Citation
Exploring cultural perceptions of illness in lay and expert social groups

Material Information

Title:
Exploring cultural perceptions of illness in lay and expert social groups
Creator:
Graham. Jeremy David
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
vi, 87 leaves : illustrations, forms ; 28 cm

Subjects

Subjects / Keywords:
Cross-cultural comparison ( lcsh )
Medical care ( lcsh )
Medical anthropology ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 83-87).
Thesis:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Anthropology
General Note:
Department of Anthropology
Statement of Responsibility:
by Jeremy David Graham.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
36405008 ( OCLC )
ocm36405008
Classification:
LD1190.L43 1996m .G73 ( lcc )

Downloads

This item has the following downloads:


Full Text
EXPLORING CULTURAL PERCEPTIONS OF ILLNESS
IN LAY AND EXPERT SOCIAL GROUPS
by
Jeremy David Graham
B.A., University of Colorado at Denver, 1993
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment of
the requirements for the degree of
Master of Arts
Anthropology
1996


This thesis for the Master of Arts
degree by-
Jeremy David Graham
has been approved
by
H/uw a,
/ ' Date


Graham, Jeremy David (M.A., Anthropology)
Exploring. Cultural Perceptions of Illness in Lay and
Expert Social Groups
Thesis directed by Assistant Professor Kitty K. Corbett.
This thesis explores the problematic underutilization of
biomedicine by United States latinos. A model of culture
is described, under which the relative merits of a
cultural orientation versus a political-economic
orientation are explored. Data were collected from
registered nurses, lay anglo women, and lay hispanic
women in a metropolitan area. The data suggest that both
lay groups differ from the nurses in their level of
biomedical understanding, but is inconclusive regarding
the presence of culturally-different explanatory models.
The implications for biomedical relationships of unshared
meanings and devalued cultural capital are discussed.
This Abstract accurately represents the content of the
candidate's thesis. I recommend its puhl i rat-.i nn _
ABSTRACT
Signed
\
ill


ACKNOWLEDGEMENT
I extend thanks to the individuals who generously agreed
to be interviewed for this exploration, and to the
community members who provided introductions. I am
deeply indebted to my advisors for their patience, their
considerable editing, and especially for encouraging me
to explore the project freely. Valuable statistical
advice was provided by Dr. Tammy Stone.
I thank especially G.L.G., L.L.G., A.B.D., C.M.G, E.M.M.,
and my family and friends without whom this thesis could
not have been completed.


CONTENTS
CHAPTER
1. INTRODUCTION...................................1
An Overview of Culture.......................1
The Culturological Orientation...............8
The Political-Economic Orientation..........16
Issues: Underutilization of Biomedicine.... 21
Issues: Unshared Meanings...................26
2. METHODS AND SAMPLES USED......................31
The Samples.................................32
Registered Nurses......................34
The Latina Sample.....................3 7
The Anglo Sample.......................42
Methods Utilized............................45
Multi-Dimensional Scaling..............45
Disease Severity Surveys...............47
Semi Structured Interviews...........48
v


3. FINDINGS AND RESULTS..........................50
Multi-Dimensional Scaling..................50
MDS Mapping: The Nurses...............50
MDS Mapping: The Latino Women.........54
MDS Mapping: The Anglos...............57
Comparative MDS Appraisal..............60
Survey Findings.............................63
Technical Term Assessment Findings....65
Semi-Structured Interview Findings..........66
4. DISCUSSION AND DIRECTION......................70
The Merits of A Cultural Orientation........70
The Merits of A Political-Economic
Orientation.................................73
APPENDIX
A. EXAMPLE OF SURVEY INSTRUMENT.............76
B. AN EXAMPLE CORRELATION MATRIX FILE
FOR MULTI-DIMENSIONAL SCALING............80
C. GUIDE FOR SEMISTRUCTURED INTERVIEWING... 81
D. LABELS FOR MDS MAPPING OF TERMS..........82
REFERENCES.......................................83
vi


