Britten (1973) studied the impact of self-administering blood products in 45 hemophiliac patients selected from a cohort of patients with 1genital coagulation disorders routinely followed through the New England Medic::al Center in Boston. The intervention consisted of theoretical instruction for palients and family members on hemophilia treatment and complications, instruction t sterile technique and venipuncture as well as suggested dosages for minor and hemorrhage, and two to four supervised infusions. The self-therapy intervention resulted in a reduction of 74%, 89%, 76%, and 45% for absenteeism, hospital days, otpatient visits and health care costs, respectively, for the 45 patients. J Most intriguing were the subjecti l evaluations completed by study participants. Levine and Britten ( 1973) r,orted that, "All4 5 patients indicated their complete satisfaction and felt that home therapy had greatly improved their lives. I None wanted to change the program .... Numerous patients volunteered the information that they were enjoying their rllationships with the physician and hospital staff for the first time. Several attributed j\. s to their own participation in health care and their decreased dependency needs" (pp. 198-199). This patient feedback tended to suggest that overall health status and qukty of life improved significantly as a result of patient participation in self care. HoweJer, except for the limited morbidity,

PAGE 40

mortality, and this informal patient satisfaction survey information, no formal evaluations of patient-assessed function1 status or health outcomes were conducted. Chronic Disease Management Greenfield, Kaplan, Ware, Yano, and Frank (1988) studied the effects on blood sugar control and quality of life in diabet cs who participated in medical decision I making. The intervention consisted of a 20-minute coaching session while the patients were waiting to see their physicians. Thj components of this session included: a) following an algorithm to review with thJ patient the last recorded medical visit, b) I assisting the patient in identifying medii decisions most likely to arise during the visit, c) encouraging the patient to focus en treatment decisions that could be affected by lifestyle and/or preferences, d) rehearsJng negotiation skills, e) encouraging patients to ask questions of their physician, f) revi1wing obstacles that can impede negotiations with their physician, and g) rehearsing h1 to overcome these obstacles. Glycosylated hemoglobin (HbAI c)l values for the experimental group decreased from 10.6 +/-2.1% at pre-intervention to 1 +/-1.9% (p<0.01) after the intervention. There was no change in (HbAic) values for the control group from pre-intervention (10.3 +/2.00/o) to post-intervention (10.6 +/-2.22%). These researchers reported that the experimental patients were approximately twice as effective as controls during post-intervention in the elicitation of infolation from physicians, reported 27

PAGE 41

significantly fewer functional limitations : obility, role, physical, lost work days, self care), and assessed their health (overall perception of health, health concern, number I of current health problems) more favorably. Mayo, Richman, and Harris ( detennined the effect of an outpatient program designed to reduce re-admissions and hospital days for asthmatic exacerbations among adults with asthma. The intervention was classified as an intensive outpatient treatment program wr.ch consisted of a vigorous medical regimen combined with an educational program wruch emphasized teaching patients aggressive self-management strategies in case of asthma exacerbation. The patient I education program did not involve any written, lecture, or audiovisual materials. Patient education was completed during rve contacts between the patient and physician. Hospital readmission rates and hospital day rates per patient per month I were reduced by 70% and 64%, In commenting on the fact that wrihen materials, lectures, and support groups I were not used in this study, Mayo et al. (1190) reported that," ... it rapidly became evident that the face-to-face interaction of lhe medical care provider and patient I resulted in the most meaningful transfer of information . .. Preliminary results suggest that a nurse practitioner is as effjve as a physician in reducing readmission rates among difficult adult asthmatics" (pp. 870-871). 28

PAGE 42

Lorig, Mazonson, and Holman ( 993) studied the effects of a four-year Arthritis Self-Management Program (ASMP) on pain, physical disability, physician I visits, and cost-benefit in osteoarthritis (IDA) and rheumatoid arthritis (RA) patients. Studies were also conducted to ascertain how the ASMP achieved the measured effects. Results indicated that change scor from baseline through four years fur the two ASMP groups were persistent and statistically significant (p
PAGE 43

versus elective surgery were more risk -a ersive than physicians. When BPH patients (n=284) at Kaiser Pennanente HMO in denver viewed an interactive video describing the benefits and risks of watchful waiting versus elective surgery for BPH, all but 30 chose watchful waiting over surgery. Tjs patient preference for watchful waiting over elective surgery approximated a 44% reduction in the rate of prostate surgery I among HMO members in the lint year. jimilarly, surgical rates initially plummeted 60% at Group Health ofPuget Sound jo in Seattle. Lennan, Brody, Caputo, Smith, Lazaro, and Wolfson (1990) constructed a Perceptions of Involvement in Care Scale (PICS), a self-report questionnaire that assessed patients' perceptions of doctor and patient behaviors that occur during a routine medical visit. Specifically, this mslment was designed to ascertain patient perceptions about doctor filcilitation of involvement, level of information exchange, and patient participation in decision making. Subjects used in the development of this instrument were pty HMO patients attending a primary care office at an academic medical center. The! subjects were predominantly female (62%) and black (800/o) with a mean age of39 years. A major finding of this study was tLt physician facilitation of patient 1 d ctual . I I d . akin "gnifi 1 mvo vement an a patient mvo vement m ectston m g were st cant y related to patients' satisfaction with care.ldditionally, physician facilitation of patient involvement and infonnation exchange wet consistently and significantly associated 30

PAGE 44

with patients' perceptions of post-visit c anges in their understanding, reassurance, perceived control over their illness, and lxpectations for functional improvement. Also, patients who reported that they wJe more involved in decision making with the doctor were more satisfied with their phJsicians' technical competence. These findings Ieinfurced previous research (WLa., hrui, Baniatua, Carter & Lippincott, 1984) which demonstrated tll important role physician facilitative behavior can and should play in producin improved therapeutic benefits for patients. An intriguing study by Brody, Miher, Lerman, Smith, and Caputo (1989) examined the relationships between patiets' perceptions about the roles they played during medical visits and their subsequenJ attitudes about their illnesses, treatments, I and self-rated improvements. This study involved 117 adult patients from a general internal medicine faculty practice which adult primary care to a predominantly HMO population. The meL age of the patients was 32 years; 62% were female, 85% were minorities, 92% HMO insurance, and a presented with symptoms suggestive of relatively minor illnesses (sore throat, vaginal discharge, musculoskeletal pain) at the Je of the study. Questionnaires assessing the roles batients wanted to play during the medical visit, the role they later perceived they plated, illness outcomes, patient attitudes about their illnesses, and patient satisfaction wi, their physician were administered befure the medical visit and then again one day and one week after the visit. Their physicians 31

PAGE 45

also completed a questionnaire which provided infonnation about the seriousness of the problem and expected alleviation of atient discomfort and dysfunction one week later. Forty-seven percent of the patijs reported playing an active role wbile 53% of the patients reported playing a passiv, role. One week following the medical visit, active patients reported statistically (p = 0.04) less discomfort, greater all f d I h. a1 edical di. evtatton o symptoms, an more 1mprorement m t err gener m con t10n than did passive patients. One day post-VIsit, active patients reported less concern with their illnesses (p = 0.04), a greater sense if control of the illness (p = 0.04), and more satisfaction (p = 0.02) with their physi,s than did passive patients. The researchers concluded that their findings were consistent with others who have demonstrated the impact that increasing a patient's participltion in care can have on improved disease controL higher quality oflife, and a more latisfactory doctor-patient relationship. They speculated that patients' increased slnse of control and self-efficacy may be possible explanations for their findings. Patient Preference for Information and Involvement in Decision Making The traditional doctor-patient relatonship has been characterized by a Parsonian, paternalistic approach the patient has played a predominantly passive, dependent role and the physician r asswned an autonomous, professionally 32

PAGE 46

dominant role in the health and medical care decision-making process. However, a variety of social, political, and cultural +s in combination with technological advances in communication and management point toward a changing doctor -patient relationship characterized r increasing consumer/patient participation in decisions affecting both health and medical care services. Although the previously referenf studies have provided evidence that the inclusion of patients in the health and medical care decision-making process can result in more appropriate use of health care returces, decreased health care costs, increased patient satisfaction, improved control of disease, and better clinical and functional outcomes; it is also important tl address the fundamental questions related to the consumer/patient's role in clinical Jecision-making. That is, to what extent do consumers/patients prefer to be more infoLed and what type of information is desired I about health and medical care decisions? fo what extent do consumers and patients want to be more actively involved in what type of health and medical care decisions once they have been afforded specific infoFation? And to what extent should physicians encourage active and collaborahve patient participation in health and medical care decisions? Cassileth, Zupkis, Sutton-Smith, ,d March ( 1980) examined the degree to which 256 cancer patients desired to be more informed and to participate in decisions regarding their care. Using an lnformatioJ Styles Questionnaire, patient preferences 33

PAGE 47

regarding involvement in medical decisiTs were measured using two alternatives one indicating a preference fur their Jr to make decisions about their medical care, the other indicating a preference to participate in decisions about their medical care. Similarly, preferences for quantity and q1ity of information regarding diagnosis and treatment were ascertained. l The most notable findings from t s study were: 1. Patients who desired detailed Jrormation about their treatment in I comparison to those who didn't seek sue! information were younger, white, better educated, and more recently diagnosed. 2. Those patients who preferred to participate in treatment decisions were younger and better educated compared tj those who prefurred to leave treatment decisions up to their doctor. 3. Even though age was the only factor that distinguished those patients who wanted more infOrmation and active involr in their own care from those who preferred a minimum of information and active involvement, a majority of patients in I each age group preferred active participation and detailed information. 4. There was a strong correlation Jetween preferences for maximum =-c. d h d I d uuormat1on an t e esrre to partiCipate m ec1s1ons. 34

PAGE 48

5. Because not all patients desired an active involvement in the medical decision-making process, physicians sholld be attentive to patient preferences for such roles. Stroll, Lo, and Charles (1984) examined to what extent 210 hypertensive patients from a conununity hospital c, an HMO, and a Veterans Administration outpatient clinic preferred more infonnJ' on about hypertension and preferred to participate in medical decisions. Cone rntly, the patients' physicians were asked to estimate their patients' preferences for more infonnation about hypertension and preference to participate in medical decisibns. The researchers also studied whether or not patient demographic and other health lcare related fucto111 affucted patient preferences. Similar to other studies, sJn et a!. utilized five different preference alternatives encompassing "active", "conaLrative", and "passive" decision making styles to ascertain how decisions are made and patient preference for how decisions should be made. 1 The major finding of this study w that clinicians underestimate patients' desire for more infonnation and discussioL about therapy but overestimate patients' desires to participate in decisions. That is\ 41% of the patients preferred more infonnation about their illness, but clinicians underestimated patient desires for discussion about therapy in 29% of cases; and while 53% of patients preferred to 35

PAGE 49

participate in treatment decisions, clinicians estimated that their patients desired to participate in decision making in 78% ofle cases. In terms of how decisions are us1y made, 63% of patients that the clinician usually makes the decision whill clinicians reported such decision making in only 20"/o of the cases. Additionally, 80't of clinicians reported that the patient participates to any extent in decisions while only 3 7% of patients reported that they d akin I participate m eas1on m g to any extent. In terms of how decisions should L made, 47% of the patients preferred that the clinician make the decisiols without patient participation. Approximately one-third of the patients Jmcated that the physician should make the decision while strongly considering the palient's opinion. Less than one quarter (19%) of the patients indicated a preference to Le equally with the physician in making decisions and 3% indicated a preference ft making decisions themselves. In contrast, physicians believed that patients wanted ti participate in decision making in 78% of the cases. In only 22% of the cases did the physician believe the patient wanted the clinician to make the decision without the batient. Because of possible physicianpatient discordance about desires fur mJ1 information and preferences fur medical decision-making, and, given the fact that enhanced patient participation may result in improved outcomes, this study acknowledied that physicians may want to find explicit 36

PAGE 50

ways to discuss these issues with patient in order to facilitate communication and decision making. Using a study population of312 ,atients randomly selected from a hospital based primary care clinic, Ende, Kazis, ASh, and Moskowitz (1989) developed the informed and their desire to make medic decisions. The API consisted of two scales --one to assess information seeking and one to assess decision making. For each of the two scales, totals were linearly adjustjd to range from 0 to I 00, where 0 corresponded to complete lack of desire rl. information seeking or for decision making and 100 corresponded to the strojgest possible desire for each of these dimensions. On the 0 to 100 scales for information seeking and preference for decision making, the mean scores were 79.5 and 33.2, respectively. There were no significant correlations between patients' desires for iDformation and their desires for decision making or with patient satisfaction. with ather studies, they found stronger preferences for active involvement in deciston making in younger, more educated, higher income individuals, and less preferelce for involvement in decision making among the more severely ill. Interestingly, \after adjusting for sociodemographic fuctors and severity of illness in a regressiol only 19% of the variance in decision making preferences could be explled. The authors concluded that patients 37

PAGE 51

had a strong desire to be well informed, ut, in support of other studies, relatively I weak preferences for participating in medical decisions. The findings of this study tended fo support the view that patients prefer a doctor-patient relationship whereby the doctor keeps the patient informed and ed d . akin bl d th fun engag m ects1on-m g processes, ut oes not reqwre e patient to ctlon as the principal decision maker. The pointed out that a great deal of the individual preference among patients for information and decision-making roles may be more related to ingrained specific culJral and ethical variables rather than the sociodemographic factors measured in thil study. To this point, they, like Stroll et al. (1984}, acknowledgld that physicians should have explicit discussions with individual patients regar g information and decision-making preferences. Sutherland, Llewellyn-Thomas, Lockwood, Tritchler, and Till (1989) used a convenience sample of 52 patients (35 woLen and 17 men) with a mean age of 48.5 years who were undergoing treatment for lancer (radiation therapy or chemotherapy) to study their desire for intbrmation and +erences for participation in treatment decisions. Preferences for intbrmation wT assessed with two instruments the Health Opinion Survey (HOS) and the Information Seeking Questionnaire (ISQ). P c. c. . . dl.. ed . re1erence 10r participatiOn m treatment cisJons was assess usmg a quest10nnarre adapted from Stroll et al. (1984). 38

PAGE 52

The results showed that although most patients indicated that they actively sought information, 63% of the them felt the physician should take the primary responsibility in decision making, 27% fit decision making should be an equally shared process, and 10% felt they should a major role. The authors pointed out that this discrepancy between patient preferences for information compared to their preference fur a limited role in decision .L may be best explained by the patient's desire ... to satisfy an aspect of 'psychblogical autonomy' that does not necessarily include participation in decision making'' 262). Like others, they suggested that I clinicians should be sensitive to situation-dependent preferences for information and d akin h f di d a1 I ec1s1on-m g on t e part o m VI u Another study assessed patient desire for information and involvement in decision making using the Autonomy Preflrence Index (API) and the Health Opinion Survey (HOS). The subjects were 167 unilersity-based primary care outpatient I patients with benign prostatic hyperplasia, back pain, or mild hypertension (Nease and Brooks, I 995). Assessments from both insLents indicated that patients' desire fur information exceeded that for involvement in decision making. Consistent with the findings by Ende et al. (1989), the API decision-making score in this study decreased as the seriousness of the clinical condition tcreased. That is, as the seriousness of the hypothetical clinical condition increased, patients' desire for involvement in decision making decreased. 39

PAGE 53

Consistent with previous studies, rhis research pointed out that there is substantial variation among patients in th ir desire for information and for their involvement in medical decisions. That is,l although sociodemographic factors (younger age, female gender, higher eduJI lion) may be associated with greater desires for information and decision making, a 1 ;ge part of the variance for patient responses to these two aspects of medical care remals unexplained. Similar to others (En de et al. 1989; Sutherland et al. 1989; Strull et 1. 1984; Cassileth et al. 1980), these researchers pointed out that even though desire for information is high, the substantial variation among individual patilnts for involvement in decision making should serve as a caution against rigid polibes and recommendations about the degree to which patients should be involved in mtcal decisions. Beisecker and Beisecker (1990) studied 106 rehabilitation medicine patients (42 men and 64 women ranging in age frol17-85 years) to determine information seeking behaviors and decision-making prJerences using the Desire for Medical I Information and the Locus of Authority in Medical Decision-making Scales. The results indicated that patients desired infoiation about a wide range of medical topics, but when communicating with their they did not engage in many information-seeking behaviors. FurthermJe, patients believed that medical decision making authority should rest primarily with the physician rather than with the patient. Finally, it was noted that patient information-seeking behaviors were more closely 40

PAGE 54

associated with situational variables such as length of interaction, diagnosis, and reason for the visit than with patient atti des or sociodemographic characteristics. The researchers surmised that inflrmation-seeking behavior was attenuated in this study because the subjects were primkruy patients with chronic diseases and disabilities. Like others, they also emphjized that situational factors (diagnosis, presence ofa companion, reason for visit surrounding the doctor -patient relationship may better explain patient infirmation seeking behaviors and desires to be involved in medical decision-making than sociodemographic variables. They suggested that future research examine thl impact of situational variables on information seeking behaviors and preferebce for involvement in medical decision-making. Bradley, Zia, and Hamilton ( 1996) evaluated patient preferences about involvement in medical decision making and factors that might influence these I preferences in a predominantly white female population. They used a survey instrument that included I 0 specific case depicting different acute and chronic medical situations that varied in severity. For each scenario, patients were asked to select one of seven responses varied levels of physician/patient decision making. These responses ranged lorn very passive ("I prefer that my doctor tell me what to do") to collaborative ("I that my doctor and I make the decision together") to very active ("I prefer that I 1e the decision without any information or 41

PAGE 55

recommendation from the doctor") prefe ences for involvement in decision making by the patient. The survey responses in this stud 1 indicated that patients filvored the physician as decision maker. However, the prefer,ces for involvement in decision making varied by scenario. The authors acknowledged that patient decision making preferences were scenario specific and p4bably illness specific. Patients wanted less control in the hypertension scenario and Jore control in the cancer scenario. Consistent with other studies ( Cassileth 1 al. 1980; Strull et al .. 1984; Ende et al., 1989), increased age and lower educational levels were related to increased emphasis on the physician as primary decision mak:J whereas lower age and increased educational levels were related to increase6 patient control. Similar to others (Nease and Brooks, 1995; Ende et al. 1989; Suthjrland et al. 1989; Stroll, 1984; Cassileth et I al. 1980), this study suggested that: a) preference for patient involvement in decision-making depends upon the situation and b) bhysicians must Jearn to be both facilitators of decisions as well as decision makers upon the patient and the situation. Hack, Degner, and Dyck (1994) eJammed the quantitative and qualitative relationships between preferences for medical information and preferences for involvement in making treatment decisions in 35 women (aged 32-83 years) with stage I and II breast cancer who were 2-6 months post-diagnosis. Using a card sort technique developed by Degner and Sloan 1992), they explored patient preferences 42

PAGE 56

for specific medical infonnation and preferences for involvement in making treatment decisions. This card sort technique pro1ded patients with five different degrees of preferred involvement in medical decisiols ranging from active ("I prefer to make the final selection about which treannent I +ve") to collaborative ("! prefer that my doctor and I share responsibility for which treatment is best for me") to passive ("I prefer to leave all decisions my treatment to my doctor''). They also developed a card sort measure to delennine patient preferences for diagnostic, prognostic, and treatment-related infonnltion desired by the patient. The information preference card sort procedure addressed the specific manner by which patients want physicians to provide them with information (quantity, media, type of descriptive terminology to be used). Overall, the findings suggested that patients who desire an active role in choosing their treatment also desired morl detailed infonnation about their medical care. That is, those patients who preferre1 a more active preference for involvement in decision making also preferred to receit more detailed infonnation regarding their care while those patients who preferred a more passive style preferred either a brief explanation or did not want any infonnatiln at all. However, consistent with other studies regarding patient variability (Ende L al. 1989; Sutherland et al. 1989; Strull et al. 1984; Cassileth et al. 1980), Hack et aljpointed out that there were nine passive patients who wanted detailed information egarding their treatment. 4B

PAGE 57

Second, all of the patients who preferred active involvement in decision making and most of the passive patients wanted tir physicians to provide them with all treatment alternatives and the degree of rii associated with each alternative. Active and collaborative style patients wanted complete information about all major alternatives while the passive patients pjerred to hear about the best single treatment or hear nothing at all. Third, education was the only sociodemographic variable significantly related to decision-making role preference. ThoJ patients who preferred an active decisionmaking role were more likely to have gradLted from high school whereas the nonhigh school graduates preferred a more pre role. In contrast to other studies, age (Cassileth et al. 1980; Strull et al. I984; Ende et al., 1989), and illness severity (Cassileth et al. I 980; Ende et al. I 989; Bjadley et al. I 989), were not significantly related to decision-making role prererencel. Fourth, the qualitative analyses disJiosed that: a) patients actively engaged in decision making wished to play such a rol, because it affords them a greater degree of control over "body, health, and life;" b) the more passive patients wanted to be kept infonned, but at the same time it was impolm "to place faith and trust in their physicians because only physicians possess the medical knowledge and expertise necessary to formulate treatment decisions'[ (p. 286); c) active patients were more like to believe their physicians wanted them to De more active in medical decisions; d) 44

PAGE 58

being pressured to play a more active role in decision-making processes was more anxiety-producing for the passive patien s; and e) both active and passive patients style and circumstances. To this point, Hack et al. noted t t several patients emphasized that physicians needed to be sensitive to and respect the information and decision-making preferences of each individual patient. Preferences for information and decision-making may differ depending on milestones related to sevedty of illness, elapsed time since diagnosis, and treatment progression. Similar to other Ldies, the younger and better educated patients expressed a greater desire for ilvement. Using the same card sort procedures, Davison, Degner, and Morgan (1995) assessed information and decision makinJ preferences of 57 men (mean age 71 years) with prostate cancer who were 0-6 montJs post-diagnosis. The majority of the men in this study were married (74%) and had lels than a high school education (54%). Twenty-nine (51%) had received no +t. Consistent with another recent study by Degner and Sloan (1992), a majority of men (58%) in this study preferred a ,ve treatment decision-making role, while 23% and 19% preferred a collaborative anCI an active role, respectively. However a high proportion of men wanted more inflation regarding likelihood of cure, types of treatment available, and disease progressiol. 45

PAGE 59

Speculations as to why the majori of men preferred to delegate treatment decisions to their doctor included: 1) old r married male patients tend to involve their wives in medical care decisions and subse uently tend to relegate to them a decision-making role; 2) older patients may be less likely to believe they should challenge the authoritarian role of the physician; and 3) older patients may actually prefer a more authoritarian model in the doctor-patient because they have a high level of trust in their personal physician; 4) there ry be a fear of assuming responsibility for decisions that may lead to poor outcomes;\ and 5) they may believe they lack the knowledge and information necessary to make a good decision. As has been true of most of the stuhies reviewed in this section, Davison and colleagues (1995) point out that informatiJn seeking and decision-making preferences are determined by a variety of sociodemo+pbic, intra-personal, and situational factors. Thus, clinicians should be explicitly sensitive to individual differences in regard to needs for information and involvJment in medical decisions. Davison and associates (1995) also sounded an approje call fur the critical and urgent need for research in determining the role thlt wives .and the family unit play in health and medical care decisions. Beaver, Luker, Owens, Leinster, Degner, and Sloan (1996) also used the Degner and Sloan (1992) card sort procedu\e to examine the degree of control desired in treatment decision making in 150 women (mean age 54.8 years) newly diagnosed 46

PAGE 60

(2.5 weeks) with breast cancer. Their r aim was to ascertain whether these women preferred to play an active (patient-controlled), collaborative Gointly\ controlled between patient and or passive (provider-controlled) role in decisions regarding breast cancer treatment. For a majority (66%) of the newly diagnosed women, lumpectomy was theJ.\lanned treatment with mastectomy the planned treatment for the remaining 34 o. Two hundred women with benign breast disease served as a comparison group. I The most significant finding of thil study was that a majority of the newly diagnosed women preferred a passive roll, leaving treatment decisions to their doctor; whereas the women with benign breast dilease (in the comparison group) preferred a I more collaborative role. For the newly diagnosed group, the variables found to be weakly correlated with decision-making pteference were age, with older women preferring a more passive role; education, lower levels of education associated with preference for a more passive role; arld social class, with lower social classes preferring a more passive role. However, regression analysis, age and level of education were not predictive of decision laking preferences. Even though social class was predictive of decision making pr1\ferences (p = 0.04), it's influence was relatively weak accounting for only 2% of he variance. In the benign group, age and social class were weakly correlated with prlferences for decision making. A,family 47

PAGE 61

history of breast cancer and marital status were not associated with decision making preferences in either group. Beaver and colleagues speculated about the explanations for these findings. Using supportive information from other cancer studies and preferences for decision making, it was argued that potentially life threatening illnesses may make patients more passive. They also noted that even tough a majority of the women with newly diagnosed breast cancer preferred a more passive role in decision making, there were still many who wanted some fum of involrent. They cite that a lack of appropriate information could also be one of the reasons that women felt uncomfortable in playing an active role in decision making. Similar to previous research cited in this review, they emphasized once again that the most prudent clinical approach is )or the clinician to be sensitive to and take into account each individual patient's prefJences for decision making involvement. They reiterated that there seems to be no sLple predictive formula for determining I patient's preferences for decision-making. o assist clinicians with such assessments, they suggested that the card sort procedure could easily be integrated as a screening tool into the everyday practice of medicine. The authors indicated that future studies were underway to determine the effect of"tune since diagnosis" on decision-making preferences. 48

PAGE 62

A particularly interesting and noteworthy study was conducted by Thompson, Pitts, and Schwankovsky ( 1993) the situational and demographic influences I on the preference for involvement in medical decision-making. Participants for the study were 459 randomly selected memt of an HMO wbo were predominantly white (87%), female (68%), low to middle income (49%), married (65%), had some college (63%), with a mean age of 54 rel.. To assess patient preference for decision making they used the Ende et al.(1989) ktonomy Preference Index (API) and the I Desire for Involvement Questionnaire (DIQ) they developed specifically for this study. Thompson et al. (1993) the significant difference between the two instruments. The API is constructed to ask about patient preference in regard to medical situations that generally fall wiJ the province of a medical professional. The DIQ is designed to ascertain desire fJ input into situations where medical expertise is not required. Four vignettes +varying severity (knee injury, high blood pressure, cancerous growth, and stress heddaches) each requiring two treatment I choices with varying side effects, trade-off! between quality and length of life, and effects on patient's lifestyle were used to preferences. As hypothesized, one situational faLor was found to be strongly related to patient desire for involvement. Respondentl expressed a greater desire for involvement in decisions that did not require medical 4ertise (DIQ) and significantly lower preferences for involvement in decisions that required medical expertise (API). With 49

PAGE 63

this finding, Thompson and colleagues suggested that the conclusions reached in other studies (Strull et al. 1984; Ende et al., 19k9; Sutherland et al. 1989; Beisecker and Beisecker, 1990) about patients not wantLg to be involved in medical decisions are misleading. Additionally, in an analysis to see if experience with a condition impacted the desire for involvement, they found that thLe who had hypertension were less likely to I want decisional involvement than those who did not have hypertension. They speculated that experience with a chronic bondition and all of its associated difficulties may lessen the desire for resplbility. From a clinical practice viewpoint, Thompson and associates (1993) pointed out the documented advantages ofhaving patients participate in'medical decisions while at the same time conceding that clinibians need to recognize that there are individual patient differences in decision-mlg preferences. They urge clinicians to be sensitive to these differences and subseJuently to the types of decisions for which patient involvement is most likely to be and beneficial. In summary, it is clear that patients and desire information about the diagnostic, prognostic, and therapeutic aspects of medical decisions. What isn't so clear is to what extent patients want to be rvely involved in medical decisions. The literature reviewed here has presented conflicting findings. The evidence has suggested that there are certain and situational factors which have 50

PAGE 64

been consistently lissociated with decisiormaking preferences: I) age, with younger individuals preferring a more active role; 12) education, with more highly educated individuals preferring a more active gender, with men preferring a more passive role than women; 4) marital stai with married men preferring a passive role; 5) prognosis, with more seriously ill patients preferring a more passive role. Two critical variables that need further studyJ \ elude time since diagnosis and the role that wives and the family unit play in health d medical care decisions. It is clear from this literature reviel that information seeking and decisionmaking preferences are determined by a Lety of sociodemographic, intrapersonal, and situational factors. The most prudent llinical approach is for the physician to be sensitive to and take into account each inkvidual patient's preferences for information and for involvement in decision-making. Finally, the intriguing work by Th,mpson et al. (1993) indicates that there may be some significant methodological issues isociated with how we are assessing patient preference for involvement in medical decision-making. That is, respondents in their study expressed a greater desire for iJvolvement in decisions that did not require medical expertise as measured by the D, fur Infurmation Questionnaire (DIQ) and significantly lower preferences for involvement in decisions that required medical expertise (API). Like the API, the Locus Jf Authority Scale utilized in the Beisecker and Beisecker (1990) study also required to answer many of the 51

PAGE 65

questions. What information should be covered in the patient's history? Whether there is a need for medication? Which drug to hse? What dosage to use? Whether to change medications? Whether to have the patienl undergo diagnostic tests? Whether to treat the patient in the hospital or on an basis? Thompson et al. 's ( 1993) opinion that misleading conclusions have been relched in other studies (Strull et al. 1984; Ende et al., 1989; Sutherland et al. 1989; rsecker and Beisecker, 1990) about patients not wanting to be involved in medical decisions warrants important consideration. Physician Perception about Patient Desire for Infonnation and Involvement There is a substantial body of evidLce regarding patient preference for information and medical decision making. However, there seems to be a coinciding scarcity of information regarding physician perception about how active patients should be in seeking information or in in health and medical care decisions. The purpose of this section is to review some of the more pertinent physician infonnation that may provide further insight on the matter. I As noted previously, Strull et al. (1984) examined to what extent 210 hypertensive patients from a community hobital clinic, a free-standing, HMO and a V Ad lini. I ed b eterans nurustratton outpattent c c preJ.err more uuormatlon a out hypertension and preferred to participate in medical decisions. Concurrently, they 52

PAGE 66

detennined how well clinicians estimated their patients' preferences about these aspects of medical care and whether or not patient demographic and other health care related fuctors affected patient In terms of how decisions are usually made, 63% of patients reported that the biinician usually makes the decision while clinicians reported such decision making l only 20'/o of the cases. Additionally, 80% of clinicians reported that the patient pjcipates to any extent in decisions while only 37% of patients reported that the patient in decision making to any extent. In tenns of how decisions should le made 47% of the patients preferred that the clinician make the therapeutic decisij without patient participation. Approx imately one-third of the patients indicated lat the physician should make the decision I while strongly considering the patient's opron. Less than one quarter (19%) of the patients indicated a preference to share eq11ally with the physician in making de-cisions and 3% indicated a preference for making themselves. In contrast, physicians believed that patients wanted to participate in decision making in 78% of the cases. In only 22% of the cases did physician believe the patient wanted the clinician to make the decision without the P,atient. Because there was relatively little demographic infonnation characterizing thl physicians in this study, vei)Ilittle else about physician perceptions regarding patilt information seeking and involvement in I decision-making can be extrapolated from this information. 53

PAGE 67

Doctor-Patient Communi tion & Relationships: Challenges/Barriers ahd Effect on Outcomes One of the very appropriate and rd concerns that physicians have when health plans begin to consider the imple,entation of24-hour telephone-based nurse counseling (TBDS) services is, "What r will this service have on the doctor patient relationship?" Preserving the physician-patient relationship in an era of managed care has become a central concJm for most physicians (Emanuel and Dubler, I 995). In studying the perceptions and olunons that physicians have about contemporary TBDS services, it's impera!ive that we have a fundamental understanding of the communication dynLcs, challenges, and success factors which comprise the doctor-patient relationship. Traditionally, the doctor-patient rrionship has been predominantly a paternalistic one in which the patient has assumed a very passive, dependent, and cooperative/acquiescent role while the phtcian has assumed a very active/directive, autonomous, and professionally dominant/controlling role. Recently, however, ethical, social, economic, and legal concert in combination with cultural shifts and advancements in communications and management technology have begun yet another era in the "social transformation of American medicine" (Starr, 1982). This era is advocating a more informed coJsumer of health and medical care; more patient .autonomy and self-determination in medical decisions; and a mutual, 54

PAGE 68

Collaborative partnership between physicians and patients in communicating with one another and in subsequently making h..k and medical care decisions. In their review of the literature ol doctor -patient coliiiiiWlication, Ong, DeHaes, Hoos, and Larnmes (1995;p. ,3) stated that, "ColiiiiiWlication can be seen as the main ingredient in medical care." In their treatise, they examined this issue in three domains: purposes of medical colunication, communicative behaviors, and the influence of communicative behavior\ on patient outcomes. This will serve as a useful framework for our review. Doctor-Patient Communications: Challenges/Barriers Ong et al. ( 1995) identified three different purposes of communication between doctor and patient: I) to develoi an inter-personal relationship, 2) to exchange information, and 3) to make diisions. Since perspectives on decision making have been previously reviewed in Ws report, attention will be focused information exchange processes. The development of interpersonal relations and the information exchange process bas included the actions giving, receiving, inteipreting, and evaluating verbal and non-verbal informati n by both patients and doctors. Physicians and patients seek and need information for obvious reasons. Physicians need different 55

PAGE 69

types of information for diagnostic, prognostic, and treatment planning purposes. As reviewed previously, patients have a stro g desire and preference to know as much about their medical situation as possible Ong et al. (1995; p. 904) have captured the patient's need for information by stating, "From the patient's point of view, two needs have to be met when visiting the doctor: the need to know and understand (to know what is the matter, where the pain comes from) and 'the need to feel known and understood' (to know the doctor accepts him and takes him seriously)." Speedling and Rose (1985; p. 116) have supported his perspective by stating that, "The I physicians' expertise in curing the illness is rarely seen to count more than his or her facility to care for the patient's psychosobal needs." Several factors have contributed tl documented discordance in information exchange between patients and physicians Observational research (Beclanan and 1984) has indicated that the physician interrupts the patient's beginning narrative after only 18 seconds; and the t complaint presented by the patient may not be the most critical one! It has been noted that 60-80% of medical diagnoses and I treatment decisions can be made without the need for expensive diagnostics just on the basis of a complete history and listening tj what the patient has to say (Hampton, Harrison, Mitchell, Prichard, and Seymour 1975; Sandler, 1980). Sir William Osler, the father of modem internal medicine, astUtely implored physicians to listen -for the I patient will reveal the diagnosis (Osler, 18 9). 56

PAGE 70

In support of the previous DiMatteo (1994) outlined fuur barriers or limitations to physician-pati,t communications: 1) physician use of medical jargon resulting in patient lack ofunderstanding and misunderstanding; 2) lack of active listening to patients by physii; 3) the desire the patient has for infonnation is usually not fulfilled by the provider; and 4) societal expectations and the I influence of medical tradition make it difficult for patients to effectively assert their needs for infonnation and participation 1 decision-making. Facilitating patients in telling their story about their illness expenence and then actively listening to what they have to say are critical success fu1 optimizing the doctor-patient relationship. As noted previously, patients alm1st always want as much infonnation as possible about their situation. However, physicians have tended to I underestimate this desire for information itrull et al. 1984), they may have subsequently provided insufficient infonnation to the patient to fully satisfy his/her needs. Ong et al. (1995), like DiMatteo (111994), have pointed out that another contributor to patient dissatisfaction with he exchange of infonnation is rooted in a lack of concordance between patient perclption and physician perception about the !l'!!ll of infonnation exchanged. often view medical infonnation as that which is objective (diagnostic test results, prognosis, type of disease, treatment procedures) whereas patients define medit infonnation in terms of its personal relevance to their illness experience (what IS the likelihood of cure? what is the effect Sil

PAGE 71

of treatment on my job? social life? farnilr? sexual activity? risk of disease to my family?). To this point, F rederikson ( has emphasized that a more effective strategy for physician consultation with is one which attends to the "concepts, perceptions, and views of the patient" raL than providing vast amounts of standardized information. Utilizing her mbdel of information exchange for medical consultations, she studied tasks in a self-selected sample of 3 5 general practice doctor-patient dyads. The consuhation evaluation lent was a questionnaire which asked the patient to rate the level ofintportance ( esr desirable, unimportant) for the following information exchange tasks: exploring patient ideas, dealing with patient concerns, exploring patient expectations diagnosis and treatment, discussing the effects of problems, explaining the etiology, explaining treatment, discussing side effects, and outlining a prognosis. The fobowing outcomes were also rated by the patients: satisfaction, patient understandink of health problems, doctor's response to h f I f :ffi ed, t e pattent, perceptions o appropnateness o treatment o er treatment mtenttons, development of the relationship between Joctor and patient, and change in concern and information provision. The information exchange tasks most strongly correlated with patient satisfaction were: dealing with patient conlems, exploring patient expectations, 58

PAGE 72

exploring ideas, and discussing elects of problems. Fredriksen (1995) noted that the quality of information provided y physicians to patients is more important than quantity and that physicians should arget information toward the specific individual needs of the patient. von Friederichs-Fitzwater, Flynn, and Williams (1991) analyzed 30 transcripts of conversations between p,cians and patients in fuur different settings (AIDs clinic, hospice program, family practice clinic, hospital room/doctor's office). The purpose of this study was to identifY Llational control patterns in the . b h . d I Th h h commurucattons etween p ystcians an patients. e pnmary emp as1s was on ow (relational dimension) physicians commuJcate with patients rather than on what (the dim ")h"" .\d .. content enston p ys1c1ans commurucate to patients. A total of 5,968 utterances were ctded fur furmat (assertion, question, talkover, noncomplete and other) across nine relational categories (support, nonsupport, extension, answer, instruction, order, disclnfirmation, topic change, and initiation. ")C .. b J\ .. d. d tenrunation ommurucations etween p ystcians an patients were concentrate into three of the nine possible relational ca egories: extension, support, and answer with predominantly symmetrical exchangeJ The researchers concluded that the interactions were primarily doctor-<:enterj or disease-centered as opposed to patient centered as the physicians frequently chanJed the topic, asked more questions and talked more than the patient. 59

PAGE 73

Further analysis also revealed thjt there was a tendency on the part of patients toward a more domineering behavior. However, further analysis of the conversations reflected a ... lack of the dialectic labt ofunderstanding, oflistening and interpreting, of clarifying and acknowledging what has been said, and responding" (p. 30) by the physicians. Additionally, whe1l the patients in this study raised an emotional issue, physicians tended to respond with question about physical symptoms or change the subject, therefore, controlling le meaning of the illness. The findings of this study support1 previous relational research and the frequent complaint by patients that they l not pennitted to tell their stories. The authors suggested that physicians could djvelop a more patient -centered partnership model of care by allowing patients to talk bout their illness experience; providing I more empathy and direct support; and facilitating patient communication around topics which are particularly meaningful to the 1ient such as feelings and expectations. Suggestions for further research in relational communications included: the significance of gender on relational and the influence of I physician gender on patient credibility; the influence of physician attitudes about dh h h I 1 hi d 1 certam toptcs; an ow t e p ystctan-patleDt re at10ns p eve ops over tune. Lerman, Daly, Walsh, et al. (1993) ldied 97 patients (mean age of 55 years) I with Stage I or Stage II breast cancer who had all undergone primary surgical treatment and who were scheduled to recei e postoperative radiation or 60

PAGE 74

chemotherapy. This study evaluated patient-provider communications in relation to patient coping styles and psychological Patient-provider communicationiwere evaluated using the three subscales from the Medical Interaction factor of th Cancer Rehabilitation Evaluation System (CARES): 1) Problems Obtaining InfoJation from Medical Team; 2) Control of Medical Team; and 3) Difficulty Commu1cating with Medical Team. The percentage of patients reporting one, two or three, fJur to six, and seven to ten communication problems were 84%, 39"/o, 35%, and 10)., respectively. The most commonly reported problems were difficulty understanding pJysicians (49.5%), difficulty expressing feelings (46.3%), desire for more controllver the medical team (45.3%),, and difficulty asking physicians questions (42.6%). ThJse patients who were more distressed, who were less optimistic and who felt more hj less about their disease were most likely to report communications problems. Even though it was not possible to determine the direction of causal relationships, Lerman et al. ( 1993) speculated that one mechanism that might be responsible fur the relationship between J unications problems and psychological distress was that patients who were able to ask questions and voice their needs and concerns were more likely to elicit informa ion and receive personalized advice, which reduced uncertainty and provided reassuralce. They suggested that both patient oriented interventions and physician-orient1 interventions which enhance physician-61

PAGE 75

patient communications may be of value in improving patients' adjustments to treatment. Patient-oriented interventions would include those intetventions that enhance patient skills in asking questions rd expressing concerns. Finally, they recommended that future research inci prospective randOOJized trials which evaluate the impact of inteiVentions to inlrease patient and/or physician communication skills. Not only are patients frustrated 1h physician-patient communications, it also appears that physicians may be just as frulated. Levinson, Stiles, Inui, and Engle ( 1993) undertook a study of the nature ol "frustrating" visits from the practicing physicians' perspective. This study exami ed feelings of frustration with patient visits in 1,076 practicing physicians who were aJending a voluntary workshop on physicianpatient communications. l Analysis from the 25-item questio aire developed specifically for this study revealed several key findings: I. Thirty six percent reported fruTtion in up to one quarter of patient visits. 2. The perceived cause of the frustirating visits was most frequently attributed to the patient in more than 50% of the frustrating visits. 3. The most significant patient-attributed communication problems were: lack f b h .\ . h o trust or agreement y t e pattent, patient presentmg Wit too many problems, lack of patient adherelce, lack ofunderstanding on the part of 62

PAGE 76

both patient and physician, r or controlling patient, patient had special problems (alcohol, dnigs, chronic pain). Another communication problem by physicians in ''fiustrating visits" was a feeling of distress on their part during or rrer seeing the patient. For each of the seven communicalion problems identified by the physicians, Levinson et al. (1993) discussed how could be addressed by developing communication skills that are oriented to lard a more -centered, collaborative, mutual participation model of clinical1. Chan!cteristics of this model incloded: use of open-ended questions, allowing the patient to talk uninterrupted at the outset, active listening, repeating and summarizinl the patient's problems, avoiding the use of medical jargon, and involving the patient 1 decision-making about treatment options. They particularly noted the danger of the tendency for physicians to blame the patient; and, concurrently emphasized the impojce of discussion and education for h nhan. h I p ystctans m e cmg p ystctan-pattent commurucattans. Wbite, Levinson and Roter (1994) Lovided insight into one of the most I common challenges encountered by practicing physicians and one of the "frustrations" identified by physicians in the Levinson et (1993) study: the "oh by the way ... phenomenon which occurs at the end ofthJ medical visit. The objectives of the White I et al. ( 1994) study were to define and describe the physician-patient communications during the closing phase of the medical visiJ and to identifY the types of 63

PAGE 77

communication during the visit that migh have contributed to the introduction of a new problem at the close of the visit. Audiotapes of 88 outpatient visits with 20 primary care physicians were analyzed for communication patterns using the Rater Interactional Analysis System. I The patients were predominantly white, highly educated (80% had some college education), females (61%). I' Closure was initiated by physicianj 86% of the time and physicians asked whether or not there were any more quesJons 25% of the time. In 21% of the closures, patients presented new problems not previously discussed. Both physician and patient behaviors throughout the visit were associated with the presenting of new problems during closure. The specific ph1sician behaviors associated with the reduction in the occurrence of new problems at closure were: orienting the patient to the flow of the visit, giving patients more about the treatment plan, and asking about patient beliefs. The specific pltient behaviors associated with reduced occurrence of new problems at closure w4: talking more about their therapy and sharing emotions and feelings. the findings of this study are closely related to the observational studies by B+ and Frankel (1984) demonstrating that patients are interrupted on average 18 seconds into their opening narrative; and, the first problem presented may not be the \rirnary problem. 64

PAGE 78

Similar to other suggestions and odels for a more participatory, collaborative, patient-centered approach to medical cJe, White et al. (1994) proposed a set of communication skills to use before and luring closure which may be helpful to physicians in decreasing the frequency oi new problems presented at closure. Before closure: orient patient to the flow of the visit, address emotions and psychological issues, explore patient beliefS, check for Lerstanding, ask if the patient has additional questions. During closure: orient the patilnt to the closure, provide a summary of the visit, complete a final check fur undemalmng, clarify the plan, provide encouragement I and reassurance, and consolidate rappo,caring. Finally, a noteworthy practical ooservation was made regarding the concern with time pressures in the everyday practJce of medicine. These researchers pointed out that instead of asking open-ended qulstions and whether or not the patient had additional questions during closure, it J be more efficient to ask these questions during the visit and before closure begins. Implicitly embedded in this observation is the possibility that by using the aforementioned communication skills, overall clinical efficiencies in terms of total encounter Je may be achieved as well. Doctor-Patient Communications: Effects on Outcomes We have entered into an era of in health care where the ultimate purchasers of care (government and emplbyers) are placing emphasis on value 65

PAGE 79

enhancement. The notion of value enhan ment recognizes that the cost of delivering medical care and the associated resource can be reduced even as quality is improved. Because quality and value enhbcement may become central to provider differentiation and competitive advantage in the market place, the importance of providers understanding the impact of phlsician-patient communications on quality, value enhancement, and outcomes is paramount. Different dimensions have been ujd to define patient outcomes: satisfaction, clinical indicators, compliance/adherence, fuality oflife, health status, patient-derived measures of functional status, etc. The puwose of this section is to review some of h fd . .1 d" t e unpacts o octor-pattent commurucattons on vanous outcome tmenstons. In their review of the literature, et al. (1995) noted some of the more salient findings regarding the relationship patient satisfaction, compliance, and halh h.. I .. e t outcomes to p ystetan-patient commurucations: Patients are frequently dissatisfied \. th the infonnation they receive and physicians tend to underestimate plent desire for information. T -C. b "d I "gnifi 1 I ed .uuormation-gtvmg y proVl ers to patients was s1 cant y re at to pattent satisfaction. Time spent in discussing preventive care was positively associated with patient satisfaction. More time spent in the medical interview was positively associated with patient satisfaction. A more dominant and controlling physician communication style was associated with less patient satisfactibn. 66

PAGE 80

Non-verbal behaviors such as freguent eye contact, appearing interested etc were positively associated with satisfaction. Positive associations have been nLed between patient-centered care behaviors and patient satisfaction, complianbe, resolution of patient concerns. C li hasb d I h :-: d omp ance een assoCiate Wit more uuormatton-gtvmg an postttve talk. Patient-centered behaviors were related to higher self-reported compliance More patient controlling less physician controlling behaviors and more information-giving by physiJians in response to information-seeking by patients were related to improved patient health status-clinically, functionally, and subjectively. In cancer patients, lack of information seems to be related to increased coping problems, depression, anxiety, and uncertainty. In her commentary on the effects of the relationship on the quality of health care, DiMatteo (1994) denoted sJar findings: I Patient satisfaction with their physician's behavior has been positively associated with the continuity of tHe relationship. More adequate histories have been associated with higher patient satisfaction and less delay in reporting symptoms. Enhanced perceived personal call over heahh with collaboration and TBDS. Increased patient responsibility for f.edical decisions have been related to lower likelihood ofblaming the ph};sician if things go wrong. Increased adherence by patients mL physicians provide more information Physician interpersonal characteriJcs and non-verbal communications have been associated with patient health behaviors. 67

PAGE 81

Patient outcomes can be effected by positive expectations about efficacy by both the physician and the patien Improved recovery from surgecy, I decreased use of pain medications, shortened hospital stays, improved clinical improved symptoms, and enhanced management of chronic conditionk have all been associated with effective physician-patient communication. Roberts, Cox, Reintgen, Baile and Gibertini (1994) studied the influence of physician communication on the psychollgical adjustment and decision-making of I 00 post-surgical breast cancer patients. The Leer Diagnostic Interview Scale (CDIS) was used to assess the patients' perceptiot of their surgeons' behavior during the discussion in which they were informed of their cancer diagnosis. The SCL-90-R Scale was used to assess psychological stabs. Results from the CDIS assessment revealed that the physicians' caring attitude was perceived as the most important factor related to psychological adjustment while information giving was a much weaker fador. Those physician behaviors most closely associated with a caring attitude were: "I fllt my doctor cared about me as a person", "my doctor understood my fears and coni\ s" and "my doctor is a warm and caring person". The authors concluded that the u of psychotherapeutic techniques by I physicians can have a significant positive effect on a patient's psychological well-being. Based on this research and that froJ the social sciences literature, they provided guidelines for what the surgeon sJould do in conducting the cancer I diagnostic interview: select an appropriate time and place with the ideal being a 68

PAGE 82

prearranged return office visit, verbalize empathy with direct eye contact, allow the patient sufficient time to absorb the news and react emotionally, provide information about the cancer and the treatment optiJns, engage the patient in decision-making about treatment. In a very comprehensive study, Fox and Stein ( 1991) utilized random digit dialed telephone interviews to investigaJ the effect of physician-patient communication on mammography u4on in a random sample of 1,057 black, Hispanic, and white women (mean age ot53 years). Seven questions from the d =-c. I ab h uali f h mternew were use to secure uu.ormatton out t e q ty o p ystctan-patJ.ent communication regarding mammography! Results revealed that 71% ofHiJanic women had never had a mammogram and that more than half (55.3%) of the 1spanic women over age 50 years had never had a mammogram. The most important of whether women of all racial groups had a mammogram, at any time o1 within the last year, was whether their doctors had discussed mammography witj them. The data also suggested that wit, the physician-patient communication process, physicians were not effectively communicating information about screening and were discussing screening less frequJtly with Hispanic women than with white or I black women. The authors offered several possible explanations for this. First, fewer than half of the Hispanic women had clinical breast exams that reduced the r

PAGE 83

opportunity for mammography referrals. Second, physicians were not using pediatric visits for their children to offer family c,e. Third, Hispanic women tended not to view the doctor as the most believable sdurce of health information. Fourth, Hispanic women cited more barriers to and anxil about the use of mammography than did black or white women. Finally, may have created communication difficulties. With the rapid population 4wth and purchasing power of the Hispanic population in many concentrated areas U.S. (Los Angeles, Houston, Phoenix, Chicago, New York and others), the implications for provider organizations are ,apparent. In summary, the research reviewed here has provided evidence that patients, and to some extent physicians, have been tstrated with sevend aspects of doctor patient communications and the physician-patient relationship. These fiustrations and dissatisfactions have been borne primarily out of traditional doctor-patient relationships characterized by a Parsonia.Q., physician-centered, paternalistic model whereby the patient has played a predo1antly passive, dependent role and the physician has assumed an autonomous, professionally dominant and directive role in the health and medical care decision-mJg process. It is clear from the evidence preseJted here that patients as well as physician and non-physician researchers continue to rvocate a physician-patient relationship and communication style characterized by informed collaboration, mutual 70

PAGE 84

participation, and patient-centeredness. Most importantly, it is also clear from the evidence presented here that such awthes can have profuund effects on societal outcomes of interest: enhanced patient satisfaction, improved clinical and functional health status for patients, cost efficienciJs, and possibly, enhanced job satisfaction for physicians. Given these implications, Ong et al. (1995) noted that an important goal would be to establish a systematic theory and odel of doctor-patient communication. Such a theoretical framework would be useful for: 1) providing a systematic, organized approach for improving physician-patient communications and relationships; and 2) generating hypotheses regarding such relltionships. Thus, the purpose of the I following section is to review some models of doctor-patient communications and relationships which have been champion1 by researchers and practitioners. Collectively, these models may point us tbward the theoretical framework and practical models needed for advancing in this area; and subsequently, for the successful reengineering of doctor-patient communications and relationships. Models for Physician-Patient Communication and Relationships l In creating interpersonal relations ps, there have been different opinions on how to define and develop an optimal inJrpersonal relationship in the physicianpatient dyad. As early as I 956, Szasz and Hollender described three types of

PAGE 85

physician-patient relationships based largely on the degree of control required by each participant for certain medical circumsJces: activity-passivity, guidance-cooperation, I and mutual participation. According to Ong et al. (1995), some believe it is primarily a social relationship; some claim that the is detennined by its therapeutic qualities or "client-centeredness;" some 1escribe the "ideal" as one which integrates both patient-centered and physician-cent Ired approaches; while others advocate a "mutuality" in the relationship. Several u eful models have been advocated for structuring the optimal doctor-patient rei tionship. von Friedericbs-Fitzwater a!. I) noted that the term "patient-centered medicine" was first introduced by Balint, Hunt, Joyce, Marinker, and Woodcock in 1970. This was a contrasting model to thl traditional illness-centered/disease centered/doctor-centered model. Patient-lentered medicine has been characterized as follows: In the patient-centered metrod, the physician attempts to Ascertain the patient's agenda ... the essence of this method is that the physiciay tries to enter the patient's world, to see the illness thrpugh the patient's eyes. This is done by behavior that in'fites and facilitates openness by the patient. The doctor'ls goal is to understand the patient's expectations, feelihgs, and fears .... bringing the patient's problems into the Wagnosis. (von Friederichs Fitzwater et al.,199l;p.I8) von Friederichs-Fitzwater et at. (1991) po ted out that there has been no operational definition of the model. T

PAGE 86

In 1977, Engel advocated a biopsychosocial model of disease and healing which viewed disease as an open syst4 process that incorporated mind, body, and environment. This model was also in cohtrast to the traditional mind-body dualism prevalent in the traditional paternalistic tomedical model. Keller and Carroll ( 1994 ;p.131) asserted that, "While EngJl' s biopsychosocial model of disease and h alin ha ed gnifi I. d h d 1 e g s recetv s1 cant attentton m pnrnary care as an 1 ea, 1t as receive ess attention as a guide for physician behaviL." Emanuel and Dubler (1995) havl characterized the "ideal" physician-patient relationship as one that encompasses: chbice, competence, communication, compassion, continuity, and no conflict lfinterest. Their "ideal" tilted in favor of a d ll b I a1 h f more pat1ent-centere co a orattve, mutu part1crpatton approac : m tenns o choice, they have advocated for an infoL patient who is provided choice among treatment alternatives; in terms of commLcation, they have advocated for listening, understanding preferences for infonnatiol and involvement in decision-making, assisting patients in dealing with the illnJs experience as well as the disease, and a facilitative role for physicians in guiding Ld directing patients through the illness experience by recognizing values and feelgs as well as life, family, and job circumstances. From a legal perspective, lhey also pointed out that this type of communication style can allay adverse evlnts which may eventually lead to malpractice litigation. In tenns of compassion, they eLphasized the importance of empathy; in

PAGE 87

tenns of continuity, they acknowledged he investment of time necessary for d hr d. I f nfl. f prevention an c orne 1sease management; m tenns o no co Icts o mterest, It was expected that the patient's well-being tJe precedence over the physicians' own personal interests. Keller and Carroll (1992) have identified three perspectives which they believe influence the physician-patient relationf: the physicians' philosophy of disease and health, the role which the physician assuies in the relationship, and the communication strategies employed by tlie physician. They have advocated the "E4" model for physician-patient communicatilns. This model goes beyond the traditional biomedical tasks of"find it and fix it" to fne which unites the communication tasks of and with the biomedical tasks of find and fix. Like the model of Emanuel and Dubler (1995),\ this model has favored a non-paternalistic approach to physician-patient communicahons that emphasizes physician sensitivity to I patient's preferences for infonnation and involvement in decision-making. I Similar to Emanuel and Dubler (1995), Delbanco (1992) has advocated for a more patient -centered view of doctor-Jent interperoonal relationships. His approach includes seven dimensions oft which should be considered when dealing with each patient as a unique individual: respect for patient's values, preferences and d ed d d I din d. f expresse ne s; commurucation an e uo1ation; coor ation an mtegrat10n o care; physical discomfort; emotional support and alleviation of fears and anxieties;

PAGE 88

involvement of family and friends; continuity and transition. Delbanco has asserted that this approach should be: woven intt the clinical encounter in bits and pieces until practitioners become more comfortable rth it; incol]lOrated into the early years of training in medical school; adopted by patients to guide encounters with physicians; used to guide quality of care feedback Jom patients. DiMatteo (1994) proposed a based on collaborative TBDS. This model was characterized as one which includes mutual participation; explicit sharing of and expected outcomes by both patient and physician; information exchange about various treatment alternatives and their Jotential effects; and negotiated treatment decisions. To effect this she prol\ eded to identify the required physician and patient behaviors: establishment of a con ucive atmosphere; allow patients to tell their stories; physicians respond to patient's ilLess experience with concern, empathy, edi d. h I f k mun acy, an mterest; p ys1c1an encouragemen t o patients to as queStions; physician explorationofpatient goals, ojtcome expectations, values and quality oflife components; patient and doctor in considering various health management strategies; physicians should assist patienl. in organizing I information given to them; and patients should be allowed sufficient time to make decisions. Ballard-Reisch (1990) advanced aJtheoretical model of participative decision making for physician-patient interactions. This model has three phases that guide the 1/5

PAGE 89

doctor-patient interaction: the phase; the exploration of treatment alternatives phase; and the treatment decision, implementation, and evaluation phase. In tum, each of these phases includes se1ral stages. The diagnostic phase includes Jonnation gathering and infunnation interpretation by both patient and physiciL. During this phase, the nature of the 1 hi b thd d .I 0 0 ed Within" thi 0 re attons p etween e octor an pattent ts negottat s negottatton, there are four relational options: patient ,bdication, patient autonomy, collaboration, or termination. Ballard-Reisch emphasized that collaboration is the ideal model for . d . makin Th 1 I f al h parttetpauve eetston g. e exp oration o treatment ternattves p ase included the following stages: exploratioJ of alternatives and criteria, establishment of al d ghin f al I Th d ternattves, an wet g o ternattves cntena. e treatment ectston, implementation, and evaluation of impleJented treatment (phase ill) included the stages of: alternative selection, decision lplementation, and evaluation of implemented treatment. Each of the thrl phases incorporates outside input from other professionals, fiunily members, and tpport groups. Ballard and Reisch (1990) emphastzed that the physician behaviors most crucial to this model include: information 1haring; gathering information about the patient's goals and expectations, values, blliefs, lifestyles, and sociocultural relationships; active listening and providini feedback about understandings and interpretations; involving family members kd friends where appropriate; and 76

PAGE 90

collaborative decision-making. The aut ors contended that viewing the doctor-patient relationship as a three-phase process wJch progresses through eight distinct stages of decision-making: 1) provides a structuJd framework upon which physician-patient communications can be improved; and tenders decision analysis researchers with a framework for analyzing the quality of Jedical decisions. F redrikson ( 1995) has proposed r input -process-outcomes infonnation exchange model of medical consultation.[ Central to the model and to the core of the process component is infonnation exchange which, like Ballard Reisch's model (1990), emphasizes a very flexible, partijipative flow and exchange of information between doctor and patient in regards to diagnostic and treatment information and involvement in decision-making. Both lodels have embodied input, process, and outcome components. The input component incorporates both patient and physician input. From the patient's perspective, this includes: :frame of reference; motivations, goals, needs, expectations, and personal infonnation. irom the physicians' perspective, this includes: motivations, goals, needs, expectations, and medical information. The outcomes component is linked to the component and encompasses specific patient satisfaction outcomes related to rabors of patient understanding, f d h .. I I. hi appropnateness o treatment, an p ys1etan re at1ons p. 77

PAGE 91

Kreps, O'Hair, and Clowers (1994) provided a conceptual model of communications and how communicatJ s influences outcomes. They have referred to this model as the Transformation Model of Communications and Health Outcomes. They've pointed out that this model is b sed on the systems theory transfonnation model of input-process-output. In systeLs theory, inputs are the antecedents which provide the raw materials for the procesles necessary to accomplish the outputs. In I their model, antecedent conditions included health problem or risk and consumer attitudes, beliefs, and expectations. The component encompassed the communication activities engaged in by patient and the physician: provider/consumer message strategies, tge used, non-verllal cues, channels, and media. The outcomes of the communicatiion and medical care activities comprised the output component of the model: cognitivl, behavioral, and physiological outcomes. Finally, Ong et al. (1995) provide1 a theoretical framework which is very to the inputs-processes-outputii models described previously. Their model mcludes three components: a) backgroun; vanables, b) actual content of communication, and c) patient outcomes. Included in the background variables component are: culture, doctor-patient relationship, types of patients and doctors, di h . c d h lctual f . sease c aractensbcs. ontame m t e l content o commurucat1ons component were the actual communicative behaviors including both instrumental and affective behaviors. Entailed in the patient outcomls component were short-term, 18

PAGE 92

intermediate, and long-term outcomes including: satisfaction, compliance, recall and understanding of information, health sta I s, and psychiatric morbidity. Diffusion of Innovatio : Constructs. Components. and Determinants Two of the primary aims of this ry were to: I) advance our understanding about how physicians view enhanced participation of patients in medical decision making through the integration ofTBDS services, and 2) glean key insights on the implications for working with physician Iganizations in optimizing the integration of such services into medical care delivery s, stems. Since TBDS and the involvement of patients in playing a more informed and Jctive role in the medical decision-making process can be considered a "new" innovbon in the delivery care services, the framework for studying the diffusion lf innovations may assist in the fulfillment of these aims. TherefOre, the purposes ofJ section are to: review the basic com ponents and premises underlying the Jsion of innovations, identify benchmark rates fur the adoption of new innovations and, there possible, delineate factors/determinants influencing these rates. Components ofDiffusion What is the "diffusion of innovati0n" and what are the major premises underlying this concept? Diffusion has beln defined as: 79

PAGE 93

The process by which an innovation is communicated through certain channels over time among the members of a social system. The main elements are the innovation, communication channels, time, and the social system. .. An innovatioh is an idea, practice, or object that is perceived as new ty an individual or an organization. (Rogers, 1995, p.I0-11).1 Innovations in health care include such things such as new drugs, treatment guidelines, preventive services, procedures, diagnolics, therapeutics, and technologies. The rate of adoption of an innovation over time + been consistently demonstrated to fullow an S-shaped curve among various audiences including physicians, public health administrators, teachers, and fanners (Rlgers, 1995a). I This epidemiological S-shaped arption curve characteristically has an early slow phase involving a few "innovators d "early adopters". This is followed by an accelerated middle phase whereby the rJe of adoption usually takes off following adoption of the idea by the early adopterl. The rate of adoption in the final and third phase is slow ending in incomplete In association with this adoption curve, Ryan and Gross (1943) have described the different populations of adopters within a given social system. They are Lcterized in five different categories. I The "innovators" generally approximately 2-3% of the group. They have been characterized as: well endowed with resources, venturesome, risk taking, fascinated with novelty, willing to expenlent and transcend geographical and social boundaries to learn. Rogers (1995a) refels to them as "cosmopolite". Locally, they 80

PAGE 94

are generally somewhat disconnected socially, and are generally not the opinion leaders. They play an important role by Jerving as the gatekeeper for the flow of new ideas into the system. The "early adopters" generally r around 12-15% of a group. They do talk with innovators and selectively new ideas they would like to try out. Locally, they are highly regarded for mlng prudent decisions in the adoption of new I ideas and they are socially well connected. They are regarded as significantly influential opinion leaders with the "early majority''. For these reasons, they are often utilized by change agents as the local evangelist for accelerating change. The "early llllliori!y" comprise rtely one third of the group and are characterized as very local in their perspectives, more focused on current needs and I local practical problems, and do not hold in high esteem lofty conceptual ideas, science, or theory. Their innovation-deci1ion making process has been characterized as very deliberate and therefore slower thL that of the innovators and early adopters. They generally have a high state of readinlss to hear about new innovations that are relevant and practical for solving local pcJblems from peers they know well and trust. They look to the early adopters for indijons of what is prudent to try in terms of new innovations and serve as an importanl interpersonal network link between the early adopters and "late majority". 81

PAGE 95

The next group, referred to as tlie "late majority'', comprise approximately one are more conservative in their views an distrust "cosmopolite" sources of knowledge. They look primmily to the early majo1 for their learning channels, local "proof', and cues in adopting new innovations. The late majority do not adopt new innovations until they become the status quo and p1 pressure eventually forces their adoption. Finally, Ryan and Gross (1943) define the last 16% as the "laggards". They are the ultra-conservative traditionalists L are the last to change and to adopt new innovations. They are isolated in their 4cial networks and are generally suspicious of change, change agents, and innovations. The label given to this group may be an iJ1iustjce. Even though slow to adopt 1 innovations, this group is often useful and valuable in making wise, systematic, calcillated choices based on a rational view of their own limited economic resources. Adoption Rate Determinants Rate of adoption has been defin1 as, "The relative speed with which an innovation is adopted by members of a social system and is measured by the number of individuals who adopt a new idea in a spJified period" Rogers (1995a, p. 250). He has identified at least five variables effectJg the adoption rate of innovations. 82

PAGE 96

First, the rate is influenced by hlw the potential adopter groups perceive the attributes of the new idea. Rogers pp. 250-251) emphasized that those innovations that have the following are more likely to see rapid adoption when compared to those innovations thal do not have these attributes. I) "Relative advantage the degree to Lhich a new idea is perceived as better than the idea that it supersedes. 2) Compatibility the degree to which an innovation is perceived a consistent with the existing values, past experiences, \and needs of potential adopters. 3) understand and to use. 4) Trialabilitythe degree to which an innovation can be experimented with on a limited basis. 5) Observability the degree to which t1e results of an innovation are visible to others." Rogers (1995a) pointed out that 50-900/o of the variance in adoption rates can be explained by these five attributes. As we b see later, these attnbutes have indeed been demonstrated to be influencing facto s on the diffusion of innovations in health and medical care. 83

PAGE 97

A second variable influencing the rate of adoption is the decision-making process I itself. In the process of adopting a new innovation, individuals go through a decision\ making process that includes the following stages (Rogers, 1995a): 1) knowledgeI becoming aware that a new idea exists; 2) persuasion-developing a favorable attitude about the new idea; 3) decision -deteLg whether to adopt the innovation; 4) implementation -putting the new idea 1o use; and 5) eotifirmation -evaluating the innovation for the purpose of continued adoption, modification, or discontinuance. Interestingly, this typology is conceptujy congruent with the transtheoretical model of behavior change that includes contemplation, preparation, action, and maintenance stages and DiClemente, 1983). According to Rogers (I995a), Jr additional variables influencing adoption rates include: 1) the nature of the communication channels within a social network, 2) the nature of the social system throuJ which the innovation is diffusing, and 3) the degree to which change agents are involJed in promotional activities. He speculates I that when considering all five of these Ties, the diffusion process is driven predominantly by how the attributes are pirceived and how this perception is conveyed through the adopter group's interpersonal networks. In addition to Roger's (1995a) clalsic social system theory of diffusion, Green (1980) has provided another typology thal holds significant potential for assessing factors which influence the diffusion of Jovations. It also provides a framework for 84

PAGE 98

targeting interventions for enhancing the diffusion of a given innovation. In his PRECEDE Green has classified lhese factors into three categories: predisposing factors (physician knowljge, attitudes, beliefs,), enabling factors (skills/competence, resources, practice I\ aracteristics), and reinforcing factors (reimbursement, peer support, positive eedback, enhanced self-esteem). The application of this typology will be explJred in a later section examining the diffusion of preventive services. Diffusion of Innovation in Health Care The gap between medical science and medical practice (Greer,l994) as well as the gap between best achievable practice1 and actual practice (Berwick, 1998) is well established. Many medical innovations (nlw drugs, treatment guidelines, preventive services, procedures, diagnostics, and technologies) may take yean or even decades to reach widespread adoptien. Ford, Kalumy, and Sondik (1990) have poignantly noted that the time involved tJ achieve the full transfer of new treatments continues to be problematic in medicine. Even though an enormous body of scholarly research exists on the diffusion of innovations, it is still not a process. A practical standardized process for success has not yet been codified. Thlrefore, it is not surprising to find wide variation in the rates of adoption for new in health and medical care services. The 85

PAGE 99

purpose of this section is twofold. The first is to review, where available, rates of adoption related to new drugs, practice and services. The second is to distill those that have been demonstrated to predominantly influence diffusion and rJes of adoption. We'll begin by first examining one of the earliest and most classic sociometric studies on the diffusion of medical innovations. I Using pharmacy prescription Jrds with information about the month in which each doctor adopted the new dru4 Coleman, Menzel, and Katz, (1966) studied the spread of an antibiotic (tetracycline) Jmong physicians in four lllinois communities. They also collected data from physician r1spondents regarding other physicians in their community with whom they had discussl the new drug. This inlormation allowed the researchers to trace the social networks trgh which the innovation was diffused. The commercial sponsor of the study, ler, provided academic detailers to extensively promote the new drug. This innovation spread rapidly reaching almost complete adoption in 17 months. Physicians reported that the acadLc detailers had little direct influence on their decision to adopt. Instead, the resejchers discovered that information about the new drug spread primarily by word of molth from physician to physician and, generally, a physician did not adopt tetraJcline until a respected peer had shared his/her experiences with the new innovatiol. Greer (1994, p. 434) insightfully noted so

PAGE 100

that what is usually forgotten in this study is that the best predictor of general adoption was not adoption by the physicians who were recognized as opinion leaders, but "adoption by those physicians who with the most other physicians the most often." She emphasized that, "the more people talk and stimulate further talk, the more there is opportunity for consensus, practical problem solving, and social pressure to change." In another prescription drug study, Peay and Peay (1988) examined the adoption process fur temazepam thro, interviews with 124 specialists and general practitioners. They found that within 13 months after its release, 71% of the physicians were familiar with the drug ( time between release and awareness was about 17 weeks), 48% had prescribed it (average time between its introduction and first prescription was 5.5 months), j 27"/o preferred it to the alternatives. I In contrast to the Coleman, and Katz (1966) study, they reported that contact with academic detailers was the Jost consistent predictor of favorable reception by physicians. Sixty-eight p4 reported that their first news came from commercial sources, 40% citing the in particular. Only 28% reported discussing this new drug with other doctot before deciding to prescribe. Eighteen percent reported colleagues as the most important information sources in their first decision to prescn"be. Doctors who had mte frequent contact with detailers (at least once per week) were aware of the drug ear;lier, prescribed it earlier, and were more 87

PAGE 101

likely to prefer it to alternatives than those who reported less contact with detailers. On the other hand, physicians who cited professional sources as their first source of infonnation became aware of it later, prescribed it later, and were more likely to prefer the alternatives than those who did not. l tenns of practice characteristics, there was d ha 1 I diffi ed fr h no evt ence to support t t so o practttilners er om group practitioners m t e reception of temazepam. Peay and Peay (1988) emphaticall noted that there was little evidence to support the Coleman, Menzel, and Katz (1966) depiction of innovators, early adopters, and physician communication n tworks. They contended that their data h th d. I infl h b were more consistent Wit ose stu Ies reportmg an uence on p ysicians y sources and factors external to the classicjsocial system theory. In this case, they contended that there may be a relationshi 1 between therapeutic risk and adoption rates. That is, since temazepam was perceived as a low therapeutic risk, physicians may have been willing to relinquish evaluation in discussions with colleagues and prescribe based on infonnation providld by the detailers. Interestingly, Coleman, Menzel, and Katz ( 1966) noted that detailJs were most often identified as sources of \ infonnation just after the drug became available, but closer to the point of adoption, more professional sources (such as colleaJes) were predominant. Peay and Peay (1988) appropriatel1 concluded by acknowledging that the adoption of new drugs is an extremely complex process and that the consistent 88

PAGE 102

prediction of which doctors will adopt new drugs has thus far eluded those in the field. Adoption rates may indeed be influencld by the interaction of many factors including: h.. 'ct/ ... hill h .. p ystctan conta commurucatton Wit co eagues, age, practice c aractenstlcs, perceived risk of the drug, and academil detailers. Lomas (1991} noted that there le more that 60 organizations in the United States and at least ten countries that technology assessments of new medical ideas, use some form of consensus devellpment in their deliberations, and pUblish practice guidelines. The predominant l of all this activity is to close the gap between the science of medicine and the bractice of medicine, between achievable best practices and actual practice, between wJat is known and what is actually done in delivering health and medical care. UltJtely, to influence physicians in improving the quality of care. Stross and Harlan ( 1979) conducted one of the earliest studies assessing physician awareness and knowledge aboJ the outcomes of a major clinical trial. They surveyed (82% response rate) 228 family and internal medicine physicians to examine their knowledge of the results of l clinical trial demonstrating the efficacy of photocoagulation in diabetic rjinopathy. The preliminary results of this clinical trial were published in April1976. Stross and Harlan conducted their study in 1978, two years following the publication of the clinical trial results in the American Journal of Ophthalmology. 89

PAGE 103

Even though 43% of the family Jhysicians were aware that a large clinical trial had been conducted, only 28% reported\that they were aware of the results. Sixty percent of the internists were aware of tie study and the results. When asked to manage two pa,ent problems mvolvmg diabetic retmopathy, only 33% chose the correct treatment both patients. When physicians who were familiar with the study were asked their source of information regarding the clinical trial results, more than two thirds repol that an ophthalmologist or colleague was the primary source. The authors conclul that this study supported the suspicions of many that valuable information from expLve clinical trials was not being brought to th f . h . dl b. ed. th d e attention 0 practicing p ystCians an ras not emg mcorporat mto e every ay practice of medicine. They suggested that professional interactions with influential colleagues may be an effective comrnunicltion channel and a key mechanism for diffusion. Jordan, Burke, Fineberg, and Hanley (1983) studied selected findings in the diffiJsion of innovations in bum care. Thel study approach was to systematically select a set of innovations in bum care and query physicians about their awareness and use of the innovations. Eight treatments wre selected from a National Institutes of Health Consensus Development Conference (1977-78) and classified according to their complOJ
PAGE 104

were randomly selected to receive a poStal questionnaire (sent out in 1979) that asked questions about familiarity, first knowltge, sources of knowledge, and utilization of the eight selected treatments (response rate of86%). After aggregating results for all treatments, the average awareness level was 75%. More physicians were consensus treatments (82%) than low consensus treatments (70"/o) and low coLple
PAGE 105

delivering CME on new innovations at teaching centers and conferences, there may be better results in spreading innovations ,ouglt local outreach initiatives. In addition, such initiatives should take into consideration three important considerations. 1) What innovations are most appropriate to incljde in such local programs? They suggested that high consensus treatments with low Less levels are prime candidates. For high complexity treatments where awareAess is low, the CME effort should be primarily directed at awareness and infolg physicians of these treatments. 2) Which physicians have the most need for Lrormation? Obviously, CME programs should identify and target those physicianl who have low awareness and use levels. 3) How should infonnation about innovatioJs be made available? It was suggested that the most popular channels of influence rJ awareness and use be identified then subsequently used for communication. Second, the authors noted that several factors tend to influence adoption rates: a) the availability of resources, b) the trJmg and skiUs necessary to use the treatment, c) the regulatory environment, ld d) the physicians' conviction (or lack thereof) that the treatment actually works. The implications are obvious. All the aforementioned observations should be taken into consideration when trying to structure more effective CME programs performance improvement strategies within physician organizations. 92

PAGE 106

Hill, Levine, and Whelton (1988) studied the awareness, use, and impact ofthe 1984 Joint National Committee Consenlus Report on High Blood Pressure (JNC ill). The objective of the study was to utilizl the diffusion of innovation framework to examine diffusion of the JNC ill consls report and to determine to what extent its recommendations had been adopted by !hysicians. They conducted a pre-post-survey (response rate of 44% to both in a stratified random sample of Maryland physicians who were likely to be caring for individuals with high blood pressure. The survey was intended to oL information regarding: awareness of the JNC lll one year after its release, channels most frequently used for I receiving the report, use of recommendations in everyday practice, and practice behavior changes preand post-release report. This research team reported that Jne year after publication, even though 62% of the physicians were aware of the 58% reported having a copy available, 81% reported familiarity with the report, lruy 17% reported use in practice. The comnnmication channels most frequently rted for receiving the Consensus Report were pharmaceutical company representatives, professional journals, and professional association mailings. Second, these researchers found that practice behaviors pre-and post-survey were not significantly different. That is, mhre that two-thirds of responding physicians were already practicing in a manner conJent with nine of the ten treatment 93

PAGE 107

recommendations studied. The authors ncluded that this particular consensus report codified rather than changed practice beLvior. This conclusion was remarkably consistent with the findings ofKosecoff, lanouse, Rogers, McCloskey, Wmslow, and Brook (1987) in their study of the effectJ ofthe National Institutes ofHealth (Nlli) Consensus Development Program on phjician practice. Through 1986, the Nlli had conducted 60 consensus development conferences encompassing drugs, devices, diagnostics] therapeutics, prevention, and public health. The primary purpose of this study was to lxamine how consensus conferences have affected physician, knowledge, attitudes ld practice behaviors. This study collected I data from several sources including a physician survey on the knowledge, attitudes, I and practices of physicians as well as a review of hospital medical records to determine changes in actual practice. This study repjrted on the medical review component and was designed to examine to what extent: 11 conference recommendations addressed areas of needed of change, 2) pre-conference practice behaviors were already in high I compliance with the conference recommendations, 3) the recommendations stimulated change in practice where none was occurriJg, and 4) the recommendations accelerated change that was already under way. The Jedical record review was designed to assess the impact of four conferences: 1) thJ treatment of primary breast cancer, 2) the use of steroid receptors in breast cancer, 3) lesarean childbirth, and 4) coronary artery bypass. 94

PAGE 108

For each conference, the researcllers studied medical records from 10 randomly selected acute, non-specialty +spitals drawn from two periods before and one period after the conference, with eaci period lasting nine to 12 months. Time period one was 13-24 months and time period two was 0-12 months prior to the conference. Time period three began Je to12 months following the conference. Compliance for each of 11 recommendatilns selected from the four conferences was studied. Results showed that the conferences mostly failed to stimulate change in physician practice. For six of the II recoLtions, the compliance level remained at less than 50% following the conferencelln addition, compliance was, in general, increasing during the year or two immedilely preceding the conference and compliance failed to accelerate following le conferences. One of the major findings in this study was that the Nll:I recommenltions addressed areas where there was already high confurmity while at the same le neglecting areas of practice that needed improvement. Consistent with the study 1 Hill, Levine, and Whelton (1988), this research tended to support the fact that the consensus reports codified rather than changed practice behavior. The research bt Kosecoff et al. ( 1987) prompted an astute sociological researcher (Greer,1987, p. 2710) to pose the very appropriate question, "Should we be surprised that the national Jonsensus has caught up to the field?" Greer's (1994) own obseiVations have doclmented that decisive change largely occurs 95

PAGE 109

in local communities of practice and receptiveness to change from sociocultural determinants. In this regard, Kosecoff et al. (1987) acknowledged that changes in practice do not inevitably occur from the disseminat+ of consensus reconunendations. Their reconunendations for changing physician iractices in the dissemination of practice guidelines were consistent with diffusion of innovation theory. First, the consensus conferences should concentrate on areas practice that need the most improvement. This generally means that topics shoold partly on the basis of data regarding existing community practice so that both the state of science as well as the state of practice can be taken into account] Second, it should be recognized that li h d I h d h comp ance Wit some recommen attons ,ay reqwre c anges not un er p ystCian control, such as the acquisition of new tools, technologies, resources, and facilities. Third, successful physician behavior changl will require dissemination and l . th 1 11 d based ilizin" tmp ementatton strategtes at are more o""""" commuruty, an practiceut g respected peers for performance feedback Ld face-to-face endorsement. Specifically, to change the state of practice, a local disjmination .and implementation program I nmst meet some minimum requirements.le information must be timely, clinically grounded, and clinically relevant at the local level. The intended audience should be targeted appropriately and assessed for readiness, willingness, and ability to act on the recommendations. 96

PAGE 110

Lomas, Anderson, Domnick-Pierre, Vayda, Enkin, and Hannah (1989) evaluated the effect of a national consenls statement on cesarean births in Ontario, Canada. Like their U.S. counterpart, the Canadian consensus statement explicitly recommended a lower rate of cesarean births. The guidelines were disseminated I extensively in Canada between March and June of 1986 in a mailing to all obstetricians who were on the mailing list of the natioJal specialty society (Society of Obstetricians and Gynecologists) and all hospitals with Lore than 50 beds. The guidelines were published in both the national medical jot and the bulletin of the society. Monthly discharge data from Ontario hospitals of obstetricians before and after the dissemination of the guidelines were used fo examine awareness, attitude, knowledge, self-reported effect, self-reported practice and actual practice. The analysis revealed that one, after dissemination 94% of the obstetricians were aware of the guidelines and 82-85% meed with them, but specific knowledge of the content of the recommendations was pbor (67% correct response). The obstetricians reported learning of the guidLes from multiple sources with the most predominant being the specialty society bjletin (94%), discussion with colleagues (56%) and the Canadian Medical AssociJion Journal (52%). Even though at two years following release of the guidelines 33l of the obstetricians and hospitals reported they luid changed their practices l a result of the guidelines, data on actual practice showed that c-section rates were 15-49% higher than the self-reported rates. 97

PAGE 111

That is, two years after dissemination, p ysicians' behavior had changed very little. An interesting secondary calculation reiarding rate of change was completed by these researchers. They approximated the n ber of years it would take for the overall c section rate in Ontario to reach 10.5%, le approximate rate in England and Wales. At the rate of change generated by the in Ontario, it was estimated that it would take Ontario more than 30 years reach a rate of I 0.5%! Lomas et al. ( 1989) concluded practice guidelines may predispose physicians to considering change, but behaviors are influenced and sustained by many things besides research evidencl. They identified some possible administrative, educational, patient-centled, and/or economic factors potentially infl h . I fl "d lin ed thr f uencmg p ystctan unp ementatton o practice gw e es: percetv eats o malpractice litigation from patients, inadluate skills, economic and socioeconomic d d"ed I din c.ed p mcenttves, an patient-me tat regar g prererr treatment. omtmg to the work of Greer (1987) and Kosecoff 1 al. (1987), Lomas et al. (1989) acknowledged the local and individualized nature of the forces shaping clinical policies. They concluded that ''unless are embedded in a broader program that addresses the need for translation and imJlementation of the guidelines locally'' (p. 1310), clinical policies will be unlikely to lffect rapid change in practice behaviors and I unlikely to effect the ultimate objective ofiimproved quality and patient well-being. 98

PAGE 112

In another study on cesarean births, Lomas, Enkin, Anderson, Hannah, Vayda, and Singer (1991) conducted a randomizld controlled trial with 76 physicians in 16 Ontario, Canada community hospitals to the effect of audit and feedback compared to local opinion leader education on guideline compliance. The practice I guideline was produced and disseminated in 1986 in collaboration with the Society of Obstetricians and Gynaecologists ofcla. The guideline explicitly recommended clinical actions to increase the trial of lablr and vaginal birth rates. The unit of randomization and intlrvention was the community hospital. Eight hospitals served as a control group and Jceived a copy of the practice guideline. Four hospitals service as an intervention group for audit and feedback (A/F) on physician perfonnance regarding compliance with tl!le practice guideline. Four hospitals served as an intervention group for the use of lojally selected educationally influential opinion leaders (OIL) who utilized a "detailing" in furmal and infurmal interactions with peers. For the experimental groups, lhe duration of the active interventions was 12 months. After 24 months (1988 and 1989), the trial oflabor and vaginal birth rates in the AIF group were significantly differnt from those in the control group. However, the trial of labor and vaginal birth rates for the OIL group were significantly I higher than the control group, 46% (P = .007) and 85% (P = .003), respectively. I Duration of stay was significantly (P<. 001) lower in the OIL group. The mean c-99

PAGE 113

section rate for the OIL group dropped b;y 11.5)0/o from the pre-study year of 1987 to the end of the two-year study period whJreas in the control and the AIF groups, the mean rate increased by 4.7% and 6.7%, rlspectively. Quality of care (measured by maternal deaths, infant deaths, complicaJons, and Apgar scores) was not "ed" "h ... \Th b thah findin compromts m e1t er Intervention group\. e aut ors note t t ese gs can not be explained by case-mix differences between the groups. It was noted that the failure of thj AIF strategy to create any change ran contrary to the literature. They that differences in methodology (change I agent in previous AIF study acted much like an opinion leader) and research environment (teaching hospital vs. commLty hospital) could be partial explanations for these differences. Also, because physlians in the AIF group would only agree to department level feedback and would not to personal individualized feedback (an essential component for AIF e:ffectiveless ), the power of the intervention to significant change may have been comproLsed. Also instructive from a diffusion plrspective was the fact that even in the OIL group, over 50% of eligible women still ulderwent an elective c-section instead of a trial oflabor. The authors noted that this Lay partly be explained by "patient factors", I another element that has been related to cJange in physician behaviors. Specifically, in follow-up interviews with OIL physicians, lhe most common reason offered for lack of 100

PAGE 114

guideline compliance was the refusal of many women to undertake an offered trial of labor. The researchers concluded that educationally influential opinion leaders with educational support can generate change among their peers when I they agree and prepare to be agents of change. The authors speculated that the I success of the opinion leaders may lie in their ability to take local circumstances and cultures into account in their educational/iommunication activities with peers. At the same time they noted that even four years after the release and widespread dissemi nation of a clinically sound practice guideline, the compliance rate among physicians encouraged by opinion leaders was no bjer than 70%. Given this observation and the aforementioned "patient factors" they Lserted that further advances in the I appropriate use of c-sections may have to be derived from patient education and regulatory or economic incentives that 4 already been proven eJrective. Grilli, Apolone, Marsoni, Nicoluccj, Zola and Liberati ( 1991) investigated the impact of a national education program baled on the dissemination of written guidelines for the treatment of breast, colJectal, and ovarian cancer in Italy. Through a survey of 770 physicians and a review o)lmedical records of I, 483 patients, this study examined whether: 1) the guidelines reached the targeted audience of physicians --those working at community hospitals] 2) they were effective in influencing physicians' opinions, and 3) the guidelines Lere being used in patient care. 101

PAGE 115

Three task furces for breast, color, and ovarian cancer were launched in Italy in 1978, 1979, and 1985, respectively. The distribution process for the breast cancer guidelines included: a mailing to t general hospitals, distribution at some national scientific meetings, availability upon request from the task forces' coordinating centers. The guidelines for I olorectal and ovarian cancer were distributed at scientific and specialists' meetings and were made available upon request from the task forces' coordinating centers. Ofthe 1,874 physicians identified or the study, 770 (41%) completed a valid questionnaire. Compared to respondents tended to be female and younger physicians with radiotherapists ,d medical oncologists responding more frequently than other specialists. The medical records review included care provided to all newly diagnosed breast (742), colorectl (641), and ovarian (100) cancer patients I consecutively seen at one of 45 hospitals from March 1986-March 1987. Compliance with guideline recommendations was by examining how closely patterns of care followed 36 recommendations (12 frdm each disease site). The percentage of doctors awareness of the guidelines were 60"/o, 47%, and 44% for breast, colorectal, and Jvarian cancer, respectively. Additionally, those who were more aware of the guideliles reported that the guidelines were brought to their attention informally by o \er colleagues. 102

PAGE 116

Significant variation in compliance with the 36 recommendations, especially in relation to diagnosis and staging, were lealed: 37% to 89%, 48"/o to 82%, and 10% to 97% for breast, colorectal, and ovarij cancer, respectively. Compliance with the 36 recommendations for treatment range1 from 300/o to 86%, 11% to 90%, and from 18% to 91% for breast, colorectal, and otarian cancer, respectively. This research team concluded thaJ guideline development and dissemination had a negligible effect on cancer care. Jey delineated a set of requirements very much like those specified by Kosecoff et l. (1987) for changing the state of practice. The message must be timely, scientificall1 grounded and clinically relevant. Consideration must be given to the context in which local clinical decisions are made. The appropriate audience must be targetel. The target audience must be both willing and able to act on the information. A novel and insightful study conducted by Grilli and Lomas, ( 1994) investigated the relationship between comlliance rates and key aspects of the specific clinical message of the practice guidelines. All studies in the English language medical . literature published between 1980 and 19i! that provided practice guideline clinical content, compliance rates, endorsement official organizations, and which taJgeted providers were eligible for the study. A total of23 studies addressing 143 recommendations across 70 different medilal and surgical procedures were selected for the study. 103

PAGE 117

For each of the 143 recommendations, the authors categorized them according to specialty area, type of procedure ( dilostic, surgical, etc.), and as being high or low on characteristics thought to influenJe diffiJsion. For diffusion characteristics, the authors adapted the five category taxonoLy proposed by Rogers (1995a): relative advantage, compatibility, complexity, tJability, and observability. Lack of information in the studies did not allow tAe authors the capability to assess the relative advantage and compatibility. Thus, each \of the 143 recommendations were classified as being high or low on complexity, trialability, and observability in a typical clinical setting. High complerity, tria/ability, and observability were respectively defined as when a practitioner with usual training and skills, working in an average setting: 1) "perceives it to be difficult to acquire the lkills for or to understand, or has no direct control over any resource changes requir1 for implementation; 2) is able to experiment with it on a limited basis; 3) is able to obtain timely feedback on any impact of the intetvention on patients' outcomeslp. 204). Overall, the mean compliance these studies was 54.5% with significantly (P < 0. 001) higher compliance rates for the specialty areas of cardiology ( 63.6%) and oncology (62.5%), than those fur preventive care ( 46.2%) or obstetrics and gynecology (43.3%). +mmendations regarding procedures with high complexity had lower compliance rates than those low on complexity ( 41.9% vs. 104

PAGE 118

55.90/o; P=0.05). Those recommendations judged to be high on trialability had a signficantly (P =-0.03) higher compliance rte (55.6%) than those judged to be low (36.8%). Interestingly, this study found no influence of observability on compliance rates. As the authors noted, this was surprising in view of the fact that the outcomes is often cited by physicians as major reason for non-compliance. Multiple linear regression was uscl:l to assess the overall amount of variability in compliance rates which could be explled by all the pre-selected characteristics of the message (data source, area ofpracti1 complexity, trialability). Overall, these characteristics explained no more than 4 7% of the observed variability in compliance rates. The authors arrived at several key \conclusions and recommendations that are instructive for informing and guiding the diffusion of innovations in health care. First, the observed compliance rates suggest thJe is still significant room for performance improvement in the current delivery of care. Second, since only about half of the variance in compliance rates could be lxplained on the basis of diffusion and other characteristics, these researchers that a larger percentage of the variance might be explained if other recognized catLries of potentially influential factors were included in the analysis: the source of the lessage, the channel of communication, the clinical setting, and the nature of the receiig audience. 105

PAGE 119

Third, complexity was the single most effective predictor of compliance. Obviously, recommendations with high ctmplexity and low trialability are more likely to result in low levels of compliance. Grilli and Lomas ( 1994) asserted that the nature of diffusion efforts may differ for high coLplexityllow trialability recommendations I compared to low complexity/high trialability recommendations. That is, for recommendations with high complexity Jd low trialability, diffusion practitioners may I effi a I de akin want to p ace greater ort mto pre zsposmg practlttoners to cons1 r m g changes in their delivery of care before laLching efforts that are more focused on enabling and/or reinforcing strategies. In contrast, for recommendations with low complexity and high trialability, less effo1 may be needed to predispose clinicians to change because they may be already recerve to this kind of change in practice. Instead, "efforts could move more to a focus on strategies such as opinion leaders to enable change at the locallevel'1 (p. 212). The authors went on to acknowledge that from a diffusion perspective, even with a highly implementable message, it is still unclear whether to focul efforts on the providers themselves (audience), the incentives and structure oflhe workplace (setting), or a combination of the two. Lawrence (1990) has examined the diffusion of the U.S. Preventive Services Task Force (USPSTF) Guidelines. He has commented that because of the number of topics reviewed and the variation in the qua.Iity of scientific data supporting the 106

PAGE 120

guidelines, the analysis and prediction o1the diJfusion of the USPSTF recommendations are complex. He goes! on to explain that because the guidelines are directed at an entire population with many subgroups with varying needs, and because some of the recommendations are more jechnical (immunizations and screening tests) while others emphasize behavioral (counseling), such differences significantly influence the rate of diffusion. Notwithstanding, he utilized Green's (1980,1988) PRECEDE model (predisposing, enabling, and reinfurcing ,ehavioral determinants) as a frameworlc for analyzing the aids and barriers to the diffiision of preventive services. Since the PRECEDE framework is complementaryrl o Roger's (1995a) and Greer's (1987, 1994) contention that diffusion is largely sociocultural process; it will be instructive to briefly review Green's (1988) and Lalence's (1990) delineation of these three categories ofbehavioral determinants. Predisposing factors for the diffusion of preventive services included physician: knowledge and attitudes toward services, personal health behaviors, confidence in counseling abilities, and beliefs about their patient's interest in health promotion and disease prevention. EnabJg factors included: physician competence to -C'. h.. .lb c. d. peu.onn preventive servtces; p ys1c1an remt ursement 10r ren enng preventive a practice environment organizj to facilitate the delivery of preventive services; time to provide services; efficient reminder systems; physician perception of 107

PAGE 121

guidelines as being clear, coherent, and scientifically grounded. Reinforcing factors included: peer support, positive feedbacJ and receptivity by patients, evidence of intermediate results such as improved he1th behaviors among patients, and an "enhanced sense of self-efficacy in one's role as a healer." This distillation of behavioral detenninants provides a framework for assessing strategies to enhance the diffusion of preventive services as well as a foundation for the review of additional studies on this topic. Three years prior to the release first edition of the U.S. Preventive Services Guidelines, Lewis, Wells, and (1986) constructed a conceptual model and an associated questionnaire to study r determinants of physician counseling practices. The questionnaire measured the agressiveness, indications and techniques used by physicians in counseling patients smoking, exercise, weight control, and alcohol use. The independent variables asLsed motivations, perceived skills and barriers, medical-specialty, and personal hLth habits. A random sample of male physicils (151) with an average age of 51 years al d "'--=1 I d b representing gener an li:U.IWY practitioners, mtermsts, surgeons, an o stetnetangynecologists from a county medical completed the questionnaire. The questionnaire also assessed the health habits of the physicians. Fifteen percent reported they were a current smoker, 58% reported they did not exercise enough, and 24% reported that drank alcohol about every other day. 108

PAGE 122

In tenns of motivations to 87% agreed that smoking is dangerous to your health, 85% agreed that counseling \was important, and 87% agreed that physicians have an obligation to counsel. Even though 74% indicated they knew how to interview patients, only 21% reported rhey knew how to counsel patients and only 12% felt they were effective in counseling. Physician perception of skill in counseling was a statistically significant (P =0. 05) +ctor of physician counseling behavior for smoking, exercise, and weight control. 'Ilhat is, those physicians who perceived themselves to be more skilled in counseJg were more likely to engage in counseling th d kin I. d gh 1 Thi err patients m regar s to smo g, exermse, an wet t contro s ts consistent with the framework provided by Lawrent (1990) who cited physician confidence in their ability to motivate behavioral change in their patients as a predisposing factor for the diffusion of preventive services. j Over half of all physicians report ; counseling all patients about smoking I (52%), wetght control (56%), and alcoho1 (55%). However only 26% reported counseling all patients about exercise. Forty-four to 55% of physicians said they counseled individuals for less than two jutes with regard to all health habits included in the study. Lawrence ( 1990) cited physician health behavior as a predisposing factor to consider in the diffusion of preventive 1= The most significant finding of this early study by Lewis, Wells, and Ware ( 1986) was that physicians who had poor health 109

PAGE 123

habits did not fully counsel their patients about those habits, and physicians attempting to improve poor health habits counseled significantly more than physicians I who were not trying to change their own behavior. The authors concluded that strategies which fucus health promotion 1rts on physicians may have a nrultiplier effect. That is, maintaining physicians' hjalth through risk factor reduction may I benefit both the physician and ultimately the patient. Schwartz, Lewis, Clancy, Kinosl Radany, and Koplan (1991) surveyed 2610 I members and fellows of the American College ofPhysicians from four geographical areas of the U.S. to estimate internists' usl of immunizations (pneumococcal, influenza, tetanus, hepatitis B), screening jests for cancer (breast examination, Papanicolaou smear, stool occult blood tel), and other diseases (electrocardiograms, cholesterol, chest x-ray), and behaviorallunseling (smoking, exercise, alcohol and seat belt use). In addition, information reJard.ing background information, personal health, and record keeping were gathered. Overall, results revealed that this group of internists used effective preventive interventions less frequently and ineffectivl practices. more frequently than recommended by expert guidelines. Y physicians and those whose practices were primarily oriented toward general intimal medicine exhibited greater compliance with expert recommendations for health 1motionldisease prevention. Attitude and lack of knowledge were associated with the use of preventive services. However, one 110

PAGE 124

of the most significant findings of this stlildy supported the observations ofLawrence (1990}, Lewis, Wells, and Ware (1986).1There was a strong and significant (P < 0.01) relationship between physicians' person 1 health behaviors and their propensity to provide the same services to their patienJs. Furthermore, this relationship. remained persistent for all three classes of health prevention activities studied: immunizations, screening, and for behavioral factors. Like Lawrence I (1990) Lewis, Wells, and Ware (1986), these authors suggested that perhaps an I appropriate strategy for improving the diffusion of preventive services is to target physicians' personal health practices. McPhee and Bird ( 1990) studied the utilization of cancer prevention guidelines in clinical practice as recommended by thl National Cancer Institute (NCI}, the American Cancer Society (ACS}, and the U.S. Preventive Services Task Force (USPSTF). Their objective was to answer three basic questions: 1) Is there a I problem? 2) Why is there a problem? 3) What can be done about it? To answer these questions, these researchers compared thl results they obtained in studying 52 physicians from a university general inteJal medicine practice with other published studies. Their results revealed that there is indeed a problem. Even though physicians tend to overestimate their own performance rates, actual performance rates for cancer prevention guidelines are consistently beldw those recommended by NCI, ACS, and 111