CHAPTER 1
INTRODUCTION
In exploring the problematic utilization of health
care services by some Latino populations in the United
States, what are the relative merits of a cultural
relevance approach versus a political economic approach?
This thesis explores some indices of whether cultural
background and socioeconomic status impact lay people's
understandings of some biomedical illness terms.
An Overview of Culture
What is meant in this paper by culture, and by
cultural difference? For the purposes of this
exploratory work, culture is conceived as the belief
system that governs how individuals make sense of and
cognize events and objects in the world of experience.
This aspect of culture involves what Spradley
described as "social rules," which are "instructions for
constructing, combining, interpreting, and otherwise
1


dealing with symbols" (1972:29) Symbols are the mental
labels applied to objects in the world. Assemblies of
cultural rules form cognitive maps and models, which
govern the ways that an individual assesses reality.
Spradley's focus emphasizes the ways in which these rules
are codified in language structure, but the concept is
equally native to the areas of systems programming and
psychology.
Casson uses the psychological term "schema," to
describe these cognitive knowledge systems. Schemata, he
writes, are
conceptual abstractions that mediate between stimuli
received by the sense organs and behavioral
responses... Schemata are autonomous and automatic--
once set in motion they proceed to their conclusion-
- and they are generally unconscious, nonpurposive,
and irreflexive (1983:430).
Where schemata are individuals' mental structures for
cognizing the world, some aspects of schemata are
idiosyncratic, and other aspects are common to all
humans. Some aspects of schemata are common to members
of some groups but not others:
Schemata differ in their distribution in
populations... cultural schemata are neither unique
to individuals nor shared by all humans, but rather
shared by members of particular societies
(1983:440).
2


Members of given social groups share many of the
same cognitive schema and cultural ideation because they
share similar experiences and life trajectories. When
people live much the same, they experience much of the
same stimuli in the formation of their cognitive
schemata. The concept is emphasized in Marx's assertion
that the material existence of a human shapes his
consciousness, making class and status the major
underlying cause of cultural difference in a mass
society.
Focusing on this "cognitive aspect of culture," Ward
Goodenough writes,
People who deal recurringly and frequently with one
another develop expectations regarding the manner of
conducting these dealings. They make some of their
expectations explicit and formulate some of them as
rules of conduct. The do not consciously formulate
others but react to a person's failure to abide by
them as a breach of appropriate behavior, saying
that he behaves or talks oddly or in a mixed-up
fashion (1970:98).
Ideation orders social behavior by providing a set of
automatic expectations regarding how people interact with
each other and the world. Not everyone in a social group
steers by the exact same cognitive map, but as Goodenough
continues,
[T]he variance in their individual expectations must
be small enough so that they are able to accomplish
their purposes with and through one another
3


reasonably well most of the time... In this respect
a people's culture is like a people's language
(1972:100).
Cognitive maps include taxonomies with which people
organize their world are built to make the world
predictable, so that human individuals can synchronize
their interactions with each other, and with the. needs of
life.
The knowledge structure within the individual is the
component of culture that is explored by most
methodology, from unstructured interviewing to cognitive
testing. However, the schema takes shape in the world
through social interaction. The determinism of knowledge
structures has been challenged and redefined in the last
two decades, based on anthropology which views culture as
negotiated and transitional.
Geertz (1973) discussed culture as a "web of
meanings" within which any one fact or experience draws
its meaning from context. Culture becomes a text to be
read and interpreted. But where literal text may inspire
human action only indirectly, culture gives human
ideation immediate physical reality, through behavior and
activity. However, Geertz's sense of shared meanings has
value in understanding how social ideas constrain human
relationships.
4


Bordieu's notion of habitus recodifies "webs of
meaning" in a manner useful to exploring cultural
differences between social groups. Habitus is:
A system of lasting, transposable dispositions
which, integrating past experiences, functions at
every moment as a matrix of perceptions... and
actions... [it is] a subjective but not individual
system of internalized structures, schemes of
perception, conception, and. action common to all
members of the same group or class (1977:85) .
The habitus concept integrates the notions of culture as
traditional knowledge, shared meanings, and dynamic
mechanisms for living. It is the environment within
which the rise and fall of cultural ideas take place, and
is the currency by which experiences and new ideas are
evaluated and appraised. Moreover, Bourdieu's habitus is
conceived not only in terms of different societies, but
of the "shared life chances" Weber described for groups
within a society divided by social class, party, and
status.
Habitus, offers a materialist theory of culture,
which credits cultural ideas with material outcomes. The
material conditions determine consciousness, but
consciousness impacts material conditions. The shared
life chances of common-group members shape their shared
meanings, and their shared meanings reproduce and enforce
their common life chances.
5