PAGE 125

the USPSTF. Consistent with the behavioral determinants framework provided by Lawrence (1990), these researchers cateJorized the major barriers to implementation into physician factors (physician disagreement with recommendations), patient factors (patient apathy, discomfoJ refusal, inconvenience), test factors (cost, risk, accessibility, uncertain cost and cliniL effectiveness), and health care delivery systems factors ( lack of reimbursement, ILk of equipment/staff, scheduling problems, inconvenience). These researchers advocated inte entions that focus primarily on enabling and reinforcing factors for both physicians an patients as the most promising for enhancing the diffusion of cancer prevention guidelines. Specifically, 1) physician audits with feedback, 2) computerized reider systems for physicians, 3) reminder systems and behavioral counseling/educatiln for patients. Battista, Williams, and MacFarlanj (1990) studied the detenninants of preventive practices with respect to cervical, colorectal, and lung cancer I through face-to-face interviews with 552 fee-for service primary care physicians in the Canadian provinces of Quebec and New BLnswick. They gathered information on over 200 items including: a) physician sociidemographics, (b) practice and organization characteristics, (c) continuing medical education activities, (d) knowledge of risk factors associated r each type of cancer, e) physician beliefs about the efficacy and effectiveness of selec:rted preventive practices, (f) perceived 112

PAGE 126

barriers to implementing preventive services into clinical practice, and (g) reported patterns of adult cancer prevention. sJilar to McPhee and Bird (1990) they grouped the perceived barriers to prevention into I physician-related (lack of knowledge, perceived ineffectiveness of procedures, risk of procedures) patient-related (lack of interest, lack of information, loj compliance, time, and cost), organization related (accessibility, time constraints, Jera11 costs). For regression modeling they grouped these predictor variables for eat cancer type into a typology conceptually congruent with Green's {1988) predisposing, enabling, and reinforcing factors: cognitive, sociodemographic and ocsamlonal determinants. Like Mcl'hee and Bird (1990), Jey fuund that when comparing reported practices to actual practices (billing recofs), physicians overestimated their preventive services practice behaviors by 10%. Multiple regression analysis revealed particular predictor variables for each cancer type. !Specifically, knowledge, continuing medical I education, beliefs regarding efficacy, gender, accessibility, and provider-related barriers were significant (P < 0. 01-0.00 I) predictors of detection scores for breast cancer. The most important determinant of breast cancer detection scores was gender, with female physicians having higher det1on scores than male physicians. I For cervical cancer, physicians with higher knowledge scores and younger aged who practices in high volume urban lenters were more likely to perform Pap 13

PAGE 127

tests according to recommendations. Fo, colorectal cancer, the most important detenninant of practice behaviors was the absence of perceived practice barriers (P < 0.001). Additional significant predilors were knowledge, continuing education, in this preventive activity. For lung cancer prevention (anti-smoking counseling), the most significant (P < 0.001) detenninant was the smoking status of the physician with non-smoking physicians being more likely to provide anti-smoking counseling. Three additional predictors were continuing belir and language (English-speaking). For lung cancer detection, the routine use of chest x-rays in asymptomatic individuals, a I test not recommended by guidelines, was r to be associated With the strength of physicians' beliefs in the usefulness of early detection of lung cancer. I Attesting to the complexity of this issue and the limitations of research I instruments, these authors acknowledged that their regression models could only explain between 5-17% of the total varianJe in practice behaviors. Even though reimbursement per se was not examined in this study, the authors cited previous research they had conducted which disclosed a lesser involvement of fee-for-service physicians in adult cancer prevention than lalaried physicians. These researchers concluded that in fee-for-service rei.mburseLent environments without specific financial incentives for preventive practicesl organizational factors were more 114

PAGE 128

important in shaping physician behaviors. Therefore, in fee-for-service environments, organizational factors (accessibility, timJ constraints) may be fertile ground for interventions aimed at increasing physiciL behaviors for adult cancer prevention. Summary. In sum, the rate ofJovation diffusion in any system is influenced by a complex and often confounding +on of many social, cullwlll, individual, organization, and environmental that may possibly include: Perceived attributes of the innovation: relative advantage, compatibility, complexity, trialability, observability. Stages of decision-making: knowledge, persuasion, decision, implementation, and confinnation. Physician predisposing factors: knowledge, attitudes, beliefs, skills/competence, and personal health behaviors. Physician enabling factors: training, resources, practice characteristics. Ph =-c. c. b I d ys1ctan reuuorctng !actors: reun ursement, peer support, positive peer an patient feedback, enhanced self-esteerh and self-efficacy. Patient filctors: patient-mediated p.l...s regarding preferred treallllent; positive patient feedback, patient receptivity/apathy. Organizational/health care delivery sykem factors: local practice norms, available equipment, technologies, facilities, tmie, scheduling, access, convenience. Communication/information attributJ: the nature of the social system and the nature of the communication channelsiwithin a social network, relevancy, timeliness, clinical and scientific grounding, source of the message, appropriately targeted audiences. Environmental factors: regulatory environment. 115

PAGE 129

Rogers (1995a) has noted that ne innovations bring with them an element of uncertainty. "Uncertainty" in the pracril of medicine seems to be a common road in the health care landscape inextricably Jng the research regarding adoption of new innovations. geographical variations (Wr. 1982), and appropriateness (Brook, Lohr, Chassin, Kosecoff, Fink & SolomoJt, 1984; Brook, Kamber, Mayer-Oakes, Beers, Raube, & Steiner, 1990). To help lesolve this uncertainty, physicians often tum to near peers for further information, subjective evaluation of new innovations, and consensus within the context ofthe local\communi{Y of practice standards, experiences, and norms. "This information exchange abou1 a new idea occurs through a convergence process involving interpersonal networks. The diffusion of innovations is essentially a social process in which subjectively perceived information about a new idea is communicated. The meaning of innovatiJn is gradually worked out through a process of social construction" (Rogers, 1995a, p. xvii). "The more people talk and stimulate further talk, the more there is opportunity for consensus, practical problem solving, and social pressure to change .... frequen.t discussion creates a sense of movement .... The problem of uncertainty is Jsolved by a new consensus" (Greer,1994, p. 434). Thus, closing the gap between medical science and medical practice, spreading innovation, and accelerating the rate ofadlption ofbetter practices may be as much a 116

PAGE 130

function of an organization's social capital as its intellectual or financial capital. The successful adoption of any new innovatiln within physician organizations will require an understanding of how to leverage and maximize that social capital. It would I behoove practitioners who are faced with the daunting challenge of successfully institutionalizing best or better practices rthin physician organizations to be cognizant of these factors and to subsequently account for them in any implementation strategy. Summary and Implications for Research First, there is a substantial body Jr evidence regarding patient preference for information and medical decision making. However, there seems to be a coinciding scarcity of information regarding physiciL' opinions about how active patients should be in seeking information or in plicipating in health and medical care decisions. The proposed research will b1d on the work ofStrull et al. (1984) in assessing physicians' opinions about how medical decisions should be made. Because there was relatively little demographic information characterizing the physicians in the I Strull et al. (1984) study, very little else about physician perceptions regarding patient infonnation seeking and involvement in drsion-making could be extrapolated from existing research. The proposed study will provide additional insight by regressing various sociodemographic factors and prabce characteristics on physicians' opinions ab . I d" al d .I. out patient mvo vement m me 1c ectstons. 117

PAGE 131

Second, TBDS services are expenencing rapid growth and integration into managed care delivery systems. Concurr1tly, there is a dearth of information regarding the opinion of physicians about such services. Even though the opinion of the physician stands as a critical variable in optimizing the ultimate impact such decision support interventions will have t physician-patient relationships, clinical efficiencies, patient outcomes, medical care utilization, and costs, reviews of published literature revealed no systematic assessmJnt of physician viewpoints about these types of services. Nor has anyone tried to ascejain the implications of their opinions on the implementation, utilization, and ultimate pact of such services on health care quality, cost, and access factors. The proposed research will be one of the first to: 1) advance our understanding about how physicians view enhanced participation of patients in medical decisionmaking and the integration ofTBDS servibes into care management; and 2) provide key insights on the implications for worJg with physicians in optimizing the diffusion and incorporation of such services across Jhe continuum of care in managed care delivery systems. I 118

PAGE 132

QUANTITATIVE JTIIODS AND RESULTS Introduction s. h iliz.d I h .. ab urvey researc was ut e to ascertam p ystcians oplDlons out: 1. More informed, actively involjed patients as partners in health and medical care decisions. 2. The impact they believe consumers/patients being more informed and taking a more active partnership role in health d medical care decisions will have on: the quality of the doctor -patient relationship, clinical efficiencies, health outcomes, patient satisfaction, health care utilization, health care costs, and malpractice liability. 3. The incorporation ofmos serces into care management and the impact they believe a TBDS service will have om the quality of the doctor-patient relationship, clinical efficiencies, health ojtcomes, patient satisfaction, health care utilization, health care costs, and malpice liability. This chapter will describe the methods, analyses, and results pertinent to the quantitative portion of this study. 119

PAGE 133

Methods I This section will describe the subjects, the instrumentation (Physician Opinion Survey), and the data collection protocol for administering the survey. Subjects The domain of the physician pop ation studied included all primary care physicians (approximately 859) and speciLists (approximately 1306) who were members of the Rochester Independent P1actice Association (RIP A) and the Rochester Community Independent Pract ce Association (RCIPA). All physicians eligible for the study (N = 2, 165) were ed a self-administered Physician Opinion Survey (POS). Physician Opinion Survey (POS) Both research and theoretical constructs supported POS question development. The questions pertinent to preferences for making decisions and patient desire for information was guitd and informed by a substantial body of survey development and associated patienJ research in this area (Cassileth et al., 1980; I Strull et al., 1984; Ende et al., 1989; Sutherland et al., 1989; Brody et al., 1989; Beisecker and Beisecker, 1990; Neufield J al., 1993; Davison et al., 1995). mo

PAGE 134

The development of questions related to the effects of patient decision and self-management support on various quality, lcost, and access parameters was guided and informed by both research (Vickery et al. 1983; Lorig et al., 1985; Greenfield et al., 1989; Mayo et al. 1990; Dentzer, 1991; Long et al. 1993;) and theory (Lynch & Vickery, 1993; Vickery & Lynch, 1995). Final questions and format fori POS were developed in collaboration with the executive and physician leadership from RCIP A and RIP A. Following the guidelines ofPieterzak et al. (1990) for dlvelopment of new measures, a modified multi-stage pretest and revision L used for the POS: I) pre-test with staff, colleagues, and experts 1iuniliar with the troject; 2) use feedback from Step I to guide adjustments and revisions; 3) pre-test with staff and physicians that are not members of guide final revisions. The final POS is illustrated in A. The survey includes a profile of demographic data and physician practice 1haracteristics. The first section of the POS asked physicians their opinions about mJr infonned and actively involved patients in the health and medical care decision-makihg process. The second section of the survey asked physicians their opinions abJut informed consent and telephone-based 121

PAGE 135

decision support (TBDS) services as wei as some open-ended questions regarding the key challenges and benefits of implemenlg TBDS services. POS Administration and Data Collection Protocol The following data collection prolocol was used for the self-administered POS: 1. Administration The survey was administered through the mail by RCIP A using the standard membership mailing list for RCIPl and RIP A. 2. Mailed Survey Packet Contents The contents of the mailing piece included: (a) a cover letter printed on IPA letterhead that included the name, signatures from IP A executive and physician leadership, IP A IUllllber fur questions about the survey; (b) the POS printed on IP 1. letterhead; and (c) a business reply envelope for returning the survey. 3. Mailing Procedures a) b) c) Initial mailing ofPOS to study subjects June 1996. Physicians who did not ..L surveys within three weeks were mailed a follow-up survey. Data collection stopped six weeks following the mailing of the second POS. 122

PAGE 136

4. Confidentiality-All surveys were returned to the University of Colorado Health Sciences Center Departmlnt of Family Medicine. Each survey was I numbered for the sole pwpose response rates. In accordance with the University of Colorado Healtli Sciences Center Research Protocol for Human Subjects, all responses wlre anonymous and confidential. We obtained li . hrl gh h f h . N unp Cit consent to parttctpate t Gu t e return o t e questtonnarre. o physician was identified by name.IAll analytical infonnation was reported in the aggregate. Analysis and Statistical Methods This section describes the selectioh of the independent variables, the dependent variables, and the statistical tests utilized L analyze the POS data. Sample Description Descriptive statistics for the independent variables were used to describe the physician subjects and their practice charaberistics. Description ofVariables Descriptive statistics and frequen distributions were calculated for each of the 30 opinion variables selected for studj. 123

PAGE 137

Selection of Dependent Variabl sand Psychometrics. From the 41 total opinion variables ascertained by the POJ, key opinion variables consistent with the aims of this research were selected for 1y. For those items where there appeared to be a similar underlying "construct" or "fJctor", inter-item correlations using the alpha coefficient for internal consistency were lalculated. The alpha coefficients for the inter -item correlations are summarized 1 Table 3 .I. Table 3.1. Cronbach alpha reliability c9efficients for composite items ofthe Physician Opinion Survey. I ---Question# (B2C,B2H,B3B,B3D) More Involvement 2 I Question# (B2A,B3!A) Minor Issue Decisions 2 I -Question# (BIA,B3C\ ) Physician Time 2 Question # (B2D,B2(J) 885 0.29 856 0.30 893 0.67 None ofthe correlations met the 1pha of 0.70, the criterion for new scales (Nunnally and Bernstein, 1994). TJerefore, 30 survey questions emerged as the physician opinion factors (dependent varilbles) of interest for this study. A summary I of the 30 survey questions utilized in the final analyses is illustrated in Table 3.2. 124

PAGE 138

Table 3.2 Physician Opinion Survey utilized in final analyses. To ic I Question No. More Infonnation Patients should ask more questions Patients want more infonnation Patients need role in asking questions Patients need role in seeking information More Involvement Patients want more involvement Patients need role in decision making Minor Issue Decisions Who should make minor decisions Patients need role in minor conditions Major Issue Decisions I Patients Need Role in Chronic Conditidns Patient Preference Physician Time Physicians don't have time to involve patient Physician don't have time to answer 1atient Physician Awareness ofTBDS Physician Opinion ofTBDS Reasons Health Plans Implemented TBDS Improve quality of care Improve member satisfaction Reduce health care costs Respond to business community Improve patient outcomes Reduce liability claims Improve efficiency Potential Impact of TBDS Unnecessary phone calls Inappropriate visits Patient demand for services Liability claims Length of clinical encounter Health outcomes Physician-patient relationship Patient satisfaction 125 B2C B2H B3B B3D B2A B3A BIA B3C BIB B3E B2F B2D B2G Dl D2 D3A D3B D3C D3D D3E D3F D3G D4Al D4A3 D4A4 D4A6 D4A7 D4Bl D4B2 D4B4

PAGE 139

Selection of Independent Variables. The independent (predictor) variables were chosen for their potential to influence opinions: gender, age, location of d al f I k, d f rest ency, spect ty, type o practice structure, practtce VISits per wee an percent o patients in a capitated reimbursement pl. Two independent variables were eliminated I from further analysis. "Years in practicer was eliminated because it closely paralleled "age" and added no new infonnation. "Salary structure" was omitted due to low priority and minor interest as a study vjable. Selection of Statistical Tests Logistical Regression Analysis. The descriptive statistics and frequency distributions for each of the 30 depend, variables were examined. This examination revealed non-normal distributions of positively and negatively skewed physicians' opinion about patients being encouraged o ask more questions about their medical care. Approximately 94% of the physicians "agreed" that patients should be encouraged to ask more questions leavinJ the remainder in the "no opinion" (3%) and "disagree" (3%) response categories. ThJefore, where there were dependent variables with highly skewed non-normal distributilns, the seven-point Likert scales were 126

PAGE 140

recoded for analyzing the real differences between any "agree" responses vs. "all other''; and "no change" responses vs. "1 other." For the dependent variables that fd not exhibit an extremely positive or negative skewness, but instead present, a more rectangular or bimodal distribution, the seven-point scales were recoded from seven categories to three categories: "agree", "no opinion", and "disagree." the five-poim scales regarding who should make minor and recoded as ''both" vs. "all other." Table 3.3 summarizes the recoded Likert categories for each of the 30 dependent variables and the frequency percentage for each ote categories. Table 3.3 Recoded Likert categories frequency percentages for dependent variables logistical regression analyses. _Dependent_ Variable #lj __ ..fategories __ Minor decisions (BlA) both (56)/ all other (44) Major decisions (BIB) both (68) I all other (32) Patients want more involvement (B2A) agree (88)/ all other (12) Patients should ask questions (B2C) agree (94)/ all other (6) Physicians can't involve patient (B2D) disagree (56) I all other (44) Patients don't want to make decisions agree ( 46)/ no opinion ( 1 0) I disagree (B2F) (44) Physicians can't answer questions disagree (65) I all other (35) (B2G) Patients want more information (B2H) agree (90)/ all other (10) Patients need role in decision making agree (82)/ all other (18) (B3A) 127

PAGE 141

Table 3.3 (Cont.) Dependent Variable (question #) Patients need role in asking questions (B3B) Patients need role in minor condition (B3C) Patients need role in seeking info Patients need role in chronic conditions (B3E) Physicians' awareness ofTBDS (D1) Physicians' opinion ofTBDS (D2) TBDS to improve quality of care (D3Al) TBDS to improve member satisfaction (D3B) TBDS to reduce health care costs (D3C) TBDS to respond to business (D3D) TBDS to improve patient outcomes (D3E) TBDS to reduce liability claims (D3F) TBDS to be more efficient (D3G) TBDS impact on phone calls (D4A1) TBDS impact on inappropriate visits (D4A3) TBDS impact on patient demand (D4A4) TBDS impact on liability claims (D4A6) TBDS impact on clinical encounter (D4A7) TBDS impact on health outcomes (D4B1) TBDS impact on phys-pt relationship (D4B2) TBDS impact on patient satisfaction (D4B4) Categories (frequency %) agree (88)/ all other (12) agree (72)/ all other (28) agree (87) I all other (13) agree (66)/ all other (34) unaware (26) I all other (74) positive (21) I neutral (44) I negative (16) unaware (18) agree (37) I no opinion (30) I disagree (33) agree (59)/ all other (41) agree (83)/ all other (17) agree (72)/ all other (28) agree (38)/ no opiniC?n (30)/ disagree (32) agree (22)/ no opinion ( 40)/ disagree (38) agree (61)/ all other (39) no change (57) I all other (43) no change (57) I all other (43) no change (58) I all other (42) no change (77) I all other (23) no change ( 69) I all other (31) no change (64) I all other (36) no change (51)/ all other (49) improve (36) I no change ( 46) I worsen (18) 128

PAGE 142

The frequency distributions for each of the 30 dependent variables detennined the selection of either binomial or mulJomiallogistical regression analysis to evaluate the simultaneous contribution of the phJsician demographic and physician practice characteristics ( mdependent variables) J predicting physicians' opinions (dependent variables). This technique is effective in analyses where categorical independent variables are used to predict ordinal dependent variables with non-normal distributions (Wolfe and Smith, 1996). The criterioJ variables were physicians' opinions. Gender, I f d al I f . age, ocatton o rest ency, speci ty, type o practice structure, practice VISits per week, and percent of patients in a capitaled reimbursement plan were used as predictor variables. The predictor variables were tegressed on each criterion variable. Odds ratios and 95% confidence intervals were calculated. All calculations were made using SPSS version 9.0 Open-ended questions. The open-ended questions were coded, segmented, and analyzed using methods of qualitativl analysis described in the Qualitative Analysis Section of Chapter 4. Results This section provides a descriptioh of the subjects, a description of the variables, and logistical regression resultJ for the POS. 129

PAGE 143

Sample Description Relevant subject characteristics are presented in Table 3.4. A total of914 clinicians retwned useable questionnairJ, a response rate of 42%. Table 3.4 Phymcian and J.!'"actice cmkeristics .* Variable I No. % Gender Female Male Age 30-39 40-49 50-59 60 and over Years in Practice <11 11-20 21-30 >30 Location of Residency Rochester Other Practice Primary care Specialists 130 197 715 212 325 210 157 214 303 199 171 447 462 379 535 21.6 78.4 23.4 36.0 23.2 17.4 24.1 34.2 22.4 19.3 48.9 50.5 41.5 58.5

PAGE 144

Table 3.4 (Cont.) Variable Practice structure Solo Single specialty group Multi-specialty group Hospital Other/combination Practice visits/week < 100 100-199 200 or more % patients capitated 0 1-49 330 20h I 50 or more 2017 % 28.6 37.4 9.3 7.2 17.5 37.5 26.5 36.0 44.5 27.5 28.0 *Total N = 914. Ns will vary by questithn due to missing data. Approximately 78% of the subjls were male and 21% were female who had been in practice for an average of20 yels (SD = 11). The mean age of the physicians was 48 years (SD = 11 ). ApJroximately 50% completed their residency training in Rochester. For the purposes of this study, primary care jncluded those clinicians in general practice, family practice, general internal medicine, and general pediatrics (Donaldson, MS, Yordy KD, Lohr KN, Vanselow NA-,1996). All others were classified as specialists. Primary care clinicians complsed approximately 40% of the subjects while nearly 60% of the subjects were specialiJs. 131

PAGE 145

A little over one third of the subjects practiced in a single specialty group and nearly one third in solo practice. The mL number of practice visits per week and the mean percentage of patients in capitated leimbursement plans were 283 visits/week (SD = 592) and 24% (SD = 30), resp -lely. Description of Variables Univariate Analyses. For a majori:tY (24) of the opinion variables examined in this study, there was a rather strong and kectionally consistent opinion held by most physicians resulting in either a positive o1 negative skew in the frequency distributions. For a small number of opinion variables (6), there was a greater variance of opinion resulting in more of a rectangular or bimJdal distribution. The frequency percentages for each of the recoded Likert scales of lh dependent variable are summarized in Table 3.3. As described previously, the distributions were utilized to guide the selection of statistical tests. Missing Values. A large number of cases with missing information can create inaccurate or biased results. The variablel with missing values are presented in Appendix D. Tabachnick & Fidell (1983) have outlined the different methodologies on how to treat missing data. For this analysis, calewise deletion of data was selected. This 132

PAGE 146

methodology requires the elimination o observations with missing values for a given variable from the analyses. Logistic Regressions Logistical regression techniques used to detennine which physician/practice characteristics were the most significant predictors of physicians' opinions, the referent criterion variables] Regressions of physician practice characteristics and practice variables on opinions about more infurmed, actively involved patients and ffiDS 1 summarized in Table 3.5. Statistical adjustments for multiple comparisons were not made. It should be recognized that when conducting multiple comparisons, lhe probability of one or more Type I errors is greater than 0. 05. That is, at least one J twenty comparisons may be significant due to chance alone. Table 3.5 Logistic regression analysis o physician characteristics and practice variables on oEinions. Opinion .Y ariables Predictor Variables and P V aloes G PT R s A v c Patients Desire Information Patients should ask more questions 0.33 0.90 0.47 0.18 0.27 0.89 0.68 Patients want more information 0.10 0.27 0.61 0.36 0.99 0.29 0.61 My patients take > role asking 0.19 0.83 0.51 0.26 0.04 0.16 0.20 questions My patients take > role in seeking inf1 0.29 0.19 0.53 0.72 0.15 0.39 0.10 Patients Desire Involvement Patients want more involvement 0.44 0.73 0.84 0.74 0.33 0.20 0.80 Eatients take > role in decisions 0.10 0.11 0.35 0.95 0.53 0.25 0.57 133

PAGE 147

02inion Variables I Predictor Variables and P Values G PT R s A v c Who Should Make Minor Decisions l Who should make minor decisions 0.30 0.60 0.78 0.27 0.32 0.47 0.99 My patients take > role tx minor illnes 0.35 0.04 0.06 0.00 0.54 0.27 0.57 Who Should Make Major Decisions 0.75 0.81 0.76 0.49 0.00 0.96 0.91 My Patients Take > Role in Chronic 0.58 0.37 0.34 0.00 0.38 0.73 0.08 lllness Physician Time MD doesn't have time to involve 0.69 0.11 0.17 0.43 0.96 0.74 0.30 Patients MD doesn't have time to answer 0.70 0.51 0.15 0.12 0.01 0.20 0.01 patients Patient Preference for Decision Making 0.73 0.39 0.61 0.08 0.08 0.85 0.70 Physician Awareness ofTBDS Services 0.10 0.26 0.00 0.02 0.02 0.17 0.06 Physician Opinion of TBDS Services Physician opinion ofTBDS (all) 0.84 0.67 0.14 0. 70 0.63 0.80 0.10 Physician opinion of TBDS (aware) 0.37 0.13 0.50 0.03 0.69 0.11 0.84 Physician opinion ofTBDS (unaware) 0.99 0.00 0.40 0.09 0.00 0.16 0.05 Reason Health Plans Implemented TBD Improve quality of care 0.45 0.40 0.68 0.02 0.74 0.42 0.84 Improve member satisfaction 0.78 0.34 0.62 0.00 0.72 0.86 0.49 Reduce health care costs 0.44 0.96 0.02 0.31 0.55 0.12 0.87 Respond to business community 0.52 0.80 0.02 0.00 0.20 0.19 0.21 Improve patient health outcomes 0.73 0.02 0.04 0.06 0.83 0.88 0.64 Reduce medical liability claims 0.03 0.09 0.04 0.69 0.48 0.76 0.58 More efficient care of capitated patients 0.41 0.74 0.60 0.06 0.05 0.86 0.25 I r

PAGE 148

Table 3.51Conq 0 inion Variables Predictor Variables and P Values G PT R s A v c Potential Impact of TBDS Volume of calls to practice 0.58 0.40 0.52 0.01 0.71 0.54 0.09 Number of inappropriate visits 0.08 0.78 0.52 0.01 0.90 0.43 0.04 Patient demand for services 0.72 0.46 0.26 0.47 0.26 0.00 0.02 Number of medical liability claims 0.27 0.37 0.10 0.69 0.05 0.87 0.13 Average length of clinical 0.79 0.30 0.46 0.99 0.65 0.31 0.21 encounter Patient health outcomes 0.10 0.35 0.83 0.97 0.68 0.76 0.25 Quality of doctor-patient 0.66 0.42 0.29 0.14 0.85 0.97 0.28 relationship Patient satisfaction with medical 0.56 0.36 0.96 0.21 0.54 0.52 0.14 care (G) gender, (PT) practice type, {R) lofation of residency, (S) specialty, (A) age, (V) number of visits/week, (C)% capitation. Patients Desire More Information. Physicians were of th,e opinion that patients do desire more information, should ask Lore questions, and should be encouraged to seek more infonnation about health and r.edical care issues in the medical decisionmaking process. The percentage of physicians agreeing with the statements: a) patients should be encouraged to ask 1re questions about their medical care, b) patients want more information regarding health and medical issues, c) I would like my patients to take a greater role in asking Juestions about their care, and d) I would like my patients to take a greater role in seeJng available information a bout a medical problem or condition was 94%, 900/o, 88r, and 87%, respectively. The relationship between age and physicians' opinions about their patients taking a greater role in asking questions about their care was siJmcant (P = 0.04) but nonlinear. Physicians 135

PAGE 149

in the 40-49 and the 50-59 year old age groups were more likely than physicians in the 30-39 and the 60 and over age groupk to agree that their patients should take a greater role in asking questions about thrl. care (89.7% and 91.8% versus 84.7% and 85.4%). Patients Desire More Involvement. Overall (N = 891), 80-90% of the physicians were of the opinion that desire more involvement in the health and medical decision-making process. The pJcentage of physicians agreeing with the statements: a) patients want more involvjment in their health and medical care decision-making process and b) I would tle my patients to take a greater role in the medical decision-making process was 8,. and 82%, respectively. There were no significant differences of opinion related to physician/practice characteristics. Decisions About Minor Health Issles. Slightly more than half (56%) of the physicians indicated both physician and pJtient together should make decisions about I minor health issues, while nearly 34% in,cated mostly the patient. There were no significant differences of opinion related te physician/practice characteristics. Almost three fourths (72%) of the physicians agreed with the statement that they would "like their patients to take a role in treating their own minor conditions." Practice type (P = 0.04) ami specialty (P = 0.002) were significantly related to decisions about minor health isles. Physicians in multispecialty group I practice were more likely to agree (85.5%1 that their patients should take a greater B6

PAGE 150

role in treating their own minor conditions whereas, physicians in solo practice were less likely to agree ( 63.7% ). Primary le physicians were more likely than specialists to feel positive about their patients a greater role in treating their own minor conditions (81% versus 65.8%). Decisions About Major Health Issues. Slightly over two thirds ( 68%) of the physicians indicated that decisions regarbg treatment for major health issues should be made by both physician and patient tlgether while 28% indicated mostly the physician. Age was significantly (P = 0. bo7) related to physicians' opinions about who should make decisions regarding treatmL fur Illl!ior health issues. Younger I physicians were more likely than older physicians to believe that both physician and I patient together should make decisions about major health issues. Older physicians (50 and over) were more likely than youlger physicians to believe that decisions for major health issues should be made by the physician. Chronic Conditions. Two thirds lr the physicians agreed that they would like I their patients to take a greater role in self-managing chronic conditions. Specialty was significantly (P = 0.000) related to this olinion. Primary care physicians were more likely than specialists to agree that they I ould like their patients to take a greater role in self-managing chronic conditions. Patient Preference for Decision aking. Physicians were asked to what extent they agreed with the statement, "patients want their physicians to make decisions for 137

PAGE 151

them." Opinions were varied on this factor with approximately 46% agreeing, 44% I disagreeing, and 10% with no opinion. There were no significant differences of opinion related to physician/practice cJacteristics. Physician Time. Slightly over hJ (56%) of the physicians disagreed with the statement that physicians do not have time to fully involve patients in the decision-making process. There were no significant differences of opinion related to physician/practice characteristics. Nearly two thirds (65%) disagreed with the statement that physicians do not have adequate time to answer patient qukstions about medical care. Age (P = 0.008) and % of patients capitaled (P 0. 007) tere significantly related to this opinion. There was a linear relationship between and physicians' opinions about having adequate time to answer patient questionl. That is, with increasing age there was an increasingly higher percentage of physiciLs disagreeing that physicians do not have adequate time to answer patient questionl. In particular, physicians 60 years of age and older were more likely than younger to agree that physicians do have d . I a equate tune to answer patient questions. There was also a linear relationsJj between the percent of patients capitated and physicians' opinions about having time to answer patient questions. W h h .I gl high f h It mcreasmg capitation, t ere was an mcreasm y er percentage o p ysiCians hah"" d h dl. . agreemg t t p ysiCians o not ave a ate tune to answer patient questtons.

PAGE 152

Additionally, there were no statistically ignificant relationships between visits/week and physicians' opinions about having time to answer patient questions. Approximately two thirds of all physi1s, regardless of the number of visits/week, disagreed with the statement that physi]. ans do not have adequate time to answer patient's questions about medical care. Physician Awareness of (TBDS) Services. Physicians were asked to report I whether or not they had heard of the TBDS service that had just been implemented within the community by the health planJ; and, to what extent they were familiar with (N = 888), 73.7% had heard of the servi e. Of those reporting that they had heard of the service, 31.3% indicated that they were familiar with its purpose and available services while 42.4% reported that even kough they had heard about the service they knew very little about its purpose and av1able services. Location of residency (P = 0.002{ specialty (P = 0.016), and age (P = 0.02) were all significantly related to physician familiarity about the TBDS purpose and available services. Physicians who had clmpleted thdr residency in the community I reported being more familiar with the TBDS purpose and available services than physicians who completed their residency elsewhere (38% versus 24.9%). Primary care physicians reported being more familiar with the TBDS purpose and services than specialists (38.2% versus 26.5%). Physijans in the 40-49 and the 60 and over age 139

PAGE 153

groups were more familiar with the service than were physicians in the 30-39 and the 50-59 year age groups ( 35.1% and 341 versus 28.4% and 26.8%). Physicians' Opinions About the (TBDS) Service. Physicians' opinions about the TBDS service were analyzed for thrl different groupings: I) all physicians, 2) only those physicians who reported familiar with the purpose and available services, and 3) only those physicians 1o reported indicating that they were unaware of the service or knew very little about its purpose. Physicians' opinions for these I three groups are summarized in Table 3.l Table 3.6 Physicians' opinions ofTBDS service. Physician Group I Opinion Neutral Negative All physicians* 2I!I% (I83) 44.5% (387) I6.0% (139) Knowledgeable ofTBDS service 32]ZO/o (89) 42.0% (II6) 25.7% (7I) Unaware ofTBDS service 36]8% (25) 48.5% (33) I4.7% (10) The sum does not total to I 00% as I8% of all physicians reported being unaware of the TBDS services. As illustrated in Table 3.5, for the overall group (N = 869), there were no significant differences of opinion related to physician/practice characteristics. For those familiar with and knowledgeable about thl mos service (N = 276), specialty was significantly (P = 0.03) related to physicils' opinions about the service. Primary care I physicians were nearly three times more likely than specialists to have a generally positive opinion about the TBDS service. I40

PAGE 154

For those physicians who reported being unaware of the TBDS services (N = 68), practice type (P = 0.001), age (P = 0.002), and% ofcapitated patients (P = were significantly related to physicians' opinions about the service. Even though these relationships were statistic1y significant, small cell sizes precluded any practical significance. Reasons for Implementation of TBDS Service. There were several primary reasons why physicians believed that the I ealth plans were implementing TBDS services. The reasons are summarized in descending order in Table 3.7. Table 3. 7. Physicians' opinions about hy health plans implemented TBDS. Reason I "Agree" Percentages* To reduce health care costs [ 83.3% (708) To respond to business community interests 72.4% (617) To more efficiently care for capitated patients 61.1% (521) To improve member satisfaction 59.1% (501) To improve patient health outcomes 38.0% (323) To improve quality of care 37.5% (317) To reduce the number of medical liability claims 21.6% (184) The sum will not equal to 100%. Each percentage is the percent of physicians who reported "agree" out of the total nJmber of physicians reporting for that particular survey question. Implementation Reason: Reduce
PAGE 155

Implementation Reason: Respond to Business. Location of residency (P = 0.02) and specialty (P = 0.001) wJe significantly related to physicians' opinions I that the health plans implemented TBDS services in response to interests of the I business community. Primary care physicians were more likely than specialists to I agree that the health plans implemented rns services to respond to the interests of the business community (78.7"/o versus r-9'/o). Physicians who had completed their residency in the local community were more likely than those who had completed their residency outside the local community tJ believe that the health plans implemented TBDS to respond to the interests of the Lsiness community (75. 7"/o versus 69.2%) I Implementation Reason: More Efficient Care. Even though there were no statistically significant differences of oplon related to physician/practice characteristics for this variable, there weJe some interesting trends. A greater percentage of physicians with the highest percentage of capitated patients (50% or more) than physicians with a lower percentage of capitated patients (zero or 1-49%) believed that health plans implemented thl TBDS to more efficiently care for capitated patients ( 68.2% versus 61. 7"/o and 57. Io/1There was an inverse relationship rh age. With increasing age, a lower percentage of physicians believed that health plans implemented TBDS to efficiently 142

PAGE 156

care for capitated patients. When compared to specialists, primary care physicians were more likely to believe that the TBD\S was implemented to more efficiently care for capitated patients ( 67% versus 57%)1 Implementation Reason: ImproJ Satisfaction. Specialty was significantly (P = 0.005) related to physicians' opinioJs that the health plans implemented TBDS to b p I h likl h unprove mem er satis1actton. nrnary care p ys1C1ans were more e y t an specialists to believe that the health plans implemented TBDS to improve member satisfaction (65% versus 54.8%). Implementation Reason: Improve Health Outcomes. Opinions were varied regarding physicians' beliefs about improLg patient health outcomes as a reason for why health plans implemented TBDS. Ejr though 38% agreed this was one of the reasons, 32% disagreed and 30% had no epinion. Location of residency (P = 0.04) and practice structure (P = 0.016) were sikmcantly related to physicians' opinions that the health plans implemented TBDS t improve patient health outcomes. Physicians who completed their residency 'in the community were more likely than physicians who completed their residency butside of the local community to believe that health plans implemented TBDS to patient health outcomes ( 40.4% I versus 35.1%). Physicians who were academic hospital based were more likely than

PAGE 157

physicians in other types of practice stlctures to believe that health plans implemented TBDS to improve patient iealth outcomes. Implementation Reason: Improve Quality of Care. There was also heterogeneity in physicians' opinions .fut improving quality of care as a reason for why health plans implemented TBDS. Even though 37.5% agreed this was one of the reasons, 33% disagreed and 30% had nl opinion. Specialty was significantly (P = 0.02) related to physicians' opinioJs that the health plans implemented TBDS to improve quality of care. Primary care ptysicians were more likely than specialists to believe that the health plans implementel TBDS to improve quality of care (45.5% versus 31.9%). Implementation Reason: Reduce Medical Liability Claims. There was a wide variance in physicians' opinions about re6ucing medical liability claims as a reason for why health plans implemented TBDS. ty 22% agreed this was one of the reasons. However, 38% disagreed and 40% had no opinion. Gender (P = 0.03) and location of residency (P =0.04) were both significanlly related to physicians' opinions that the health plans implemented TBDS to redule the number of medical liability claims. I Males were more likely than females to that this was a reason (39.9% versus 29.8%). Physicians who had completed their residency within the local community I were more likely than physicians who had completed their residency external to the local community to disagree that this wj a reason (40.8% versus 35%). 144

PAGE 158

Potential Impact ofTBDS Service. A consistently predominant percentage of physicians were of the opinion that th1 would be no change (versus positive or negative impact) on certain health care variables as a result of implementing TBDS services. The selected health care variablls and the percent of physicians reporting no change for each are summarized in de.Jming order in Table 3 .8. I Table 3.8. Physicians' opinions about the potential impact ofTBDS services on selected health care variables. I Variable I %(No.) Believing No Change* Number of medical liability claims 77.4% (639) Average length of clinical encounter 69.1% ( 570) Patient health outcomes 64.0% (529) Patient demand for services 57.5% (476) Number of inappropriate visits 56.9% (468) Volume of phone calls to practice 56.6% (467) Quality of doctor-patient relationship 51.5% (425) Patient satisfaction with medical care 45.9% (380) The sum will not equal to I 00%. Each percentage is the percent of physicians who reported "no change" out of the number of physicians reporting for that particular survey question. Potential Impact: Number of Medical Liability Claims. Age was significantly (P = 0.05) related to physicians' opinions rout the potential impact ofTBDS on the number of medical liability claims. Physicians in the 30-39 year old age group were more likely than older physicians to belieje that the TBDS would have no impact on the number of medical liability claims. 145

PAGE 159

Potential Impact: Length of Clinical Encounter. There were no significant differences of opinion related to physicibpractice characteristics. Potential Impact: Patient HealtJ Outcomes. There were no significant diffi f . I ed h .1_, h . erences o opinion re at to p ysici&:uu practice c aractenstlcs. Potential Impact: Patient Dejd for Services. Number of visits per week (P = 0.004) and percent of patients capJated (P = 0.02) were significantly related to I physicians' opinions about the potential pact ofTBDS on patient demand for seiVlces. Physicians with less than I 00 vis ts/week were more likely than physicians with I OO-I99 visits/week or greater than 200 visits/week to believe that the TBDS would have no impact on patient demand for services (63.I% vs. 53.90/o vs. 54.3%). Physicians with I-49% of capitated patijnts were more likely than physicians with zero capitation or greater than 50% capitatioj to believe that the TBDS would have no impact on patient demand for service (61.6% vs. 53.9% vs. 54.3%). Potential Impact: Number oflnappropriate Visits. Specialty (P = 0.014) and percentage of patients capitated (P 0.,) were significantly related to physicians' opinions about the potential impact of TBDS on the number of inappropriate visits. Specialist were more likely than primary lare physicians to believe that TBDS would have no impact on the number ofinappro\ riate visits ( 62.7% versus 48.7%). I46

PAGE 160

Physicians with 1-49"/o of the their patiTs capitated were more likely than physicians with zero capitation or greater than 50% capitation to believe that TBDS would have no impact on the number ofinappropria visits (62.2% versus 53.5% and 52.3%). Potential Im act: Volume ofPho e Calls. Specialty was significantly (P = 0.008) related to physicians' opinions about the potential impact ofTBDS services on phone call volwne. Speciali,s were more likely than primtuy care physicians to believe that TBDS would have no impact on phone call volume (62.5% versus 48.6%). Potential 1m act: Doctor-Patient Relationshi There were no significant differences of opinion related to physici practice characteristics. Potential Impact: Patient Satisfaction with Medical Care. Even though there were no statistically significant differencel of opinion related to physician/practice characteristics, physicians with 50% or Jbre capitation were more likely than physicians with with zero or 1-490/o capi tion to agree that the TBDS service would increase patient satisfaction with medical bare ( 43% versus 33% and 33%). Logistic Regressions SIIIIllll3!Y. lpproximately 90% of physicians were of the opinion that patients desire more infonnahon about health and medical issues and generally desire more active involvement in the medical care decision-making process. A nearly equal percentage (82-88%) weri of the opinion that they would like to see their own patients take a greater role in asking questions, seeking information, and 147

PAGE 161

participating in the medical decision-making process. Both of these opinions were l allh"" \d. h .. CODSlStent y strong across p YSlClanr practlce C aractenstlCS. When considering patient invol Iemen! in decision making for minor, major, and chronic conditions, approximately thirds of the physicians were of the opinion that the physician and the patient should collaborate in making decisions about minor as well as major health issues. Similarly, two thirds were of the opinion that they would like their own patients to take a more active role in the management of chronic conditions. However, for minor health ilsues, there was a tendency for physicians to believe that patients should take a role and for major health issues there was a I tendency for physicians, especially the older physicians, to believe that physicians should take a more predominant role.,early two thirds of the physicians, especially older physicians and physicians with no capitated patients, believed that physicians have adequate time to answer patient qulstions about care and to fully involve them in the decision-making process. l Even though a consistent majori of physicians were of the opinion that, in general, patients desire more in the decision making process, that they want their own patients to take a more aLive role, and that physicians have adequate 1 th I h .. b h t1n1e to mvo ve patients, ere was a greater vanance m p ys1e1ans opm10ns a out t e patient's preference for decision making] Slightly less than one half were of the 148

PAGE 162

opinion that patients wanted their physicians to make decisions for them while an equal percentage disagreed and 10% hJ no opinion. Approximately three fourths (73 j7%) had heard of the 24-hour TBDS service that had been implemented within the coLunity in the two months prior to this research. Of those reporting that they hld heard of the service, 31.3% indicated that they were familiar with its purpose and services while 42.4% reported that even though they had heard about the seh.ice they knew very little about its purpose I and available services. Depending _upon r.iarity and knowledge about the service, 20-40% of physicians held a generally ,itive opinion, 40-50'/o a neuttal opinion, and 15-30% a generally negative opinion. Ph sicians reporting a higher number of patient visits/week were generally more positive about the TBDS service than were those seeing fewer patient visits. As summarized in Table 3.7, physicians believed that the primary reason the health plans implemented TBDS services were for reasons of cost control in response to the interests of the business community, as opposed to reasons related to clinical quality as measured by patient health outJomes, quality of care and malpractice. As summarized in Table 3.8, physicians predlminantly believed that the TBDS services I implemented within the community would have virtually no impact on selected health care delivery parameters associated with lfficiency and effectiveness. 149

PAGE 163

Of all the predictor variables, logistic regression analysis detennined specialty to be significantly related to the greateJ number (9) of physician opinion variables. Specifically, when compared to specialiks, primary care physicians (general practice, family practice, general internal medi4 and general pediatrics) were more likely to: 1) agree that their patients should take J greater role in treating their own minor conditions; 2) agree that their patients slould take a. greater role in self-managing chronic conditions; 3) be familiar with ld knowledgeable about the TBDS services; I 4) have a more positive opinion about lDS services; 5-7) believe that the health plans implemented mos to respond tor of the business community, to improve quality of care and to improve member satisfaction; 8-9) believe that TBDS services would decrease the volume of cbs and the number of inappropriate visits. L a1 di I ed th 1 f d regresston an ysts m cat at ocatton o rest ency was significantly related to five physician variables as well. When compared to I physicians who had completed their residency external to the local community, those I physicians who had completed their residency within the local community were more likely to: 1) be familiar with and about the TBDS services; 2) believe that the health plans implemented TBDS to reduce health care costs, 3) believe that the health plans implemented TBDS to respond to interests of the business community, I 4) believe that health plans implemented rs improve patient health outcomes, and 5) disagree that the health plans implemented TBDS to reduce medical liability claims. 150

PAGE 164

Logistic regression analysis indicated that age was significantly related to four physician opinion variables: (1) decisionl for major health issues, (2) time to answer patient's questions, (3) their patients tlg a greater role in asking questions, and I (4) awareness ofTBDS services. Specifically, older physicians were more likely than younger physicians to believe that: a) dlisions for major health issues should be made mostly by the physician and b) physiaJ. bave adequate time to answer patient questions about medical care. The relatibnship between age and physicians' opinions about their patients taking a greater role in asking questions about their care was nonlinear. Physicians in the 40-49 and t e 50-59 year old age groups were more likely than physicians in the 30-39 and the 60 dover age groups to agree that their patients should take a greater role in asJng questions about their care. The relationship between age and awareness lfthe TBDS services was also nonlinear. Physicians in the 40-49 and the 60 and ot age group reported being more familiar with the service than physicians in the 30-39 and the 50-59 year age groups. Even though there was a statistically significj relationship between age and opinion about TBDS services (60 and over physicians brg more positive) in the group of physicians who were unaware of the TBDS service just implemented in the community, the small number of physicians reporting that they Lere ''unaware" of the service precluded any practical significance to this finding. 151

PAGE 165

Finally, logistic regression analysis determined that the percent of capitated patients was significantly related to two physician opinion variables regarding the potential impact of TBDS on utilization. Physicians with zero or greater than 50% capitation were more likely than physicians with some capitation (1-49%) to believe that TBDS services would have an impabt on patient demand for services and the number of inappropriate visits. Similarly J physicians with a higher number of visits/week(> 100) were more likely tJ physicians with fewer visits/week(< 100) to believe that the TBDS would have an imract on patient demand for services. There was also a linear relationship between %1 of patients capitated and physicians' opinions about having adequate time to answer patient questions. With increasing capitation, there was an increasingly higher percent1e of physicians agreeing that physicians do h d . I not ave a equate tmle to answer patient questions.

PAGE 166

c I TER4 QUALITATIVE .METHODS AND RESULTS In1uction Qualitative inquiry (key inteJViews), was used to explore more in depth physicians' opinions about: I) Co,sumers/patients being more informed and taking a more active partnership role in Health and medical care decisions. 2) The impact they believe a more informed, actLe consumer/patient will have on the quality of the doctor-patient relationship, clinicj efficiencies, health outcomes, patient satisfaction, health care utilization, healtJ care costs and malpractice liability. 3) The incorporation of telephone-based decisiot support (TBDS) services for patients/consumers into the care manage, ent process. 4) The impact they believe the incorporation of TBDS in the health care delivery process will have on the quality of the doctor-patient relationship, clinical e ciencies, health outcomes, patient satislilctioo, health care utilizatioo, heal, care costs and malpractice liability. This chapter will describe the methods, analys s, and results pertinent to the qualitative portion of this study. 153

PAGE 167

Met ods This section describes the instrumentation, and process used in conducting the key informant interviews\ Sample Selection Patton (1990) has pointed out that, "Qualitative inquiry typically focuses in depth on relatively small samples, even sLgle cases (n=l), selected purposefully .... The logic and power of purposeful samplg lies in selecting information-rich cases for study in depth" (p. 169). In contrast probability sampling used in quantitative research, sample size in qualitative res,ch is not dependent upon probability statistics and power calculations to compute sample size. Instead, "The validity, meaningfulness, and insights generated from qualitJive inquiry have more to do with the I infonnation-richness of the cases selected and the observational/analytical capabilities I of the researcher than with sample size" (Jatton, 1990; p. 185). Because the purpose of this aspect of the study was to obtain lonnation-rich "thick description" of physicians' opinions about the key researlh questions, a fonn of purposeful sampling known as maximum variation sampling wL used to select the study subjects.

PAGE 168

"Maximum variation sampling occurs when one seeks to obtain the broadest range of information and perspectives ol the subject of study. Guba and Lincoln (1989) claim this is the preferred stratel for qualitative inquiry" (Kuzel, 1992; p. 3 7). This approach is one of deliberate!{ seeking out diverse variations and coounining important common patterns in their experiences. Initial data are then elaborated upon and given more breadth and strength by teliherately searching for confinning and disconfinning cases that will ... suppoi or challenge the investigator's understanding ofthe topic of study" (KuzeL 1992; p. 41 ). When conducting maximum variation/confinning-discoofinning sampl subject selection continues to the point of saturation or redundancy with 12-20 Jata sources recommended (Kuzel, 1992). Therefore, 21 clinicians were seljcted for key informant interviews from the domain of the physician population to be studied ---all primary care physicians (approximately 859 family practice, generalists, internists,) and specialists (approximately 1,306) that were memberl of either the Rochester Independent Practice Association (RIP A) and the Roclester Community Independent Practice Association (RCIP A). Seven key informLt physicians each were selected from those f h d I ali h d b a1 groups o pnmary care p ys1Cians an speet sts w o reporte emg pos1t1ve, neutr or negative about the TBDS service. 155

PAGE 169

"Key informants are individuals io possess special knowledge, status, or communications skills, who are willing to share their knowledge and skills with the researcher and who have access to or observations denied the researcher" (Goetz & LeCompte as cited in Gilchri11992). They are "trustworthy, observant, reflective, articulate, and a good story teller" (Johnson as cited in Gilchrist, 1992). Snowball sampling 1992; 1990) was used to select the physician key informants who met the above criteria. r is, using the key informant criteria and the requirement to identifY physicians who held preconceived opinions (positive, negative, neutral) about the TBDS setv:ice, +live director and the physician president of each IP A were asked, "Who should we t81k to?" The IP A executive directors and I physician presidents began the process of identifYing initial key i.Dformants who in turn directed them to additional physicians wJo were thought to be "information-rich" informants. This iterative process continjed until the full complement (21) of key informants was obtained. Physician Opinion Interview (POI) Guidel Following the guidelines for quali ative interviewing, the standardized openended interview approach (Patton, 1990) r used to conduct the key informant interviews. The questions comprising th,Physician Opinion Interview (POI) were extracted from the Physician Opinion Survey (POS). The questions were formatted to

PAGE 170

accommodate the interview process and to acquire more in-depth information about the topic of interest. Like the POS, final questions for the standardized open-ended interviews were developed by the research team in collaboration with the executive and physician leadership from RCIP A Jd RIP A by following the guidelines of Pieterzak et al. ( 1990) for the developmlnt of new measures. The final POI guide is illustrated Appendix B. The survey includes a profile of demographic data and physician practice characteristics. The first section of the POI asked physicians their opinions about informed and actively involved patients in the health and medical care decision-ma.lfutg process. The second section of the survey asked physicians their opinions atut informed consent and telephone-based decision support (TBDS) services as well as some open-ended questions regarding the key challenges and benefits of implementLg TBDS services. Interview Structure and Administration Upon arrival at the scheduled tim and place for administering the interviews, the following agenda was used with each physician: a) complete introductions; b) review of the purpose of the interview, the interview process including tape recording, what will be asked in the intelews, how the information will be used, and the voluntary, anonymous, and confidentt nature of the interviews; c) answer any questions or concerns and obtain consent; d) conduct the interview and 157

PAGE 171

e) conclude appropriately. Physicians were given the opportunity to decline the interview, stop the interview at any poinl or refuse to answer any questions with which they were not comfortable. The face-to-face structured inte ews were conducted by the investigator during the week of August 22-26,1996. The interviewer took notes during the interviews and all interviews were audio 1 aped. Audio tapes and notes were anonymous and confidentialidentified inly with an identification number. The avernge length of each interview was rtely 60 minutes. AD interviews were transcribed, coded, segmented, and anal}lzed as detailed in the following analysis and statistical methods section. AnJ yses All audio-taped, standardized op ded interviews were transcribed, coded, segmented, and analyzed using methods lr qualitative analysis (Patton, 1990; Crabtree & .Miller, 1992). Methods of qualitative rs generally encompass three primaiy functions: I) preparing, organizing and ceding the data in order to identify key data I areas and to facilitate future data retrieval and analysis; 2) segmenting the text data into analytically meaningful components; land 3) interpreting the data through self-158

PAGE 172

immersion in retrieving, sorting, making ronnections, examining patterns, and testing interpretive models of the coded and segptented components (Crabtree & Miller, 1992; Patton, 1990; Reid, 1992). Transcription, Coding and Segmentation All audio-taped standardized ope -ended were transcribed into data files using standard word processing. All text files Lere converted to ASCIT format in preparation for use with NUD-IST 4 (QLntative Solutions and Research Pty Ltd. I 1997) text-analysis software. Using NUID-IST 4, all text files were imported, numbered, and printed for initial coding. The first step in coding was the development of a template in the form of a codebook used fur organizing and interprg text. A codebook fur tbis study was created by using a template analytic approach, ". . beginning with a basic set of codes based on a priori theoretical understandinks and expanding on these by readings of the text" (Crabtree & Miller, 1992; p. 95). The beginning set of a priori codes was based on the investigator's I understanding of the issues and initial review of text files and field notes. Using an I iterative process that encompassed multiple readings, interpretation, and recoding of the text files, the primary investigator indlpendently expanded and refined the initial set of codes. Collaboration and consensJ among the two primary investigators I 159

PAGE 173

determined the "final" codebook to be used in text analysis and interpretation. All text files were computer coded for future retl.eval, sorting, and analysis using NUD-IST 4.0. Data Analysis and Codebook The goal of data analysis in qualitative inquiry is to provide infonnation-rich descriptive data ("thick description") by rbing salient patterns, themes, and categories that emerge from the data. Iligenous typologies are classification systems comprised of verbal categories used by the key infonnants. In contrast, analyst conslructed typologies are classification comprised of patterns, categories, and themes constructed by the investigators (Patton, 1992). "Negative case analysis involves addressing and considering altitive interpretations of the data, particularly noting pieces of data that would tend to Lme the researcher's reconstruction of reality" (Erlandson, Hartis, Skipper, All+ 1993; p. 121). In this study, particular attention was given to negative case analysis and to the development of both indigenous and analyst-constructed typolbgies in developing the codebook and in analyzing the data. Using an iterative process ofmultip1e readings and interpretation of the text files, the initial analysis was expanded 1 refined by the primary investigator. The outcomes of the iterative analytical process included: a) refinement of coding 160

PAGE 174

constructs, b) construction of a profile summary for each case document (interview), c) construction and import of a case ,table for linking quantitative and qualitative interview data, and d) question-by-que,on analysis delineating direction or how the key infonnants felt about a given topic negative, neutral) as well as why they felt the way they did. The final codebook used for text analyses and interpretation is summarized in Appendix C. Inter:pretation The goal of interpretation in inquiiy ... involves going beyond the descriptive data ... attaching significancj to what was found, offering explanations, drawing conclusions, extrapolating lessoJs, making inferences, building linkages, attaching meanings, imposing order, and 1ealing with rival explanations, disconfuming d d 1 f I. th abil. f cases, an ata lffegu annes as part o testmg e VI tty o an mterpretat1on .... The emphasis is on illUmination, understlding, and extrapolationrather than causal determination, prediction, and generalizatlon" (Patton, 1992; p. 423-424). For this study, data analysis and lerpretation incorporated a modified furm of I member checking to enhance methodological soundness. "Member checking provides I for credibility by allowing members of stake holding groups to test categories, interpretations, and conclusions" (Erlandtn et a!., 1993; p. 142). A modified furm of member checking was conducted by pro 'ding a copy of the preliminary analysis and

PAGE 175

interpretation to three independent phys cian reviewers who were not a part of the study population for critique and feedba k. One of the physician reviewers was David B. Nash, M.D., M.B.A, Associate D1 and Director of Health Policy and Clinical Outcomes at Thomas Jefferson Universir Hospital and Jefferson Medical System in Philadelphia, Pennsylvania. The second !physician reviewer was Mark Earnest, M.D., Internist at the University of Colorado Health Sciences Center General Internal Medicine Clinic and doctoral student in lhe Health and Behavioral Sciences Program, University of Colorado at Denver. physician reviewer was Zach Gemarg, M.D., Internist, ChiefMedical Officer of a 140,000 member mixed model HMO and President/ CEO, Gerbarg and Associ1 In order to guide and to provide a more uniform, standardized approach to the checking process, each of the physician r were asked to test the data analysis, interpretations and conclusions by answ,ring the following questions: I) From a physicians' viewpoint, did you detect a different pattern of key themes and issues emerging from the data? If so, they? 2) From a physicians' viewpoint, were any key themes or issues missed? tr so, what were they? 3) From a physicians' viewpoint, did you arrive at different int rpretations and/or conclusions about the data? If so, what were they? 162

PAGE 176

Minor adjustments and revisions from the independent physician reviewers were incorporated into the final analysis. Collaboration, and consensus among the two primary investigators guided the "final" analysis and interpretation. Results In the information that follows, dtcriptive characteristics of the interview subjects are provided. Each section beJs with a specific interview question followed by a balance of analysis and thick d4on utilizing representative participant quotes to illustrate the pertinent themes that emerge through in-depth analysis. Each section closes with a summary of the .key themes lthat emerged for that question. A llll!ior purpose of this analysis las to present and organize key infurmant responses in a way that provided more in depth insight to each question as well as a more clear overall pattern of emerging themes and issues. A concluding synopsis of common overarching themes that emergl across all questions is provided. Subjects As previously described, criterion-based, maximum variation snowball sampling was used to select 21 (20 males land I female) key clinician informants from the approximately 2, 000 IP A physicians bls studied. Relevant subject characteristics are presented in Table 4.1. 163

PAGE 177

164

PAGE 178

The mean age was 46.6 years (SI = 8.3. Fourteen completed their residency Rochester. There were 13 primary care hysicians (internal medicine= 9, family practice = 2, pediatrics = 2), seven specia.Ity physicians (ophthalmologists = 2, gastroenterologist= 1, orthopedic surg,n =I, nephrologist= 1, allergist/immunologist = I, clinical psychiatrist = 1 ), and one clinical psychologist. Opinions About Patient Involvement in Medical Decisions The first series of questions. ( Q 11) were intended to explore clinicians' opinions about patient involvement in hfh and medical care 1. What do you think about patients w ling more information regarding health and medical issues? Twenty out of 21 clinicians had a ery positive opinion about patients wanting more information regarding health and medical issues. Physicians placed major emphases on: a) better decisions leading Jo improved health outcomes and quality, b) the doctor -patient relationship, and c) Lnsumerism and patient preference. Better decisions, outcomes and Physicians said that a more informed consumer/patient would be more likely to make better decisions generally resulting in superior quality, more efficient care and iproved health outcomes. Generally I think that's a idea. I think informed patients do take better care of themselves. If they understand what's going on, they make bettd decisions. (Pediatrician) 165

PAGE 179

In my experience thus far, it makes the care a lot easier. When it's one-sided it mhlces it very difficult to convey the importance of taking medications, following a specific treatment program, espetially if the patient doesn't ask I questions or doesn't get involved. Because if they're having any difficulty with the prbgram they may not be that likely to voice that and will just nbt follow the program. I tell patients outright from the front tHat I'll always offer them the best treatment plan, but if thete' s something about that treatment I plan that doesn't jive with what they expected, or doesn't fit with their life style, we do. find a middle ground. I find that a lot easier, it puts the patient at ease, and it allows them to speak to me. And thai has a lot of pluses, not only in just management of the but, I think it allows better rapport between the patient and tre physician so they're likely to tell me something if something's not working or something goes wrong. (Internist) I think it's great. From my perspective, the more patients know in terms of their medicalibowledge base the easier it is for me to take care of them because they understand the rationale, the treatment. And it makes thy job easier and it makes good medical care easier. (Internist) In general I think it's a gopd thing. Patients that are better informed are more likely to have a good result and do well. (Gastroenterologist) Although most physicians supported and endorsed patients being more informed, one physician expressed a concern that more rormation may lead consumers to wrong conclusions. He said that he didn't believe one could generalize that a patient having more information was a good idea. The concern is that they use a little bit of information and think they understand it aAd come up with totally wrong conclusions . .. like with a lot of things, a little knowledge can create big problems. Then 166

PAGE 180

you have to spend a lot of time undoing what they think they know. (Internist) However, the same physician, fully supported patients desiring more active participation: j I tend to find that as the c1atient gets more involved in making decisions, things usually run more smoothly. (Internist) Doctor-patient relaJionship. Not iy did physicians think that patients wanting more infonnation about health and medical issues was positive, most indicated they actively encouraged information seebg by patients. They saw themselves as I 11teacher11, 11facilitator11 of decisions, and the most reliable source of information for patients in the context of the doctor-pati1t Patients need to be educatL as much as possible. The word doctor means (Orthopedic Surgeon) I think one of the ways rl useful to my patients is being able to provide informatiob in an appropriate context. (Internist) I like to stay away from a paternalistic practice and I like to shoot for a collaborative modeled medical practice as much as can be done. (Fanilly Practice) I don't see my role as pateLalistic. I see my role purely as an advisor and to give niy best possible professional opinion. (Internist) J I find myself frequently g to a patient. I had a conversation yesterday wit a man who had been recommended a surgical p+cedure by a specialist surgeon. And I asked him what the surgeon explained to him about 167

PAGE 181

the potential risks. He e lained the benefits, the patient could explain the to me but he knew nothing about what the potential risks 9r what the failure rate was. And I said, 'well, I think that's more information you need from I that surgeon, and certainly it's information I would want to know if I were in your pdsition before the procedure itself' I find that comes up quitJ a bit. Getting patients to go back to their specialist and sayl 'I need more information to make a proper decision here.' It's interesting how frequently they come to this office after donsulting with somebody and wanting more information saying that they didn't get enough. (Internist) Consumerism and patient preferlnce. Informants acknowledged that health and medical care have entered into an erl of consumerism characterized by patients desiring more information about health ld medical care issues and often desiring more active participation in health and ,dical care decisions. They pointed out that physicians will need to be prepared to work with the strong and growing trend of health care consumers who might have a strong preference for more health and medical information. It's clearly the trend where things are going, more consumerism, and I think its what we all see in the office. (Internist) It's a natural part of having a more educated population. It's natural and it's an empbwering thing that's happening. I'm not surprised it's I think if it's done-it will happen. The question \is whose going to provide the information. (Family Practice) I just think that the days the doctor being dogmatic I and saying, 'Here's the pill, you take that and you'll get better are over.' People arb understanding, listening to their own bodies, their owh environment, trying to make 168

PAGE 182

their own decisions with a little bit of guidance from the physician. (Internist) I have no problem with tHat. In this day and age with I modem technology I think it's perfectly appropriate for them to want to have mol information to base their decisions. (Pediatrician) Summary. Overall, clinicians held a positive opinion about patients and consumers wanting more information abJut health and medical issues. They acknowledged that there is a growing coLmer trend characterized by patients who have a preference for more information rlther than less. Informants viewed themselves as the best source of health and medical lormation for their patients. Many conceded that with an increasing number of informed patients, their communication style would need to be more predominantly ''teacher" and "facilitator" rather than paternalistic decision maker. Physicians Jndamentally believed that a more informed patient would result in better health and ledical care decisions leading to improved quality of care, health outcomes, and dodor-patient relationships. 2. What do you think about patients wanhng more active participation in the medical care decision -makingprocess? Twenty out of 21 clinicians positively endorsed a more active patient in the medical care decision-making process. PJysicians most frequently emphasized issues related to the dynamic of the doctor-patieht relationship, recognition of the variation in 169

PAGE 183

patient preference for active participation and the subsequent need for physicians to be I sensitive to varying patient preferences or involvement in the medical care decision-making process. Doctor-patient relationship. Ph sicians once again noted that more active involvement of patients in the medical dJcision-making process is a significant trend in today' s health care, a trend they must b1 prepared to deal with in the doctor -patient relationship. They believed that generally more active involvement by patients in the doctor -patient relationship would lead J increased patient compliance, and therefore satisfaction with care. It's a real phenomenon. If-egarding the importance of patient participation in the diagnostic and therapeutic process, I think it's probably the most significant change in the dynamic of the dodor-patient relationship. Whether physicians like it or not, they think it's important enough, isn't really the issue. It's that it's a reality that needs to be addressed. (N I think if they take a personal interest in their care rather than relying on the to do everything for them that they're much better, first 0f all, to bring things to your attention that you might otherwise have missed; to be compliant with their therapy because they understand. (Gastroenterologist) I think that when patients get to own more of the decision making process they're mbre comfortable with whatever I the results might be of their clinical choices. I think that a I lot of the chagrin and the feelings of disaffection that patients have of their medical care is when they feel like they were led down some garden pa$, some complications occurred and if they had been more !actively solicited or set up that 170

PAGE 184

path, I think they're more willing to deal with the consequences. I give them active I involve my patients actively. They're given alternatives to make a decision. I put the ball back in their court. I to find that problem, explain it to them, give them alternatites. (Orthopedic Surgeon) One alternative to enhancing doctor-patient relationships was explained by a clinical psychologist. He pointed out that, increlsingly, he assumes the role as one who facilitates a better understanding and communications between patients and other physicians. Patients that I work with, part of my role that I serve and play is to help them become more active with their physician. It works out better, overJ.n, in terms of clients otherwise, they don't understand why probedures are being done, they don't understand the rationale. The more they understand the better. It's not written down that that's what rm supposed to do, but that's ultimately what I try to do. The doctors here are very willing to go along with that as well. I can give you an example of one patient who felt lack of control. Her personality was one of needing a lot Jf control. She's in a high power job and used to a lot of contrbl. Now she's sick and the control is out of her hands. She getting really angry at the doctors for prescribing all these mediations. 'What are these for?' she didn't understand it. And \after doing some work with her she was able to articulate this to the doctor and the doctor was able to give her contrbl or some say about when they'll increase the dosage, whed they won't. And they're working more cooperatively. Ultuhately, she's much happier. That's what a lot of it is. I think :doctors, often times, medical doctors don't realize that Jl>atients don't understand what they are saying. It really takes burturing a person through the process. I went through this yesterday with another patient. She was much more willing to tak9 the kind of medication and under stood why they wouldn't give her the Tylenol with codeine in it by the end of our discussion. She was much more ready to be a partner. To, kind of, 1eally take part in the treatment. 171

PAGE 185

Patient preference. At the same time, however, it was noted that patient involvement in medical decisions is somewhat dependent upon patient preference. That is, some patients clearly do not wJt to be involved in the decision-making process. Instead, they would rather theJ physician make the decision for them. Interestingly, several physicians noted tJt they were not comfortable with patients who preferred that the physician take a lore paternalistic, directive approach. I It bothers me a bit _when get that line from the patient, I propose choices to them and they say, 'well, you're the doctor, tell me what to drl.' They just want you to tell them. I don't like that. I feel with that. Patients like to be told what to do rather than participate in decision making. Being told what to do and feeling comfortable that the doctor is able to tell when not feeling fully equipped to participate in those ch ices. But I think it is my job to help them be comfortable with participating rather than just telling them what to do. Some perhaps, aren't even capable of doing that. Some patients, for whatever reason, whether it be their illness or reasons, some patients just aren't able. You can try and a proxy who can participate for them or you can just plan role and make the decision that you think, is the best without their input. You have to work with what you have to wdi with. (Internist) Physicians also acknowledged that for those patients who do prefer to be more 1 1 d. h d akin I :fi active y mvo ve m t e eetsxon m g process, patient pre erence or consumer choice is paramount, even if the physicii tends to disagree with the patient's decision. Something that patients want is to know my opinion as the medical professional. I really think is best even though there may be a range and they can have some choice and they 172

PAGE 186

can ask questions about ach choice. They want to know what I think is the best option bd what I would recommend doing. So I think this all comes into being part of the proper communi cation about treatment enabling the patient to make an active choice. And when a disagrees with me, even about accepting any treatment at all, I explain to them that I understand that and that I'm not to take away their autonomy, that that is paramount. They iwml always have the right to choose not to accept treatment. And after I've explained to them potential benefits risks of treatment, nontreatment, then I know that they can truly make an informM decision and that my conscience rests I easy because they've made a decision with the best available infonnation. (Internist) I'm all for it. They're the one's who are taking the medication. They're the one's who getting the tests. They have the right to make decisions, irrational I may think they are and however frustrated it makes me when they come in with their herbal preparations or desire for this test or that test. All I can do is tell them what Iithink is the right way to go and why I think that, and if my reasons are valid to them, which they usually are. And sometimes reasons are valid to me and they con vince me that we should choose another course. It depends on the severity of the 4n. (Internist) Summary. Overall, physicians heltl a positive opinion about patients and con-. .. I. h ed'cal d makin sumers wantmg more active partictpatJ.on m t e m 1 care eCJ.Stong process. Physicians contended that a patient who was more actively involved in medical decisions would be given a greater sense oi control and self-confidence culminating in greater compliance with agreed upon diagnostics and therapeutics. This would result in higher overall patient satisfaction with lcare. At the same time, it was recognized that patient involvement in medical deciilns is somewhat dependent upon patient preference. That is, some patients clearl}l have a preference for more involvement in 173

PAGE 187

medical decisions whereas others do not want to be involved. Additionally, for those patients who clearly prefer an actle involvement, physicians generally respect consumer choice, even if the physician tends to disagree with the patient's decision. 3. What do you think will be the impacl on health status/outcomes o(those patients who seek out infonnation and more hcnvelyparticipate in medical decision making (better. same. worse)? Sixteen physicians representing t)[oth primary care physicians ( internal medicine, family practice, pediatrics) and specialists (gastroenterology, ophthalmology, nephrology, allergyfunmunology, and pr) believed the impact on patient health status and outcomes would be better. Two (orthopedic surgeon and pediatrician) believed the outcomes would be the saml, two (internists) believed that patient satisfaction with treatment would increaJe regardless of the outcome, and one (internist) believed that the impact on he1th status/outcomes would be dependent upon the disease. I think those patients shoti.ld be healthier in the long run, because they will and, as I said earlier, comply with their recommendatiops better. If they understand better that a certain medicine has been prescribed, and the importance of it, then they would be rhore likely to take it than if they didn't know. 'The said do this, but I don't think it helps me very much. I don't why he wants me to do it.' I think in that situation theYt're much less likely to comply with the medication. If perhaps, are some side effects, and they understand that the side are something they have to put up with for a greater goo4, they would be more likely to take it. Whereas if they don't really understand, 'that damn stuff gives me 174

PAGE 188

diarrhea' or 'it makes feel tired, what good is it? I'm not going to bother with it.' So, I it would help them to comply better and then they should be ih better health. (Pediatrician) Th "lh 1 .. "hd"" akin e more active y t ey m t e ectston m g process with respect to their health care the more invested they are in fully participating. When patiehts feel more invested in the process they are more likely to with whatever decisions are ultimately made .... I think anybody is more willing to do something that they participate in. If the patient feels as though he or she is being forced to do something that have had no input into, then they're probably less likely to comply with whatever the recommended regiment is. (AIIergiTologist) The best example I o: off the top of my head, would be to look at diabetic control, which I think there's lots that shows when a person is actively fngaged, and takes responsibility for their management. That you can achieve a lot more than by blinded directing of a patient towf,rd achieving something that they're not even sure what it is. (Intjst) I think they are more to be compliant, not only with appointments but with meldications and with sudden changes in direction ofhealth care. If something goes wrong, if they have an understanding this is a possibility, or have an I understanding why this is happening, they won't get frustrated and feel like they're being excluded from the process. I think, overall, it's going to have a very positive impact, much better than the old paternalistic approach to health care. (Internist) Th I ed th f =-.t: b ere was some very mmor concern exprss at nnsmterpretatton o uuormatton y patients may lead to poorer decisions and thus worse outcomes. Probably it will be much bltter. I think occasionally you may I find a person who will misinterpret some of the information or will direct some of the ormation they get in the wrong way. (Internist) 175

PAGE 189

In other illnesses it may be a hindrance to appropriate care. Things like chronic syndrome, or fibromyalgia's where the request for more infopnation leads them to places that physicians' traditionally liave not felt comfortable or have not found any clinical utility . . (Family Practice) Patient satisfaction. Even in thTe few physicians who believed that clinical and/or functional health outcomes would be equivocal, they acknowledged that a more active, informed patient would be a morl satisfied patient even if the outcomes were not as positive as hoped for. I think it's going to be positive. I've seen some of the research I that's come out so far it looks like patient's that are actively involved in their health calre, not only feel more satisfied, are more happy with their heJith care overall. (Internist) Ihink bin h 'dl. .. p. t one t g t at It oes 1s It mcreases satistactiOn. at1ents feel more satisfied with outcomes. And even if something doesn't go to wishes. And people who spend the time and effort thi.nkiDg and teaching themselves and gaining knowledge about illness apd health to be an active participant in decision making begin to 1eam, 'hey, this is not black and white exact science.' There's a lot of judgment, interpretation and there's a lot of dealing with probabilities, and dealing with uncertainty so that decisions are often that, well in 60% of the time this treatment works but 40% [doesn't, or 70% doesn't. And they come to appreciate that so that f.hey're not as likely to be disappointed, upset, or have negative feelings when things don't work out. You just come back and you'll go through the process again. So I think that people who are more 1active in their medical decision making, once again, are going to happier. More satisfied with outcomes and less disappointed in nbgative outcomes. (Internist) Medica/liability. Physicians belilved that increased patient responsibility and I accountability in the medical decision-making process would lead to improved patient

PAGE 190

satisfaction and health outcomes resulting in less potential for litigation-even if the outcomes were not as positive as hoped for. I think, in general, it will robably decrease if they are more involved in the decision making. But I don't think it goes I away because, unfortunately, they can always say that they didn't understand the expianations, and you're still the doctor, you're still the expert. Sq I don't think it will go away. But I think, I've always found tpat if you work with the patient to gether that there's much lfss possibility of them saying, after the fact, 'how come, this,! this and this', if they see that they're involved in helping you re that decision. (Internist) Patient preference was recognized as a factor influencing outcomes. As previously described, some patients d, more infonnation and want to be involved in decision making whereas others do no Also, it was acknowledged that there may be a selection bias in those who seek out more information and who actively participate in medical decisions. That is, such individuals may already have this type of coping style and will naturally do better the more information and the more I involvement they have in health and medical decisions. My bias would be to say J.uer, but I also think that people that seek out information have a coping style that information will be beneficial. But people that won't seek it out given the information, it will onl make it worse. I think some people are information seekers add some people are deniers. Those I deniers have a hard time if given too much information. They get overwhelmed. If they lare seeking out information I think it's probably better to givel it to them. I don't think it's so hard and fast and the people wlio are deniers, if you do some work with them, they come aroJnd eventually to use information, but they need a little workf They take in information as they're ready to take it in. You can give it to people, if they're ready. 177

PAGE 191

They'll ignore it or it 'Will increase their anxiety, or just not pay attention to it if they're 1ot. (Clinical Psychologist) Outcomes would be better because that's the personality of the I person. If you had the same person who wanted to seek the information, I don't knoJ whether you can do that, but generally the person who wants to lseek out the information is one who wants to be more It would be more likely, that they're involvement in learning about different quality of life issues, that things may change. If thJy tend to be more active in their care. But is it because they the information or is it the personality of the person? I don't know what would be affecting that more. (Ophthalmologist) Summary. Providers believed that those patients who actively seek out -S:. d h 1 .I. edi al d . "11 h b huormat1on an w o more act1ve y partlapate m m c ec1s1ons W1 ave etter clinical and :functional status health outjmes. They thought this was particularly true for patients with chronic illnesses. The lechanism for better health outcomes and reduced potential for litigation was relat1 primarily to improved understanding, . d b I. I din d li d part1c1pat1on, an uy-m to treatment regunens ea g to 1mprove comp ance an overall satisfiu:tion regardless of wheth, the outcomes were negative or positive. Patient preference was again acknowledied as a fitctor influencing outcomes. Some patients desire more infonnation and want to be involved in decision making whereas others do not. Health outcomes will be iLproved if physicians are sensitive to patient preference for active involvement and jtch their response to that preference. If physicians fail to match their response to preference, then health outcomes and overall patient satisfaction may be diminilhed. It was pointed out that selection bias 178

PAGE 192

might be contributing to better outcomes of those who more actively participate in medical decisions. J 4. For a minor health issue (sprains. 1ains. cold/flu $Jl!11ptoms). who would you want to make decisions about Please provide the reasoning supporting your opinion. Explain how this would i#tpact: clinical efficiency. the doctor patient relationship. patient self-eQicacy. healtH outcomes. health care costs. and medical liability. Table 4.2 illustrates provider's opinions about who they would want to make decisions about treatment for a minor ( sJrains, strains, cold/flu symptoms) health issue. One internist reported somewhere between "only your patient" and "mostly your patient with input from you", depen
PAGE 193

Clinical efficiencies and tpproprateness. Questions about clinical efficimcies and appropriateness generated the most Physicians emphasized that there would be a vecy positive impact on clinic efficiencies for patients who participated in decisions about treatment for minor heal issues. That is, the number of medically inappropriate visits, the volume of unne ssacy phone calls, and clinical encounter length would all be reduced over time. Anything that's minor is to be self limited and therefore would not require muca put from a physician. In these times of physicians, especially s ecialists, seeing more patients with more severe diseases, t they have in the past, ultimately we will have less time to deal patients with minor complaints. complaints the more effici nt will be our utilization of our time. (Allergist/Immunologist) As far as a demand man ement issue that, certainly, we've been talking about for years. e need to change patient demand or it's going to be impossible to ontrol costs. A well-informed patient who demands care approp ately, that's going to turn out to be a good thing. (Internist) Probably it would have a ositive effect on clinical efficiency. Hopefully, there would be less time spent in the office, lower bills and so on and so because mostly patients can do that kind of stuff at home. lot ofthat from the practical point of view becomes easier for tile patient, saves them an office visit, saves the time in me for thf office for other more serious problems and so forth. So it's probbly good. (Ophthalmologist) I think it's tremendous. I t in terms of giving patients the right tools to understand things understand what's serious and what isn't, you can avoid a lot o the inefficiencies that occur as a result 80

PAGE 194

of mundane complaints co g in to an office and clogging up an offices precious time and aking away from patients who have much more serious complaints t at necessitate direct patient contact. (Internist) I think it probably would rfduce, to some extent, the number of office visits, and, the number of hospital emergency visits as well, that are inappropriate. I think in that, if patients learn about illnesses and can re'*ember that, and use that the next time that illness arises then theYj don't have to go through the same scenario all over again, su e, it should make it more effective -things that really are self "ted, and could be handled at home. One of my pet peeves, in articular, is sprains and strains going to emergency rooms, and sitt g there being untreated, no first aid, a patient with a sprained e sitting there with no ice, no ace, the ankle is down on the floor getting worse for two or three hours because they're not going o be seen early in the emergency room, they're put down on the b ttom of the triage list. It's a total waste oftime. (Pediatrician) To be a little nihilistic, for r lot of these things it doesn't matter much what particular you choose. So, I guess, from a notion of resources I'd spend more of my time advising people about matters of gr ater importance that would have a greater effect on their heal h outcome than to advise them about relatively minor matters. amily Practice) I think that if you have pe pie who can basically do first aid with minor illnesses it does help with office efficiencies in that it enables the internal office to keep the schedule open. The maintenance visits lthe chronic illnesses, a large part of what we do, the physical e s that people need to help maintenance in that regard And then the urgent day to day illnesses that come up than to have them clogged by people who are coming down cold symptoms or early flu symptoms or minor strains that they clm treat themselves. (Internist) 181

PAGE 195

Doctor-patient relationship. Ph sicians believed that patient involvement in minor health decisions would contribute o an enhanced, more positive, collaborative doctor -patient relationship. I would hope it would m e it better. I think this whole idea of the doctor being on a ped stal and mandating what the patient has to do, I think that wa a bad relationship. I just think it's the give and take and a far m re pleasant relationship to do it this way. (Internist) Probably an improvement because for minor little things that they could have taken care of emselves. For me to ask them to come in here and wait an hour i my waiting room would probably ruin my doctor/patient relationship. Not ruin, but they'd be pissed. I would be too. (Family Pr ctice) Even though physicians had a positive in ,Jination toward more significant patient involvement in making decisions about treatment for minor health issues, they also viewed patient encounters for minor issu s as an opportunity to build the doctorpatient relationship. The relationship between patient and a physician is, I trust my physicians' style. I like the way he approaches these problems. You don just talk about their cold or their sprained ankle when they come in. You talk about other issues, not just their lillness, but what's going on in their lives. The more contact you have with a patient, the more you develop a and a trust, and a knowledge base. So I look at encounters over minor illness as beneficial in those ways as seri us illnesses in building on our relationship in some fashi n. (Internist) 182

PAGE 196

But I think they're getting more than just the treatment for their minor health issue w en they come in. If they're anxious about their illness, what e er is bothering them, then you can put their anxiety to rest. d that encourages the doctor/patient relationship, it enforces t t. (Internist) Although physicians were predo antly positive about patient involvement in decisions regarding minor health issues, t ey expressed two concerns that patient involvement may have a less than optimal impact on the doctor-patient relationship. First, some were concerned that, while gi g the patient a more prominent active role in the decision making process the physi an might appear to be a "wet noodle" or "indecisive", ''uncaring" or "unconcerned in the eyes of the patient. I think it's definitely a plus It all depends -lots of times you'll get patients when you say, 'well, what do you want to do?' and immediately they're g ing to say, 'well, you're the doctor.' I think that puts them off. I think that's kind of a poor way out of it. If you couch it prope ly, where you don't get too -a lot of people are busy, they wan to come in and get out and get back to their life. I think if you ouch it in a sense that, okay, this is what the problem is. I thi the best way to deal with this is to do this. What kind of ork do you do. Do you have any responsibilities that could ave impact on? 'Yes, I have to do this, that, or the other and that could really be a problem if I do that.' Well, here arr some of the other options, it may effect its ability to it may effect the outcome, but it may allow you to do those! things. 'Well, I'm not interested in doing that if it's going to affect the outcomes,' or, 'okay let's do that and maybe I can back and see if it's progressing too slowly, we can try somethirg else.' I think it has definite, positive impact on the relationship, as long as you don't look like a we noodle, like, 'well, you make the decision.' I think it's got t be the type of thing where you really 83

PAGE 197

have to approach it from consumer point of view. This is the best option, this is the w rst option, what's your lifestyle? (Internist) Patient preference. Another fear expressed by some physicians was that many patients may not prefer to take on an a ve decision-making role, even with minor health issues. If forced to do so, there ay be a negative impact on the doctor-patient relationship and an accompanying redu on in patient satisfaction. That may be a little bit p ient dependent. I think one of the things that we're seeing i that many patients do want to take more of an active role in g for themselves, or their family members, and they woul like this. However, I think there are some patients who still p efer kind of the paternal physician. They really just want to I ve everything up to the doctor, and if the doctor's trying to them to become more active in decision making, and they're not terested in that, that's an imbalance. They may resist that and e less satisfied. They may feel the doctor's asking them to dttoo much. (Pediatrician) Those people who are us to having physicians tell them what to do in a situation with health problem would probably be less satisfied with a sy tern or an approach that causes them to be a little bit more self. reliant, but I think most patients would not have a problem (Allergist/Immunologist) There will be a lot of pati ts, we know this to be true, patients who want to be told wha to do who are very frustrated by having the options presented to them. (Clinical Psychiatrist) Patient satisfaction and Physicians believed that a more collaborative approach would provide p lents with more confidence and control in managing a minor health issue and woul generally result in higher patient satisfaction. 184

PAGE 198

I don't know what the real significance is, but certainly it empowers the patient a bit and I think most people feel positive about that and gi fS them more control over their own health. (Ophthalmollgist) I think that improves as I said earlier, because informed participating patients feels that theyre empowered and that they're in which is what many people want. (Internist) Health outcomes. Many physici s noted that patient involvement in decisions about treatment for minor health issues w uld have neither a positive nor an adverse effect on outcomes. Most physicians hel the opinion that most minor health issues were self-limiting, would run their natural course, and that patient participation in treatment decisions would have no signi:fi ant effect one way or the other on clinical or functional status outcomes. I don't know that health o comes are at all effected in the minor category. I think that my ew of these categories is that people, in are not going t 1 have adverse outcomes, or better outcomes whichever is taken. (Internist) Since most of the minor illilesses are going to get better on their own anyway, I don't know as it's going to make a big difference on that. (Pediatrician) I guess the way I feel abou this for minor health issues is most of these minor issues are s If-limiting no matter what therapy is selected. (Family Practice) At the same time, physicians were conce ed that some patients might mistakenly per-ceive a major condition as a minor one, d would not appropriately benefit from physician input. They frequently comment d that patients should always have the 85

PAGE 199

option and feel comfortable consuhing wtth their physician at any time about "minor'' health issues. So, I think they should hare some involvement with their physician. I think, after a "fhile, they should have the -knowledge to make a fair number of -fuose decisions without input from their physician. But, they shouid always have that option if they have a question, for whatever rbson, that they can call a physician. I think as patients become ;mewhat more educated about illnesses, minor illnesses, and under tanding them, that they could make decisions on their own. B t, hopefully, would still recognize that there are certain times whfn it would be important for them to check with their doctor. fthey're not sure about what seems to be a minor illness, and the want to be sure about it, they should still have that availability o check with their physician. (Pediatrician) But I think that if patients can take care of things themselves and they are minor, and they ay need help in deciding that they're minor, then they should. I think that would only improve our relationship. (Internist) Minor, which is there, is open to interpretation. I've seen sprains and strains that are more serious than fractures, and we've all seen what sJem to be cold and flu symptoms, which tum out to be seriohs illness. I think that an informed patient can make bad decirions. They need some input. I think that it should be a partnership. (Internist) To me it's always the decision ultimately. But I'd like them to make that decision with me. If they're decision is different from mine I'll respett their decision. (Clinical Psychologist) Health care costs. Many physiT postulated that a more active involvement of patients in minor health issues would reduce health care costs. 186

PAGE 200

Generally, if people take e of things themselves it costs less. If they come in it costs a Httle more. (Internist) I believe they would go Less office calls fur minor problems. I think also in the ofthings that ifyou're thinking about cost of health care involving time away from work and avoidance of accidents and so forth that that should also decrease. Generally I would think the overall cost to society of health care should decrease. Health care costs, I think, ould be less when the patients would be able to make decisions n minor health issues with relatively little input from physicianJ. (Allergist/Immunologist) I think it will decrease hejth care costs. With managed care, it's funny because if there's a co-payer, or a $15 co-payer come in here, you would think twiee if you had some sniftles and you were going to pay $50 to see mf. as opposed to $5 to see me. So, the copay has made it such that is no financial liability for coming in here, or no significant finahcialliability. But I think if I had to pay I $50 to come in for a cold I'd think, "gee, what can I possibly do. Should I look up in my Cdlumbia physicians' and surgeon's guide what I should do for this?'! It's where managed care has actually, I think, tremendously physicians financially. I tell people, this is something you _can handle at home. You don't need to see me for this. (Internist) I think it definitely should feduce health care costs because it should reduce some of visits to either physicians, or emergency rooms for things that really are self limited, and could be handled at home. (Pediatrician) However, one physician was adamant that patient involvement in minor decisions and the implementation of a patient support slrvice like TBDS would cost physicians money. Well, of course we give away billions of dollars of telephone time, we collectively. We discussed it at a meeting last night 187

PAGE 201

as a matter of fact. The loss. We're the only profession in the world that doesn't charge r1 or telephone time. We will continue to lose money, hand over oot, as we have in the past. It will never change. If they don' have to come into the office it will save them money. It will lose us money. In terms of the big picture, having them take bare of it themselves or with some telephone advice is least to the system. But somebody pays for that and that's me. So whenever we have any telephone advice it will always be frck and that's just the way it's going to be. So healthcare costs, it will have no impact other than we will lose money. It won't cost the any money, it just means we lose money for every minute spend on the telephone. But we have been willing over the years to give that up and I guess will con tinue to be willing to do that. (Pediatrician) Another physician expressed the view tjt patient involvement in minor decisions was being driven solely by a cost managemenl objective without due consideration for patient well being. Economic reasons, this is product that's being driven by, 'how can we lower health care costs'' rather than necessarily always what's best for the patient! I got an invitation to come to a meeting and this is what Jere going to do, saw a lot of business leaders in Rochester, and I have to say I think this is being viewed as business issue rather thkn as a patient welfare issue. I think the decisions are pretty driven by the bottom line, that's what business is about. It's beeh seen for a couple of years that dollars are more important than dutcomes sometimes. (Internist) Negative!disconfirming cases. J contrast to the majority opinion, two physicians, held the opinion that decisionl about treatment for minor health issues should be made mostly by the physician 1th input from the patient. Part of the rationale behind their opinions centered ound the relationship of what they 188

PAGE 202

considered "minor'' and the implied resp nsibility of the physician to take a more predominant decision-making role in rellvely more complex clinical conditions. I don't take care of that J ..... y minor things. Minor things to me is like a that doesn't need to be taken care of tlris week, but can wait a couple of weeks from not(Gastroenterologist) I don't treat a lot of cold d flu symptoms, but all sprains and strains are not! minor, at least the ones I get .... I'm the next of treatment of the minor things. So I don't see as fuch of the minor stuff Usually when it gets to mr it's more severe. (Orthopedic Surgeon) Patient preference was another part of tHe rationale for the gastroenterologist's opinion. Even when the condition was Led to be minor, the gastroenterologist believed that patient preference is for thj physician to tell them what to do. When I go to my doctor flr minor things I tend to want to be told what's to be done and to have it straight forward and simple, and I don't want to involve discussion and all pf the options for minor things. Often times the patient population I see anyway, they want to kno-lv, "Is it a minor thing or is it a major problem?" I They want you to answer that question. And then if it's a minor thing then they mostly just want me to tell them, I "what should I do about this?" They want to know about alternatives some'tirites, but it's mostly just ... it's much more along the fmes, ''tell me what I should do about this to rltake this better. I want it to be fixed and go away and be done with it." 189

PAGE 203

Even though not as straight forward, th orthopedic surgeon implied that for health issues that were clearly determined to b minor, the treatment decisions should still remain mostly that of the physician with put from the patient. The really minor I would tend to go with number 4 (mostly physician with from the patient) .... But if I do see the minor, I'd like a more infonned patient. If they have a contusion o they put ice or heat on it? They should put ice on There are still a lot of people out there putting heat on and their ankles swell even more. So yeah, the need for ed cation is there. But at my ... I Guess when it comes to rp.Y level, my own problems, I don't invest a lot of time, defining the problems. Summary. Physicians differed in terms of their preference for involving patients in minor health decisions. How ver, the primary tendency was towards the patient having a more prominent role in oecisions regarding minor health issues. I Respondents emphasized that there would be a very positive impact on clinical efficiencies for patients who participate in decisions about treatment for minor health issues. That is, the number of medically inappropriate visits to their office and the volume of unnecessary phone calls woul be reduced. This would, in turn, allow h . :fi fth . I . h illn M p ys1etans to ocus more o err t1me on pat:J.ents Wlt more senous ess. ost physicians indicated that having the patilnt more involved in minor health issues would reduce health care costs. Although physicians were predo antly positive about patient involvement in decisions regarding minor health issues, here were two distinct concerns expressed 190

PAGE 204

about patient involvement having a less t an optimal impact on the doctor-patient relationship. I) The appearance or perion that by giving the patient a more prominent role, the physician would appT indecisive or lack understanding and concern. 2) Many patients may prefer not to take on an active decision-making role, even with minor health issues. Physicians felt that a more collaborative approach would provide patients with more confidence and control in managing a minor health issue and would generally result in higher patient satisfaction. They also believed that patient involvement in decisions about treatment for minor health issues would have neither a positive nor an adverse effect on outcomes. An exceptioJ to this may occur when a patient mistakenly perceives an illness or an injury as minor rl hen in reality it may not be minor at all and would appropriately benefit from physicia input. 5. For a major health issue thospitalizatlon. chemotherapy. surgery). who would vou want to make decisions about treatment? \Explain how this would impact: the doctor patient relationship. clinical efficiency. hkalth outcomes. patient satisfaction. health care costs. and medica/liability. I Table 4.3 illustrates physicians' oJinions about who they would want to make decisions about treatment for a major chemotherapy, surgeiY) health issue. One clinical psychiatrist indicated Js opinion would fall somewhere between I "mostly your patient with input from you' and ''you and your patient together." Two internists, one allergist/immunologist, on ophthalmologist, and one pediatrician 91

PAGE 205

indicated their opinions would fall some here between ''you and your patient together'' and "mostly you with input r. your patient. Table 4.3. Physicians' opinions aboutJ:ho should make decisions about treatment for a major (hospitalization, chemoth tpy, surgery) health issue. Only your Mostly your y(] u and your Mostly you Only patient patient with pa tient with input you input from you to1 ether from your patient (61 Internists (2) Internists (2 FamPrac (1) Opthalmol ( 1 Gastroent (1) Nephrolog (1 Clin Psychi (1) Peds Similar to decisions about treatm ed for minor health issues, physicians differ terms of their preferences for involvin! patients in decisions about major medical making for major health decisions trend primarily towards the physician having a more prominent and paternalistic role. Doctor-patient relationship. Ph sicians generally agreed that patient participation in decisions about major he th issues would enhance the doctor-patient relationship. I think it definitely impro es it. If you sit down with your patient and you discuss oz;tions, which is what we're supposed to do --providing info ed consent before someone undergoes a surgical procedure or c emotherapy. Discuss options, then it goes a long way in terms 1 f developing trust. (Internist) 192

PAGE 206

The doctor/patient relatio ship, again, I personally feel is enhanced. I guess you have t ask the patient that, but from my point of view this enhance the doctor/patient relationship. I think it's better understan g of each other's needs, and a deeper level of respect when one i willing to take the time to talk to the patient and recognize their needs and their desires to be involved in a relationship. I think t at enhances the relationship. (Ophthalmologist) At the same time, there was a strong sen ent that the physician should play a more prominent and directive role in these type of decisions. Many physicians believed that severity of illness influences the degree o patient and physician involvement in major medical decisions. It was a widely held op on that the severity of major medical issues required greater physician involve ent. The obvious difference is tpat involving the physician is going to be more irnportanf as the illness becomes more complicated and more critifal. The patient is going to have less ability to think clearly, and to think rationally and to remove fear and thr at from the decision-making process as the illness beco es more severe. (Internist) Some patients do not perc ive the severity of their asthma. Those patients need more input from the physician. (Allergist/Immunologist) I think that in major illness most physicians are going to find themselves, not necess1arily equal participation but more significant contribution from the physician. The key word here is acuity. You don't bkve time to coax the patient into a participant's role. If thinks are changing rapidly, there has to be a fairly prompt decisibn made as to what the next step is. (Nephrologist) The level of information n ed is higher. The level of sophisti-cation. the amount of inpu from physicians that you need is 93

PAGE 207

greater. But at the same time there are clearly different points where the factors in makfug decisions are not medical, but are more personal. (Family Ptactice) Patient preference and sati.ifactibn. Physicians were of the opinion that patients generally like to be involved in ljor medical decisions; and that, when in volved appropriately, patients should be Lore satisfied. However, physicians made a subtle but significant distinction regardink the relationship between patient satisfaction and patient preferences for involvement L the decision making process. Many phys icians emphasized that there may be a rJge of major medical decision-making preferences by patients. Some patients Jay prefer a more passive role ("I prefer that my doctor tell me what to do") while sot may prefer a more collaboralive approach ("I prefer that my doctor and I make the (iecision together''). It really depends on the pLent. Some patients, which are really older patients, may kant you to make the decision for them. And I still want inp t from them but I have a mostly younger patient populatio so they want to make a lot of the decisions. They are relati ely educated, so they want to make decisions. (Internist) It's a Welbyish way ofbe' g to say that patients want-there are a minority of patients -kho would really want me to take an iron hand and just makf decisions for them. But in the long run even those patients be dissatisfied if they felt like all the decision was taken away from them. (Internist) I I think those issues are going to fall under either category 3 or 4 depending, a little bit, ort the medical sophistication, and willingness of the patient tb become involved in the decision making. It's the --how the patient feels about becoming actively involved in decisi 1 n making versus just saying, "you're 194

PAGE 208

the doctor, go ahead" and "you tell me what to do." I've been in practice for 25 years, and when I first started practice I think a far greater numbJr of my patient's were in the, "you're the doctor, tell me t do" mode, than I see now. But I still see a significant percentag who feel they would like me to do most of the decision g. But more and more who would like to be partners in the decision making. (Pediatrician) Several physicians conceded that if clinicir fililed to be sensitive to the patient's own preferences in decision-making preferences for major medical decisions, physician patient discordance may occur, resulting i lower patient satisfaction and a less than optimal doctor-patient relationship. I think in most instances people are satisfied when they're involved. If they feel they're getting f fair evaluation of where they're at. lfthey feel that the advice.rou've given, ifyou laid forth the reasons why you've given it, are going to be satisfied as opposed to not knowing. (Internist) Again, for most patients, satisfaction should be improved for those patients who want a collab rative relationship. For patients who won't want a collaborative elationship they wouldn't like it. (Clinical Psychiatrist) There's nothing worse a patient who really wants to partici pate, trying his best to with an authoritarian physician of the old school that says, 'don't ask me any more questions. I've told you exactly what you should do.' And that was the way our grandfathers, to some practiced a lot of medicine. (Nephrologist) Some subtleties here. I think patients revert a little bit to the, 'you're the doctor, I'm the patient' mode, as they get sicker. It's kind of a regression, it's almost a parent/child kind of a role, 'I don't know what I'm do g, you do, I'm sick. This is scary, threatening to me, tell me hat to do.' They may be more satisfied with a physician w o doesn't keep throwing it back in 95