Bourdieu's notion of "cultural capital" is a useful
tool in understanding how habitus impacts class relations
in biomedicine. Built on scientific discourse and the
ideal of objectivity, biomedicine devalues the
explanatory models of lay people, whether they are based
on distinct folk models or on vague shadows of
biomedicine.
Cultural capital relates to social class
reproduction-- how the rich beget rich and the poor beget
poor-- as a series of processes, which contribute to
overall denial of most social mobility. Cultural
capital, as delineated by McLeod, involves the general
cultural background, "knowledge," disposition, and skills
passed on from one generation to the next (1987:12). It
incorporates common use of sentence structure and vocal
accent, dress, and posture, as well as familiarity with
museums, arts, sports, and other canonized social
activities. Bourdieu suggests that children of higher
socioeconomic classes inherit different "capital" than do
children of lower classes. Authoritative institutions
reward a performance specific to a valorized, upper-
class, set of social.technologies. Schools therefore
reward those who enter with more prized cultural capital
with higher academic credentials, which then serve as the
credentials which grant one access to higher-class
6


prosperity in the future. The strata of academic
achievement are the strata of social class, covered over
by a veneer of scholastic achievement.
Similar processes occur in biomedical relationships.
Non-biomedical explanatory models are actively (if
quietly) devalued in clinical relationships. Like an
upper class setting the value of different sets of
cultural capital in schools, the biomedical establishment
serves as an upper class determining and rewarding only
select cultural capital in clinics.
Biomedical relationships are hindered by (and
perhaps exploit) precisely the same processes of cultural
capital and devaluation that act in education.
Biomedicine may in fact play a covert role in the
reproduction of social class through this mechanism.
Valued cultural capital is turned into greater power to
access health resources (including simple awareness of
health services), which directly and indirectly allow
greater control of material and economic resources. This
control of material aspects of life translates back into
access to higher social class. The role of the
biomedical model as a social control mechanism has yet to
be fully explored.
The problem of non-participation in biomedical
relationships hangs between economic inaccessibility on
7


one end, and a lack of shared meanings on the other, a
question of how and whether culture and class impact
utilization. Do United States latinos lack shared
meanings with biomedicine by virtue of their different
cultural traditions, or simply by lack of access to
cultural capital, the authoritative knowledge valued by
the biomedical enterprise? The former notion suggests
that an exploration should focus on cultural
understanding and appropriateness, while the latter
suggests that the problem be viewed as an aspect of class
marginalization.
The Culturological Orientation
A cultural appropriateness approach favors the
reconciliation of conflicting "explanatory models" as a
key to understanding some group differences in health
care utilization. Incongruous explanatory models can
impede shared understandings, and thereby limit the
patient-caregiver relationship. The basic idea of a
cultural approach can be stripped of some implications
and reduced to a central theme: that patients go to
health resources that are consistent with their belief
systems about the way the world works; by implication,
8


they tend to avoid using care resources that deny their
worldview and explanatory models.
Explanatory models are defined by Kleinman as "the
notions about an episode of sickness and its treatment
that are employed by all those engaged in the clinical
process." The relationship between the explanatory
models held by patients and caregivers is a "central
component of health care" (1981:105).
Kleinman separates the idea of pathological disease
as a somatic event from the semiotic progression and
relationships within the cultural construction of an
illness as a personal and social experience. Where the
ethical landscape appears differently to different
parties involved in a clinical relationship,
communication depends on recognition of meanings of
actions for all involved (1988) .
Kleinman (1988) defines five major areas where
explanatory models seek to make sense of an illness
event:
1) etiology
2) how symptoms appear
3) pathophysiology
4) predicting course and outcome of sickness
5) defining sensible treatment
Of these, agreement on the likely outcome of an illness,
9


and feeling comfortable with the treatment regimen, are
major elements in cultural compatibility. Phrasing the
patient-caregiver exchange in terms reasonable to the
patient's explanatory model can be crucial. The formal
organization of these elements is more precise within the
biomedical model than for most lay persons, and lay
people's models may not attend to all the categories:
Lay explanatory models disclose the significance of
a given health problem for the patient and his
family, along with their treatment goals...
Vagueness, multiplicity of meanings, frequent
changes, and lack of sharp boundaries between ideas
and experiences are characteristic of lay
explanatory models (1981:106-7).
Within the cultural approach, the patient-caregiver
relationship is a central component of health care,
because people go to healers because of the illness that
they experience,- not the pathological disease behind it.
Using the term "symbolic healing" can have "the
unfortunate connotation that there are other forms of
healing which are not symbolic" (Csordas and Kleinman
1990:11). Some level of sharing of meanings is necessary
to fulfill the need for the therapeutic relationship.
Explanatory models are not the same as a person's
general views and beliefs about health and illness. As
Kleinman defines the term, an explanatory model is a set
of ideas and expectations related to a particular illness
10