PAGE 209

their lap, but rather takes a more active role in, trust me. I'm going to help you throu this. We're going to get you better. I'm going to make the decisidns, and you're going to do well. I think that you kind of start to in to that as you go along as that patient's physician. How :rpuch, how able they are at that point in time to participate, but satisfaction, obviously, is going to vary from patient to patient. patients are going to want to be told, and some patient's are going to want to be involved. It depends on desires and the patient's (Internist) Many patients want in the decision making process. And I don't think it's the 1ght thing to do to just hand that off to people who are, in many eases, less prepared to deal with even the concepts. It isn't that I to take their right to decide away from them, but in major illness,! considering the short time factors that are often involved, it doesn't seem to be fair to just look at the patient and say, 'okay, it's your chll.' That's the dark side of patient empowerment. (Nephroldgist) If they're involved in that lhoice and they feel comfortable with the choice that they'+ making and they know that they can always back up and go another route, then they're I satisfaction will be high. is a minority of patients who would really want to take an iron hand and just make decisions for them, but in the long run even those patients would be dissatisfied if thY felt like all the decision was taken away from them. I can prbsent my cases forcefully to a patient about what I think needs tb be done and I can render my opinion, but they still nee to be involved in that process. I think they'll feel better, ev the patients who want me to make I decisions for them. And they'll feel more satisfied. (Internist) l Additionally, physicians frequently comm nted that they would respect the decision of the patient even when they didn't agree lth the patient's decision. It's really 50/50, me and tJe patient. Where, really, the reality is my opinion is very valuJd by the patient and that there's 196

PAGE 210

probably more of a directire component there than I feel com fortable with. (Family Prafice) When we go to the major pealth issues, I think we're dealing with issues where we may[ have therapy that is truly more effec tive that we need to push for in the medical point of view. For example, this week I had a patient who had a heart attack at home, didn't want to go td the hospital, and I had to tell her I think it's the wrong decisibn. But if that's what they want, this is what we need to do what your choice is. I think in this situation I'm much more likely to say, these are the things that have been shown to improre your longevity, or your outcome. I think you need to be these. What do you think about it? And then let them modify my treatment plan by their expectations, rather than put the out and say, 'it doesn't make any difference, you decide.' So, rm more likely to direct situations there. I still want them iniolved. I think people still need to understand that they need o be part of it, and they need to buy on with it. But, I think, where a major health decision where I can give people solid evidence of change and outcome, I have an obligation to tell them that this is the treatment plan that has the best outcome. When I do that, it is more directed than throwing out the options t saying, 'choose one.' (Internist) I think I have a responsibility to my patient's to tell them what I feel is clearly a best treatpent. Or to tell them that I don't know, and let them know in making sure they have a chance to modify that plan based 9n their values system. I think that that's beneficial, rather than being absolutely directive saying this is what you have to dot or being totally open ended and saying, what do you want to do about things. I think they need to see that you have dlear feelings on it, make it known that that's what you want, you're willing to modify. People respect that, I think. (IntenUst) I think there it becomes a 5[10/50 decision ---actually the decision is going to be the patient's. The patient has the final eventual decision making -going to make the decision, let's face it, I'm not going to make the decision but I think it's a partnership and we're both going to do it. They can make the 197

PAGE 211

wrong decision, in my eyes, and sometimes they do. But they sometimes have valid rea.kons. A woman was here today, she keeps coming back rotator cuff tear. She's 84 years old and believes she is too frJil for surgery. I have to respect her opinion on that. I point dut to her though that just last week a I 78-year-old man who's come back 5 months after surgery has full range of motion, no dam what so ever. Perfect result. I point that out to her. if she makes the choice not to have surgery, that's her choice.\ I'm not going to force her into it. There's some people I don't respect their decision. They're making a foolish decision! But it's their decision. I continue to support them. It's their to make that decision. It's their body. I tell them I wouldh't advise them making that decision, but I respect their right t1 make it. (Orthopedic Surgeon) One pediatrician aptly summarized the whole perspective about doctor-patient re lationships and patient satisfaction in to patient preference for involvement in major medical decisions: I think, and this is part of rhat I think is what the art of medicine is. And that is, is the physician appropriately reading the patient's desire for\involvement, and matching his response to that desire? And if the physician is able to do that, then that should enhJnce the relationship. If the physician and patients aren't cdmmunicating that need appropriately, then there will be problemk. Let's say, if the patient really I wants to be more actively decision making, and the physician doesn't pick that up and adopts a more paternalistic approach, then the patient's not going to be satisfied. If, on the other hand, the physician feels tliat it's important for the patient to be completely involved in the decision making and starts I providing the patient with F,ormation and expecting patient's to make decisions, and they're not ready to do that, and they don't want to do that, thenj that's a miscommunication in the other direction and that coWd cause problems also. So, I think, the key here is for physiciah and patient to sort ofbe able to 198

PAGE 212

read each other, and to each other where they are in that desire to be in1olved in decision making, and then there would be a good fit and a good match. Health outcomes and complianc Most physicians contended that outcomes They cited improved patient compliance, control, and empowerment as the mechanism largely responsible for improved outcomes. And, I think that the style of care that I like to deliver to my patient's is such that they feel a little more empowered. They feel that whether I can affect the circumstances of their illness directly or not, I will be thb-e with them through that illness. And I think that when spend the time with your patient that way, I think, that improves outcomes. (Family Practice) An informed patient, an i4olved patient, is a more satisfied patient, and emotional outcomes, and clinical outcomes are improved if the patient is iltvolved. (Internist) I Generally, I would expect that outcomes should be improved. If patient's their major illness better and the implications about the treatment and how important it would be to continue, often times, these treatments would long term medications, and the use of those, and tlle patient uses them because they understand their illnesk better, then their health I outcome would be better. rediatrician) I would hope that health outcomes improve in this situation. If you give peopie enough information, and I that they buy on to a plan that has medically proven efficiency, I think you're better off. I think if people are engaged in their decision ,th proper direction that you'll get better compliance and buy on to a treatment program. (Internist) 199

PAGE 213

It very much improves it. ren you're in the hospital, you're wearing that you don't have your clothes on, you feel you can't do !htything for yourself In order to do this that doctor has sit down with you and go through a set of options that a patient has to make a decision. There are choicds, so that's empowering for them. It really helps peoplb. They have some control over their illness rather it just happening to them. (Family Practice) Clinical efficiencies. Physicians ixpressed conflicting opinions regarding whether or not involving patients in decisions about major medical issues would be more efficient. The gamut of opinions jged from those who believed it would definitely be inefficient and would take mire time to those who believed it would definitely be more efficient Others beliejled that involving patients in these types of decisions may take more time initially, bu in the long run it would be more efficient and would result in higher patient satisfjion. This may be less efficient in terms of dealing with patients rather than absolutely diredtive approach. It's going to take more time in the procbss. I don't see how this gets you away quicker than a only' approach to things. (Internist) It's more efficient to get somebody involved early that has the knowledge base than it \would be in something that may well be self limited. I think it's always more efficient to get the patient involved early That's true. across the spectrum. (Internist) But often times, I think, pafients have a lot of good common sense. If your job is to presf.nt it to them in an understandable way, and if you succeed in hat at all, then I think sometimes 200

PAGE 214

it does improve clinical e, ciency. You can hone right in on productive. (Gastroenter logist) That should be improved, because the patient should be able to -if they're more invol ed they should be able to make some of the decisions, or hnderstand the decisions more I quickly. Now, there may be sort of an investment oftime up front on a major issue this, where it may take more time to educate the patient. But that, I would think, would pay dividends later on becaus5 then the patient would be better able to manage the care and you wouldn't have to be trying to constantly re-explain or get the patient involved later on. (Pediatrician) 1 Health care costs. Physicians eXIjressed uncertainty and mixed opinions about what effect involving patients in decisionk about major medical issues would have on health care costs. The spectrum ranged Jom those who were of the definite opinion that patient involvement would reduce heLth care costs to those held an equally determined opinion that patient involvemt would increase health care costs. Well, unfortunately, it's of unclear how that's going to play out. Because, I thiilk, patients want everything done, or a lot of patients ,ant everything done for things that aren't necessarily Needing an MRI scan for their headache. (Internist) I I am not so confident that it does anything for health care costs. I think it may health care costs. I think the more informed people are more likely they are to seek access to good care and the less likely they are to be satisfied with less care. The more ?}formed people are the better informed they want their d0ctors to be, the more specialists r

PAGE 215

they're going to want to see. The more they're going to want things like cholesterol scrJening and PSA determinations to be done. (Gastroenteroloii.st) For a major health you're talking about something like hospitalization, or surfery the cost is going to be the same whether the patient has made the decision, whether I've made the decision, or whether made the decision together. I find that many patients, when they understand the situation, in fact, will comfortably defet certain problems, that would be in the elective situation. In non-elective situation, I don't know whether there's a difference or not. There may be no difference whatsoever in hbalth care costs in that area. With cataract surgery, for example, as an elective procedure, many patients who are :functiorurtg satisfactorily, as long as they understand that the surgerY is available but there isn't a reason to do it as long as they legally drive, and their visual needs are being met every day. :rytany patients, I find, are happy to defer surgery. (Ophthalmologist) I think health care costs be reduced because it will provide for better treatment, is, hopefully, more optimal treatment which usually would be cost effective, because patient's presumably under the should recover more quickly and should have less complications. (Pediatrician) Health care costs, again, I lhink there's no doubt that it's going to improve it. Again, are less likely to fly into the ED if they know what to expect from a treatment course from their I illness. People are less Iikefy to be ringing the nurse's bell and being sent to the ICU because no one really knows what's going on with the patient over weekend. When they're nervous about something that they llave in regards to their heart failure, their heart attack, their pnehmonia, their cancer. So I think that would help drive down heaith costs. Anything where the patient is playing a role in the and having an understanding in their disease, I think is going to decrease cost. Conveying some information to a patient doesn't cost a cent, it really doesn't. Everything else does. Havihg them come into the ED because I they're nauseous from the chemotherapy medicine That's going to 202

PAGE 216

cost money. Or having sent to the CCU because they're having some twinges of chest pain and they're not sure that it's their reflux or if it's their eart and the doctor never explained which is which so, 'we'll, send them to the CCU.' That's going to cost money. Somebod who has pneumonia, that's on a home treatment and was never what to watch out for the signs of, suddenly gets very dizzy and passes out and is found cyanotic at home, where they werejfeeling sick three days earlier but really didn't want to bother That's going to cost more money. Any situations where you intorm the patients about what to expect with, not only about a but what the complications from the disease can be short and term, and what the complications of the therapy can be short and long term. It's always going to save money because there will less phone calls, less panic reflex with regards to their illness. (Internist) Medica/liability. Physicians affi1ed that involving patients in decisions about major medical issues would definitely redtce medical liability. However, "involvement" was defined as more than j st the rote process of informed consent. Involvement meant being able to establis a genuine relationship with the patient as a whole person. In included a relationship t at would allow the physician to understand patients and their entire and, to empathize with them as well as educate and infurm them about treatment rptions, benefits, and risks. When physicians practice with this style, medical liability is reduced, even when the outcomes are not as positive as expected. Well, my own sense about that is, medical liability decreases, almost exponentially as an inverse to the degree of investment in the relationship between he doctor and the patient. (Family Practice) 203

PAGE 217

I think it lowers it. I thinl(: that it's trust building. It's getting the patient and you on thJ same team and participating in decision making as a I think it really lessens your likelihood of any malpractfce because you're working as team rather than butting II think a lot of malpractice issues probably could be headed off with more of a team approach. (Internist) J The informed consent do rine is that patients need to be explained the risks and bertefits of any treatment and the risks and benefits of any alternative treatment that might be available. If you fail to provide that your patient you're failing in your obligation as a physician and you're subject to medical liability I under the law." Providing informed consent can be a very emotionless event that really help you connect with your patient. I think it's not so much involving the patient in decision making as emoting with your patient. Because there are lousy doctors who have bad outcomes who never get sued because they're good to patients. They're just nice people who listen to their patientsJ And there are great doctors who I have the best gifted hands in the world who will get sued because they're asses. So, I think that more has to do with -not so much sitting down and risks and benefits, but sitting down and empathizing and emoting with your patient. (Internist) As I tell our medical studeJts, 'you should make a friend and I partner' of your patients.' .t}nd that, generally speaking, that patient will be a satisfif. patient who will not sue you even if a goof-up is made. Gener:ally they won't do that. (Orthopedic Surgeon) The more you inform your patients, the more they have an understanding of what options are and why they chose an option, there's going to less concern for liability. No one can really hold you liable if they were involved in the treatment plan, if they werel involved in the decision process. How can somebody sue yoJ if .. 'Yeah, I'm upset with the doctor because I had my entire removed instead of just the lump removed' 'According to thJse notes the doctor discussed with you the options including 1Jmpectomy, radiation, and mastectomy. 204

PAGE 218

Your words at the time ,ere, 'take the whole thing off: I don't care.' 'Apparently you made the decision with him.' 'Well, yes that's true.' 'Do you feel tou were infonned? What our options were?' 'Yes, but I was vef.y upset at the time.' 'It says here also that you didn't have to the decision right now, you could call him back in two or three days.' I think as long as you're infonned I and you understand, you literally put yourself in the shoes of the patient and say, 'how woilld I feel in this situation. What would I want to know?' That's situation where you reduce your liability dramatically. You're going to reduce it to zero. There's always going to be somebody whb wants to sue, that's going to happen if somebody slips on your :frbnt step. If you're terrified of it, you can be paralyzed in providing care. The most you can do is docu ment what you do and be as informative as possible to your patients. That's the most you can db. (Internist) I think there's a very big blnefit here for medical liability reduced problems, because, I when patient's start thinking about suing most of the time it's because of communication problems between I them and their physician. I think if physicians involve and engage patient's in decision then the communication process, of necessity, has to be there's much less likelihood for misunderstanding, and the patient then later on saying, 'well, the doctor never told met s or that' and 'if he had, such and such, wouldn't have happehed.' Then you're more likely to get a suit. (Pediatrician) Summary. This group of physicians differed in their preferences for involving I patients in decisions about major medical issues. In contrast to minor health issues, preference for decision making of major hLth decisions primarily trended towards the physician having a more prominent and palemalistic role. Physicians contended that . 1 . d" al d .I. ld h b d pattent mvo vement m maJor me 1c eciStons wou ennc t e octor-pattent relationship and enhance patient satisfactiJn if clinicians were sensitive to individual patient preferences for involvement in declsion making. They believed that health 205

PAGE 219

outcomes would be improved and the po entia! for medical liability would be reduced if k 1 d .\. b .t:'. di al patients too a more active ro e m eas1ons a out treatment J.Or maJOr me c Ph'' d ld nfli' .. b hth tssues. ys1c1ans expresse uncertamty an co cttng opiruons a out w e er or not I patient involvement in major medical decisions would improve clinical efficiencies or reduce health care costs. j 6. What do you think about your patien -s taking a more active role in the self management of chronic disease conditiofzs (e.g. asthma. arthritis. diabetes. hypertension. low back pain)? Explain hdw this would impact:.the doctor-patient relationship. clinical efficiency. health Diltcomes. patient satisfaction. health care costs. and medica/liability. This group of physicians were unanimous in their belief that patient involvement in the self-management of cJonic disease was absolutely essential for optimizing health outcomes, patient satisf1ction, doctor-patient relationships, clinical efficiencies, cost management, and in redjcing the potential for medical liability. Health outcomes. Physicians _j emphatic that a more active patient role in the selfmanagement of chronic disease jould result in better clinical and functional status health outcomes. I think this is the area where it can substantially increase the health outcomes results. Diabetes in particular. Asthma secondarily. But, yeah, the patient does affects the I outcomes more than what the physician does. Sometimes I'll have a patient that's got k poorly controlled chronic condition who I'll send to a \specialist and I find that they don't control them any better than I do because the problem is not the knowledge base ofthe medical provider, the problem is the patient doesn't have t e knowledge or the motivation to 206

PAGE 220

do what they need to do. So chronic diseases most of the management is self management. (Family Practice) The only way that patientl do well with chronic disease and that you get good health Jutcomes is if the patient's become dramatically involved with their weight loss, with their diabetes, with their hypertension. Bfcause, usually, with these diseases they are major life changes they have to do to take affect and the patients have to be in bharge of it. Otherwise, it doesn't matter how many times ydu tell them to do it, if they're not actually doing it and gettirlg involved realizing themselves what needs to be done, just not going to work. There's no way you can get good outp>mes unless patients are dramatically involved with their own clironic illnesses. (Internist) A chronic illness is for the rest of their lives. It's not something that's going to go and it's largely dependent on their compliance, and insight, and their understanding, and their buy-in to a treatment plan for them to do well and I minimize complications. li'he more they know and the more they participate, the better they're going to do in general. (Internist) Now, if they've got a proolem such as diabetic retinopathy, that may be a different but to the extent that they're educated about the and about the types of things that they can do in terms of control, proper medication use, physical exercise, follow-up with the doctor who's managing their diabetes, thbse kinds of things, I think, those I patients clearly do better. fhe health outcome, I think on these patients it's better particularly with the chronic problem. Obviously, diabetes for exJmple, someone who doesn't under stand the nature of their is likely to be a poorly I controlled diabetic. With the types of chronic eye problems that we see, these patients ho far better with an active role in self management. (Ophthalinologist) I think of chronic disease, as arthritis, where management techniques are largely techniques, methods to control symptoms, to control their functional capacity. Medically 207

PAGE 221

their joints may be shot in the same amount of time no matter which approach is taken, but if a person has learned to cope with their disabilities, witH proper access to arthritis association I programs and techniques for managing their disease, they may ultimately need a hip at the same time, but between here and there they may have functioned a lot better. (Internist) They're going to be less Jely to be in the ED, less likely to be in the hospital, less likely to get their medications confused and more likely to have a course of the disease. I think, in the outcome, have better morbidity and mortality in the long run, especially with astruJtics, diabetics, COPD'ers, and even Alzheimer's patients. satisfaction, same. (Internist) Patient satisfaction. These physidians had consensus that patients taking a more active role in the Jf chronic disease would result in higher patient satisfaction. It's enhanced, once again. Diabetes, for example, 'when Patients have an understanffing of their illness, know how to use their glucometer, adualiy know what to eat and what not to eat and are not tethered by the telephone with I every blood sugar that's abnonnal. They actually know what to do. They love it. l.\nd they'll be grateful for it. They feel much more in cohtrol of their illness. (Internist) I would think that would Jcrease because, again, there's a great satisfaction of doink something yourself and realizing that it has been ydur own input that has helped and made something betterl Whether it's fooling around in the stock market or cuttfug your lawn or taking care of yourself. (Internist) It's higher. lllnesses are disempowering. Decision power is empowering. So I think beople feel better about their illnesses no matter where they are in the extent of their illness. (Family Practice) r

PAGE 222

Doctor-patient relationship. M y physicians indicated that, when patients take a more active role in the self-managlment of chronic disease, it improves the doctor-patient relationship through enh ced trust and the removal of barriers to communication. No doubt on the doctor-patient relationship because when someone has a chronic those are the patients you're going to be seeing the most of And they're the ones that are going to get very aisgusted with the medical system unless they feel like a true advocate out there for them. I think that's where the physician really has to play a role As their advocate, an und9rstanding what it is that they want. So I think it has a very positive impact on the patient physician relationship beca!use they can really feel like, 'this is my doctor, he knows that's going on.' They don't feel like they have to ljuggle things between six or seven different doctors or eight Hundred different medicines saying, 'well, I have to trust that s guy or this gal knows everything that's going on.' I think it as nothing but a positive impact on the patient-physician re ationship. (Internist) I think it's very positive. I I hink the more the patient is educated the more they talk to the doctor with comfort. The more the physician treats the patient like a peer the more the communication fall between them. It substantially enhances the doctor-patient relationship. (Family Practice) I Patient preference. Individual lees in patient preference for in-volvement in chronic illness care were mentioned by only two physicians who made b bib" lc. .. I hr "illn nunor comments a out t s emg a concern LOT patient mvo vement m c omc ess care. 209

PAGE 223

In most cases it will enhance it. It's going to worsen it in those situations where people iant to be treated as opposed to involved. Some people d1 n't respond to that approach. Some people are very much, the doctor, tell me what to do' approach. I think the majority of patients are going to respond to it positively once they buy into that system. (Internist) For a majority of the physicians, it was Lost as if it were a given, a necessary com ponent in the continuum of care that patiLs be involved in the self-management of chronic illness. Clinical efficiencies and cost management. Many physicians remarked that this is definitely one area where a more Jtively involved patient can have a significant impact on health care delivery system effilencies and cost management without sacrificing quality of care. That's one place where I tlililk that you clearly can show a benefit. Studies show a rltore efficient delivery if you empower the patient to mJnage their care, their chronic disease. That's one that I don't think there's a lot of argument with. (Internist) Clinical efficiency, this is Pfobably where I think the biggest savings are. If you have somebody who has an understanding of their discltse process, they're not going to call you every time their blbod glucose is 205. They're not going to show up in your dffice every time they cough up some yellow sputum. not going to fly into the ED every time they get a little i.vheeze, if they're an asthmatic, or the family's not going tol rush them into the hospital because grandma's getting confused again and left the stove on and almost burned the house ddwn. These are the patients, that if they have a better understanding of what the disease process means, that it's chronic, it's probably going to get worse with 210

PAGE 224

time but we can control you can avoid diabetic acidosis, you can avoid hypoglycemia _1 hich are both things that might require more clinical cost of time and effort and so forth. So I would think it would be better. (Internist) I think that would lower care costs. Fewer office visits, fewer complications. (Internist) What I've seen in studies Jays improvement. Less dollars because of recurrent acceJs to a higher cost things by allowing people to it own their own. (Internist) This is one lruge area for health care costs. I was just looking at an article on diabetes management, and a very, very large percentage of care costs go to deal with the complications of diabetes. \Its been shown that if you can improve the patient's of their diabetes you can reduce these health care costs by huge, huge amounts, billions of dollars for the (Pediatrician) Self efficacy. Physicians attributJ improved health outcomes, patient satisfuction, doctor-patient relationships, tlinical efficiencies and cost management primarily to the mechanism of enhanced patient self-efficacy. They explained that when patient takes a more active role in re self-management of chronic disease, they feel more in control of their lives and the disease and therefore more empowered to deal with the disease in a positive fashion. Confident, knowledgeable, and in control, which is, with chronic disease, largely what determines a patient's perception of their disease. (Internist) If they have an active role ip their health, 'look, you have this illness, it's a chronic illriess, it's not going to go away, but you can take an active in controlling it and this will allow you to control your life.' So, the diabetic who 211

PAGE 225

knows that when the sugar moves up to 250, they have a sliding scale, they can adfi.Ist them, they don't have to pick up the phone for the doctor and they can still go out with their friends or they can still gb to work. They don't have to end up in the ED. I think can handle it, that's very empowering. (Internist) So they're more productire and, once again, they feel healthier, they feel happier that they're in charge, that they're managing an illness. (Internist) I They're much more satisfied if they can control their own outcomes. Because they \feel as if they're in control of their own body. People, especially with chronic disease, like to feel that way. (Ophthalmologist) Medica/liability. Numerous physicians asserted that the potential for medical liability would be reduced significantly +en patients are more actively involved in the self-management of chronic disease. Patients who have a better understanding of their when things do go wrong. I think the more people have bought into a system on their own, and are given choiccls, the less likely they are to accuse you of being the cause a bad outcome. People aren't going to sue themselves because they didn't manage their disease properly. (Internist) The medical system, you mow, just treated her as a disease. Once she felt like she was 1\somebody who had a disease and she had a role in her care, she was hardly ever in the ED. And I knew in my heart of hearts no matter what happened to this lady she would never kue because she was involved in her health care. (Internist) 212

PAGE 226

If they're taking a more aC1:ive role in taking care of themselves then liability would go doWn. They wouldn't be hanging on every decision that I gospel and wouldn't, if there was a bad outcome with a decision, they wouldn't always I be looking to me. They'd be looking towards themselves. I think it would, in all these casesi I think a more active patient involvement would decrease liability. (Internist) I think that the relationshiJ is enhanced and the potential for liability is less. Patients tatkmg an active role is a positive all the way around. It's the Studies show that patient's who have a good rapport with their doctor are much less likely to sue. They may well be better care. I find that patients, for example, lot's of diabetics who are interested in their disease may know much more abobt their condition than a number of the health professionals they 'fme into contact with, and will evaluate any recommendation you r-ake on their fund of knowledge, and may suggest somethinp to you which you hadn't thought of, or makes more sense iq their particular situation. They just end up getting better dare, I think. The likelihood of a problem is somewhat lessdned by that. I don't know what studies would show in this\ area, if it's ever--ifliability costs versus patients taking an active role have ever been looked at, but I would strongly that you would see a lower liability cost. (Ophthalmologist) Summary. Physician informants 4ressed unanimous support for the utility of patient involvement in the self-managei of chronic disease for optimizing health outcomes, patient satisfaction, doctor-patient relationships, clinical efficiencies, cost management, and reducing the potential fJ medical liability. Out ofthe three d ab I. 1 <. illn scenanos use to measure opwons out anent mvo vement mmor ess, maJor illness, and chronic illness), physicians asized that patient involvement in the

PAGE 227

management of chronic illness held the most value and would have the most significant overall impact on health care system qu1ty, cost, and efficiency factors. This, personally, I think il where something lnce patient involvement is going to have thej biggest impact on all of these things. With the minor stuff it's going to be a minor impact, with the major stuff it's going to bb a pretty good impact, but this, I think it's going to have the biggest impact. (Internist) Opinions about Telephone-Based DecisiL Surux>rt The next series of questions (Q 11 0) included more specific questions regarding physicians' opinions and perceptions of a 24-hour telephone-based decision support (TBDS) service that had been iJplemented for 80,000 families two months prior to the completion of these intervie s. Eighteen physicians reported that they had heard of the TBDS service and were familiar with its purpose and availal)le services. The remaining three physicians I reported that they had heard of the TBDS service but knew very little about its purpose and available services. Physicijs expressed a range of opinions about the TBDS service: generally positive (7), netral (11), and generally negative (3). I 7. Please describe how you think the telephone-based decision support service will impact vour patients and your practice. Physicians differed in their opinions about the potential impact of implementing l h b dd .. .I h. d o a te ep onease ec1S1on support servtce on t err pattents an practice. ne opmton group (n=9) indicated they were unceJ or "not sure." 214

PAGE 228

. My patients know about it, theyre talking about it, they're accessing it, they're in and telling me that they've used it. Not sure of the nrtpact on my patients. Whether it's been a good thing, or a bad thing, because it's just getting started. (Internist) I don't know that much ab0ut it so it's a little hard to know how it will impact. It's ... I'rtt not sure. It would be nice if it took away some of the phone calls. (Internist) I b .I . am not sure ow tt s gomg to unpact my patients m my I practice. For eye issues, I don't have a sense at this point, for example, what percent bf my patients may call with ophthalmology issues to a like this. My guess would be that it wouldn't be a number and I would guess there wouldn't be a large impact (Ophthalmologist) Another opinion group (n=9) said that the service would have "no impact" on their patients and their practice. Many in this group were of the opinion that they personally fulfilled the role of the service Jy providing their patients with the information and counseling they needed tjl be more active, involved consumers. They viewed themselves, as one physician put it, as a "TBDS supplier" for their own patients. I don't think it will. I get a lot of phone calls. They kind of pile up during the day and I renlm them after hours. I think of myself as a "TBDS" supplier becadse the patients call me up and they ask these questions that I'm they ask the people at the TBDS service. Perhaps the only thing is if there was a 1-800 number, patients might feel freer to call. TbJy may feel somewhat inhibited to call me because I'm busy. I haven't\found that to be the case though. They still seem to call. I talk with them five, ten minutes on the phone for 2 5

PAGE 229

what ever problems theyt'e got and we make a decision whether they should come in or if we can take care of it by phone with just some simple advice. (Internist) I think it will have a minimal impact for our patient's in. our practice because, I suspebt if we're providing the kind of information that we should be providing for our parents, they would have I little need to call the TBDS service for further information. I And, we have for each ofJ our well-child visits a one page handout talking about hea.Ith and safety issues and feeding I issues. We also have probably at least a dozen disease related handouts how to manage \children with colds, how to manage children with fevers, diarrrea. head injuries, what ever. We have handouts that explain aboft those things that we give to those patients when they make the visit with us. We also make fairly hberal use of the Academy of Pediatrics pamphlets on a whole range of subjects, and we\have some that we handout specifically, and a whole bunch that we have in both the waiting rooms and the examining rooms for to pick up and peruse or bring home. (Pediatrician) There are four of us that share on-call and we've always encouraged patients to ca.h us. I've received several calls this weekend. In other wbrds, one of us covers for the whole practice. Much of }vhat we do is give out advice on the telephone. I'm not so rure it will make any tremendous impact on our practice. O*ternist) Patients with real big we see to it that they're pretty well informed, if they're to that, through getting those patients connected with Alnerican Liver Foundation Local Chapter. We think the TBDS service is a commercial venture, as I see it, fulfilling a need\ that has arisen that is fulfilled largely by volunteer organizations. The TBDS service does some of the same things volunteer orgkuzations have been doing for patients. Patients have sought this of thing out for quite a while. Patients get information frbm these organizations about their disease, past what they siJply get in the doctor's office. (Gastroenterologist) 216

PAGE 230

The third opinion group was a minority of physicians ( n=3) who felt the service would have a "positive impact" on their Jatients, mostly related to clinical efficiencies. Their positive reactions to such a service \generally clustered around the issues of clinical efficiencies and the doctor-patienJ relationship. I Clinical efficiencies. Two physicians in the "positive impact" opinion group I said the service would improve clinical efficiencies by reducing the number of unnecessary phone calls and visits as well as the average length of the clinical encounter. I think overall it's going to be a plus just, hopefully, to cut down on the number of phbne calls, and as an educational tool for the patients. I don't thirik: the TBDS service's goal is to make I radical medical decisions. [ think they're just there to basically to inform you of options are available for particular illnesses or to provide you with with regard to particular illnesses or questions, and I think that will enhance it. Average length of the clinibat encounter it may decrease that mainly because if they \did contact the TBDS service and they say, 'you really should see the doctor because there's a concern for this', they'll be I much more definitive for what they're coming in for. Thet called up and they said 'this is what I should do.' I actujy had an experience with somebody who did use the TBDS seryice, or used the book that came along with it, and looked up something about their eye and it wasn't getting better and called me and were very clear on what they wanted. It actually reduce the length of the encounter. The volume of Calls to the office -it probably will reduce that. That's probablt the biggest impact there. It's those little annoying calls that yolii get about, 'I was bit by a mosquito and it still itches', like that. (Internist) I I think you may be able to decrease the amount of triage, probably half of the triage questions are problems that really 217

PAGE 231

could very easily and ju 1 as effectively be handled by just by it. half that would call be able to really focus it down to the important information. Nkses that triage would be able to get at the problem in a more fashion. (Family Practice) There was an interesting dichotomy in tJe views expressed about clinical efficiencies mong physicians of the same opinion + On the one hand, there were several physicians who expressed a "not sure" or a "no impact" opinion who also expressed a hope that the service would have a impact on clinical efficiencies by reducing inappropriate or unnecessary phone calls I I dont know that much aoout it so it's a little hard to know how it will impact. It's .. .Ijm not sure. It would be nice if it took away some of the nuisance phone calls. (Internist) On the other hand, the same physicians J the "not sure" and "no impact" groups commented that there was no such thing L inappropriate or unnecessary phone calls or visits. Furthermore, they viewed thesJ so called "nuisance calls" as opportunities to build on and enhance the doctor-patient rllationship. If the doctor and the pati.L are meeting over a concern that a I patient has, it's an appropryate visit. No. I don't think there's such a thing as an inapprofriate medical visit. As a country and a society, you have to [decide how they want to spend their dollars, but is it an appropriate visit to a doctor if a patient has a concern, yes. It's like saYmg there's a stupid question. If you have a question that needs\ an answer, you ask it. A lot of visits that maybe weren't there are a lot of interactions that didn't necessarily fix somdhing that needed to be fixed, but they address that need for the at that point in time, and those limited dollars and limited hme, and limited access, you have to 218

PAGE 232

decide what's more appro1priate and less appropriate, or more important, and less impottant, but not inappropriate, and appropriate. It's a continuum. I'm sure the calls that would have t.o me will to as I've before. Over mmor illnesses I think mjh of the time.tt will turn out to be okay. (Internist) I'm a finn believer that 95f, 98% of patients are very reasonable, they're responsible and it's the 2% or 5%, probably 2% that call for everything. It's hard to ctlange their behaviors if their anxiety based behaviors. It's hard to kn?w what type of impact this service would have on the nuisance phone calls too. Example of that would be, often times when a patient with a nuisance phone call during the day, it's simple dvice, a 'yes, n0' or 'do this', typically I'd give it back to one of our senior staff to ban the person back. Inevitably, if it's one I of the people who call a lot I get the chart back on my desk, 'answered question, they kill want to talk to you.' And you tell them the same thing and it's likb, 'okay, I just wanted to hear it from you.' (Internist) So, in my population, I don't want to discourage patients from calling me. I don't calls my front desk and they say they want to talk fo the doctor about a little pain in their leg. I don't think I want my secretaries to say, 'have you tried the TBDS service?' to try to decrease my phone load at the end of the day. I consider those to be key to insuring our relationship and providing patient \satisfaction. There's a possibility it could be detrimental to out relationship because it kind of throws another provider in there, tmd I don't want to inhibit my patients access to me. (Internist) I think that anytime you interact with a patient you're building a doctor-patient and then if the interactions went down with certain patients, beca'lilse of using the TBDS service, that could I perhaps negatively effect the doctor-patient relationship. Perhaps it would be positive, at for the doctor's side of it. I've always thought that's one of the rriain reasons for doing a physical exam on patients, is the health mhlntenance issues you work through. But a lot of the value in physical exam, in my estimation, is the 219

PAGE 233

time you spend with the patient interacting with them, building a relationship. (Internist) Doctor-patient relationship. Other physicians reiterated that any positive or negative impact on the doctor-patient rjationship would fundamentally resuh from: a) how the physician positioned the se]-le with his/her patients and b) whether or not the information provided by the service 1 ould support the physicians' viewpoim. It depends on how the phrsician handles it. If the physician says, 'oh, what the hell did you1call them for?' 'They don't know what to tell you. It's going to 1 ave the patient more confused than anything. (Internist) I think it all depends on h wit's used on the other side of the phone. I think, if it's used as an e ucational tool it could be helpful and supportive of the doctor afd it could improve outcome. If it's going to create confusion in a Pjtient to create second guessing, it could obviously drive a wedge llletween the doctor and the patient. (Internist) Summary. Physicians differed in feir opinions about the impact of implementing a telephone-based decision surport service on their patients and practice. There were three distinct opinion groups rh a majority (n=l8) of the physicians falling into the first two: I) not sure, 2) no impact, and 3) positive impact. Many physicians viewed themselves as a decisiol support "supplier" for their own patients. Reoccurring comments clustered around lhe issues of clinical efficiencies and the I doctor-patient relationship. There were divergent opinions about the impact on clinical efficiencies. Some physicians in + "not sure" and "no impact" opinion groups stated that clinical efficiencies would be improved through a reduction in unnecessary 220

PAGE 234

calls and office visits. Other physicians the same opinion groups believed that there was no such thing as an unnecessary call or visit. Furthermore, these same physicians viewed calls and visits for simple self-c i issues as opportunities to enhance the doctor-patient relationship. Many physicians reiterated that a positive or negative impact on the doctor-patient relationship would primarily result from: a) how the physician positioned the service with his/her patients and b) whether or not the information provided by the service wo+ support the physicians' viewpoint. 8. What will be the most significant challenges vou anticipate following the implementation ofatelephone-based decision SUJ!POrf service for patients? Physicians asserted that the two lost significant and related challenges facing h I f I I d . 1) h .. t e unp ementat1on o a consumer: pat1ent ec1s1on support servtce were: p ys1c1an endorsement and utilization in the everyd y practice of medicine and 2) patientconsumer utilization. Physician endorsement. Resolute concern was expressed that physician buy-in and support for the community telephone-based decision support service had not been sufficiently obtained. I think adoption by the physician community. If we embrace it and refer to it and encmftage our patients to use it, then I think it would only be subcessful. If your physician didn't refer you to it, patients be distrustful of the information otherwise. If an insurance company said to call this number in an effort to avoid emergency room visits and avoid medications and avoid doctors visits, I think that people would be very distrustful with that infontiation. So, a challenge would be, the physician community e1 bracing and using the service. 221

PAGE 235

Referring their patients I told you before, I don't feel comfortable with that yet because I'm not sure I trust the information yet. I think t at's the biggest challenge. I think that there's no advertising that would work other than physician referral. That's the one kr.1 element for the success of the service. (Internist) I think at the community evel, for physicians in general, I think there is a challenge for them to understand the role of the service. Some physicians' are comfortable with patient involvement in decision making than others, and they -some physicians' are I not going to be happy with being 'second guessed', and particularly if they make some for a course of treatment and the patient's may then service and learn about some other options and then call physician back and say, 'hey, you said do this, I heard threugh tlie service that I could do that.' It may not sit well with some ptcians'. (Pediatrician) Community wide the mos' significant challenge has been sus picion, professional about inappropriate or un professional interference in the physician/patient relationship. Along with a concern for people in that does the information meet community standards? Is it in line with community standards? That it's not a one size fi, all. (Gastroenterologist) If I see my patients getting advice that I have a conflict with, then I would be unsupportive. (Internist) It all really depends on I suppose the reason that this is being done. In other is it to just help patients and educate patients, on the one hand, rhich may be the most benign use of it, which I think there still may be potential for interfering with the doctor-patient (Internist) Just the one other thing tht physicians' in the community are wondering about is, how this being paid for? Everybody's talking about saving healtH care dollars. How is this project being funded? I hear a lot !of rumors, I don't think I really have any facts. I don't that being discussed in any explicit way at that informational meeting I went to. In an era where 222

PAGE 236

we're trying to contain cof!s, what is the cost of this and what is the benefit of this, needs to be evaluated, not just by the medical leadership that you're talking about, but I think the physicians' who are buying into it neJd to know what they're paying. If in fact they are. (Internist) I don't want to alienate pJtients in the sense that if you're sitting there and talking to them bd at the end of the conversation you I say, 'and there's the decision-support service now and you can call them.' They're goingjto feel like, 'he's trying to give me the brush off here. Is he sayiqg that I shouldn't have called him for this problem?' You don't \want to have that disjointedness. I'm trying to figure out how te promote to my patients without making them feel alienated, I thinlt: that's a concern. If you say, 'now you can call this 1-800 numbe{ instead of speaking to me.' That would make me feel like 'ph my god, a 1-800 number.' You can talk to me, but this is an option. It's got to be couched in a certain way. It's got to bel something where it's informational, general information. Shouldn't be something that's touted by the I physician as, 'Are the HMO's making you do this?' That type of thing. I think it's got to_bJ couched in a way that if you have any health concerns, things that. (Internist) I will make one very and affirmative statement right now. And that is, the to which this succeeds or fails is going to be determined 'o a large extent by the enthusiasm and participation of the phsicians to take care of their businesses. If the care givers say, 'oh, this is a bunch of crap.' It will be very difficult to the patient's the message of support that you're really trying to Again, I'd make the point that the key dynamic is how reeeptive the physician is to empowering the patient by sharing kno.J.rledge. The limiting factor of this is still going to be to what eJtent the physicians feel that they are able to participate and thefre going to be able to facilitate the I patients in learning and helping. In other words, the potential is that both patients can and physicians' but they both have to come to some understandibg. If I had any reservation about this project it would be that it rbight be just an incredible benefit to have, even at this point, a +ore focused physician education component. I could tell yo that the success or failure of the 223

PAGE 237

whole thing is going to hilige on that. And if it fails it isn't going to be because the concept[s wrong, it's because you haven't got the kind of quality participation that you need to make something I like this show its merits. rephrologist) Patient/consumer utilization. Physicians emphasized that without physician advocacy for the service and without intJgration into the daily practice of medicine, patient utilization of the service would njt be optimized. With sub-optimal consumer/patient utilization, the overall -vllalue of the service to all key stakeholders (physicians, patients, insurers, employers may be compromised. Possibly Patients are used to being able to contact their physicians offices and some people will have a hard time accepting the idea of advice coming from somebody with whoT they have not had any contact. I think that would probably be a potential problem that would have to be overcome before it was widely accepted. (Allergist/Immunologist) The most significant challenge that I see is getting people To understand what the is and to keep it up front in their mind when they begin to have questions. So, I guess it's consumer education about the service is the I biggest challenge. Maybe I should be saying physician education about it, but I tHink the bigger challenge is to have people plugged in and physicians would have to fullow under those (Internist) Probably the biggest one is going to be getting patients to actually pick up the phone land use it. If you think about yourself, if you fell and twisted your ankle, you had a cold and it's been 4 days and not quite getting better, who are you going to call? If you Have a very good relationship with your physician, you're gouig to want to call your physician. (Internist) 224

PAGE 238

First I think it's getting patients to use it. Getting patients to use a valuable source infonnation. It's new, it's different. It's not something that's Pfrt of their health seeking behavior. Frankly I think that the biggest boundary is just that, getting people into the habit of thlnking about this as one of the things they do when they feel a medical issue instead of flexibly calling their doctor, whicH is what they're used to doing. (Family Practice) Summary. The two most significant and related challenges facing the 1 f I d I. . I) h .. unp em.entatton o a consumer pattent easton support semce were: p ystetan endorsement and integration of the servif into the day-to-day practice of medicine and 2) patient/consumer utilization. The primary reasons cited for a lack of widespread physician buy-in and support included: 1. Lack of extensive participation in the development, and decision to implement the consumer decision support service. 2. Insufficient information or misunderslding about the service. 3. Inadequate experience with the servil. 4. Unfilmiliarity with the protocols 1 by the telephone-based nurse counselors. 5. Potential conflicts between the information provided by the telephone-based service and the physicians' opinion abbut appropriate utilization of medical services. 6. Suspicions regarding the primary aim of the service. 7. Uncertainty regarding its value in tenJs of impact on patient health and overall cost-benefit. 8. Uncertainty regarding how to "positior" the service to avoid alienating the patient within the context of the doctor-patient relationship.