event. However, the information to build an explanatory
model comes largely from the backgrounded, general set of
beliefs an individual uses to interpret such events:
culture, varying across ethnicity and class.
Since cultural schemata give rise to explanatory
models, people of different cultural backgrounds can
experience diseases in different ways, and assign them
different meanings depending partly on social group
membership. Crimm and Greenberg (1981) have suggested
that biomedicine must increasingly temper paternalistic
beneficence with a cross-cultural vision of autonomy.
Guarnaccia and Angel (1989) have examined
somaticization among United States latinos, suggesting
that latinos may translate psychological stresses into
physical pain and symptoms in ways that differ from anglo
patterns. Lock and Kleinman (1996) suggest that some
patterns of pain experience may relate more to social
expression than to pragmatic biomedical problems or
psychological processes.
Rankin-Hill and Bates (1994) related patients'
beliefs regarding control of pain and illness to ethnic
background and to forms of expression of chronic pain.
The pain experiences revealed were different for groups
in Puerto Rico from groups in New England. The authors
recommend the development of culturally appropriate and
.11


relevant programs which help patients feel in control of
their illness experience.
Kleinman, Good and Guarnaccia (1990) reviewed
literature on the mental health status of latinos in the
United States. Among many latino subgroups (especially
Puerto Ricans), the ailments nervios and atagues de
nervios are poorly understood. Primarily psychosomatic'
symptoms are expressed mostly as culturally-appropriate
times, and occurrences more upon levels of acculturation
than upon clinical symptoms. The reviewers suggest that
the benefit of biomedicine to these patients is limited
and resisted because current methodologies do not attend
to cultural factors.
Latinos in the United States are a large category
containing a great deal of variation, but some workers
have attempted to describe some major unifying cultural
traits useful to shaping biomedical services directed at
latinos. Acosta and Hamel (1995) provide a simplified
index of cultural traits associated with latino identity.
Among the central characteristics they propose are:
1) Collectivismo- a sense of the community as the
common overarching good; willingness to forego
individual gain to maintain a role in the
community.
2) Orgullo- a strong sense of local pride,
specific to both ethic identity and the
immediate community.
12


3)
Espiritualismo- commonly-held concept of life
issues (personal success, illness) as
representations of spiritual forces.
4) Machismo- culturally "official" male authority,
value of family males as providers.
5) Marianismo- ideal value surrounding women's
roles, favoring submissiveness, nurturing,
devotion to marital and extended familial
networks, and children.
Acosta and Hamel see major cultural features as social
terrain to which biomedical services must be made to fit.
They propose their generalizations as guidelines to
locating "Natural Support Systems" within ethnic
communities, particularly in the areas of community
mental health and disease prevention. Their predictions
are highly generalized and do not reveal the workings of
any particular latino individual or subgroup, but do
point out some potentially important community values.
Valle and Vega (1982) identified similar traits as parts
of social support networks utilized in latino
communities.
Other cultural-appropriateness approaches in
biomedicine have used similar typologies. The primary
text used to train medical students in physical
examination illustrate the typological mindset of
cultural romanticism. The text states that:
13


To assume homogeneity in the beliefs, attitudes, and
behaviors... is to court error and to miss the mark
in the effort to understand the individual. The
stereotype, a fixed image of any group that rejects
the potential of originality or individuality within
the group, is itself to be rejected. People can and
do respond differently to the same stimuli (Seidel
et al 1995:34) .
The value of this passage is belied by the reductionism
of the following text; Seidel et al go on to present a
table of cultural traits associated with various ethnic
backgrounds. Latinos are described as people who "have
little regard for the past" among whom "males are
considered big and strong." They live with a "relaxed
concept of time-- may be late for appointments" (1995:39-
49) .
Boulette (1978) has warned against confidence in
stereotypical cultural determinism, especially regarding
the values of destino, machismo, and espiritualismo,
which fails to understand the agency of the individual.
Typologies can themselves impede cultural understanding.
A cultural relevance approach, she suggests, may often be
better served by a negotiated model of exchange between
caregiver and patient, than by a typological model.
Typologies are also problematic because some latino
cultural traditions have faded in recent decades. In
some subgroups of Untied States latinos, traditional
folk-models of illness have less impact on people's ideas
14


than they have had in the past. Saunders (1954) focused
on the creation of guidelines for field health personnel,
suggesting that ideas of personal and spiritual fate
(destino) made "faith and fatalism... the first
ingredients in Spanish-American folk medicine"
(1954:152). Scheper-Hughes and Stewart (1983) measured a
marked decline in the importance of curanderismo in rural
New Mexico, compared to older ethnographies, and describe
a cultural landscape shaped by mainstream biomedical
values:
We would have to conclude that curanderismo in
Northern New Mexico today has moved from being a
primary and important source of medical care to an
alternative and very occasional source in cases of
chronic pain and illness or for those puzzling
psychosomatic and stress-related afflictions for
which biomedicine has little respect and virtually
no cure and therefore for which it has none of the
creative names and faces that the folk medical
system readily supplies (1983:884) .
Ray, Trevino, and Higginbotham (1990) also report that
use of traditional folk medicine (curanderismo) by United
States latinos relates to dissatisfaction with
biomedicine (they found also that curandero use did not
relate to access to biomedical care or to income).
George Foster has suggested that although
traditional humoral folk medicine occurs among United
States latinos, humoral healing practices are often empty
15


cultural artifacts, "shorn of theory" and lacking a
generally accepted humoral theory of the body (1994).
Looking at latinos even outside the United Stats,
Weller has described a decline of the humoral model in
Guatemala, concluding that:
[A] simple rule may be neither possible nor
desirable. Service providers need to be aware that
[a folk model] may exist, but that if a particular
patient/client does not subscribe to it, that actual
interpretation can be highly idiosyncratic and need
not interfere with the delivery of health care
(1983:256) .
The Political-Economic Orientation
The political-economic concept of health care
underutilization emphasizes the relative inaccessibility
of biomedical understanding to the poor and marginalized.
This emphasis focuses not on the conflict of meanings in
the biomedical relationship, but upon lack of control of
information. Cultural difference may at times cover over
the political-economic issues in biomedicine:
A focus on "culture"-- hence on traditions, values,
and attitudes-- rather than on class relations can
be and is widely used to mask and distort the brutal
realities of power and exploitation (Leacock
1982:257).
From the political economy perspective, latino
underutilization of health care is a case of
16


marginalization by a dominant social class, whose
meanings (the biomedical model) are outside the economic
grasp of marginalized groups in general. Baer (1989) has
described United States biomedicine as a "dominative"
institution which mirrors class relationships in ways
that cover over class exclusion; such a system falsely
places the responsibility for utilizing available
resources with the marginalized patient. A critical
social science in biomedicine
[C]annot take biomedicine at face value as the
scientific medical system, as some are inclined to
do, but neither can it stop with merely declaring
biomedicine to be yet another ethnomedicine, as
others would have it (Singer et al. 1990:183).
Rather, individual biomedical relationships and the
health care utilization practices of social groups result
from underlying class-stratified control of biomedical
knowledge and resources.
In building a methodological perspective, Singer
suggests that the empowerment of communities and social
classes is crucial to their members' individual power in
biomedical realtionships (1994). In practice, Singer's
concept translates into a "community-centered praxis," an
applied anthropology agenda which advocates communities'
rights to shape their own relations with larger social
systems (the individual patient's biomedical
relationships are antecedent to the class position of his
17


community, and are not addressed directly by the
political-economic orientation).
Biomedicine is the dominant model of health and
illness in the United States. It involves a great deal
of specialized knowledge and interpretive frameworks
which are understood as objective truths by members of
biomedical institutions. Full competence in the
biomedical model is reserved for the few individuals who
are charged with its use.
The biomedical model within the mainstream of United
States populations is similarly regarded as privy to
objective truths about how illnesses will proceed and be
reconciled, and how pathologies relate to the health of
the individual. Mainstream anglo ideation about body
praxis posits bioscience as the arbiter of truth, and
generally validates biomedical knowledge; they generally
rely on. the biomedical model, but have less-expert
understanding of the model than biomedical professionals.
Coherent belief in a non-biomedical model is the
exception in mainstream anglo groups.
A political economy of medicine suggests that the
degree of agreement between the explanatory models of
caregivers and of patients (and their families) varies by
degree of economic power the patient holds. Individuals
from relatively higher social classes have more access to
18