PAGE 239

The primary reasons cited for lack ofpa ient/consumer utilization included: 1. Lack of personal physician support, ltilization, and referral to the service. 2. Lack of full understanding by patien\ s about what the service is. 3. Introducing a new mind set and a ne behavior to patients in regards to accessing health and medical infonnation from someone other than their own physician. 9. How do you propose these challenges be addressed? Recommendations for addressing the challenges anticipated following the implementation of the telephone-based decision support service naturally fell into two primary categories: a) how to augment Jhysician endorsement and integration of the service into the day-to-day practice of mldicine and b) how to enhance patient utilization of the service. Physician endorsement/support. Physicians frequently conceded that creating change and spreading a new il among physicians is often "complex and problematic." They also acknowledged tJat compounding the problem in this community initiative was the monumenJ challenge of communicating with and obtaining a critical mass of support from Jhe approximately 2, 000 physicians who comprised the two independent practice Jssociations. They pointed out that mailings I and meetings don't generally work for informing physicians and for obtaining participatory buy-in and support for a nej idea. Additionally, they acknowledged that there will always be a certain subset of who will be reluctant to change irrespective of how well you infonn or en1age them in the change process. However, 226

PAGE 240

informants made several practical recommendations for amplifying physician endorsement and support of the decision support service. I. The most consistent recommendatim1. was to establish strategies for tapping into the already existing formal and infoAnat network relationships within the existing physician organizations. Physicians Jlked about nurturing these fonnal and infonnal networks in more meaninJl ways to engender more physician d &: h I. p () partiCipation an support .10r t e serwce. nmary attention was gtven to: a working with and through the exiJ IP A subcommittee structure in the design, development, implementation and ev1uation of the service and (b) identifying and working with a few key influential p1sicians who are recognized as opinion leaders among their peers. The value If informal physician network communications such as the hospital doctor's lounge should be recognized as a b1 h.&: din I. ab th VIa e mec antsm .10r sprea g commurucations out positive expenences WI the decision support service. j I think we've learned man times the hard way that with any kind of a new program, ydu need to have physician involvement in the design and of the program for it to become accepted. We fun into this all the time as we're trying to improve of providing care, and managing care. You may have a idea but ifyou sort of try and say to the docs., 'do this', doesn't work. But if you say to the physicians, 'we think that fuaybe we can improve the way patients with asthma, or dikbetes in this community are treated, and that a lot of the are not individual physician related, but more systems related. \How do you think the system can help you provide better c 'e.' Then, they get involved and make '2.27

PAGE 241

suggestions. Then you can say, 'we had a group of primary care physicians, and asthma sdecialist's who sat down and they worked on this and these where their suggestions. And then the physician community says, 'okay, I\know, so and so, the internist', and 'I know this asthma specialist, and if they think this is the way we should do it, and that's gding to make things better, okay.' But if it comes from the insurbce company, or some other nonphysician source, it's proBably not going to be accepted. (Pediatrician) It's real hard to make an argument to put the manpower into trying to get out to eight+n hundred physicians on some kind of an individual basis. I think you have to think about what are your efficient kind of like a telephone situation. Say there's a bomb down the street and Joe Blow I calls somebody else and clills three or four other people. I think that if you think ab9ht the existing structures that are there in the medical community and whether you work through medical staffs or whether work through chiefs of departments, departmental meetings; ways that you can give a presentation to 30 physicibs, 15 physicians or 50 physicians. That's the way that I wouid do it if I had to make some choices I around resources. The other piece of that is that I would decentralize that as much as !possible. And in my own work with networks, even just the fact that I was willing to go ten miles and five miles and fifteen fnes to meet with docs at their parent hospital and not expect to come to me went a long way at showing good faitli. (Family Practice) Where we find that we gel the best payoff is when we get down I to the sub-committee level. The 30 sub-committee's representing different physician specia!ks. That, generally, breaks it down into a small enough group\ so that they can understand it and they communicate it to their colleagues, that works pretty well. The only one group that we hare a little difficulty with is the internist's, because there are 400 of those, the sub-committee may be a dozen or so, it's just not big enoJgh to communicate to the rest of them. Most of the other sub-conlmittee's --they represent specialist's that vary in size from 30 to 100-150, and then a sub-committee of 10 or 12 can communicate to many, but to communicate to 400 is a 228

PAGE 242

little trickier, and we're working on ways to try and do that. But I think the sub-committee tbvel, for them to truly understand what the new strategy is, and buy Jto it, that seems to be our biggest payoff. (Pediatrician) I don't know, we tried, sp 1 eaking as a medical director for the IPA, we tried hard to communicate that sort of infonnation, and we had reasonable turn outs.\But, communicating any kind of infor mation to 2,000 physicians is not easy. As a people, they're hard to get together for You know darn well that mailings only accomplish so much, and ffyou can't reach them through the mail and you can't get them to \come out to meetings, how do you communicate to them? AS usual, the best way is through word of mouth, and probably the cbmmunication that takes place in the doctor's lounge in the is probably one of your more effec tive methods, but it's also 1extremely time consuming because you can only reach one or two I docs at a time that way. I think you know what will happen is that solme physicians will become hopefully more comfortable with this type of service and actually see that it helps them take better care patients and say to a.colleague, 'boy, you know I had this patient who had diabetes and they just wouldn't do anything I told them to\do. The decision support service kind of laid it out to them how imE. ortant it was that they manage their sugar closely, and, by golly, this atient really now is working with me and wants to do that.' I think ose kinds of experiences repeated often enough will spread the wotd and that's when the service will become more valued by physicians (Pediatrician). 2. Establish a physician oversight committee with key physician opinion leaders to help guide the initiative. l The continuing communic tion to the physician community I in terms of having a physician oversight committee. That's such an important thing. I \got at least a half a dozen nasty letters about this demand program. Everyone of them after a personal phone call was, 'oh, I understand. Okay, let's see how it goesJ' So having meaningful ties with authentic physician or physician leadership is very, very important. Docs can overtly or covertly consciously 229

PAGE 243

or unconsciously do a lot to influence this. The emphasis is, more can be done to encourage the docs. (Pediatrician) 3. Educate and inform as many doctors in the field as possible about the nature, purpose, benefits, and value of the service to them and their patients. I I think that you'd have to some kind of an educational campaign directed at the that would encourage their trust in the service, and provide concrete examples of how physicians in othet communities have better, happier lives and get home and can actually see their kids because the service has influenced the number of calls they receive. (Internist) 4. Arrange opportunities for key physicians to "experience" the service first hand. Provide examples of how certain cls from patients would be handled. My concern is probably 4 same concern of other people. Even though we've been told a little bit of what's going to be said (by the nurse ), we just haven't heard it ourselves, if you lmow I mean? It is as if we'd like to call up and pretend we're patient, have a conversation, hear what they tell us and, j I think, unless I've kind of heard it myself how this is worldng, I'm not sure I'll feel totally comfortable with it. (Interhlst) 5. Afford key physicians and subcommibees the opportunity to review, approve, and even participate in deVeloping the tot 10-20 protocols that would be most frequently used by the nurse counsel0rs. I have not seen any protolls for the most common things that I would run up againsi. I'm curious as to what they are. You're going to have to lobk at the data and be sure they have data on the commondst calls that have been encountered. Then they have to take the! top 10 commonest calls and provide the protocols to t e physicians. You may only have 130

PAGE 244

to send the top 10, maybe back pain or something to the orthopedist, or a primary care physician and let them know what information is being given out. It will I be a lot of work but it will be worth it in the end. Credibility is the challenge. (Orthopebic Surgeon) 6. Provide frequent physician-specific rback about the perl'onnance of the decision support service in terms of rumber of patients, type of call, and patient satisfaction with the service. I Try and give people feedbrck on how it's being accessed in the community. I think that's going to make me feel better to know the volume is going if you're able to measure any impact on patient satisfaction, or on medical costs in the community that you give drompt, periodic reports back to the practitioner's in the comm'bty about how this product is effecting those areas. (Inttst) 7. Provide physicians and subcommittees not only the opportunity to participate in approving and developing the counJling protocols, but also an opportunity to influence the training and education lf the nurse counselors providing the service. One way of addressing thJ would probably be if physicians in various specialties, as as primary care physicians, took part in the education of nurses that gave advice. If the advice could be made, to seme degree, specific for each patient's physician, that may be difficult, but if somebody called in and identified herself as Dr. I patient, that some how there could be information some of the information that the nurses had at their could be specific for Dr. __ In other words, so the patient could be reassured that Dr. has approved\ofthe information that they're going to be handing to patients the phone. For example, if there was a block of information having to do with asthma, if that could be shared with those of us who take care of a lot of asthmatics and we could s 'well, that's good', or 'perhaps, 231

PAGE 245

I think that could be done a little bit differently, or a little bit I better', and I think it coutd be done on the individual physician basis, or in the tase of the IP A, it could be done by the allergy subcommittee. Each major practice, almost every allergy pradice in the area has a representative I on the subcommittee of IP A, on the allergy subcommittee. I think that would be a of making sure that those physicians who see most bf a given problem could have some input as to what kirld of advice is going to be given. Then I think they would more comfortable with it, and then patient's, in turn, would know that the information that theyve been given been approved by the IP A allergy subcommittee, or if they iCientify themselves as a patient of a pulmonologist, they can y the same thing about the IP A pulmonary subcommittee (Allergist/Immunologist) 8. Assist physicians in how to "positio '"the decision support service to avoid alienating patients in the doctor-pari nt relationship. To actively promote it yo 've got to be careful. It's kind of like advertising yourself in the DNC, it turns people off. Patients expect a certain of professionalism and advocacy from their physifian and if they feel like at the end of it you're saying, 'and now you can call this 1-800 number and get your taken care of without ever talking to me and you won't have to bother me any more', is how it will come across\ It's going to come across like an MCI commercial or something. I think it's got to be couched in a certain way l The glitzy. stuff can be sent to the house but for the physfcians' office, it has to have a professional appearance to it, something that says, 'TBDS I for our patients with concerns.' (Internist) 9. Involve and educate nursing and office staffs who play a key role in patient education. rl32

PAGE 246

I think a way to operationalize this might be to connect with Office managers as much more as with physicians because they are usually the people who kind of do that leg work and get those kinds of things Jut where they belong. Nursing staff would be another key of people to connect with because nurses often get involved in doing patient education activities. So if they sort of know this is a way that they can connect with something,_which wi make job easier, that might be another good thing. (Family Practice) I 0. ClarifY what exactly it is physicians +t and expect from such a service. I think it does need to be elear what the physician organizations want from thls program. If we want it to be strictly, 'okay, you're of twisting your ankle, here are the options that ybu have available to you,' they can give advice of, stay of it, elevate it, put ice on it, that's an option of what you can do yourself It would be listed in the book that was distributed to patients. Versus, another option is, you can call your physician. Call the emergency room and rela)"'l your symptoms to them. Is it strictly options based, and that's what the physician organiza tions want? Fine. If they want to be more directive saying you want people giving that's going to resolve challenges by giving clear guidance as to what physicians want. I see my perceptioimay be different than what the physician organizations a ally asked for in this. Because I do know that physicians, physicians in the community have been clamoring for a ,ong time about how to get a re duced number of phone calls, reducing the amount of telephone based management. Once there may be mixed messages about that. It's kind of, don't want the volume of phone calls but we also don't want faceless strangers dispensing information to our And yet a service like this is just that. The physicians have a conflict here of how to efficiently manage patient dare and not be tied down to the telephone day and and yet giving patients access to all the information (Internist) 233

PAGE 247

11. Complete _an opinion survey of all field physicians regarding their perspectives about patient decision support serviLs and suggestions for successful I implementation prior to program dtopment. I suspect before this was implemented it probably would have been a I idea to have had a significant questionnaire go around all the physicians, clearly the pediatricians, the family the internists who represent basically all thejdoctors that they would be calling about. I think tha they should have surveyed the Rochester physician group to find out how they felt about this sort of thing. Would they support it and if they support it would they have any suggestions on how to do it. (Pediatrician) 12. Assist physicians and employers in forming alliances and partnerships that will c. h r I opturuze success 10r t ese types o commuruty truttatlves. When you start with the in health care, it just changes all the rules and doctors are very worried about this because they're\ not business men. It may seem to everyone on the outside that they're business men, because they make a lot of money they're not business men. They're not used to how functions. There are different ways of forming 1alliances in business, and friendship means different things in ousiness than it does in the personal relationship or than it in the medical society. What is friendship, what is an alliahce? How do partners function? Same way that doctor's dJnt understand patients, I think, there's that same lack ofubderstanding between doctors and I business people. (Clinical Psychologist) 13. Develop a local consumer/patient dLsion support service franchise that is physician owned, directed and operatld.

PAGE 248

Probably, the best solutio? to this challenge would be that the local health care 9ommunity comes up with its own locally based program, protocol's developed and designed by local and staffed by locally based people. I know therr's a joint committee that looked at this program and then gave it the go ahead, both physician and based. individuals, and decided to go ahead It as the pilot trial. (Internist) Patient utilization. Recommendations for increasing patient utilization clustered around two basic strategies: a) physician support and integration ofthe service into the day-to-day doctor-ratient relationship and b) direct marketing and education campaigns to consumers/patients. Patient utilization is the and if you don't do it through the physician, if RCIP A, which I know will go through the physicians. But if their insurance company wants to bypass fue physician, it will back fire, because the physician will be distrustful. Patients will always call the next day to make sure that the advice they got from TBDS was dorrect. If they sense any distrust on the part of the physician, they just won't call again. They'll call the dhysician. (Internist) I'm sure utilization is a ch].enge. The question is what do patients know. What of patients know about it, and those who know about it, where's the number posted? Lot's of times when you nef
PAGE 249

do that kind of stuff. All the types of things that one would think of to utilize Ia new service that patients are not familiar with. ( Ophthalnlologist) Need to have a structure in the workplace that's reinforcing to peo:Rle. I think there ought to be a communicative in the physicians' office, in the physicians' waiting rootd. Otherwise, it's kind ot: well I work is one part of my life and my health care is another part of my life. You ;eally need to integrate that. (Family Practice) Summary. Suggestions for addressing tie challenges anticipated following the implementation of the telephone-based decision support service naturally fell into two . ) h lh . d d. f h pnmary categones: a ow to augment p ystCian en orsernent an mtegratlon o t e service into the day-to-day practice of edicine and b) how to enhance patient utilization of the service. Recommendations for amplifying physician endorsement and support included: I. Working through the existinJ formal and informal network relationships within the existing physician \organizations, especially the IP A subcommittee structure and Key physician opinion leaders. 2. Establishing a physician overlight committee of key physician leaders. 3. Educating as many in the field as possible about the riature, purpose, benefits, and value of the service to them and their patients. 4. Arranging opportunities for Jey physicians to "experience" the service first hand and providing examples of how certain calls from patients would be handled. 5. Affording key physicians and 1subcomrnittees the opportunity to review, approve, and even in developing the top I 0-20 protocols that would be most frequently usetl by the nurse counselors. 236

PAGE 250

6. Assisting physicians in how tq "position" the decision support service to avoid alienating patients in the doctor-patient relationship. 7. Providing frequent physician_\ pecific feedback about the pelfonnance of the decision support service l terms of number of patients, types of calls, and patient satisfaction with the service. 8. Providing physicians and the opportunity to participate in approving and developing counseling protocols as well as an opportunity to influence the training and education of the nurse counselors providing the service. 9. Involving and educating nurs1 g and office staffs who play a key role in patient education. 10. Clarifying what exactly physijans want from such a service. 11. Completing an opinion all field physicians regarding their perspectives about patient support services and suggestions for successful implementation prir to program development. 12. Assisting physicians and empleyers in forming alliances and partnerships that will optimize success for these types of community initiatives. 13. Developing a local decision support service franchise that is physician owned, directed and operated. Suggestions for increasing patient utilization included: (l! enhancing physician support and integration of the service into he day-to-day doctor-patient relationship and (b) direct marketing and ucation campaigns to consumers/patients. 10. What will be the most significant bejfits you anticipate (pi/owing the implementation of a telephone-based deJision support service for cnsumers!patients? What else would you like me to know abdut your thoughts regarding the value of the telephone-based consumer decision support service to the community? I Physicians indicated that the most significant benefit to the community would 237

PAGE 251

be increased consumer/patient access to good quality, uniform information about health and medical concerns. Access to good informati1 n, is the biggest benefit. If this can improve the quality o the information that's readily accessible, it's going to b a benefit. (Internist) And I would just say that this TBDS has that one strength. And that is t least you know the informa tion that they are giving is uniform. (Pediatrician) Physicians suggested that increased infoii ation access would ..;pport patients in making better decisions about use of he th care resources resulting in: a) improved clinical efficiencies, b) higher patient sati faction, c) enhanced health outcomes and d) reduced health care costs. Clinical efficiency. Physicians bllieved that clinical efficiency would result from a reduced volume of phone calls an medically inappropriate visits. As it stands now, reducing the phone call volume. That's the biggest problem in my office and probably most offices. We keep adding phone lines and people are answering them and they just keep cdming up. A lot of it is, hopefully, to reduce the phone volume the patients that are having chest pain, or coughing up blood can get through. That's a lot of it, is trying to get a better ha4dle on handling that phone volume. In the long run, I don't know if it's going to have much impact on anything else. (Int4 From the doctor point of'?ew, to the extent that it educates patient's and may perhaps save them time If patients are being educated by this ser+ce, and come in and actually the doctor spends less time with them, and the time is a higher quality time in the sense thltt they get right to the issues that 238

PAGE 252

are pertinent to that patient, that would clearly be a benefit. Obviously, both to the pa1 ients and to the doctors. (Ophthalmologist) Patient satisfaction. Physicians sserted that a more informed, self-directed patient would result in higher patient sati faction. I would like to see happie and more self directed patients. Would like to see less cal s for self limited disease. I would really like to bump-up the sophistication and tenacity of selfmanagement in patients "th chronic disease. To me that's the largest opportunity here. (Family Practice) more uniform health and medical inform ion would be a more informed patient which in tum would foster better illness manalent and improved health outcomes. We're hoping as patients nderstand both their minor and their major illnesses bett they will make better use of the medical care system, and oe less demanding of inappropriate services and more able to I manage their illnesses to provide better outcomes and less domplications. (Pediatrician) For those physicians who L.t providing their patients the information, it might uhprove health outcomes . But if there's no other access tb that information, then it would improve health outcomes. 1The patient would be more satisfied. (Internist) Health care costs. Physicians reiterated that more informed and involved patients would result in improved health Jare cost management without sacrificing quality of care and patient satisfaction. 239

PAGE 253

I would be happy if the bfnefit that we saw was a cost neutral improvement in P!ltient satisfaction. rd think it was a very successful prdgram if it didn't cost us money, and people were happier] I would also consider it a great benefit if we improved clst outcomes. (Internist) But if what they are doing is to support the doctor/patient relationship, then, by getting patients more involved in their own h th care could improve outcomes, and thus also, obviously prove costs.(Intemist) Doctor-patient relationship_ P,sicians held conflicting views about the potential benefits to the doctor-patient rionsbip. Some physicians were of the opinion that patient access to outside their personal physician, including a telephone-base decision support service, rould potentially disrupt the doctor -patient relationship while others held an equally strong opinion that patient access to information would support and stimulate higher quality physician-patient communications. Anything like that, any flow of information outside of the venue of physicians, I'm rlot saying we're a deity, but it can effect that relationship bedause the patients going to say to himself, 'I got this from Internet. I got this from Prevention Magazine. I got this from the decision support service.' The information fge, I think, in lots of ways can be the physicians' greatest enemy or greatest ally. It depends on how it's going to be (Internist) At least I was somewhat ebused about the service. I think it will help open up between physicians and patients. It's not going to tnake somebody who's totally non assertive and unable to the simplest question of their physician into a strong person that comes in and says, 'here, I want to have a distussion about what's wrong with 240

PAGE 254

me today.' They'll still be timid and meek and difficult to I deal with and easily rolle9 over. But, sometimes there are alternative ways of getting information. People may get information in that route. I It may tum a very ineffectual physician/patient relatioruJhip into a very good one. I'm very positive about it. ( oftroenterologist) I think that if the patients to understand that this is something that can be valf able to them they can sort of insist that their physicians become better listeners. It isn't a power struggle of who' going to instruct whom, but in fact, the patient walks in fd says, 'doctor, I'm not critically ill but I have certain concerns about this area and I want to pay for your apt attention \for about 7 minutes, not interrupted by phone calls or anything else.' The doctor is going to respond to that. The docter's not going to immediately jump up and say, 'I can't fulfill tease expectations.' But on the other hand if you go in and watch a doctor in the course of a day you wonder if they evrr think of that in the whole process. And I would hope that one of the benefits might be that the physicians will come to thoughtfully reflect a little bit more on the quality of communications that they receive and partici pate in. Let's say that I several patients who have made phone calls, and had absolutely excellent information and have come away with the perception that whoever was on the other end of that line really was to what I had to say and gave me their best effort, tp.oughtful individual, consideration. And they go in and they tell their doctor that. That would be incredible communication. The doctor would be unlikely to ever have it considered that the patient got a better quality communi cation on a phone line to a decision support service than right here in his office. So I would think that the possibility exists that there may be ... that patient's may drive more interest and receptivity from the pHysicians in what's really a very important two-way commJnication. (Nephrologist) Summary. Physicians indicated thjt the most significant benefit to the community would be increased consumerJr atient access to good quality, uniform 241

PAGE 255

infonnation about health and medical concerns. Physicians purported that increased infonnation access would support patils in making better decisions about use of health care resources resulting in: a) imJroved clinical efficiencies, b) higher patient satisfaction, c) enhanced health outcomjs and d) reduced health care costs. In contrast, physicians presented mixed vieLs regarding the benefits to the doctor-patient relationship. Closing Questions The last two questions of the interview were intended to provide one final opportunity for physicians to provide additional thoughts and opinions that may not have come forth during the intervie"l 11. What else would vou like me to know about your thoughts regarding a more active involvement ofpatients in health imd medical decisions? The issue that physicians most fr1quently mentioned was the fact that medicine had entered into a new era. It was their ew that this new era is one that is more consumer centric. They emphasized that this new era of consumerism will be characterized by advancing sophistication and rapid growth of communications and infonnation management technologies. cbnsumers will accelerate the access of health and medical information from various soirees such as the Internet. A majority of the h dhi d I ed =-c. ed d p ystctans vtewe t s as postttve an SURPOrt a more uuorm an active pattent while recognizing that informed and actiJely involved patients may 42

PAGE 256

require them to make changes in the wa they manage the doctor-patient relationship and how they approach physician-patieJ communications. Specifically, many believed that J would become increasingly important for them to hone their physician-patient colunication skills. Many talked about the need to take the time necess8I}' to have 1 higher quality dialog with patients that would include discussion about not only the biological aspects of health and disease, but also about the social and culturaJ Jors affecting patient's health and health care utilization. Many conceded that with smJe patients they will need to become less paternalistic in their approach. Physicianl will have to become more sensitive to patient preference for involvement in h.Jth and medical care decisions and adapt their style to that preference. They acknowl ged that with many patients, this may require patient relationship. The dilemma they regularly spoke of was how to do all of this in practice environments and health care that are becoming more volume dependent and productivity-driven every Ly. Second, they reasoned that with the advancing sophistication of communi4n technologies, providers will be required to expand their communication modalities :6:-bm the traditional face-to-face encounter to include interactions with patients via the Jntemet, telephone, and voice mail. One of the most :frustratinl things to me is the patient who comes to see me and rays, 'I have this pain, what is it?' Its like they don't understand the necessity of the questions and answers, gi ng as much information as 243

PAGE 257

possible. It's :fiustrating in that they seem to be, the expectation is, 'if I tell y9u I have a pain, obviously this pain can mean only one thing.' Where as, it's only through the process ef communicating, asking the questions, listening carefully to the answers and then feedback my analysiJ of the infonnation that I I can really help a person. jT o me, communication, active patient participatioh is the only effective way to practice medicine how a days. (Internist) I really think tbat I'm modem in my view that patients have a right to kdow what's going on in their medical life. And I think that it's our job to provide good education or at lead have literature and/or things that we can talk about to tnake sure that they're educated. So, I clearly think that the!1 days where the patient comes in and gets a prescription and doesn't even know what the prescription is, are gone. [hat's the way is was when I was growing up. (Pediatribian) And it's the same when thl bring something in to me. The latest newspaper this week was on this lipo protein subset A, or it was. I said, 'gee, you know as usual the newspapers got it before JAMA does.' I don't know about that ndw. Thirty years ago it would have been hard for me to 'I don't know.' I would have probably whiffed wafted and said, 'oh, it doesn't amount to anything.' But I think it's important that they understand that we don't tdnow a lot of this stuff either. But we're going to find out and we're going to do the best that we can to advise the,. (Internist) It should be about a broad [array of things that are far beyond the fact that doctors JUSt use symptoms as the excuse that gets the patient in that you can do what you want, what you need to d9 with a patient. All of those need to be expanded. expansion is, well, how do you communicate? Do yol communicate in person? Do you do it by phone? Is it by TV? Is it by videos? Where is it? How is it? So, I'd be r open minded and experimental I 144

PAGE 258

with regards to that. Different ones are great for different people. Think that there brobably won't be one modality that's going to work. If ytl>u want to talk to my father it's going to be through his cbmputer and not through the tele phone because he'll be on! his computer as many hours as he can. And he hates to on the phone. He would never call someone on the telephone. But if he can down load information about a health care issue he has he would love to do that. It should be multi-modality. (Family Practice) The way that we used to lransmit information was one to one about one issue. The tloctor to the patient about the symptom. That's the The doctor. The patient. The I symptom. We've got to eJg>and all of those. It doesn't have to be the doctor. are a lot of other possibilities than the doctor. It doesn't have to be face to face. There I are a lot of possibilities otrer than face to face .. (Family Practice) 12. What else would you like me to know about your thoughts regarding the value of the telephone-based decision support serwice to the community o(Rochester? The responses to this question +minimal and did not include any substantial content. Physicians responded to this question in two principal ways. Most indicated they had no additional coLents as they felt their opinions about this had already been captured in other parts Jfthe interview, especially question 10 regarding the "most significant benefits fjllowing the implementation of telephone based decision support. For those few lo did provide minor additional comments, their inputs were integrated into the analylis and summary of question 10. 245

PAGE 259

A major purpose of this analysis as to present and organize key informant responses in a way that provided more depth insight to each question as well as a more clear overall pattern of emerging t emes and issues. This summary provides a synopsis of common themes that emergeCi across all questions in the analysis. Where appropriate, each thematic section comples and contrasts physicians' general opinions and perceptions about that issuJ with their opinions about that issue in relation to the implementation of a decision support (TBDS) service for patients. Theme 1: Support Idea of More lnformeCi Active Patients Physicians fundamentally support d and actively endorsed patients' and consumers' desire for more information out health and medical issues. They acknowledged that there is a growing co sumer trend characterized by patients who have a preference for more information rather than less. Physicians principally believed that a more informed patient woJld result in better health and medical care decisions leading to improved quality of cLe, health outcomes, patient satisfaction, clinical e:fficiencesand doctor-.patient relatJonships. 246

PAGE 260

Similarly, physicians held a fundamentally positive opinion about patients and consumers wanting more active particip tion in the medical care decision-making medical decisions would be have a great r sense of control and self-confidence culminating in greater compliance with r upon diagnostics and therapeutics. This would result in improved health outcomei and higher overall patient satisfaction with care. \ TBDS: Physicians indicated that the most significant benefit ofTBDS services to the community would be increased coJsumer/patient access to good quality, uniform information about health and mjical concerns. They suggested thai increased access to this type of informati1 would support patients in making better decisions about use of the health care Theme 2: Improved Health Outcomes: Chronicl!vlajor Dlness Physicians believed that, in gen,, a more informed consumer/patient would be more likely to make better decisions usually resulting in superior quality, more efficient care and improved health outcomls. They believed this to be particularly true for patients with chronic illnesses. Also, jost physicians contended that outcomes would improve when patients are involved in decisions about major medical issues.

PAGE 261

They explained that the mechanis responsible for improved outcomes was related primarily to elements of self-efficlcy: improved patient self-confidence and control over body, life, and disease leaJg to improved compliance and overall satisfaction. Physicians clarified that evel if outcomes are less than what was expected, a patient who has a higher undlrstanding of their illness and who more actively participates in treatment decisionl and self-management will still have a higher satisfaction level, and, will be less likely tllitigate despite poor outcomes. Physicians reiterated that patient Jreference for active involvement in medical decisions was a key factor influencing oLes. They contended that the probability of improved health outcomes would be .Jgmented if physicians were sensitive to patient preference for active involvement Ld matched their response to that I preference. If physicians failed to match their response to patient preference, then health outcomes and overall patient sati+on may be diminished. Several physicians noted that patient involvement in decisions about treatment fur minor health issues would have neith, a positive nor an adverse effect on outcomes. Many physicians reasoned that most minor health issues were self-limiting, ld th. al dh I. d .. wou run err natur course, an t at patient partic1patton m treatment eas1ons would have no significant effect one way dr the other on clinical or functional status outcomes.

PAGE 262

TBDS: Physicians purported tha one of the benefits of patient access to more I uniform health and medical information t ough the TBDS service would be a more informed patient. In fact, they believed t at this was one of the most significant strengths of the TBDS service. They pos ated that this would favor better patient d .. db illn lui ... d. c.. d ecisions an etter ess management res tmg m Improve patient satis1act10n an health outcomes. Theme 3: Improved Patient Satisfaction (Patient Preference) Physicians acknowledged that a ore active informed patient in decisions about minor, major and chronic illness tould be a more satisfied patient overall. Consonant with health outcomes, they ,lained that the mechanism responsible for higher patient satisfaction was related primarily to elements of self-efficacy: improved patient self-confidence and control over bldy, life, and disease leading to improved compliance and overaD satisfiu:tion with ttment. Physicians reiterated that patient prence for involvement in varying types of medical decisions also influenced patient satisfaction. That is, if clinicians failed to be sensitive to individual patient differences l decision-making preferences for minor, d hr il1n edicald I h di d maJor, an c orne ess m ecrsions, p ysician-patlent scor ance may occur, resulting in lower patient satisfaction. Thi\y also believed that satisfaction was affected by the patient's perception about the phys ian being "indecisive", "unconcerned" or r

PAGE 263

"uncaring" if the patient was encouraged to take a more active and prominent role in the decision-making process as describ in Theme 4. TBDS: Physicians indicated tJ the most significant benefit ofTBDS to the community would be increased consumJ/patient access to good quality, uniform information about health and medical coLems. Many believed that increased access to this type of information by patients J indeed translate to higher patient satisfaction with health plans and medical services in general. Theme 4: Improved Doctor Patient Relationship (Patient Preference) Physicians acknowledged that a more active involvement of patients in the medical decision-making process is a trend in today' s care environment and one that physicians need to be prepared to deal with in the dynamic of the doctor-patient relationship. Physicians contended that, generally, a more informed active patient in decisions about minor, major, and chronic illnesses would enrich the doctor -patient relationship and Ice patient satisfaction, if clinicians were sensitive to individual patient differences in preferences for involvement in decision making as previously described. Although physicians were mainly about patient involvement in decisions, some expressed concern that mLe informed actively involved patients have the potential for negative impact on the ddctor-patient relationship. Some physicians 250

PAGE 264

expressed an apprehension about proactively involving patients in decisions and giving them a more prominent role. They percer that from the patient's viewpoint, the physician may appear to be "indecisive," !'uncaring" or "unconcerned," especially if s/he referred the patient to a TBDS servibe. One physician compared it to coming across like "an MCI commercial." TBDS: Physicians held conflictin views about the potential benefits ofTBDS to the doctor-patient relationship. Some physicians were of the opinion that patient access to infonnation outside their persor physician would potentially disrupt the doctor-patient relationship. They were clncemed that the information provided by the TBDS service could potentially be in conflict with their viewpoint about a given medical issue. Many physicians contendt that if the infonnation dispensed by the TBDS service was in agreement with their opinion, they would be inclined to support I the service. If the information dispensed by the service was not in alignment with their view about a given issue, they would not for the service. Second, as described above, phyj didn't want to be perceived by patients as "indecisive", "uncaring" or "unconcerned" if slhe referred the patient to a TBDS Thi . .lh h h h sernce. s vtewpomt was m concert Wit t e contention t at one way to strengt en I physician support and utilization of the TBDS service was to provide physicians with assistance in how to "position" the servi, to avoid alienating the patient within the context ofthe doctor-patient relationship (Theme 11). 251

PAGE 265

Other physicians were of the eq ally strong opinion that patient access to telephone-based decision support wouldlstimulate higher quality physician-patient communications. They were comfortabll knowing that patients are accessing infonnation from a variety of sources (nlwspapers, self-help materials, friends, Internet) and that the infonnation being 1spensed by the 1BDS service was at least uniform, standardized, high quality inforon. These physicians optimistically viewed patient inquiries about infonnatii they had obtained elsewhere as opportunities to open constructive dialo and to enhance communications within the doctor-patient relationship. Theme 5: Improved Clinical Efficiencies for Minor and Chronic mnesses l Physicians contended that more a, tive patient involvement in decisions about minor and chronic disease conditions woLd definitely have a positive impact on clinical efficiencies and cost management sacrificing quality of care. Out of the three categories for patient involvl. (minor ilhtess, major illness, and chronic illness), physicians emphasized that involvement in the management of chronic disease held the most value and would hate the most significant overall impact on health care system quality, cost, and efficilncy factors. Most were of the opinion that the number of medically inappropriate vi) s to their office and the volume of unnecessary phone calls would be reduc 252

PAGE 266

However, conflicting opinions w re expressed regarding whether or not involving patients in decisions about maj!l r medical issues would be more efficient. The gamut of opinions ranged from thos who believed it would definitely be inefficient and would take more time to who believed it would definitely be more efficient. Others were of the opinion that involving patients in these types of decisions may take more time initially, but in the long run it would be more efficient and would result in higher patient satisfaction. TBDS: Physicians were uncert and differed in their opinions about the impact and benefit they thought the implementation of a telephone-based decision I support service would have on clinical ejciencies. Some stated that clinical efficiencies would be improved through a eduction in unnecessary calls and office visits. Other physicians contended that thte was no such thing as an unnecessary call or visit. Furthermore, these same physicii viewed calls and visits for simple, selflimiting, self-care, minor issues as to enhance the doctor-patient relationship. One of the major driving forces fo The Rochester Community Initiative for Demand Management and the eventual TBDS service was an identified need by formal physician leaders to reduce the mnnber of Ledically inappropriate and wmecessary "nuisance" calls and visits to their practicJ because of increasing pressures to manage costs and optimize productivity. 2153

PAGE 267

As far as a demand mana ement issue that, certainly we've been talking about for yers. We need to change patient demand or it's going to control costs. A well-informed patient who demands appropriately, that's going to tum out to be a good trurlg. (Internist) However, it was evident that th+ was conflicting opinion and lack of consensus about the specific need and there was lack of clarification regarding exactly what a majority of practicing physicians from a TBDS service in regard to improved clinical efficiencies. On the onl hand, many physicians believed that a TBDS service would help manage patient demJd and appropriate use of the health care system. This would allow physicians to janage costs and improve productivity. On the other hand, there was a group of phyll icians who contended that the TBDS service was unnecessary because physicians shol!lld be their own "TBDS supplier" for their own patients (ThemelO). I do think it does need to be clear what the physician organizations want from this program. I see my perceptions may be than what the physician organizations actually askJd for in this. Because I do lmow that many in the community have been clamoring for a long time how to get a reduced number of phone calls, redhcing the amount of telephone based management. Once \again there may be mixed messages about that. It's Kind of, 'we don't want the volume of phone calls, but\we also don't want faceless strangers dispensing information to our patients. And yet a service like this is just that. The physicians have a conflict here of how to effibiently manage patient care and not be tied down to the telbphone day and night, and yet giving patients access to all the information. (Internist) 254

PAGE 268

Several physicians pointed out that what exactly a majority of practicing physicians wanted from a TBDS service was key to physician endorsement and use of the service (Theme 11 ). Theme 6: Reduced Health CareCosts for Minor and Chronic Conditions Congruent with improved clinical fficiencies, most physicians were of the opinion that involving patients in miTKIT ih issues and chronic conditions would help reduce health care costs without compromising quality of care. They believed I this to be particularly true for chronic coniitions. However, physicians expressed uncertainty and mixed opinions about the effect of involving patients in decisions about major medical issues on health care bosts. At one end of the continuum were those who were certain that patient invowlnent would reduce health care costs because patients would employ more appJpriate use of medical care services. On the other end of the continuum were those whl held an equally determined opinion that patient involvement would increase health 1are costs. Some cited that more informed active patients may actually request and inJ st on more services rather than less. TBDS: A small number ofphysici s expressed hope that the TBDS would have a positive impact on health care costs or the community. A predominant number of physicians were either uncertain or belie ed there would be no impact on health care costs for their particular practice. 255

PAGE 269

Theme 7: Reduced Potential for Medic Liability Physicians affirmed that medical bability would definitely be reduced by appropriately involving patients in decisi I ns about major medical issues and in chronic disease management. However, many expounded that "involvement" meant being able to establish a genuine relation hip with a patient that would allow the physician to understand, communicate, and relate to the patient and their illness in a more holistic approach. To emote and e I pathize with them as well as educate and inform about treatment options, benefits, d risks. Physicians contended that with this type of physician-patient communica on, medical liability is reduced, even when the outcomes are not as positive as expeJed. Patients who have a better understanding of their disease and particitte in diagnostic and therapeutic decisions are less likely to sue when things do go Jong. TBDS: A predominant number o physicians indicated that either they were "not sure" or that there would be "no irnplct" of the TBDS service on reducing the "al.C'. di alliabili" h .L.l potentl 10r me c ty m t etr practice. 256

PAGE 270

Theme 8: Self-Efficacy Improved with ore Informed Active Patient Physicians frequently attributed ,proved doctor-patient relationships health outcomes, patient satisfaction, clinical em!ciencies and cost management primarily to I the mechanism of enhanced patient self-efficacy. They explained that when a patient refers to take a more informed active role lin decision making and in the self management of disease, they feel more in Lntrol of their lives and the disease, more confident to deal with the disease a po) tive fashion. Summary: Themes 1-8 Physicians fundamentally support the concept of more informed, actively involved patients in health and medical carl decisions. Physicians were of the opinion that a more informed, actively involved palients would result in better health and d.cal d 1 d. I ali d me 1 care ec1s1ons ea mg to unprovements m numerous qu ty, cost, an utilization variables. Physicians' opinions Lding the potential impact of more informed actively involved patients in medibal care decisions are summarized in Table 4. 4. Physicians frequently attributed the uJderlying mechanism of improvement for these potential impacts to enhanced patienJ self-efficacy. 2fJ7

PAGE 271

Table 4.4. Summary of physicians' op ons about the potential impact of more informed active! involved atients in various es of medical decisions. Variable Type of Medical Issue Minor Major Chronic Health outcomes ..1 t t Patient satisfaction t t t Quality of doctor-patient relationship t t t Volume of unnecessary calls J.. ? J.. Number of inappropriate visits J.. ? J.. Length of clinical encounter J.. ? J.. Health care costs J.. ? .J, Medicalliability ./ .J, .J, t = increase, .J, = decrease, ./ = neutrai!no impact, ? = uncertain Theme 9: Variance in Physician and Patil Preferences Physicians differed in terms of t1ir preference fur involving patients in decision making about minor as well as major health issues. For minor and chronic health issues, the tendency was primarily lowards the patient having a more prominent role in the decision-making process. In lntrast, preference for decision making fur major health decisions was primarily incled towards the physician having a more prominent and directive role. l They reasoned that two primary :fi ctors contributed to the need for physicians to have a more prominent and directive Jle in major medical decisions. First, many physicians believed that severity of illness is a variable influencing the degree of patient and physician involvement in major medichl decisions. They tended to equate severity of illness with the need for a higher level f clinical knowledge about the condition; 258 I

PAGE 272

therefore, the perceived need for a highe level of physician involvement in influencing and guiding decisions about treatment. Physicians said that as patients grow sicker, they generally tend to place more trust in reir physician and tend to want the physician to take a more prominent and directive role in decision making because of his/her clinical expertise. Second, as physicians varied in their preference for involving patients in decision making, physicians recurrently elphasized that patients also differed in their preference fur involvement in the decisiimaking process. That is, some patients may prefer a more passive role ("I prefer that my doctor tell me what to do") while some may prefer a more collaborative approach ("I prefer that my doctor and I make the decision together''). Several physicians ccmceded that if clinicians failed to be sensitive to individual patient differences il decision-making preferences for minor and llll!ior medical decisions, physician-patient rscordance may occur, resulting in lower patient satisfaction and a less than optimal (ioctor-patient relationship. In gl .. I b J:'. J:'. terestm y, m contrast to opm10ns a out patient pre1erence 10r actlve d d' a1 d I h mvo vement m mmor an maJor me IC ecistons, p ysiCians were not as mtense nor as incessant in their opinion about the need to recognize individual differences in patient preference for active involvement in the management of chronic illness. Only h .. d. ablhib' J:'. two p ystCians rna e mmor comments out t s emg a concern 10r patient involvement in chronic illness. For a majojty of the physicians, they recognized that 259

PAGE 273

differences in patient preference for acti e involvement in managing chronic disease existed. However, regardless of patient preference, it was almost as if extensive patient involvement in treatment decisioJs and self management was a given, an uncompromising and necessary componr\ t in the successful management of chronic disease. That to achieve better outcome patients had no choice but to become more actively involved in decisions about and for chronic conditions. Worth repeating here is how one summarized the essence of how important it is for both physician and paJent to be aware of this issue in the doctorpatient relationship. I think, and this is part ofl:hat I think is what the art of medicine is. And that is, is the physician appropriately reading the patient's desire for inv lvement, and matching his response to that desire. And if the hysician is able to do that, then that should enhance the relatio'Fhip. If the physician and patients aren't communicating that need appropriately, then there will be problems. Let's say, ifthe/1 atient really wants to be more actively in decision making, and th physician doesn't pick that up and adopts a more paternalistic\ approach, then the patient's not going to be satisfied. U: on the hand, the physician feels that it's important for the patient t9 be completely involved in the decision making and starts providing the patient with information and expecting patient's to make decisions, and they're not ready to do that! and they don't want to do that, then that's a in the other direction and that could cause problems J.lso. So, I think, the key here is for physician and patient to\sort ofbe able to read each other, and communicate to each other where they are in that desire to be involved in decision nWang, and then there would be a good fit and a good match. (Pediatrician) 260

PAGE 274

The take home message for clinicians is that for any given patient at any given time, there may be a variety of sociodemigraphic, intra-personal, and situational factors that determine a patient's preference for more information and/or more active involvement in different types of medicallare decisions. The most prudent clinical approach for the physician is to be sensitile to each individual patient's preference for I information and to the types of decisions for which patient involvement is most likely to be preferred and beneficial. Failure to 1o so on the part of the clinician may result in physician-patient discordance and a IesJ than optimal doctor-patient relationship with ensuing negative impacts on patient jarisfuction, health outcomes, and medical liability. Theme 10: TBDS -Uncertain or No Impact on Physicians' Practices and Patients Physicians differed in their opinions about the impact they thought the I implementation of a telephone-based decision support service would have on their patients and practice. A majority (n=18) physicians were either uncertain or I believed there would be no impact of the 1BDS service on their practice. Many noted they have not had enough experience with lhe service yet to really know what type of I impact it might have on their patients and their practice. Others candidly stated they weren't familiar enough with the service to know whether it would have a substantial impact on their practice. 261

PAGE 275

Another group of physicians said that the service would have "no impact" on their patients and their practice because t ey personally fulfilled the role of the service by providing their patients with the info arion and counseling they needed to be more active, involved consumers. They themselves, as one physician put it, as a "TBDS supplier'' for their own patients.l Many physicians were somewhat decisive and contradictory in their views about the impact of the TBDS service wt it came to speculating about the impact of the service on their practice. Many physicians stated that either they were unsure of the impact or that there would be no imp \ct ofTBDS services on their particular practice. Theme II: TBDS Lack of Support by Practicing Physicians Will Diminish Its Value Even though there was near unanimous support and endorsement for the concept of a more informed and active coAsumer of health care services, there were mixed opinions about the actual implemenlation of a community telephone-based decision support service for patients/consuLers. Three, eleven, and seven of the key informant physicians held generally neutral, and generally positive opinions about such a service, respectively. Even tllough only three physicians held an explicitly negative view about the service, seemed to be a prevailing attitude of 262

PAGE 276

"let's wait and see" skepticism or uncertainty even among the more positive physicians. It was the resolute opinion of these physicians that extensive physician endorsement and support fur the mos tee bad not been obtained. Because of this, they asserted that the two most signijficant and related challenges facing the implementation of the TBDS service werl: 1) physician endorsement and utilization in the everyday practice of medicine and 2) latient/consumer utilization, primarily due to lack of endorsement by the practicing phjsician community. The primary reasons cited fur a lack of widespread physician 1y-in and support included: 1. Lack of extensive participation in the design, development, and decision to implement the consumer decision support service. 2 lnsufli :_c. . d \ d" b th . etent uuormatton or mtsun erstan mg a out e sernce. 3. Inadequate experience with the servicJ 4. Unfamiliarity with the protocols used the telephone-based nurse counselors. 5. Potential conflicts between the inform1on provided by the telephone-based service and the physicians' opinion abo'ut appropriate utilization of medical services. 6. Suspicions regarding the primary aim or the service. 7. Uncertainty regarding its value in tenJ of impact on patient health and overall cost-benefit. 8. Uncertainty regarding how to "position" the service to avoid alienating the patient within the context of the doctor-patient relationship. 263

PAGE 277

Practical suggestions for ampl.ifyih.g physician endorsement and support included: 1. Working through the existing formal and informal network relationships within the existing physician organizations, e] ecially the IP A subcommittee structure and key physician opinion leaders. 2. Establishing a physician oversight rttee ofkey physician leaders. 3. Educating as many physicians in the f.eld as possible about the nature, purpose, benefits, and value of the service to fhem and their patients. 4. Arranging opportunities for key to "experience" the service first hand and providing examples of how certain calls from patients would be handled. 5. Affording key physicians and the opportunity to review, approve, and even participate in developing thJ 1 top I 0-20 protocols that would be most frequently used by the nurse counselor 6. Assisting physicians in how to the decisions support service to avoid alienating patients in the doctor-patie11t relationship. I 7. Providing frequent physician-specific (eedback about the performance of the decision support service in terms of mhnber of patients, type of call, and patient satisfaction with the service. 8. Rendering physicians and the opportunity to participate in approving and developing the counselihg protocols and the opportunity to influence the training and education o1the nurse counselors providing the service. 9. Involving and educating nursing and office staffs who play a key role in patient education. 10. Clarifying what exactly it is physicians want from such a service. 11. Completing an opinion survey of all field physicians regarding their perspectives about patient decision support serviceJ and suggestions for successful implementation prior to program deve opment. 2'4

PAGE 278

12. Assisting physicians and employers f forming alliances and partnerships that will optimize success for these kinds of community initiatives. 13. Developing a local consumer/patienl decision support service franchise that is physician owned, directed and The primary reasons cited for a pbtentiallack of patient/consumer utilization included: 1) lack of personal physician utilization, and referral to the service, 2) lack of full understanding by patients kbout what the service is, and 3) introducing a new mindset and a new behavior to pa1ents in regards to accessing health and medical information from someone other than their own physician. Suggestions for increasing patient utilization included: a) lnhancing physician support and integration of the service into the day-to-
PAGE 279

c \TER5 DISISSION This chapter presents a discussion of key quantitative and qualitative results including practical implications and limijons. Recommendations for future research and conclusions are provided. Key Quantitative an Qualitative Results Using a multi-method (quantitative Ld qualitative) approach, the specific aims of this study were to: I. Ascertain primary care and sp,alty physician opinions about more informed actively involved patients as partners in health and medical care decisions. 2. Assess primary care and specialJ physician opinions about the impact they believe consumers/patients being more info ed and taking a more active partnership role in health and medical care decisions will have on: (a) the quality of the doctorpatient relationship, (b) clinical efficiencies, outcomes, (d) patient satisfaction, (e) health care utilization, (f) h th care costs, and (g) malpractice liability. 3. Assess primary care and specialty physician opinions about the incorporation of TBDS services into care management and the impact they believe a 266

PAGE 280

TBDS service will have on: (a) the quali of the doctor-patient relationship, (b) clinical efficiencies, (c) health outcomt (d) patient satisfuction, (e) health care utilization, (f) health care costs, and (g) malpractice liability. Ph sicians' 0 inions about Shared Decisit Makin Physicians in this study fundamen y supported the idea of more informed and actively involved patients in the health medical decision-making process. Overall, approximately 90% of physicians rieved that patients generally desire more information about health and medical issue!; and 80-90% of physicians believed that patients generally desire more active invo/v, ment in the medical decision-making process. Regression results indicated that tLse opinions were consistently strong across all physician and practice characteris1cs. Unknown is to what extent these I beliefS translated practice behaviors i what these opinions were concordant With patient preferences for more information and/or active mvolvement m the medical decision making process. Exlation of physician-patient concordance variables is suggested for future studies. Not only did physicians believe that latients generally desire more information and more active involvement, this group of physicians endorsed the fact that more informed and actively involved patient decisiln making was good for patients and for medicine. Specifically, they believed that a ore informed patient would result in 267

PAGE 281

better health and medical care decisions and a greater sense of control and selfconfidence fur the patient ultimately 14 to improved quality of care, enhanced health outcomes (for chronic conditions Jd major medical issues), higher patient satisfaction, enriched doctor-patient relati1nships, potential reduction in medical liability; and especially for minor illnesses and chronic conditions, improved clinical efficiencies (reduction in the nwnber ofm+cally inappropriate visits and volume of unnecessary phone calls) and cost controllable 4.4). Physicians' opinions from both the quantitative and qualitative results of the present study substantiated that patient involvement in the self-management of chrorc disease was absolutely essential for optimizing health outcomes, enhancing patir satisfaction, enriching the doctor patient relationship, improving clinical efficiencies, cost management, and reducing the potential for medical liability. Physicians' opinions about the poten "al impact of more informed and actively involved patients are in agreement with Jous studies that have illustrated the positive impact of more informed and involved patients in decision making (Ong, DeHaes, Hoos, & Lammes, 1995; DiMatteo, 1994; Lerman, Brody, Caputo, Smith, Lazaro, & Wolfson, 1990; Lorig, Mjonson, & Holman, 1993; Brody, Miller, Lerman, Smith, & Caputo, 1989; Greenfield, Kaplan, Ware, Yano, & Frank, 1988; Wasserman, Inue, Barriatua, Carter, & Lippincott, 1984; Vickery, Kalmer, Lowry, Constantine, Wright & Loren,1983). 268

PAGE 282

When physicians were asked mor specifically their opinions about patient involvement in decisions regarding uea4ent, 56% and 68% of the physicians were of the opinion that the physician and the patibnt should collaborate and take a shared d . aki h :fi d \ edi al. I ecision m ng approac or nunor an maJor m c Issues, respective y. Additionally, 66% of physicians were of ,e opinion that their patients should take a greater role in the self management of chronic illnesses. However, both the quantitative and qualitative findings of the study corroborated that physician preference varied for involving patients in specific types of decisions and that physicians varied in their opinions about th patient's preference for decision making. For minor health issues and chronic conditions, there was a tendency for physicians to believe that patients should take a more predominant role. Although physicians were generally positive about palient involvement in decisions regarding minor health issues, an intriguing qualitativ, finding in the present study was physicians' perceptions about how patients rght view doctors who gave them a more prominent role in decision making and the perceived subsequent negative impact on h d I hi S h \. d h . t e octor-patlent re ations p. orne p ysicians perceive t at patients may VIew doctors who give them a more prominent roie in health and medical care decision making or who encourage them to use altemltive information sources such as TBDS as "indecisive", "uncaring", "unconcerned", t a "wet noodle". Unknown is to what extent these perceptions are actually held by \ atients. It is suggested that this variable 269

PAGE 283

be examined in future research regarding f.hysician-patient communications and patient preference for information and active involvement in decision-making processes. Secondly, these observations suggest that some physicians may benefit from enhanced physician-patient communication skills. srch training should assist physicians in "positioning" appropriate patient involveJent in decision making without alienating the patient within the context of the relationship. For major health issues, there was a tendency for physicians, especially older physicians, to believe that physicians shoul1 take a more predominant and somewhat paternalistic role in the decision making pcJcess. Two primary factors contributed to the perceived need fur physicians to believ, that they should have a more prominent and directive role in major medical decisioni. First, many physicians believed that severity of illness is a variable influencing ttJe degree of patient and physician involvement in major medical decisions. T ly tended to equate severity of illness with the need for a higher level of physician invo vement in influencing and guiding decisions about treatment. Physicians said at as patients grow sicker, they generally tend to place more trust in their physician t tend to want the physician to take a more prominent and directive role in decisi, making because of his/her clinical expertise. These opinions were concordant with previous research (Beaver, Luker, Owens, Leinster, Degner & Sloan, 1996; Nle and Brooks, 1995; Ende, Kazis, Ash, 270

PAGE 284

and Moskowitz, 1989) indicating there wr less patient preference for active involvement in decision making with incr \ sing severity of illness. Second, as physicians varied in their preference for involving patients in decision making, both qualitative and qujtitative analyses substantiated that making. In the qualitative portion of the p esent study, physicians frequently pointed out that there is a range of medical decisiormaking preferences by patients. Some patients prefer a more passive role ("I prefer that my doctor tell me what to do") while some prefer a more collaborative approach "I prefer that my doctor and I make the decision together''). In the quantitative po on of this study, physicians' opinions were varied regarding or not "patients r their io make decisions fur them." These observatiOns are consistent 1th numerous studies documenting variation in patient preferences for information and involvement in decision making (Cassileth, Zupkis, Sutton-Smith, and 1980; Ende, Kaxis, Ash, and Moskowitz, Sutherland, LlewellynThbmas, Lockwood, Tritchler, and Till, 1989; Nease and Brooks, 1995, Bradley, Zi and Hamilton, 1996; Hack, Degner, and Dyck, 1994; Davison, Degner, and Morgan, 995,) as well as variation in physician perception about patient desire for information and involvement (Stroll, Lo, and Charles, 1984). 271