legitimized cultural capital, and are able to trade more
powerfully in the currency of dominant knowledge.
In the case of biomedicine, simply being in control
of dominant cultural capital does not bring biomedical
understanding. Biomedicine's highly specialized form of
knowledge places all lay people in one category; neither
poor latinas nor anglo businessmen are generally privy to
biomedical training. The difference is that the latter
participates from the position of more highly-valued
cultural capital. Lay people in general have a limited
inventory of biomedical understandings, but poorer lay
people have less access to valued cultural capital, with
which to gain more understanding or to organize further
information into agreement with the dominant model.
By this view, the issue of latino underutilization
stems primarily from their marginalization from the
dominant explanatory model of health care. Where their
models do not reflect biomedical understandings, the
situation is not necessarily to be interpreted as the
presence of an alternative explanatory model, but as a
lack of access to (and control of) the dominant
discourse.
Studies have indicated relatively low levels of
biomedical knowledge in latino populations. Elder et al.
(1994) examined the use of Breast-Self-Examination (BSE)
19


among latinas in two southwestern cities. Even in
populations where a high percentage (approximately 80%)
of latinas had been taught BSE by health professionals,
and a high number (62-63%) reported performing a BSE in
the past month, only 0.7 percent were found adequately
proficient at preforming BSE. English language ability
and acculturation were the most significant predictors,
suggesting that control of dominant cultural capital
affords a better grasp of biomedical knowledge to lay
people.
Lantz et al. conducted focus groups among migrant
farm workers in the American Midwest, finding that
lack of knowledge and information regarding the
causes of cancer, its prevention, and its early
detection and treatment was evident among
participants (1994:512).
Lantz et al. ascribe latino fatalism regarding cancer to
the advanced disease states with which they frequently
seek medical assistance.
Perez-Stable et al. (1992) compared latino
impressions and misconceptions regarding cancer to those
of anglos. Latinos were more likely to have a
"fatalistic" attitude regarding cancer, and to have less
knowledge than their anglo counterparts of the risks and
symptoms of cancers. All subjects were members of the
same pre-paid health plan. Latinos ascribed some
20


individual cases of cancer to a form of punishment by
God. Latinos also suspected a number of worldly
phenomena as causes of cancer, including microwave
appliances, breast feeding, spicy foods, and pork.
Similarly limited understandings of infectious
disease have been identified, especially regarding AIDS
and HIV transmission. Winn (1991) asserts that minorites
are less able, to gather information regarding AIDS and
are also more susceptible to myth and misinformation.
Individuals who were over 50 years of age, or did not
complete high school held the most misconceptions
regarding HIV infection.
Issues; Underutilization of Biomedicine
Latinos in the United States do not utilize health care
resources at the same rates as the mainstream anglo
population. It is unclear from current literature the
degree to which underutilization is due to ethnic
identity, as opposed to the low socioeconomic status that
characterizes many United States latinos. Pettitt et al.
(1994) examined the children of employees of a large
corporation, and found that children were most likely to
have received all of their immunizations if their parents
had higher incomes or were more educated. Parents with
21


the most comprehensive insurance options were also more
likely to have their children fully immunized. Latino
(and African-American) children were the least likely to
be fully immunized at all levels of parental insurance or
pay level.
Intracultural variability within the category
"latino" is great, and not all subsets of latinos show
the same degree of utilization (Aday 1984). The degree
of utilization of latino subgroups varies by both ethnic
assignment (e.g. Puerto Ricans utilize services more than
ethnic Mexicans), and by socioeconomic status. Overall,
latinos in the United States have relatively low averadge
income and educational levels. The median latino family
income in 1988 was $21,800, compared to $33,900 for
anglos (Ginzberg 1991).
Ray et al. (1990) suggest that "cultural barriers"
beyond low socioeconomic status act as contributing
factors to the underutilization of medical services by
Mexican Americans. However, access factors play the
primary role in their analysis (culturally, Ray et al
recommend reduction of waiting time in clinics and
provision of transportation as responses to the problem).
Underutilization is compounded by disproportionately
prevalent risk factors in some latino communities. A
review by McManus et al. (1993) showed that heart
22


disease, cancer, and perinatal health problems are more
likely to kill latinos than anglos. Minority children in
general were more frequently exposed to heavy metals and
other environmental toxins, and more frequently suffer
from asthma than their anglo counterparts. Individuals
who smoke, drink, and have a poor diet are the least
likely to use health care services, with Mexican-American
and Puerto Rican males the most likely of all latino
groups to consume large quantities of alcohol (Solis
1990). Among elderly people in urban areas, latinos
utilize physicians and hospitals at significantly lower
rates than anglos. At the same time, urban latinos are
at higher risk for diabetes, hypertension, and several
other major diseases than are their anglo counterparts
(Dietrich et al 1989) Taken together, the
underutilization of inpatient care by aging urban latinos
is substantial.
A major factor in latino underutilization is lack of
regular primary care insurance coverage. Roberts and Lee
(1980) found that those latinos lacking regular primary
care coverage were also the least likely to utilize other
more specialized medical care. This is especially true
of those latinos at the lowest levels of socioeconomic
status (Aday 1980). Current Population Survey (CPS) data
showed that 10 percent of anglos were without insurance
23


in the mid-1980s, compared with 32 percent of latinos.
Within the latino category, Mexican-Americans were the
least insured subgroup, with 37% uninsured. Those
without insurance were the least likely to see a
physician; nearly 40 percent of Mexican-Americans lacked
a regular primary doctor. Stroup-Benham (1991) reasons
that a high number of United States latinos live with
undiagnosed and untreated medical problems.
Schlesinger et al. (1994) reviewed 1.2 million
insurance claims, examining two different insurance
packages, of differing degrees of coverage. Latinos used
the services significantly less than anglos. Among
women, greater education was related to more frequent use
of services.
The number of latinos lacking primary care coverage
increased in the last decade at nearly three times the
rate of anglos, and their lack of coverage is expected
remain high into the next century (Andersen 1986, Trevino
1983). Where lack of primary coverage increasingly
precludes use of many other medical services, low
coverage is a significant factor in overall patterns of
latino health care underutilization.
Taussig (1987) analyzed appointment records to count
the number of broken, kept, and canceled appointments for
therapy in public mental health clinics. A number of
24


indices, such as primary home language and ethnic self-
identification, were built into scores for acculturation,
to define Mexican-American and Anglo-American samples.
Taussig found no significant difference between
appointment keeping and breaking between the two ethnic
groups. However, socioeconomic status related
significantly to all broken appointments, regardless of
ethnic group assignment.
Snowden and Cheung (1990) describe United States
latinos as having among the lowest rates of use of
psychiatric resources. Young et al. (1986) examined
factors in non-compliance with psychiatric medication
regimens. Their review of 21 studies, without attention
to ethnic background, found the greatest predictive force
in two factors: living alone, and low socioeconomic
status.
Shelton et al. (1990) determined that the most
important factors in latino health care utilization were
logistical access to services and acculturation. Ethnic
identity was a relatively weak predictor of health
resource usage compared to access to services and the
ability to speak English.
25


!>
Issues: Unshared Meanings
Divergent explanatory models impact the patient-
caregiver relationship in a number of ways. One of the
most relevant contemporary issues is in the area of
informed consent as a hallmark of medical ethics and
legal issues.
Informed consent draws its medicolegal validity from
the positivist legal tradition of the rational,
"reasonable individual," positing a natural law by which
any fully informed person would come to her decisions.
Contemporary clinical ideas of informed consent do not
adequately appreciate the impact of unshared meanings.
Informed consent relies on a basic premise of
positivistic, rational science: information is simply
information, which has meaning on its own. Such implies
that the meaning of information (the knowable) is
independent of the person dealing with it (the knower).
Informed consent means telling patients all the raw
information that a reasonable person would need in order
to make a health decision, aware of the possible outcomes
of each option. The assumption of an ideal "reasonable
person" means that any person who does not accept or
interpret the information in the biomedical manner is
26