PAGE 285

Physicians in this study noted that the most sensible approach for the clinician is to be sensitive to each individual patien 's preference for information and to the types of decisions for which patient invol }ement is most likely to be preferred and beneficial. Several physicians acknowledrl !bat if clinicians fililed to be sensitive to individual patient differences in decision making preferences, physician-patient I discordance may occur resulting in a less than optimal doctor-patient relationship and low patient satisfaction with ensuing nega,ve impacts on heahh outcomes and medical liability. These observations are consistenl with previous studies (Beaver, Luker, Owens, Leinster, Degner & Sloan, 1996; Davison, Degner & Morgan, 1995; Hack, Degner, and Dyck, 1994;Ende, Kaxis, Ash, and Moskowitz, 1989; Stroll, Lo, and Charles, 1984) suggesting that physicians jve explicit discussions with individual patients regarding information and decisionlmaking preferences. With the continued increase in the utilization of alternative medlcine (Eisenberg, Davis, Ettner, Appel, Wilkey, Rompay & Kessler, 1998), these are also consistent with recommendations about the importance of shared decision making and physicians I having explicit formal discussions ofpatientls preferences and expectations about the use of alternative therapies (Eisenberg, 1991. Not surprisingly, physicians viewed themselves as the best source of health and medical information for their patients. Prese1ng the physician-patient relationship in an era of managed care has become a central concern for most physicians (Emanuel 27

PAGE 286

and Dubler, 1995). Traditionally, the do or-patient relationship has been predominantly one in which the patient has assumed a very passive, dependent and coopera!ive/acquiescent role while the phtcian has assumed a very active, directive, autonomous, and professionally dominant1ontrolling role. However, physicians in this study acknowledged that medicine hal entered into a new era one that is more consumer centered and characterized by a!ancing sophistication and rapid growth of communications and information managem nt technologies. It was recognized that consumers will accelerate the access of heJth and medical information from various sources such as the Internet and that this in tum may require physicians to make changes in the way they mange the doctoratient relationship and how they approach physician-patient communications. Specific y, they were of the sentiment that it would become increasingly important for ph! sicians to hone their physician-patient communications skills. Many conceded that with some patients they will need to become less paternalistic in their approach. rey will need to become more sensitive to patient preference for involvement in heal h and medical care decisions and adapt their style to that prefurence. The observati+ from the present study -that physicians believe patients desire more information, thalphysicians support and advocate a more collaborative doctor-patient relationship, tha physicians view themselves as the best source of health and medical information to patients for shared decision-making, and h h . d ak hift" \. . I h t at p ystctans may nee to m e a s m p :tient commurucattons sty e prompts t e 273

PAGE 287

question, "Are contemporary physicians dequately equipped to communicate and collaborate effectively with the new bree of consumer (Herzlinger 1997) to preserve the physician-patient relationship in an era of consumerism and managed care, to enhance health outcomes, to optimize + satisfuction, and to ultimately sustain competitive advantage in the marketplacet Observations from the present study and current research in physician-patient communications indicate they may not be rg, DeHaes, Hoos, & Lammes, 1995; DiMatteo, 1994; Beckman & It is recommended that physicians objectively assess their strengths and Ts in patient communications and "position" themselves for the demands of tie emerging consumer-driven marketplace by improving physician-patient communica1on skills. Physicians should benefit from patient communications training that emph1izes active listening, dealing with patient exploring expectations/idt effects of problems (Frederikson, 1995), usmg open-ended quef1ons, alloWing the patient to talk uninterrupted at the outset, repeating and sJmma.mmg patient's problems, avoiding I the use of medical jargon (Levinson, Stiles, Inui, & Engle, 1993), orienting the patient to the flow of the visit, addressing emotions td psychological issues, exploring patient beliefs, checking for understanding, asking if the patient has additional questions, providing a summary of the visit, Lmpleting a final check for under-274

PAGE 288

standing, clarifying the treatment plan, pr viding encouragement/reassurance, and consolidating rapport/caring (White, Levinson, & Roter, 1994). Contrary to expectations, a majorit of physicians in this study indicated that they had enough time to answer patient's juestions about medical care and to fully involve patients in the decision-making pr cess. In a largely productivity-driven, predominantly fee-for-service marketplace, we expected these physicians to provide the regression analyses highlighted a statist cally significant linear relationship between the percent of capitated patients and physicians' opinions about having adequate time I to answer patient's questions. With increasing capitation, there was an increasingly higher percentage of physicians agreeing wi1h the statement that "physicians don't have adequate time to answer patient questilns about medical care. Also, even though there was not a statistically significant relatilnship, with increasing capitation there .. "glhigh fh\.. "hh h was m mcreasm y er percentage o p ylictans agreemg Wit t e statement t at "physicians do not have adequate time to fully involve patients in the decision-making process." There may be several factors that Ler into possible explanations for this finding. First, physicians practicing under managed care contracts with capitated panels are definitely under pressure to see patients i a timely manner. Certain perfonnance standards and guidelines for patient access d for seeing patients in a timely manner 275

PAGE 289

have been instituted by most managed care plans to "protect" the consumer/patient from potential access issues inherent in a system where the physician does not have to see the patient for reimbursement. Health plans and IP As monitor physician performance on these access standards d report results to patients and consumer groups. Therefore, physicians with capita ed panels may have the tendency to "squeeze in" more patients, especially in a mixed fee-for-service and capitated reimbursement environment. Second, capitated patients under care plans can tend to be more transient and more demanding than the more well-established long term patients a practice may have. When patients under aged care contracts are capitated to a given provider, they have no choice but to access care in that office if they want to be covered under the health plan; whereas, a ee-for-service patient can just go to another practice. Therefore, when capitated patienL are assigned to a given provider with whom they are not familiar and have not established a quality doctor-patient relationship, perceptions of restricted access and scheduling delays augments suspicions that they are not getting fair vall. under the plan. Thus, even though one would surmise that the fee-for-service pati+ would have better service at a physicians' office, they may be "shut out" by an already full schedule of a more highly regulated capitated group of patients the ph sician is also responsible for. 270

PAGE 290

Third, some physicians still hold a d set that capitated patients are poor reimbursement patients ($5/$10 co-pay fo the visit). Subsequently, this may create the sense that one cannot spend as much tLe with them considering the reimbursement immediately realized. In cJntrast, a fee-for-service patient requiring more time can be "billed" accordingly for J higher acuity level visit. Thus, a physician may feel more justified in billing a patient lore because they spent more time with them, but could not do so with a capitated batient. Further research is warranted to examine to what extent reimbursement + serve as incentives or disincentives to physicians talking with patients and the iffect on physician-patient communications, the doctor-patient relationship, patient healih outcomes, and physician productivity. Physicians' Opinions About Telephone-Bl Decision Suooort Physicians were asked specific about their opinions of a 24-hour telephone-based decision support (TBDS) sLvice that had just been implemented hi h "d .\ h c: b :_c: ed Wit n t e community to provt e pattents Wit a resource .tor econung more nuorm actively involved consumers in medical care becisions. The information regarding physicians' opinions about TBDS services is among the first to be reported in the research literature. Results from the present study are co sistent with findings from adoption rate studies in the d.i:ffusion of new innovations in health care (new drugs, findings from 277

PAGE 291

large clinical trials, national consensus treatment guidelines). In previous studies, rates of physician awareness with a "new inno1on" ranged from 44 94% approximately one year following their release (Peay & 1988; Lomas, Anderson, Domnick-\ Pierre, Vayda, Enkin, & Hannah, 1989; Hill, Levine & Whelton, 1988; Grilli, Apolone, Marsoni, Nicolucci, Zola & Liberati, 1991r.ln the present study, physicians were asked to report whether or not they had heard of the TBDS service (awareness) and to what extent they were familiar with its and available services. In the quantitative portion of this study, tr fuurths (73.7"/o) indicated that they were aware of the service. Of those reporting that they were aware of the service, 31.3% indicated that they were familiar wi hits purpose and available services while 42.4% reported that even though they had eard about the service they knew very little about its purpose and available services. Of the 21 physician key informants who participated in the qualitative portion ofthl\ study, 18 (86%) reported being familiar with its purpose and services. Given the fo sed intensity of the formal communications fur the physician communil inunediately prior to implementation of the TBDS service, it was anticipated that pJysician awareness levels would be high. Explanations for the finding that about one third who had reported being aware of the service were familiar with its plrpose may be partially explained by the fact that even though a very comprehensive Ld intensive formal physician communication plan was deployed, largely t \ ough mass media channels, more 278

PAGE 292

informal physician-to-physician communirtion and discussion necessary for higher levels of understanding and consensus abo t the service may not have occurred. That is, even though some studies in the diffusiJn of health care innovations have shown that the most frequently cited channel of lareness were staff conferences, Hill Levine & Whelton; Lomas, Anderson, omnick-Pierre, Vayda, Enkin & Hannah, 1989; ), others have suggested that information about a new innovation and receptiveness to change may be spread prilarily by word of mouth through informal Menzel, & Katz, 1966; Kosecoff, Kanouse, Rogers, McCloskey, Winslow & Brook, 1987; Lomas, Anderson, Domnick-Pierre, ayda, Enkin & Hannah, 1989; Greer, 1994). A higher rate of physician familiarity with and understanding about the TBDS purpose and specific services may have bee achieved through strategies that focused on the more informal interpersonal diffusion networks and consensus-seeking processes as described by Greer, (1987, 199 ) and Rogers (1995). Even though there was strong support and endorsement for the concept of a more informed and actively involved patient in the medical care decision making. process, physicians' opinions about the irnpl,mentation of a TBDS service to provide decision support services for patients were vred. Depending upon awareness and familiarity with the service, most held a neutral opinion ( 42-48%) with the remainder 279

PAGE 293

divided between generally positive (21-31%) and generally negative (15-26%). Comparable results were obtained in the qualitative portion of the study with neutral, generally positive, and generally negative expressed by 52'Yo, 33o/o, and 14% of the key informants, respectively. In cdnducting the key informant interviews, a observation of this investigator that even though only three physicians held an explicitly negative view about the rDS service, there seemed to be a prevailing attitude of uncertainty or "let's rand see" skepticism among even the physicians with positive and neutral opinions. It is believed by this investigator that the high percentage of physicians a "neutral" opinion about the TBDS in the quantitative portion of this study reflects tJe prevailing attitude of uncertainty revealed h d th k !-C. \ m t e more m-ep ey uuormant mtervters. The convergence of quantitative anl qualitative opinions of neutrality about the TBDS services may be somewhat explamed by the principle problem of uncertainty. That is, the physicians in this Ldy were exposed to an innovation (TBDS services) that they were uncertain iout. Rogers (1995) has emphasized that uncertainty is generated by innovations (defined as an idea, practice or object that is perceived as new by an individual or an orgLation) and that this uncertainty motivates individuals to seek information ablut the advantages and disadvantages of the innovation in order to reduce the uncert1nty. Furthermore, he adds that: 280

PAGE 294

Information about an innovation i 1 often sought from near-peers, especially information about theirbjective evaluations of the innovation. This information exc ge about a new idea occurs through a convergence process in lving interpersonal networks. The diffusion of innovations is essJntially a social process in which subjectively perceived information rbout a new idea is communicated. The meaning of innovation is thus gradually worked out through a process of social construction (Rogers, 1995; p. xvii). Likewise, Greer, (1994; p. 434) hal emphasized that "In discussion with colleagues, novelty is tamed, made tractab l. and joined to a familiar framework . .. The problem of uncertainty is resolved by new consensus .... In the diffusion of new innovations, members of a social syste create and share information with one another and pass through an innovation-d1sion process characterized by five stages. In the innovation-decision process, individuals pass from first awareness of an innovation (knowledge) to forming an o,n or attitude toward the innovation (persuasion), to a decision to adopt or reject (decision), to utilization of the new idea (implementation), and finally to evaluation jfthe decision (confirmation) (Rogers, 1995a). It may have been possible that the p ysicians in this study were somewhere in the continuum between the first two stages of the innovation-decision making process where there was still much uncertainty abou1 the mos and opinions (positive/negative) were still being formed. That is, the time period between the introduction of the TBDS service and the of data for this study may not have been sufficient time for significant info \ arion exchange among the interpersonal 281

PAGE 295

diffusion networks of physicians and a su sequent convergence and social construction of either a definitive positive or negative linion about the 1BDS services. Unfortunately, the design of this study did not allow us to examine this possibility. Researchers and practitioners interested in the diffusion of TBDS services need to examine more closely the innovation-decis1on process and its effect on physicians' opinions over time. Regression results indicated there were no significant differences of opinion about the 1BDS service related to physiaalpractice cbaracteristics. Interestingly, in the classic diffusion investigation studying le nature of diffusion networks for the spread of a new drug among physicians, Katz & Menzel (1966) found that various indicators of network communications and interconnectedness with a1 k I d. f d mterperson networ s were more important pre 1ctors o pos1t1ve opmmn an subsequent adoption of the new drug tban Le individual physician/practice characteristics. Specifically, affiliation with I hospital as a regular staff member, more frequent attendance at hospital staff meetinJs, sharing an office with one or more doctors, being named sociometrically as a Jurce of information and advice by other doctors as someone with whom they discusL their patient's cases, and as a best the new drug. The present study was not d signed to explore these issues. It is recommended that future research examine uch network interconnectedness factors 282

PAGE 296

as independent variables in predicting phy icians' opinions about TBDS and/or the adoption ofTBDS services. l Uncertainty may have also played\ partial role in explaining physicians' opinions about the potential impact ofTBDS on selected medical care delivery variables. In the key infonnant interviews, lpproximately 43% of the physicians indicated that they were ''not sure" or werl "uncertain" about the impacts ofTBDS. Many physicians emphasized that they had not had enough experience with TBDS services to really know what type of impa it might have on their patients and their practice. Supporting this qualitative resulj as the finding that a consistently predominant percentage (51.5-77.4%) ofp ysicians in the quantitative portion of the study were of the general opinion that TB t services would not have an impact on selected medical care delivery variables. Two possible explanations might be considered for these findings. First, the consistently predominant p\ercentage of physicians believing that the TBDS services would not have an impact selected medical care delivery variables may be a reflection of the uncertainty about lhe impact of TBDS services revealed in the more in-depth key infonnant interviews.lThis uncertainty may have once again resulted from a lack of clear understanding a out the purpose and the services of the TBDS; and, insufficient time for significant exchange among the interpersonal diffUsion networks of physici \ for convergence and social construction 283

PAGE 297

of a definitive common opinion (positive or negative impact) about the impact of TBDS. Supporting this possibility was e quantitative finding that a majority of the physicians did not believe that TBDS se ces would have an impact on patient demand for services, the number of inapp opriate visits, or the volume of calls. Ironically, one of the major driving fur the implementation of the TBDS service in the community was an identified need 9Y physician leaders to reduce the number of medically inappropriate visits and unneceJsary phone calls because of increasing pressures to manage costs and optimize p \oductivity ("For a Healthy Future Together," RCIP A Provider Newsletter, 1995). It appears that there was uncertainty within the physician community about exa ly what it was physicians wanted from a TBDS service. The second possible explanation is fhat physicians truly believed that the TBDS services would not have a significant impadt (positive or negative) on the selected medical care delivery variables. Evidence ipporting this possibility was gleaned from the key informant interviews. These interviews revealed that some physicians truly believed that the TBDS service would havj no impact on their patients and their practice because they believed that they plnally fulfilled the role of the 1BDS service by providing their patients with the and counseling they needed to be more active, involved consumers. They L1 ewed themselves, as one physician put it, as a "TBDS supplier" for their own patients. In either case, each of these possible

PAGE 298

explanations holds significant relevance f4 r the practitioner interested in implementing TBDS services. These conflicting views lack of consensus, once again, seem to point to uncertainty about the TBDS puJose and available services. If uncertainties are prevalent regarding physicians' opiniols about the purpose and potential impact of TBDS, then it is highly probable that the lervice will not be well-received and that physician resistance will deter and ultimately optimal utilization. Although studies to investigate the true Jpact of TBDS on selected medical care quality, cost, and utilization variables may be difficult to conduct, this may be an area fruitful for further study. It is worthwhile to note that even tiough an overall majority of physicians believed there would be no impact on pati nt demand for service's, the number of visits/week and the percent of capitated +ents were significantly related to physicians' opinions about the potential impact ofTBDS on patient demand for services. Physicians with a greater numberlof visits per week and physicians with greater than 50% capitation believed that t e TBDS service would have an impact on patient demand for service. Consistent w4 this observation was the finding that a greater percentage of physicians with the hikest percentage of capitated patients (> 50%) when compared to physicians with a lower percentage of capitated patients believed that the health plans implemented the TBDS to more efficiently care for capitated patients. It may be that the experiLces of physicians with high volume

PAGE 299

practices and a higher percent of capitatioi have led them to believe that TBDS services may indeed assist them in being eren more productive while maintaining high quality patient care. In an era of managed care where physicians are being asked to assume more financial risk in the care of al defined population of patients, a critical issue for future studies is assessing the trul impact of TBDS services on utilization, physician productivity, and costs. I It is also worthwhile to note that J. only exception to a majority of physicians believing that there would be no impact frbm the TBDS services was "patient satisfilction with medical care." In fact, oL one third of the physicians believed there would be in increase in patient satisfaction medical care because of the TBDS service. Supporting this result was the finding that approximately 60% of the physicians believed that the health plans TBDS to improve member satisfaction. These quantitative results werJ consistent with findings from the key informant interviews whereby physicians inkcated that the most significant benefit of the TBDS service for the community woulJ be increased consumer/patient access to good quality, uniform information about htth and medical concerns. As a result, many believed that increased access to this type of information by consumers/patients may indeed translated to higher patient with health plans and medical services in general. These opinions are conJistent with anecdotal reports that TBDS services enhance patient satisfaction (Barn \ 1995). This observation also provides a 286

PAGE 300

key implication for practitioners interestea in successfully implementing TBDS services. That is, TBDS services may be Lore acceptable to and endorsed by physicians if the primary purpose of the Jrvice is to enhance patient satisfaction. The primary reason physicians belited that the health plans implemented TBDS services were for reasons of cost c ntrol in response to the interests of the business community, as opposed to reaso1s related to clinical quality as measured by patient health outcomes, quality of care, malpractice (Table 3.6). This was not surprising in view of the fact that during j time of this study, a significant amount of attention was being given in this communi to employer concerns about rising health care costs. These observations are consiste t with Patrick and Wickizer's ( 1995) contention that the socioeconomic context of the local healthcare marketplace can significantly influence opinions and subsequent behaviors related to the introduction of new health care strategies. l Additionally, the large range and v ety of opinions regarding why the health plans implemented TBDS seems consistenth the theorem of uncertainty described earlier. Several physicians in the key info ant interviews pointed out that clarifying exactly what a majority of practicing physiclans wanted from a TBDS service was key to physician support and endorsement ofthl service. The qualitative portion of the study revealed that sufficient physician buy-Land support of the TBDS service had not been obtained. Most of these reasons wlre related to a lack of physician

PAGE 301

participation, information, experience, faririliarity, and overall uncertainty regarding the primary purpose, value, and impact of the TBDS service. These reasons, along with the practical suggestions given for amp g physician endorsement and support, have significant implications for practition rs faced with the daunting challenge of implementing TBDS and optimizing value. It is suggested that practitioners familiarize themselves with for diffilsing innovation in the medical conununity and assisting physici,s in dealing with uncertainties that are inherently involved in making decisions ab ut new alternatives. It is the opinion of this investigator that particular attention shoul be given to: (a) attributes of innovations that effect their rate of adoption, (b) varia es influencing the innovation-decision process, and (c) techniques for identifying d working with physician interpersonal communication networks. A significant finding of this study as that of all the predictpr variables, logistic regression determined "specialty" to be +cantly related to he greatest number of physician opinion variables. Specifically, pFary physicians (general practice, family practice, general internal medicine, ahd general pediatrics) were more likely than specialists to: 1) agree that their patiets should take a more active role in treating their own minor conditions; 2) agr1 that their patients should take a greater role in self-managing chronic conditions; 3) be familiar with and knowledgeable about the TBDS services; 4) have a more positiv opinion about TBDS services; 5) believe 288

PAGE 302

that the health plans implemented TBDS o respond to interests of the business community, to more efficiently care for capitated patients, improve quality of care, and to improve member satisfaction; 6) believl that TBDS services would decrease the number of medically inappropriate visits I\ d the volume of unnecessary calls. These results are somewhat di:ffic It to explain but may be rooted in the relationship between the innovation-deciJion making process and the very nature of primary care. In the innovation-decision Rogers (1995) has pointed out that social, cultural, and physical proximity and homophily (the degree to which two or education, and social status) are more lik ly to result in: (a) more frequent transfer of ideas and (b) more effective communicatiol. "When they share common meanings, a mutual sub-cultural language, and are alikd in personal and social characteristics, the gain, attitude formation and change, and o ert behavior change" (p. 19). It may have been plausible that this group of primary c e physicians in this particular community comprised a specific homophilous interperJonal social network of physicians whose frequency and means of interpersonal cojunications augmented the consolidation of consensus opinions and enhanced the probJbility of significant relationships to a number of physician opinion variables. FutLe studies should be designed to investigate this possibility.

PAGE 303

Second, and closely related to the concept ofhomophily, is the probability that by the very nature of their training and th ir personal decision to become a primary care physician, primary care physicians wire more interested in and possessed a more fuvorable predisposition towards the co+ of more infonned actively involved patients and 1BDS. Indeed, Donaldson, Yordy, Lohr, and Vanselow (1996; p. 80-81) have emphasized six distinguishing attribules of primary care that would tend to support this contention: 1. Excellent care is grounded in both the biomedical and fhe social sciences. 2. Clinical decisiory making in primary care frequently differs from in referral specialties. 3. Primary care haf at its core a sustained personal relationship benreen patient and clinician. 4. Primary care dofs not consider mental health separately from physical health. 5. t9 promote health prevent disease are mtnns1c to care practice. 6. Primary care is ihformation intensive. Furthermore, they go on to emphasize that the social sciences and humanism are core to primary care and that "beyond the knowledge of disease is knowledge of the patient as a human being" (p. 81). Thus, the very lature of primary care physicians and their training may have contributed to a very holophilous, interconnected group of physicians who possessed a more favorable predisposition towards the concept of more informed actively involved patients d TBDS. 2 0

PAGE 304

Rogers (1995a) has noted that indlvidual predisposition influences behavior toward communication messages, often r suiting in individuals exposing themselves to ideas that are in accordance with their oj interests, needs, and attitudes. When compared to specialists, primary care phylicians may indeed have taken the time to keep themselves more infonned about the !concepts of active patient participation in medical decision making and TBDS servils. It would behoove practitioners to recognize the distinguishing chantcteristi1 of primary care, to be cognizant of the concepts ofhomophily and the implicatioJ for working with physicians in the implementation of social science and humanities-based innovations in health care like TBDS. Researchers interested in the dilfuion of social science and humanitites-based innovations in health care like TBDS may rnt to investigate how and why the differentiating characteristics of primary care vs. referral specialties and homophily effects the innovation-decision making proless and the subsequent rates of adoption for these innovations. Commentary on Implications for Working with Physician Organizations This study bas attempted to idenf 1 and to provide insights on physicians' opinions about patients as partners in medidal care decisions; the use of TBDS technology to support a more informed, aJively involved patient; and the impact of patient involvement and TBDS on various \ ealth care delivery parameters. In 2 1

PAGE 305

consideration of these opinions, I would like to take the opportunity to provide the reader with some personal reflections ab+t what I think these findings mean in relation to the health care marketplace and some implications for working with physician organizations. In brief, it was my overall impressJon that even though physicians in this community genuinely supported more infc rmed and actively involved patients as partners in medical decision making, most physicians: a) were ill-equipped to convert these beliefS into meaningful clinical Ices and behaviors; b) did not have a sense of urgency about this being a priority for responding to the marketplace; c) were resistant to change and skeptical about the motivatirns driving the implementation of 1BDS, d) were very uncertain about 1BDS ha, any significant value for them. For those practitioners working with physician organizations to implement new "innovations" like patient involvement and supporting terologies like 1BDS, I would thus like to share several key insights that might be heT:ful in these endeavors. 1. Recognize that timing and local\ market dynamics can potentially have a significant impact on the attitudes, belii and opinions of physicians. During the time of this study, the local community going through a significant period of change and turmoil within their entire health care delivery system. A health . d "d I . I d hi .c. th comnusston was activate to provt e cont,umg commuruty ea ers p 10r e planning, promotion, and development of tile local health care system. The whole

PAGE 306

concept of community rating, which had ieen in existence for the past forty years was being questioned and challenged. Major rmployers who were faced with increasing cost management pressures in a competitil e global marketplace and who could no longer afford to play the role of the benevrlent industrialist were changing health care benefit plans from largely fee-for-servicedlto more tightly managed physician panels comprising preferred provider, point of service, physician-hospital and self-insured insurers and physicians were battling over ssues of control, finance and clinical autonomy. Hospital consolidation had begun to occur. A more competitive health care market place was beginning to take ,rm. For practitionen; who want to iduce demand-side strategies like consumer involvement and TBDS into physician organizations within the context of a changing and dynamic local marketplace, there must be clear and consistent information and communications regarding the primary objective( s) and intent of such strategies as well as clear indicators of achievement for lhe objective(s). This communication and information should flow not only to the folal executive leadership of the physician organization, but should also be strategijy targeted to the key informal physician opinion leaders and subcommittees as wen.l 2. Particular attention should be gijen to the art and the science of innovation diffusion. First, practitioners should consil mapping physician interpersonal 2 3

PAGE 307

communication networks within the phy ician organization. The process of mapping influence networks will assist the practitilner in identifying key physician opinion leaders and influential subcommittees wiL the IP A or group practice structure who can be specifically recruited, trained, and \mobilized to serve as influential change agents when introducing new innovatioj into the organization. Second, in working with influence networks, practitioners sbluld pay attention to the attributes of an I innovation that effects their rate of adop,on. Practitioners should recognize that physicians go through several stages of decision making in deciding whether to adopt an innovation or not. Specific, targeted, strategies should be developed for moving physicians through Lch of these stages toward implementation. Practitioners should recognize thJ the perceived attributes of a given innovation effect the rate of adoption. Jh a new innovation like TBDS, practitioners should work with identified influence ne,orks and key physician opinion leaders to: a) show how TBDS provides relative advantage over the current process for informing patients, b) how compatible the technoloJ can be in supporting physicians as their own best ''TBDS supplier" for their patit c) provide the opportunity for key rank and file physicians to trial run the TBDS, and d) provide key influential physicians with the opportunity to observe how TBDS cJ work for them. The initiative for introducing TBDS into the physician of this study fuiled to undertake a methodical, systematic diffiJsion approach }or introducing TBDS as a new innovation. 2 4

PAGE 308

One wonders if such an approach had bee1 taken, whether or not the results of this study would have been different. 3. Most, not all, but most physici s will state that they genuinely believe patients generally desire more infonnatio, about health and medical issues; and, that many patients will prefer a more collaborative decision making approach in the doctorpatient relationship. However, when it coLes to converting these beliefs to changes in bh d. I Ibli practtce e aVIors an mvestment m resources as a top pnonty, e eve most physicians and physician organizations, exlept for the most enlightened ones, will I continue to struggle with this issue and will continue to be somewhat reticent about addressing it. Why? Because most provit are currently entangled in a web of change, turmoil, instability, and financial cnallenge. They are uncertain and confused as to what kind of priority to give to this ifue, how to position themselves for the new consumer, and what kinds of direct and investments to make in resources and technology. They are primarily consumed with a focus on internal administrative and financial issues rather than on external forces that are and will continue to shape and effect their eventual survival. It is the opinion of this investigator\ that, at best, most physician organizations are only in the very early stages of really anticipating, and preparing for the new emerging marketplace. It is my belief that the new emerging marketplace is 2 5

PAGE 309

being shaped by three primary forces thai that have previously shaped other sectors of the U.S. economy: consumers, technoloi, and changing organizational structures. Specifically, the baby-boomer ge eration that has shaped virtually every other sector of the U.S. economy will emerge l. the new consumer, eventually shaping the health care industry as well. This new cd sumer of health care will be more informed and will want even more information abo t quality and costs. Access to health information on the internet will continue o support this new era of consumerism. Collectively, the top five most-visited health-related Web sites are recording nine to ten million "hits" per month. This new consumer will be more ,emanding about access (Including access to alternative medicine), convenience, and service quality; will be more assertive in the doctor-patient relationship; and in general will want to take charge and be more autonomous and independent. Pending legislation regarding patient protection and the loss ofERISA protection in combination more and more companies moving to defined contribution plans rather than deflhed benefit plans may mean that more corporations may be turning over the direr purchase of health care to individual employees. With the latest predictions that annual increases in health care costs are going to return to at least the 7-10% level for most employers, coupled with the recent backlash against liMOs, market pressures r continue to encourage insurers to get out of the business of managing healthcare and return to providing traditional r

PAGE 310

insurance functions like marketing and unrerwriting. If these bold events indeed it will only serve to fuel the emerglnce of the new well-informed, more assertive consumer. As provider organizations responi to marketplace demands fur access, convenience, cost-effectiveness, service Jd clinical quality, we will begin to see more and more provider organizational structur s move to horizontally integrated structures similar to the horizontal integration that hL occurred in other sectors of the U.S. economy. These horizontally integrated Jructures will allow provider organizations to focus multidisciplinary efforts on core bress processes fur improving clinical and financial performance in specific disease states and/or procedures. The continued evolution and deployment of communicatilns technologies involving various b . f h l h II . . com mat10ns o t e mtemet, te ep one, te eVIston, computers, mteracttve vmce d d will I d response, an mteracttve VI eo contmue to encourage an support a more informed patient/consumer and shared dec ton making between patients and physicians. It is the opinion of this investigator that most physician organizations are at best in the very early stages of understandiBg and preparing for such a market. I believe it would behoove those working wilh physician organizations to clearly articulate what this new emerging marketprl ce might look like, to create a sense of urgency in preparing for these changes, an to provide the means by which physician 297

PAGE 311

organizations can be "early movers" in tegrating appropriate actions for providing information and active involvement of clnsumers/patients in the medical care process. In doing this, it is my opinion that they Jn be positioning themselves for competitive advantage in this new marketplace. 4. Physicians contend that their esire to heal and to do what's in the best interest of the patient is at the heart ofwlily they practice medicine. Twenty-five years of research on the effects of physician-palient communication has suggested that ffi . dd .\ . .. infl e ecttve commurucatton an octor-patlent mteractlon exerts a pos1t1ve uence not only on the emotional health of the pati, but also on symptom resolution, pain functional and physiological healr status. There is so much supporting and emerging psychosocial evidence for the tlierapeutic effects of the provider within the provider-patient relationship, but so little lwareness, understanding, and practical application in the real world of traditional medicine! It is the opinion of this investigator that we must do a better job ir informing and educating physicians that there is a significant amount of research sJpporting the notion that the primary detenninants of health, disease, health ices utilization, and health outcomes have their basis in the social, cultural, psychological, and psychosocial domains of the patient. This type of research data and Jrmation needs to be integrated more frequently and more in depth into non-,tional practice-based continuing medical education (CME) for physicians. If physicians are genuine in stating that their desire 298

PAGE 312

to heal is at the heart of why they are in rcine, then it will be very difficult for them to ignore the information supporting the r lative advantage that the therapeuticallyenhancing doctor-patient communications can have on the healing process. However, since there is little evide ce that traditional CME delivery methods have direct impact on improving professional practice (Davis, Thomson, Oxman, and Haynes, 1995), education and information alone will not be sufficient to change physician behavior. The practitioner may ni. to view this as a cillfusion of innovation issue with specific, systematic practice-based strategies for predisposing, enabling, and reinforcing physician behavior sJce excellent primary care is grounded in both the biomedical and the social sciences diffusion strategies that include working first and foremost with key influential care physicians should be seriously considered. 5. Even though physicians acknow edged that patients have different preferences for involvement in decision making and that physicians should be sensitive to these preferences, it is the opinion of thiJ writer that many physicians are not that well-equipped to: a) adequately address the patient preference issue and b) effectively and efficiently provide therapeutically-enhancing interactions within the doctor-patient relationship. The therapeuti1y enhancing strategies that I'm talking about include: asking a wide range of not only about the disease, but also about the illness experience as related to the patient's feelings, concerns, and

PAGE 313

understandings of the problem; encourar patients to be active participants in care, encouraging patients to ask questions; ploviding clear verbal, written, and emotional support in response to patient's questionk; sharing power; and providing psychosocial support in collaborative decision making] I believe that the annual continul medical education (CME) requirements for physicians, at a minimum, should include a physician-patient communication training component that sensitizes physicians to i.e patient preference issue, informs them of the research about therapeutically-enhancr connnunicarions with patients, and gives them the skills and training that enables tliem to put it into practice. This type of training would also address the concern t some physicians have expressed about appearing to be "uncaring" or a "wet noodle" in the eyes of the patient when the physician attempts to involve them as a in the decision-making process. I've often wondered, what might be the of organizations like the American Medical Association or the American Aca(iemy of Family Physicians requiring so many hours of non-traditional practice-based C [ in "therapeutically-enhancing" doctorpatient communication skills? 6. Physicians in this study viewe, themselves as the best source of medical information and the "mDS supplier of chLce" for their patients. It is my perception that some physicians responded this way bLuse they genuinely believed that they do a good job in informing and actively involing their patients while other physicians 300

PAGE 314

responded this way because they were skrptical of a mDS system that they had no ownership in creating. Nonetheless, I tbif that most physicians realize that they are limited in how much time they can spend in direct contact or in phone contact with patients and that if they truly want +nned, actively involved patients they must look outside themselves for resources theJ are comfortable with. I want to be clear that TBDS doe not necessarily need to be the technology of choice. We need to keep in mind that it's not a specific technology per se that is the most important issue here. Rather, physiaan organizations need to understand that informed consumers are not going to go aLay. The new consumer wants access and convenience to health and medical infoion and they will continue to access information from various sources. They i be utilizing those sources that are of the highest quality and that are the most time and cost-effective. To this end, I would I suggest that provider organizations take a leadership role in responding to the new consumer by working with purchasers, insters, and consumer advocacy groups in structuring an information chassis fur medire that strategieally integrates and bundles various fonns of information and communications for patients using various b . f h I h I . com mat1ons o t e mtemet, te ep one, mteract1ve vo1ce response, computertelevision, and clinical support groups that LY positions them as the "supplier of choice" for decision support with their patir. I believe that the physician organizations who are truly passionate and enlightened about this issue will find 301

PAGE 315

creative ways to differentiate themselves in the marketplace by proactively establishing technology linkages with their patients t at will allow patients an "interconnectivity'' to them and health/medical information J4-hours a day. 7. Last but not least, it is the opt on of this investigator that there has not yet been a strong business case for both a more informed, actively involved consumer and/or TBDS in health care. A strong buliness case would include several components. First, there would be ditrable increases in patient health and/or satisfaction with concomitant decreases in costs and resource utilization. Second, purchasers would be willing to pay a hi+ premium and insurers would be willing to pay higher fees to physicians who were alle to demonstrate improved quality and decreased costs using these strategies. Third, increased market share would go to h I d d d ed p ys1c1an orgaruzat1ons m response to Improve qu 1ty an ecreas costs usmg these strategies. Given the financial pre+ that physicians are currently undergoing, it seems to me that Sl.ich a business case must be created before physicians will passionately embrace these strategies. I bllieve that physicians' skeptical opinions about TBDS and it's potential impact as in this snuly were also influenced by the fact that not all of the components for l strong business case were in place. As previously described, at the time of this Jdy, physicians were operating in a marketplace that was heavily involved in eLning supply-side cost control strategies. It is my opinion that many physicians resistd the TBDS service and were uncertain 302

PAGE 316

about its value because they believed tha TBDS seiVices were just another strategy being implemented by business and insur rs to control costs their own costs without concern for patient or physician welfare. In retrospect, the Rochester community initiative failed to fully develop, impleme t and effectively communicate all three components of the business case above. Also, the community initiative fell short in assuring that the most significant larts of the "business case" that had been developed for the initiative, with signifil input from the fonnal physician leaders, I had been shared with the most influential informal physician opinion leaders. What's needed is a shining examp e of a care management system that demonstrates what it means to have more informed, actively involved patients working with providers who structure therapeutically-enhancing doctor-patient relationships. With the previously described forces that ke shaping the new healthcare marketplace, I would encourage physician organizationr to heed the call for leadership in this arena by developing some beta sites and pilot programs that truly demonstrate the clinical and financial efficacy of this approach. Thl timing is ripe for such clinical process I improvement initiatives. Reiser p. 016) captures the essence of this important time in history by stating: As we reach the last decad of the 20th century, it is possible to recognize the emergenc9 of a medical era that engages the perspectives of patients lin the transformation of health care. In many guises and Pflicies, our experiences with illness are becoming keys to changing the health care system . . the passive roiJ of patients and consumers in 3 3

PAGE 317

the health care system sh uld be replaced by an active one .... a patient and coysumer-based perspective that draws on personal experi9nce with illness that can improve health care practice, education, research, and policy, and demonstrates the developfuent of a new stage in health care the era of the patie t. It is my hope that within this new era of consumerism, physicians will recognize the leadership opportunity the have to launch the single best platform we have for advancing a more effective, and humane paradigm of health and health care --one that is built on the foWJ.dation of an integrated biomedical social science approach. If physicians take the lLd in collaborating with each other and with patients/consumers to take full advantage of the historical window of opportunity that has been granted, early indicators signal a great opportunity for structuring the future logic of health care reform; and in shapin the next, and possibly the most impacting and rewarding social transfonnation of tcan medicine and public healt:lt LimitatiOnS Interpretation of the results of the study warrants caution. Our response rate for the Physician Opinion sJrvey (POS) was not high ( 42% ). Unfortunately, low response rates have beln historically ubiquitous in surveys of physician populations (Cykert, Hansen Lison & Joines, 1997). As with any population survey, factors affecting self se ection may have played a disproportionate 304

PAGE 318

role. Demographic information obtained from the IP A organizations involved in this study indicated that there was no dice among responders and non-responders in the distribution of primary care versus specialist physicians and male physicians versus female. However, since information aboJ the remaining non-respondent practice characteristics was unavailable, response bias may have affected the results. Even though we have no reason to believe that survey respondents were any different than non-respondents, the interpretation of an specific trends warrants prudence until consistent results can be obtained in adJonal study groups. Even though the development of Juestions for the POS was guided and informed by the substantial body development and patient research in this area (Cassileth, Zupkis, Sutton-Smith, Mlch, 1980; Strull, Lo & Charles, 1984; Ende, Kazis, Ash & Moskowitz, 1989; Suther! d, Llewellyn-Thomas, Lockwood, Tritchler, & Till, 1989; Brody Miller Lerman Smith Caputo, 1989; Beisecker and Beisecker, 1990; Neufield, Degner & Dick, 1993; Darson, Degner, & Morgan, 1995), measures of test-retest reliability and validity were nbt conducted on the POS. A I b f h .I :_c. . arge num er 0 cases Wit nussrg uuormatton can create maccurate or biased results. The results related to pract1ce visits/week and % of capitated patients should be interpreted with extreme cautioJ since each of these POS variables had a high percentage (19%) of missing values. 305

PAGE 319

The use of qualitative methods lay be controversiaL Patton ( 1990, p. 461) has emphasized that the credibility issue rr qualitative inquiiy is dependent upon how well the following elements have been rsed: What techniques and were used to ensure that rigorous techniques and Dfethods for gathering high quality data that were carefully analyzed, with attention to issues of validity, reliabtlity? What does the researcher bring to the study in termk of qualifications, experience, and perspective? What pradigm orientation and assumptions undergird the study? Steps taken in this study to address each of these elements and associated limitations will be discussed. Patton (1990) and Kuzel ( 1992) ie provide some guidelines and criteria for sample selection and sample size when conducting qualitative research. Patton (1990) has pointed out that, "Qualitative inquiry lypically focuses in depth on relatively small samples, even single cases, selected pwp,sefully .... The logic and power of purposeful sampling lies in selecting information-rich cases for study in depth" (p. 169). To enhance sampling rigor and identification of information-rich cases for in-depth study, the present study empltyed a combination of purposeful sampling techniques: maximum variation sampling and snowball sampling. "Maximum variation sampling occLs when one seeks to obtain the broadest I range of information and perspectives on the subject of study. Guba and Lincoln (1989) claim this is the preferred strategy ;or qualitative inquiry'' (Kuzel, 1992, p. 37). 306

PAGE 320

Snowball sampling is an approach for lo ating information-rich key informants by asking well-positioned people within the opulation under study "who knows a lot about ?"and "who should I talk to?" In most groups of people, a few key names are mentioned repeatedly. Kuzel 1992) has noted that when conducting maximum variation sampling, subject selection continues to the point of saturation or redundancy with 12-20 data sources reco ended. perspectives on physicians' opinions abol' TBDS services and patient involvement in decision making. That is, through snowb sampling, peer physicians purposefully identified seven key infurmant physicians from those groups of primary care physicians and specialist who reported being positive, neutral, or negative about TBDS I services. This provided a total sample of 21 key physician informants for the qualitative portion of the study. One poteltial limitation of this sample may have been gender bias. Out of the 21 key physician lronnants, only one (5%) was female. This is in contrast to the overall physician popjation for this study which was comprised of approximately 25% females. Future stud+ should be undertaken to determine if female physicians hold significantly differelt opinions and perspectives than male physicians on the issues of concern in this tudy. Such findings could have significant ami.fi h I. d h ed d akin r cat10ns 10r p ystctan-patlent comm cat10ns an s ar ectston m g. 3(i)7

PAGE 321

A standardized open-ended inte ew was designed to collect the qualitative information for this study. A standardizj open-ended interview was selected as the instrument of choice for this study because of it's inherent strengths: comparability of responses are increased because respondfts answer the same questions, reduces interviewer effects and bias, permits othel researchers to review and use the instrument, facilitates organization and of the data (Patton, 1990). Like the POS, the question development for the st j dardized open-ended interview was guided and informed by the substantial body of s rvey development and patient research in this area (Cassileth, Zupkis, Sutton-Smith, March, 1980; Strull, Lo & Charles, 1984; Endc; Kazis, Ash & Moskowitz, 1989; srland, uewellyn-Thomas, Lockwood, Tritchler, & Till, 1989; Brody Miller Lelan Smith & Caputo, 1989; Beisecker and Beisecker, 1990; Neufield, Degner & 1993; Davison, Degner, & Morgan, 1995). Patton (1990;p. 347-348) has stated that: No matter what style of is used, and no matter how carefully one words interview questions, it all comes to naught if the interviewer rails to capture the actual words of the person being interviewetl. The raw data of interviews are the actual quotations spokef by interviewees. There is no substitute for these data .. \the purpose of each interview is to record as fully and fairly as possible that particular interviewee's perspective, Some method for recording the verbatim responses of people being interviewed is, therefore, essential. For this study, the suggested standard for r cording the interview data as described by Patton (1992) was addressed. All intervie s were audiotaped for subsequent

PAGE 322

transcription, coding, and Additionally, Patton (1990; p. 349) notes that, "Because the raw data of interviews are uotations, the most desirable data to obtain would be full transcription of interviews.' All interviews in this study were transcribed verbatim. Methods for enhancing the rigor f qualitative analysis were addressed in this study. The qualitative data were coded + analyzed utilizing a template analytic approach (Crabtree and Miller 1992). 'J1fs involved beginning with a basic set of a priori codes based on theoretical and previous research and then expanding on these through iterative rea gs and interpretation of the text. Crabtree and Miller (1992; p. 108) have stated that one of the limitations with the template analytic approach is the "potential for mit information, especially if the codebook is produced in a completely a mannr analyst runs the danger of not looking beyond the codes." For this steps were taken to "look beyond the codes" and to addresJ this potenttallimitatton. First, using an initial set of a priori [codes, the primaiy investigator in this study expanded on the initial codes through an itl.nv. process of coding, analyses, inteq>retation, and recoding of salient pattLs, themes and categories. As a part of this iterative process, a second investigato1 independently expanded and refined the set of codes established by the primary Collaboration and consensus between the two primary investigators detJ ed the "final" codebook for the text 3@9

PAGE 323

analysis and interpretation. Overall, the iferative process encompassed ten passes of coding, analysis, interpretation, and recoding for the final codebook. Second, analysis of the data included testing for rival expllations (Patton 1990) and negative case I analysis (Erlandson, Harris, Skipper & 1 en, 1993) which involved examining alternative interpretations of the data that might refute the initial interpretation Third, a question-by-question analysis was cond cted to delineate directionality of opinion or how the key infOrmants felt about a topic (positive, negative, netural) as well as why they felt they way they did. Fourth, the two primary investigators independently coded a sample of questions to test for intlrcoder reliability. The intercoder reliability was 93%. Finally, a funn of modified m+er checking was conducted. "Member checking provides for credibility by allowing members of stake holding groups to test categories, interpretations, and conclusionl (Erlandson, Harris, Skipper & Allen, 1993 p. 142). Notwithstanding, a limitation of this study was that various logistical, resource and financial constraints did not allow us Jo conduct final member checking with the key informllllt physicians who were a part rfthis study. However, we were able to conduct a modified form of member checking or peer debriefing (Erlandson, Haris, Skipper & Allen, 1993). A copy of the pr liminary analysis and interpretation was provided to three independent physician re ewers who were not a part of the study 310

PAGE 324

population for critique and feedback. ior adjustments and revisions from the independent physician reviewers were into the final analysis. Patton (1990; p. 187) has stated that, "One way to strengthen a study design is through triangulation, or the combination of methodologies in the study of the same phenomenon or programs." This study u ilized methods triangulation to verify the consistency of and to validate the finding generated by the quantitative and qualitative portions of the study. Throughout the and discussion sections of this study, we have attempted to provide a comparative 8na.lysis of the consistencies and discrepancies that existed between the quLtitative and qualitative portions of the study. l Even though the primary investiga or of this study has a very strong appreciation for naturalistic inquiry and qjalitative methods and has had experience in conducting physician interviews, focus grl\ups, and action research, another limitation of this study may be the inexperience of th primary investigator in conducting formal qualitative research. A second is related to the fact that during the data collection phase of this study, the p+ investigator was a strategist with the local health care community working closely with employers, providers, and insurers in the development of health care cost coJainment and managed care strategies. He had intimate knowledge of the organizatioJal, cultural, economic, and political issues surrounding the significant transitions the e\ ire medical delivery system in this 3 1

PAGE 325

community was undergoing at the time. n the one hand, this intimate engagement with the local health care community mal have provided this investigator with increased access, as well as vital insights egarding the issues associated with the aims of this study that would not be apparent r available to researchers external to the system. On the other hand, unknown is tolwhat extent this involvement may have biased the data collection, analysis and int rpretation. Acknowledging these limitations t the potential for introducing personal bias, every attempt was made to adhere tJo the standards, guidelines, and criteria for conducting qualitative research as outlin by some of the most prominent experts I (Patton, 1990; Crabtree and Miller, 1992; Erlandson, Harris, Skipper & Allen, 1993) in the field. Additionally, oversight of th1 primary investigator was provided by an independent team of seasoned researchers with experience in both quantitative and qualitative research. Independent reviews and critique of data collection, analysis, and interpretation were provided by this team as a standard component of the research process. And, as previously noted, this + utilized methods triangulation to verifY the consistency of and to validate the find!. gs generated by both the quantitative and qualitative portions of the study. Finally, s ould other investigators have an interest in testing the transferability, dependability, con:firmability of these results, all audiotapes, transcipts, and records are avai able for conducting an independent audit (Erlandson, Harris, Skipper & Allen, 1993). 3 2

PAGE 326

The present study is limited to mostly middle-aged male European American volunteers in a solo or single-specialty grJup IP A practice structure with a mix of fee for -service and capitated reimbursement Jom a single urban community in one region of the country. Generalizability to male or female physicians in other ethnic groups, practice structures, compensation arrangerents, or rural environments is not known and should be done with caution. Patton (1990; p. 471) has noted that, "The importance of reporting both methods an results in their proper context can not be overemphasized." It is worthwhile to note that data 1 llection for this study occurred during a time period when this particular health care community was in significant turmoil and transition. Specifically, considerable was being given to health care cost containment by the employers. Physicians and insurers were in the midst of reacting to cost containment pressures through the de elopment of new provider arrangements and insurance products. For the first time 1n four decades, this community was experiencing significant change and upbeat in its entire health care delivery system. It is plausible that physicians' opinions ma!have been influenced by the socioeconomic context of this time period. Even though the socioeconomic. conditions of this healthcare marketplace y have paralleled those in other communities, generalizability of results to I\ther time periods and contexts should be done with caution. Future studies should e amine to what extent the context and the 3 3

PAGE 327

socioeconomic dynamics of the local h thcare marketplace effect physicians' opinions about shared decision making d TBDS services. Recommendation for Future Research This research was designed to berl rimarily descriptive and hypothesis generating. The results of this study hav suggested numerous recommendations for further research and evaluation. Ho4 it is the opinion of this investigator that in order to advance a more prominent role !r patient involvement in health and medical decisions into 21st century health care, to priority should be given to the following two types of studies: l I. Determine if there is a "busine s case" for more informed, actively involved consumers/patients in health and medical care decisions in both fee-forservice and capitated reimbursement ("Business case" is defined as having three components: a) demonstrabl increases in patient health and/or satisfaction with concomitant decreases in costs and_ resource utilization, b) purchasers would be willing to pay a higher p emium and insurers would be willing to pay higher fees to physicians who were able, demonstrate improved quality and decreased costs utilizing patient preference and involvement strategies, c) increased I market share WQuld go to physician organizations in response to improved quality and decreased costs using patient preference \ d involvement strategies). Given the 314

PAGE 328

financial pressures that physicians are ,ently facing, if at a mininrum we can not demonstrate that more informed and actilely involved patients can indeed have a positive impact on clinical quality while rdducing costs and resource utilization, then I believe that providers, purchasers, and pa ers will be less than passionate about embracing this philosophy as a sustainabl strategy. 2. Even after we demonstrate the j fficacy and/or effectiveness of more informed, actively involved patients in hr and medical care decisions, the challenge that still remains before us is how to accelerate the adoption of more informed, actively involved patients as a "best practiJe" innovation. The gap between medical science and medical practice (Greer, as well as the gap between best achievable practices and actual practice (Berwick, 19b8) is well-established. Many medical innovations may take years or even decadjs to reach widespread adoption. Therefore, I believed we need to conduct diffusion rerh to assess how the structural, organizational and systems characteristics of physicians' practices effect the relative speed at which a formal focus and emphasil on patient preference and involvement in decision making are adopted. Particular a,emion should be given to: a) how the perceived attributes (relative advantage, compatibility, complexity, trialability, observability) of more informed and involved patients effect the relative speed of adoption; b) how this perception is convr through physician interpersonal communication networks and the effect on adoption rates; and c) what characteristics 3 5

PAGE 329

ofkey influential physician opinion leadr are the most predictive in determining adoption rates. It would also be ofvalu1to: a) conduct comparative analyses of successful vs. unsuccessful diffusion, b) tilize systems vs. individuals as the unit of on diffusion rates. For those more interested in the DS services as a consumer health management and decision support per se, it is the opinion of this investigator that top priority should be giren to the previously described types of studies as well. However, it is worthwhile to point out that with the advent of widespread access to health infonnation i the internet, TBDS as cwrently structured may become somewhat obsolete. It may he of more value to conduct comparative studies that assess the clinical performancl improvement and resource effectiveness of TBDS vs. other forms of consumer decisitn support technologies such as the internet, interactive voice response, and clinical support groups. Other recommendations for future research and evaluation that have emanated fiom this study include (not necessarily 'T ordered for priority): 1. Determine if the findings of this study are generalizeable to physicians in other communities and regions of the cojtry with the same physician/practice characteristics and with differing physici,practice characteristics (male vs. female; Caucasian vs. African American vs. Mexican American, vs. Asian; multispecialty 316

PAGE 330

group practice vs. single specialty group ractice vs. IP A:, inner city vs. suburban vs. rural). 2. Investigate the epidemiology o discordance between patient and physician preferences for information and active invllvement in medical decisions. Evaluate the effect of physician-patient discordance ont patient satisfaction, patient health outcomes, patient retention, physician resource utilization, costs, clinical efficiencies, the doctor-patient relationshi,, and malpractice. 3. Utilize qualitative inquiry to determine if female physicians hold significantly different opinions and perspectives than jale physicians about patient involvement in medical decisions and TBDS for patients. 4. Compare physicians' opinions bout more informed actively involved patients to actual physician behaviors and wractice style within the doctor-patient relationship. j 5. Examine the effect of physici opinions about patient involvement in medical decision making and associated prlctice behaviors on: patient satisfaction, health outcomes, utilization, costs, clinicallfficiencies, physician productivity, h f1 al I d h d p ysiClan satis action, m practice, patient etent10n, an t e octor-patient relationship. 3 7

PAGE 331

6. Explore how patients view piysicians who give them a more prominent role in the health and medical care decision L or who encourage them to utilize alternative sources of information such a;s TBDS and the internet. 7. Examine to what extent structures serve as incentives or disincentives to physicians talking with l\ tients and the effect on physician-patient communications, the doctor-patient rela onship, patient health outcomes, and physician productivity. 8. Assess to what extent physicians' opinions about patient involvement in medical decisions and TBDS ovJ time with each stage of the innovationdecision making process. 9. Determine what patient infonnation and decision-support technologies (TBDS, internet, clinical support groups, personal coaching, etc.) are the most effective and efficient in altering physician behaviors relative to patient preference and involvement in medical decisions. j 1 0. Using the data from this stud conduct a path analyses utilizing physician/practice characteristics and op4on scores to further develop, refine, and explain the theoretical and logical relatioJhips among these variables. 318

PAGE 332

Concl sions Multi-method triangulation corroJorated a number of key findings in this study. A majority of physicians believed patients generally desire more information about health and medical issfs and that patients generally desire more active involvement in the medical decisioj"making process. They believed that a more informed patient would result in better health and medical care decisions and a greater sense of control and self-confidence fur thl patient, ultimately leading to improved quality of care, enhanced health outcomes (for chronic conditions and major medical issues), higher patient satisfaction, enricheti doctor-patient relationships, potential reduction in medical liability; and for minor illnesses and chronic conditions improved clinical efficiencies (reduction in the number of medically inappropriate visits and volume ofunnecessary phone calls) ,d cost control. However, physicians varied in their preference for involving patients inl decision making and, believed that patients have varying preferences fur involvement r decision making. Even though there was strong and endorsement for the concept of a more informed and actively involved patieht in the medical care decision making th . d dl . d" h "al process, ere were uncerta.J.nt1es an vane opm10ns regar mg t e purposes, potent1 impacts and reasons for why health plans hL implemented a TBDS service for nh . -C. d I I d" al d . e ancmg patient 11uormat10n an mvo ve ent m me IC ecisions. 3 9

PAGE 333

Primary care physicians (general 1 ractice, family practice, general internal medicine, and general pediatrics) were mrre likely than specialists to: 1) agree that their patients should take a more active r 1 le in treating their own minor conditions; 2) agree that their patients should take a feater role in self-managing chronic conditions; 3) be familiar with and knowledgeable about the TBDS services; 4) have a more positive opinion about TBDS servils; 5) believe that the health plans implemented mDS to respond to interesl of the business community, to more efficiently care for capitated patients, impii ove quality of care, and to improve member satisfaction; 6) believe that TBDS service would decrease the number of medically inappropriate visits and the volume of+ary calls. To what extent are to physician organizations IS unknown. Further quantitative and qu,tative research IS warranted by those interested in consumer health shared decision making, physician-patient communications, TBDS, and the diffusion of health care innovations. 320

PAGE 334

APPEND A PHYSICIAN OPINIO SURVEY (POS) 321

PAGE 335

PHYSICIAN qPINION SURVEY PATIENT INVOLVEMENfl IN MEDICAL DECISIONS and DECISION SUPPORT The purpose of this survey is to collect lormation about RCIP A and RIP A physicians' perceptions and opinions patient involvement in medical decisions and about a 24-hour telephone-based decfsion-support service that begins on June 1 for 80,000 families in Rochester. This counseling service is a non-directive, non diagnostic program that provides decisioJ-support, self-management support, and information to patients. Please respond tb these questions as candidly and directly as possible. Your responses are confidenti and will be known only to researchers at the University of Colorado, who will interpre the aggregated results. Please return the sunrey by June 7. Thanks for your help. A. Background 1. Gender: [ ] Male [ ] Female 2. Year of birth: __ 3. Year completed medical or osteopathic training: __ 4. Where did you do your residency traioiog? [ ] I am not residency trained [ ] Rochester [ ] Other; Specify: _______ ____,,___ ____ 5. Within the IP A structure are you considered: [ ] A Primary Care Provider _Family Medicine Internal Medicine OB/GYN General Pediatrics _Other, please specify-----+-------[ ] A Specialist What is your medical specialty?--+---------

PAGE 336

6. Please indicate practice structure: a. b. c. Solo practice Group practice, single specialty Group practice, multi-specialty Corporation (hospital-owned) [ ] Yes [ ] Yes [ ] Yes [ ] Yes []No [] No []No []No d. e. Other, specify:--------+-------7.Are you: [ ] Salaried 8. Excluding daily hospital rounds, approximat ly how many visits does the entire practice conduct in a typical week? __ 9. Approximately what percentage of your pa nts come from a capitated plan? __ % B. General Opinions and Perceptions The following section includes questions on youm general opinions and perceptions about patient in volvement in medical care. 1. Who would you want to make decisions abou treatment for your patient's medical issues? Only your Mostly your You and your Mostly you with patient patient-patient input from your Only you with input together patient from you A. For a minor 1 2 3 4 5 Health ISSUe B. Fora major I 2 3 4 5 Health ISSUe 323

PAGE 337

2. Please indicate the extent to which you agree or disagree with the following statements: Circle the number that best describes your views. I Strongly I No I Agree O(!inion Disagree Patients want more involvement in their health and medical care 1 3 4 5 6 7 decision making process. I Patients should be provided with a 1 2 3 4 5 6 7 range of treatment options available to them. Patients should be encouraged to 1 1 3 4 5 6 7 ask more questions about their medical care. Physicians do not have adequate 1 2 3 4 5 6 7 time to fully involve patients in the decision-making process. Patients expect their physicians to 1 3 4 5 6 7 discuss the risks and benefits of all available treatment options. Patients want their physicians to 1 3 4 5 6 7 make decisions for them. Physicians do not have adequate 1 2 3 4 5 6 7 time to answer patient questions about medical care. Patients want more information 1 2 3 4 5 6 7 regarding health and medical issues. Medical information obtained by 1 2 3 4 5 6 7 patients through newspapers, televisions, and lay literature is reliable. Patients who are fully informed 1 2 3 4 5 6 7 about the risks and benefits of all treatment options generally choose options that are less invasive, less risky, and less expensive. 324

PAGE 338

3. The following questions ask you to indicate t e extent to which you agree or disagree with the following statements: Circle the number that best describes your views. I would like my patients to take a greater role I Strongly No I in.-Opinion Agree Dill agree the medical decision making process. 1 2 3 4 5 6 7 asking questions about their care. 1 2 3 4 5 6 7 treating their own minor conditions. 1 2 3 4 5 6 7 seeking available information about a medical 1 2 3 4 5 6 7 problem or condition. self managing their chronic conditions (e.g. 1 2 3 4 5 6 "7 diabetes, asthma). C. Opinions About Informed Consent For the following two questions please check w ich of the following statements best describes your views about informed consent. 1. Check one of the foDowing: [ ] Informed consent is mostly for the protection fthe physician. [ ] Informed consent is mostly for the protection lfthe patient. [ ] Informed consent is for the protection of both l physician and the patient. 2. Check one of the following: [ [ ] Informed consent means the physicim descri s all potential risks of a chosen treatment. [ ] lnfonned consent means the physician the patient oftre-alternative. [ ] Informed consent means discussing with pati ts all treatment alternatives md allowing them to state their treatment preference. 3(25

PAGE 339

D. 0 inions and Perce tions About Tele hone-Based Decision Su Services The following section includes more specific questions about your opinions and perceptions of a 24-hour telephone-based decision-support service that begins on June 1 for 80,000 families in Rochester. This counseting service is a non-direc:tive, non-diagnostic program that provides decision-support, self-management support, an information to patients. 1. Please check which of the following statemen s best descn"bes your knowledge and awareness of the TBDS Service. [ ] I have never heard of the TBDS service. [ 1 I have heard of the TBDS .ervice but mn:f little about its pmpose and available services. [ ] I have heard of the TBDS service and am ar with its pWJ>ose and available services. 2. Please check which of the foUowing statemen s best describes your opinion about the TBDS senice. [ ] Generally Positive [ ] Neutral [ ] Generally Negative [ ] I have never heard of the TBDS service. 3. In this next section we would like you to indi te the extent to which you be6eve the following are reasons why Rochester health plans are implem nting the TBDS senice. Circle the number that best describes your views. TBDS is being implemented to ... No I Opinion Dasagree Improve the quality of care. 1 2 3 4 5 6 7 Improve member satisfaction. 1 2 3 4 5 6 7 Reduce health care costs. 1 2 3 4 5 6 7 Respond to interests of the Rochester 1 2 3 4 5 6 7 business community. Improve patient health outcomes. 1 2 3 4 5 6 7 Reduce number of medical liability 1 2 3 4 5 6 7 claims. More efficiently care for capitated 1 2 3 4 5 6 7 patients. r

PAGE 340

4. In this section we ask your views about the impact of the TBDS service in several areas. Circle the number that best describes your views. I I I No I Change following as a result of implementing Incjase the TBDS service? Volume of calls to your practice 1 2 J 4 5 6 7 Number of patient questions 1 2 3 4 5 6 7 Number of inappropriate medical visits to 1 2 J 4 5 6 7 your practice Patient demand for services 1 2 J 4 5 6 7 Number of patients avoiding necessary 1 2 J 4 5 6 7 treatment Number of medical liability claims 1 2 J 4 5 6 7 Average length of clinical encounter 1 2 J 4 5 6 7 Physician reimbursement 1 2 3 4 5 6 7 b. What do you think will happen to the I+ No foUowing as a result of implementing the change Worsen TBDS service? Health outcomes 1 2 3 4 5 6 7 Quality of physician-patient relationship 1 2 3 4 5 6 7 Continuity of patient medical care 1 2 3 4 5 6 7 Patient satisfaction with medical care 1 2 3 4 5 6 7 Retention of members within the health 1 2 3 4 5 6 7 plan E. Additional Comments =...,_ ..... ottbecballeogeo yoo urfollowing .. implelaeal-oflheTBDS 2. Please describe some of the benefits you antici ate foUowing the implementation of the TBDS service. 3 7

PAGE 341

APPENDIXB PHYSICIAN IN'IERVIEW (POI) 28

PAGE 342

Physician Interview Guide Patient Involvement in Medical Decisions Telephone-Bas Decision Support Introduction to Interview/Opening Statements a. Overview/background b. Purpose of interview c. Process and interview content d. Tape recording e. Confidentiality; use of information Informant questions/concerns A. Consumer/patient Information and volvement General I First, I would like to ask you some generai questions regarding your opinion about patient involvement in health and medical r decisions. I. What do you think about patients wanting more information regarding health and medical issues? 2. What do you think about patients wanting more active participation in the medical care decision-making process? 3. What do you think will be the impact Oljl health status/outcomes of those patients who actively seek out information and more actively participate in medical decision making (better, same, worse)? [Probe rationale] B. Consumer/patient Information and InvolvementSpecific Decisions We've been talking generally about patient involvement in health and medical care decisions. Now I'd like to ask you questions that are more specific in regards to patient involvement in health and medical care decisions. 329

PAGE 343

4. For a minor health issue (sprains, str s, cold/flu symptoms), who would you want to make decisions about treatmer? Only your Mostly your You an!d Mostly you Only you patient patient with your with input input from togeth \r from your _y_ou _patient 1 2 3 I 4 5 Please provide the reasoning your opinion. (Probe impact on: doctor patient relationship, clinical efficiendY, health outcomes, patient satisfaction, health care costs, medica/liability). 5. For a major health issue (hospitalizati?n, chemotherapy, surgery), who would yo want to make decisions about treatmert? Only your Mostly your You and Mostly you Only you patient patient with your pat ent with input input from together from your you patient 1 2 3 I 4 5 .I Please proVIde the reasorung supportmg your oplillon. (Probe unpact on: doctor patient relationship, clinical efficiencf,, health outcomes, patient satisfaction, health care costs, medica/liability). 6. What do you think about your patients taking a more active role in the selfmanagement of chronic disease conditions (e.g. asthma, arthritis, diabetes, hypertension, low back pain)? (Probe Jnpact on: doctor-patient relationship, clinical efficiency, health outcomes, pJtient satisfaction, health care costs, medica/liability). 3130

PAGE 344

C. Telephone-Based Decision Support Service (TBDS) The following section includes more specrfic questions about your opinions of the 24-hour telephone-based decision suppotf service that began on June 1, 1996, for 80,000 families in the community. This service is a non-directive, non-diagnostic program that provides decision-support, Jelf-management support, and information to patients. 7. Please indicate which of the following statements best describes your knowledge and awareness of the TBD : I have never heard of the service. I _I have heard of the TBDS se]'ce but know very little about its purpose and available services. _I have heard of the TBDS service and am familiar with its purpose and available services. 8. Please indicate which of the following est describes your opinion about the TBDS sernce: Generally positive Neutral _Generally negative 9. In this section, we would like to know our opinion about the potential impact of the TBDS service. Please describe hof you think the TBDS service will will impact your patients and your (Probe impact on: doctor-patient relationship, health outcomes, patient satisfaction, contiflllity of care, patient demand for services, flUJzber of inappropriate medical visits, volume of calls, average length of clinical encounter, medica/liability). 10. What will be the most significant challe ges you anticipate following the implementation of the TBDS service? 3 1

PAGE 345

11. How do you propose each of these challenges be addressed? 12. What will be the most significant benlfits you anticipate following the implementation of the TBDS service?. D. Closing Before we end this interview 13. What else would you like me to kno about your thoughts regarding a more active involvement of patients in heal and medical decisions? 14. What else would you like me to kno about your thoughts regarding the value of the TBDS service to the communi ? E. Demographics 15.1 Gender: Male Female 15.2 Year ofbirth: __ 15.3 Year completed medical or osteopat c training: __ 15.4 Where did you do your residency trai g? I am not residency trained. Rochester _Other (Specify): ____ -+----------3 2

PAGE 346

15.5 Please indicate your practice stru e: Solo practice 2-person partnership Family practice group Pediatric practice group Multispecialty group Corporation (hospital owned) Yes Yes Yes Yes Yes Yes No No No No No No Other (Specify):-----+------------15.6 Years of practice in Rochester: __ 15.7 Approximately what percentage fyour patients come from a capitated plan? % Oosing Statements a. Express interviewer's interest in other oughts, opinions, feelings infonnant would like to share following this interview (provide with telephone number and address). I b. Inquire about other possible key infonifants to interview. c. Review feedback process and d. Express appreciation/thanks to info t. 333

PAGE 347

APPEriXC QUALITATIVE AN YSIS CODEBOOK

PAGE 348

Node Location (1) (2) (2 1) (2 2) (2 4) (2 5) (2 6) (2 7) (2 8) (2 9) (2 10) (2 11) (2 12) (3) (3 1) (3 2) (3 3) (3 4) (3 5) (3 7) (3 8) (3 9) (3 10) (3 11) (3 12) (3 13) (4) (4 1) (4 2) (4 3) (4 4) (4 5) (4 8) (4 10) (4 11) (4 12 (4 13) (4 14) Node Titl s /CASEDA'l'A /MORE :IN!' mG.'l':ION /MORE INFQRMATION/DR-PATIENT RELATIONSHIP /MORE INF0RMATION/OUTCOMES & QUALITY /MORE INF0RMATION/PATIENT SELF-EFFICACY /MORE INFqRMATION/EFFICIENCY/APPROP /MORE INFORMATION/MEDICAL LIABILITY /MORE INFqRMATION/PHYSICIAN AS TEACHER /MORE INFqRMATION/CONSUMERISM /MORE INFdRMATION/CONSUMER INFORMATION /MORE INFgRMATION/BETTER DECISIONS /MORE INFORMATION/CHRONIC ILLNESS /MORE INFdRMATION/PATIENT PREFERENCE /P.AR'l':IC:IPJ'l':ION RELATIONSHIP OUTCOMES/QUALITY /PARTICIP$ ION/PATIENT SATISFACTION /PARTICIP ION/PATIENT SELF-EFFICACY AS TEACHER /PARTICIPAkiON/PATIENT PREFERENCE /PARTICIPAkiON/SEVERITY DECISIONS /PARTICIPAriON/ROCHESTER QUALITY /PARTICIPATION/INFORMED CONSENT /PARTICIPAtiON/CONSUMER CHOICE /HEALTH AND OUTCOMI!lS /HEALTH AND OUTCOMES/DR-PT RELATNS /HEALTH AND OUTCOMES/PT SELF-EFFIC /HEALTH AND OUTCOMES/PATIENT SATIS /HEALTH ANDOUTCOMES/MED LIABILITY /HEALTH AND OUTCOMES/DR AS TEACHER /HEALTH AND OUTCOMES/CIA MEDICINE /HEALTH STtTUS AND OUTCOMES/PT PREFERENCE /HEALTH AND OUTCOMES EFFIC/APPROP /HEALTH STATUS AND OUTCOMES/OUTCOMES /HEALTH ST TUS AND OUTCOMES/DISEASE DEP /HEALTH ST TUS AND OUTCOMES/CONSUM-CHOICE 33

PAGE 349

Node Location (5) (5 1) (5 2) (5 3) (5 4) (5 5) (5 6) (5 7) (5 8) (5 9) (5 10) (5 11) (5 12) (5 13) (5 14) (5 15) (5 16) (5 17) (5 18) (6) (6 1) (6 2) (6 3) (6 4) (6 5) ( 6 6) (6 7) (6 8) (6 9) (6 10) (6 11) (6 12) ( 6 13) (6 14) (6 15) (6 16) (7) (7 1) (7 2) (7 3) (7 4) (7 5) (7 6) (7 7) Node Tit.f.es I /MINOR DECXSXOHS /MINOR DECISIONS/DR-PATIENT RELATIONSHIP /MINOR DECISIONS/EFFIC/APPROPRIATENESS /MINOR OUTCOMES /MINOR DECISIONS/PATIENT SATISFACTION /MINOR DiCISIONS/PATIENT SELF-EFFICACY /MINOR DECISIONS/HEALTH CARE COSTS /MINOR LIABILITY /MINOR D CISIONS/PATIENT PREFERENCE /MINOR D CISIONS/CONSUMERISM /MINOR /MINOR D CISIONS/CONSUMER INFORMATION /MINOR DECISIONS/PHYSICIAN SUPPORT /MINOR DE\CISIONS/PHYSICIAN AS TEACHER /MINOR DECISIONS/CONSUMER CHOICE /MINOR CONSENT /MINOR DEEISIONS/REIMBURSEMENT /MINOR DECISIONS/CIA MEDICINE /MINOR DEFISIONS/SELF-MANAGEMENT DEbXSXOHS /MAJOR RELATIONSHIP /MAJOR DECISIONS/EFFICIENCY/APPROP /MAJOR DEtiSIONS/HEALTH OUTCOMES /MAJOR DEtiSIONS/PATIENT SATISFACTION /MAJOR SELF-EFFICACY /MAJOR CARE COSTS /MAJOR LIABILITY /MAJOR PREFERENCE /MAJOR CONSENT I /MAJOR INFORMATION /MAJOR DEdiSIONS/SEVERITY /MAJOR DEctiSIONS/PHYSICIAN AS TEACHER /MAJOR DEGISIONS/CONSUMER CHOICE I /MAJOR DEqiSIONS/INFORMED CONSUMERS /MAJOR DECISIONS/REIMBURSEMENT /MAJOR DEdiSIONS/CONSUMERISM /CHRONIC J:rSEASE SELF-Mmm.GEMEN'l' /CHRONIC giSEASE SELF-MGMT/DR-PT RELATION /CHRONIC giSEASE SELF-MGMT/EFFIC/APPROP /CHRONIC DISEASE SELF-MGMT/HLTH OUTCOMES /CHRONIC ti1ISEASE SELF-MGMT/PT SATIS /CHRONIC D 1 ISEASE SELF-MGMT/PT SLF-EFFICAC /CHRONIC D SEASE SELF-MGMT/COSTS /CHRONIC D SEASE SELF-MGMTT/MED LIABILITY 3 6

PAGE 350

Node Location (7 B) (7 9) (7 10) (7 11) (7 12) (7 13) (7 14) (7 15) (7 16) (8) ( 8 1) (8 2} (8 3} (8 4} (8 5} (8 6) (8 9) (8 10) (8 11} (8 12) (8 13) (8 14) (8 15} (8 16) (8 17) (8 18) (8 19) (8 20) (8 21) (8 22) (8 23) (8 24) (9) (9 1) (9 2} (9 3) {9 4) (9 5} (9 6} (9 7) (9 8) ( 9 9) (9 11) (9 12) Node Titl s /CHRONIC SELF-MGMT/PATIENT PREF /CHRONIC piSEASE SELF-MGMT/SOCIAL DETERM /CHRONIC piSEASE SELF-MGMT/DR AS TEACHER /CHRONIC SELF-MGMT/REIMBURSEMENT /CHRONIC !SEASE SELF-MGMT/SEVERITY /CHRONIC !SEASE SELF-MGMT/CONSUMER INFO /CHRONIC DISEASE SELF-MGMT/BETTER DECISNS /CHRONIC DISEASE SELF-MGMT/MAJOR IMPACT /CHRONIC IDISEASE SELF-MGMT/COMMUNITY /OPINiON iC /OPINION OUT IC/DR-PATIENT RELATIONSHIP /OPINION OUT IC/EFFIC/APPROP /OPINION IC/HEALTH OUTCOMES /OPINION IC/PATIENT SATISFACTION /OPINION IC/PATIENT SELF-EFFICACY /OPINION IC/HEALTH CARE COSTS /OPINION ASOUT IC/ACCESS /OPINION IC/CONVENIENCE /OPINION IC/CONFIDENTIALITY /OPINION ABOUT IC/INSUFF KNOWLEDGE IC /OPINION ABOUT IC/PHYSICIAN SUPPORT /OPINION ABOUT IC/CONSUMER INFORMATION /OPINION ABOUT IC/COST-BENEFIT /OPINION ABOUT IC/PHYSICIAN AS TEACHER /OPINION ABOUT IC/IC SERVICE QUALITY /OPINION ABOUT IC/PERFORMANCE FEEDBACK /OPINION ABOUT IC/PATIENT UTILIZATION /OPINION ABOUT IC/PATIENT PREFERENCE /OPINION ABOUT IC/BETTER DECISIONS I /OPINION IC/ROCHESTER QUALITY /OPINION IC/COMPETING SERVICES /OPINION OUT IC/PATIENT ED/AWARENESS / D.IPACT OF IC ON PATiENTS AND PRACTICE /IMPACT ON DR-PT RELATIONS /IMPACT ON ON EFFIC/APPROP /IMPACT ON OUTCOMES /IMPACT ON PATIENT SATISFACTION /IMPACT ON PATIENT SELF-EFFICACY /IMPACT ON COSTS /IMPACT ON MEDICAL LIABILITY /IMPACT ON CONTINUITY OF CARE /IMPACT ON CCESS /IMPACT ON PHYSICIAN AS TEACHER /IMPACT ON PHYSICIAN SUPPORT

PAGE 351

Node Location (9 13) (9 14) (9 15) (9 16) (9 17) (9 18) (9 19) (9 20) (10) (10 1) (10 2) (10 3) (10 4) (10 5) (10 6) (10 7) (10 8) (10 9) (10 10) (10 11) (10 12) (11) (11 1) (11 2) (11 4) (11 5) (11 6) (11 7) (11 8) (11 9) (12) (12 1) (12 2) (12 3) (12 4) (12 5) (12 6) ( 12 8) (12 9) (12 11) (12 12) (12 13) Node Titles I /IMPACT 9N SERVICE QUALITY /IMPACT PATIENT USE /IMPACT N INSUFF KNOWLDGE /IMPACT N PERFORM FEEDBACK /IMPACT N SEVERITY /IMPACT dN PT PREFERENCE /IMPACT dN UNDERSERVED POPS /IMPACT 1N COMMUN-MARKETING I CHALLENGES /CHALLENGES/COST-BENEFIT SUPPORT UTILIZATION SERVICE QUALTIY REIMBURSEMENT /CHALLENGfS/CONSUMER INFORMATION /CHALLENGES/PERFORMANCE FEEDBACK /CHALLENGfS/DOCTOR-PATIENT RELATIONSHIP AND APPROP /CHALLENGES/ROCHESTER QUALITY /CHALLENGES/INSUFFICIENT KNOWLEDGE IC /CHALLENGES/HEALTH CARE COSTS I /ADDRESSCFHALLENGES /ADDRESS qHALLENGES/PHYSICIAN SUPPORT /ADDRESS aHALLENGES/PATIENT ED AWARENESS I /ADDRESS qHALLENGES/COMMUNIC & MARKETING /ADDRESS qHALLENGES/PERFORMANCE FEEDBACK /ADDRESS qHALLENGES/LOCAL FRANCHISE /ADDRESS gHALLENGES/COMMUNITY ORGANIZ /ADDRESS dHALLENGES/IC SERVICE QUALITY /ADDRESS \HALLENGES/CONSUMER INFORMATION /BENEFI:TS OF :IC /BENEFITS b 5F IC/DOCTOR-PATIENT RELATIONS /BENEFITS [F IC/CLINICAL EFFIC AND APPROP /BENEFITS OF IC/HEALTH OUTCOMES /BENEFITS PF IC/HEALTH CARE COSTS /BENEFITS pF IC/CHRONIC DISEASE /BENEFITS 9F IC/ACCESS /BENEFITS bF IC/PERFORMANCE FEEDBACK /BENEFITS :F IC/BETTER DECISIONS /BENEFITS 'F IC/PHYSICIAN SUPPORT 33'8

PAGE 352

Node Location ( 12 14) (12 15) (12 16) (12 17) (13) (13 1) (13 2) (13 3) (13 4) (13 7) (13 8) (13 9) (13 10) (13 11) (13 12) (13 14) (13 15) (13 16) (13 17) (13 18) (14) (14 1) (14 2) (14 3) (14 4) (14 6) (14 8) (14 10) (14 11) (14 13) (14 14) (14 15) (14 16) (14 17) Node Tit es /BENEFITd OF IC/INSURER BENEFIT IC/CONSUMER /BENEFIT OF IC/UNDERSERVED POPULATIONS /BENEFITS OF IC/INSUFFICIENT KNOWLEDGE IC /ACTrvE OLVEMENT /ACTIVE IrvOLVEMENT/DR-PATIENT RELATIONS /ACTIVE IEOLVEMENT /HEALTH OUTCOMES /ACTIVE I OLVEMENT/PATIENT SATISFACTION /ACTIVE I OLVEMENT/PATIENT SELF-EFFICACY /ACTIVE INvOLVEMENT/ACCESS /ACTIVE INvOLVEMENT/PHYSICIAN AS TEACHER /ACTIVE OLVEMENT/SUPPORT GROUPS /ACTIVE I OLVEMENT/C/A MEDICINE /ACTIVE I OLVEMENT/CONSUMER INFORMATION /ACTIVE I OLVEMENT/UNDERSERVED POPLTNS /ACTIVE DISEASE /ACTIVE IVOLVEMENT/BENEFITS /ACl.IVE INVOLVEMENT /REIMBURSEMENT /ACTIVE IiJrfoLVEMENT/BUSINESS OF MEDICINE /ACTIVE I OLVEMENT/PREVENTION /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF /VALUE OF IC IC/INSUFFICIENT KNOWLEDGE IC IC/PERFORMANCE IC/PATIENT PREFERENCE IC/PATIENT IC/PHYSICIAN SUPPORT IC/COMPETING SERVICES IC/INFORMED-INVOLVED PATIENTS IC/IC SERVICE QUALITY IC/ACCESS C/PATIENT SATISFACTION C/DOCTOR-PATIENT RELATIONSHIP C/HEALTH CARE COSTS C/EVALUATION 3 9

PAGE 353

APPENDIXD VARIABLES WITH MISSING VALVES T

PAGE 354

Independent Age Gender Variable Location of residency Specialty Practice structure Practice visits/week % of patients in capitated plan Minor decisions (B 1A) Major decisions (B 1B) Patients want more involvement (B2A.) Patients should ask more questions Physicians can't involve patient (B2D Patients don't want to make decisions (B2F) Physicians can't answer questions Patients want more information (B2H) Patients need role in decision making Patients need role in asking questions 3B) Patients need role in minor conditions (B3C) Patients need role in seeking info (B3D) I Patients need role in chronic conds (B3E) Physicians' awareness ofTBDS (DI) Physicians' opinion ofTBDS (D2) TBDS to improve quality of care (D3A) TBDS to improve member satisfaction\ (D3B) TBDS to reduce health care costs (D3C) TBDS to respond to business (D3D) \ TBDS to improve patient outcomes (DIE) TBDS to reduce liability claims (D3F) TBDS to be more efficient (D3G) TBDS impact on phone calls (D4A1) TBDS impact on inappropriate visits (R4A3) 34 1% < 1% < 1% <1% <1% 190/o 19% 5% 5% 2% 2% 2% 3% 2% 2% 2% 2% 3% 2% 3% 3% 5% 7% 7% 7% 7% 7% 7% 6% 9% 9%

PAGE 355

Variables with missing values (Cont.) Variable 1 TBDS impact on patient demand (D A4) TBDS impact on liability claims (04 6) TBDS impact on clinical encounter 7) TBDS impact on health outcomes (D B 1) TBDS impact on phys-pt relationship (D4B2) TBDS im act on atient satisfaction 4B4 0/o Missing Cases gofo 9% 9% gofo gofo 9%

PAGE 356

RE :RENCES Barnett AA. Is knowledge really power for patients? Business and Health 1995; May:29-36. Ballard-Reisch, DS. A model of Pfticipative decision making for physician patient interaction. Health ComTicalion 1990;2:91-104. Battista RN, Williams n, MacFarlfe LA Determinants of preventive practices in fee-for-service primary care. Am J Prev Med 1990;6:6-11. Beaver K, Luker KA, Owens RG, Leinster SJ, Degner LF, Sloan JA. Treatment decision making in wo en newly diagnosed with breast cancer. Cancer Nurs 1996;19:8-19. Beckman H & Frankel R. The of physician behavior on the collection of data. Ann Intern Med 1984;101 :6\2-696. Beisecker AE & Beisecker TD. Patient information-seeking behaviors when communicating with doctors. Med Care 1990;28: 19-28. Berkanovic E, Telesky C, Structural and social psychologic factors in the decision to seek medical car1 for symptoms. Med Care 1981; 19:693709. Berwick DM. (1998). As good as il should get: making health care better in the new millenium. Washington, DC.: National Coalition on Health Care Report. Bradley, JH, Zia MJ, Hamilton N.1atient preferences for control in medical decision making: a scenario-based approach. Fam Med 1996;28:496-501. Brody DS. The patient's role in cJcal decision-making. Ann of Inter Med 1980;93 :718-722. Brody DS, Miller SM, Lerman CE, Smith DG, Caputo GC. Patient perception of involvement in medical care: rei tionship to illness attitudes and outcomes. J Gen InternMed 1989;4:506-511.

PAGE 357

Brook RH, Kamberg CJ, Mayer-Oakes A, Beers MH, Raube K, Steiner A Appropriateness of acute medical pare for the elderly: an analysis of the literature. Health Policy 1990; 14:.225-242. Brook RH, & Lohr KN. Will we to ration effective health care? Issues in Science and Technology 1986;l1-10. Brook RH, Lohr KN, Chassin M, Kosecoff J, Fink A, Solomon D. Geographic variations in the use of services: do they have any clinical significance? Health Affairs I 98413:63-73. Cassileth BR., Zupkis RV, Sutton-Smith K, March V. Information and participation preferences among ca!ncer patients. Ann of Inter Med 1980;92:832-836. j Coleman JS, Menzel H, Katz E. dical Innovation: Diffusion of a New Drug. Indianapolis: Bobbs-Merrili, 1966. Cook TD, & Campbell DT. Quas/Experimentation: Design and Analysis Issues for Field Settings. Boston, Houghton Mifflin Company, 1979. Crabtree, BF & Miller WL. A Tjlate Approach to Text Analysis: Developing and Using Codebooks.\. In: Crabtree BF, Miller WL, eds. Doing Qualitative Research. CA: Sage 1992:93-109. Cykert S, Hansen C, Layson R, J. Primaiy care physicians and capitated reimbursement. J Gen Intern Med t 997; 12: 192-194. Davis, DA, Thomson MA, Oxman Haynes RB. Changing physician performance: a systematic review or the effect of continuing medical education strategies. JAMA 1995;274:700-70,. Davison BJ, Degner LF, Morgan TR. Information and decision-making preferences of men with prostate ccilicer. Oncology Nursing Forum 1995;22:1401-1408. l Degner LF & Sloan JA Decision g during serious illness: What role do patients Really want to play? J Epi 1992;45:941-950. 344

PAGE 358

Delbanco TL. Enriching the doct r -patient relationship by inviting the patient's perspective. Ann oflntemMed Dentzer S. The allegory of the Jnveyor belt. Dartmouth Alumni Magazine. 1991; 5:23-25. DiMatteo :MR. The physician-patient relationship: effects on quality of health care. Clin Obstet and Gyn 1994;3 : 149-161. Donaldson, MS, Yordy KD, Lohr KN, Vanselow NA. (Eds.). (1996). Primary Care: America's Health n a New Era. (Institute ofMedicine, Committee on the Future of Care). Washington, DC: National Academy Press. Eddy DM. Connecting value and costs whom do we ask, and what do we ask them. JAMA 1990;264:1737-1739. EddyDM. Rationing by patientc oice. JAMA 1991;265:105-108. Eddy DM. Three battles to watch in the 1990s. JAMA 1993;270:520-526. Eisenberg DM. Advising patients ho seek alternative medical therapies. Ann IntemMed 1991; 127:61-69. \ Eisenberg DM, Davis RB, Ettner JL, Appel S, Wilkey S, Rompay MV, Kessler RC. Trends in alternative use in the United States, 19901997. JAMA 1998;280:1569-1575. Emanuel EJ & Dubler NN. Preservmg the physician-patient relationship in the era of managed care. JAMA 1995;j73:323-329. Ende J, Kazis L, Ash A, Moskowil MA Measuring patients' desire for autonomy: decision making and preferences among medical patients. J Gen Intern 1989;4:23-30. Engel GL. The need for a new med1cal model: a challenge for biomedicine. Science 1977;196:126-136. Equitable Securities Research. Access health, inc.: taking the next step. I Nashville, TN: Seth R. Frank. (Ap "122, 1996). 3 5

PAGE 359

Erlandson DA, Harris EL, Skipp r BL, Allen SD. Doing Naturalistic Inquiry. CA: Sage Publications, 1993. Association Provider Newsletter, anuary/February, 1995; 8:4, 7. Ford L, Kaluzny AD, Sondik E. 1iffusion and adoption of state-of-the-art therapy. Semin Oncol1990;17:48S-494. Fox SA & Stein JA. The effect of hysician-patient communication on mammography utilization by diffe ent ethnic groups. Med Care 1991 ;29: 10651082. Frederikson LG. Exploring info tion-exchange in consultation: the patient's view of performance and outcomek. Patient Educ and Coun 1995;25:237-246. Gilchrist VJ. Key Infonnant In: Crabtree BF, Miller, WL, eds. Doing Qualitative Research. CA: Sage Publications; 1992:70-89. Green LW, Eriksen MP, Schor EL Preventive practices by physicians: behavioral detenninants and pot1eial interventions. Am J Prev Med 1988;4:Sl01-107. Green L, Kreuter M, Deeds S, P "dge K. Health Education Planning: A Diagnostic Approach. CA: Mayfie!d Press; 1980. Greenfield S, Kaplan SH, Ware JE, Yano EM, Frank HJL. Patients' participation in medical care. J Ge. Intern Med 1988;3:448-457. Greer AL. The two cultures ofbioledicine: can there be consensus? JAMA 1987;258:2739-2740. Greer AL. Scientific knowledge an1 social consensus. Controlled Clin Trials 1994;15:431-436. Grilli R, Apolone G, Marsoni S, Ni olucci A, Zola P Liberati A. The impact of patient management guidelines on t 1 e care of breast, colorectal, and ovarian cancer patients in Italy. Med Care 991;29:50-63. 3 6

PAGE 360

Grilli R & Lomas J. Evaluating die message: the relationship between compliance rate and The subject bf a practice guideline. Med Care 1994;32:202-213. Hack TF, Degner LF, Dyck DG. Relationship between preferences for decisional control and illness infotmation among women with breast cancer: a quantitative and qualitative Soc Sci Med 1994;39:279-289. Hampton J, Harrison M, Mitchell J, Prichard J, Seymour C. Relative contributions ofhistory-taking, physician examination, and laboratory investigation to diagnosis and m11 agement of medical outpatients. Br Med J 1975;2:486-489 Herzlinger, R. Market Driven Health Care: Who Wins, Who Loses in the Transformation of America's Service /Jidustry. MA: Addison-Wesley Publishing Company, Inc.; 1997.1 Hill MN, Levine DM, Whelton P11-. Awareness, use and impact of the 1984 Joint National Committee report on high blood pressure. Am J Public Health 1988;78:1190-1194. Jordan HS, Burke JF, Fineberg H, Hanley JA. Diffilsion of innovations in bum care: selected findings. Bums 1983;9:271-279. Keller VF & Carroll JG. A new midel for physician-patient communication. Patient EdJic and Couns !994;23 Kosecoff J, Kanouse DE, Rogers 'f.H. McCloskey L, Winslow CM, Brook RH. Effects of The National InstitUtes ofHealth consensus development I program on physician practice. JAMA 1987;258:2708-2713. Kreps GL, O'Hair D, Clowers M b influences of!nnnan communication on health outcomes. Amer Behav Scien 1994;38:248-256. Kuzel .AJ. Sampling in Qualitative\ Inquiry. In: Crabtree BF, Miller WL, eds. Doing Qualitative Research. CA: Sage Publications; 1992:31-44. Lawrence RS. Diffusion of the US Preventive Services Task Force: recommendations into practice. J en Intern Med 1990;5:99s-103s. 347

PAGE 361

Lennan CE, Brody DS, Caputo OC, Smith DG, Lazaro CG, Wolfson HG. Patients' perceived involvement ih care scale: relationship to attitudes about illness and medical care. J Gen lrltern Med 1990;5:29-33. Lennan C, Daly M, Walsh WP, Jesch N, Say J, Barsevick A, Birenbaum L, Heggan T, Martin G. betweenpatients with breast cancer and I health care proVIders. Cancer 1913;72:2612-2620. Levine PH & Britten AFH. Supek.ised patient-management of hemophilia. I Ann of Intern Med 1973 ;78: 195-!20 1. I Levinson W, Stiles WB, Inui TS, Engle R. Physician frustration in communicating with patients. Med Care 1993;31 :285-295. Lewis CS, Wells KB, Ware J. A L fur predicting the counseling practices of physicians. J Gen Intern Med 1r86;1: 14-19. Lomas J. Words without action? The production, dissemination, and impact of consensus Recommendations. Amk Rev Public Health. 1991;12:41-65. Lomas J, Anderson GM, DomnicJ-Pierre K, Vayda E, Enkin MW, Hannah WJ. Do practice guidelines guide bractice? The effect of a consensus statement on the practice of physicians. NEJM 1989;321: 1306-1311. Lomas J, Enkin M, Hannah WJ, VaydaE, Singer J. Opinion leaders vs. audit and feedback to implement practice guidelines. JAMA 1991:265:2202-2207. Lorig KR, Kraines RG, Brown BS, Richardson N. A workplace health education program that reduces oJtpatient visits. Med Care 1985;23: 1044-1054. Lorig .KR, Mazonson PD, Holman HR. Evidence suggesting that health education for self-management in patients with chronic arthritis has sustained health benefits while reducing healfh care costs. Arthritis and Rheumatism 1993;36:439-446. Lynch WD & Vickery DM. The pptential impact of health promotion on health care utilization: an introduetion to demand management. AJHP 1993;8:87-91. 348

PAGE 362

Mayo PH, Richman J, Harris HWj. Results of a program to reduce admissions for adult asthma. Ann of Inter M'ed 112:864-871. McPhee SJ, & Bird JA. lmplemeAtation of cancer prevention guidelines in clinical practice. J Gen Intern MeU 1990;5: 116s-122s. National Committee for Quality 1surance. Annual member health care survey manual (version 1.0). Washingtol1 DC: Author, 1995. Nease RF & Brooks WB. Patien desire for information and decision making in health care decisions: the autollj my preference index and the health opinion survey. J Gen Intern Med 10:593-600. Neufield KR, Degner LF, Dick Jk. A nursing intervention strategy to foster patient involvement in treatment ecisions. Oncology Nursing Forum 1993;20:631-635. Nunnally JC and Bernstein IH. P, chometric Theory. 3rd Ed. New York: McGraw-Hill, 1994, pp. 264-265L Ong LML, DeHaes JCJM, Hoos Lammes FB. Doctor-patient communication: a review of the li1erature. Soc Sci Med 1995;40:903-918. Osler, SW. Lecture to medical stuaents. Albany Med Ann 1899;20:307. Patrick DL & Wickizer TM. and health. In B. C. Amick, S. Levine, AR Tarlov, & DC Walsh 1 ds.), Society and Health ( pp. 46-92). New York, NY: Oxford University Pre s, 1995. Patton MQ. Qualitative EvaluatiJ\n and Research Methods. (2nd ed.). CA: Sage Publications, Inc., 1990. Peay MY & Peay ER The role of rmmercial sources in the adoption of a new drug. Soc Sci Med 1988;26:1183]1189. Pietrszk J, Ramler M, RennerT, Ford L, Gilbert N. Practical Program Evaluation. CA: Sage Inc., 1990. 349

PAGE 363

Prochaska JO & DiClemente CC. Stages and processes of self-change in smoking: toward an integrative model of change. J Consult and Clin Psych 1983;5:390-395. Reiser SJ. The era of the patient: using the experience of illness in shaping the missions ofhealth care. JAMA 1993;269:1012-1017. Reid AO. Computer managemJ strategies for text data. In: Crabtree BF, Miller WL, eds. Doing Qualitatite Research. CA: Sage Publications; 1992:125-148. I_ Roberts CS, Cox CE, Reintgen Baile WF, Gibertini M. Influence of physician communication on newtr diagnosed breast patients' psychologic adjustment and decision-making. Cancer 1994;74:336-341. Rogers EM. Diffusion oflnnovajons. (4th ed.). NY: The Free Press, 1995a. Rogers EM. Lessons for guidelinJs from the diffusion of innovations. Joint Comm J Qual Imp 1995b;21 :324J328. Ryan B & Gross NC. The of hybrid seed com in two Iowa rural communities. Rural Sociology Sandler G. The importance of the !hlstory in the medical clinic and the cost of unnecessary tests. Am Heart J 1980;100:928-931. Schwartz JS, Lewis CE, Clancy Kinosian MS, Radany MH, Koplan JP. Internists' practices in health prolotion and disease prevention: a survey. Ann /nternMed 1991;114:46-53. Speedling EJ & Rose DN. Building an effective doctor-patient relationship: patient satisfaction to participation. Soc Sci Med 1985 ;21: liS-Starr P. The Social of American Medicine. USA: Basic Books, 1982. Stross JK & Harlan WR. The dissemination of new medical information. JAMA 1979;241 :2622-2624. 350

PAGE 364

Stroll WM, LoB, Charles G. Do patients want to participate in medical decision making? JAMA 1984;25F:2990-2994. Sutherland HJ, LlewellynThomaJ HA, Lockwood GA, Tritchler DL, Till JE. Cancer patients: their desire for inlformation and participation in treatment I decisions. J Royal SocMed 1989;82:260-263. Szaz PS & Hollender MH. A colrution to the philosophy of medicine: The basic model of the doctor-patient telationship. Arch JntemMed 1956;97:585-592. Tabachnick BG & Fidell LS. Using Multivariate Statistics. New York NY: Harber & Row Publishers, 1983. Thompson SC, Pitts JS, L. Preferences for involvement in medical decision-making: situational and demographic influences. Patient Educ andCouns 1993;22:133-140. Vickery DM, & Fries JF. Take Care of Yourself. (5th ed.). Reading MA: Addison-Wesley, 1994. Vickery, DM, Golaszewski TJ, Wljight EC, Kalmer H. A preliminary study on the timeliness of ambulatory care utilization following medical self-care intervention. AJHP 1989;3 :26-31. Vickery DM, Kalmer H, Lowry D, Constantine M, Wright E, Loren L. Effect of a self-care education program on medical visits. JAMA 1983;250:2952-2956. Vickery DM & Lynch WD. Demand management: enabling patients to use I medical care appropriately. JOi 1995;37(5):1-7. von Friederichs-Fitzwater MM, Callahan EJ, Flynn N, Williams J. Relational control in physician-patient Health Communication 1991;3: 17-36. Ware, JE. SF-36 Health Survey Marrual and Interpretation Guide. Boston, MA: The Health Institute, New Enkiand Medical Center Hospitals, 1993. 351

PAGE 365

Wasserman RC, Inue TS, Barria a BS, Carter, WB, Lippincott P. Pediatric clinicians' support for parents m es a difference: an outcome-based analysis of clinician-parent interaction. PJfiatrics 1984;74:1047-53 Wennberg JE & Gittelsohn A. vJnatins in medical care among small areas. Sci Am 1982;246:120-134. White J, Levinson W, Roter D. "?h by the way ... ": the closing moments of the medical visit. J Gen Intern 1994;9:24-28. Wolfe JS & Smith DS. Practical iomedical tests: a guide to the selection of statistical tests. Urology 1996;47: -13.