less than reasonable, swayed by superstition or
irrational fear, or is impeded in their clarity by
psychological stressors of illness. Participation in
becoming informed may be seen as compliance with, and
acceptance of, the meanings assigned to information by
biomedicine itself, compliance which can deny the social
meanings inherent, to the persona of the patient
But humans act on information by virtue of its
meanings when translated into the lives of persons. Its
meaning cannot be divorced from the web of habitus
through which it is read. It depends upon the matrix of
experience and meaning with which the information is
assimilated into an individual's meaningful life
possibilities. By virtue of being sick, a person might
temporarily assign unexpected meanings to information; by
virtue of habitus and psycho-cultural determinants, that
person might also develop lifelong understandings which
depart from the'objective intended meanings (which are
themselves shaped thoroughly by cultural expectations and
medical socialization). Supposedly objective data cannot
properly inform when divorced from semiotic context.
With the populations examined here, numerous difficulties
emerge.
One difficulty is the problem of the individual.
Informed consent is about informing an individual, who
27


makes his own decisions. In the event that this one
person cannot make decisions for himself, biomedicine
turns to a rather prescribed set of medico-legal rules
for determining another individual, who has the authority
to make decisions for the patient.
But the individual, or a proxy, is not always a
valid agent of decision processes in all social contexts.
Decision making may be shared among numerous individuals.
Within latino cultural traditions, where an individual's
identity is often understood in terms of family and
community, familialismo and collectivismo are potentially
inconsistent with the informed consent of the individual
patient.
An individual can appear to be the agent of
decision, especially when viewed from a biomedical model
built on this very assumption. Caregivers can be
distracted from the needs of the actual social units of
decision. Conflicts arise in the care of the patient
which compromise the patient-caregiver relationship.
Notwithstanding a medicolegal foundation of informed
consent as a centerpiece of bioethics, a more universally
valid goal of participation, consent, and clinical
relationships is needed. It is useful to call this goal
valid consent, rather than informed consent. The patient
in a consent covenant must be knowledgeable, voluntary,
28


and competent. The implications for cultural
appropriateness are that being knowledgeable and
competent to interpret medical information assumes
acculturation to the biomedical model, which is not
always the case (Smith, 1981) .
The matter of cultural compatibility carries
significant weight in some issues, but discussion of
cultural appropriateness too often centers on somehow
validating the ways in which the Other sees the world.
At worst, this strategy asserts "folk" models as equal to
biomedical perspectives-- a strategy doomed to fail
within biomedicine. To a biomedicine rooted in
scientific discourse, this is a direct assault on a
sovereign claim to truth, a threat to a monopoly on valid
explanatory models. This has not helped make "culture"
palatable to biomedicine.
This contentiousness need not exist, for cultural
appropriateness which validates other systems of
meanings, which elevates these systems to an equal
position of authority on illness, misses the mark itself.
The issue at hand is not that biomedicine must relinquish
its authoritative stance on the meanings of illness;
biomedical respect for culture need have little to do
with political recognition (suspending for a moment, the
normalizing force of biomedicine as discussed by
29


Foucault, Scheper-Hughes, and others). The central issue
is this: that people will live inside their (non-
scientific) systems of meaning, regardless of whether
these systems are recognized as truths. To communicate
effectively across or among explanatory systems,
acknowledging and negotiating points from the non-
biomedical systems is essential, regardless of any
question of whose systems more closely approximate
"truth."
30


CHAPTER 2
METHODS AND SAMPLES USED
The data used in this exploration are in three sets:
pile-sort data, subjected to Multi-Dimensional Scaling
and cluster analysis; survey data regarding the severity
of some biomedical disease terms; and semi-structured
interviews about respondents' beliefs about illnesses and
biomedical relationships. Each of the three data types
were collected with all respondents. Samples were
composed of twenty-two latinas, twenty-two female
registered nurses, and twenty-two anglo women.
These forms of data collection were chosen because
of the willingness of the subjects to participate in this
form of research; one of the goals was to determine how
these techniques might be utilized in relatively rapid
anthropological studies. All of these data sets were
collected in one sitting with each informant. Unlike
pure survey research, these techniques can be applied to
only a few subjects at any one time. Most of the
collection involved one informant at a time, but some
were surveyed in small groups (up to four individuals),
31


with the individuals interviewed one at a time during the
sitting.
The Samples
Women church members were targeted as "purposive
samples" for three main reasons. First, they are members
of communities to whom others turn for health care
decisions, and they are often charged with the care of
children and aging family members. Second, their
interest in being participants was higher than members of
other organizations; leaders of the women's ministries at
the churches readily provided introduction and
legitimization of the inquiry. Third, relatively
cohesive social groups are less variable than random
samples: "when intracultural variability is high, it
makes little sense to report aggregate data... results
are not only inaccurate, but misleading" (Weller,
1983:256).
The nurses were selected by some related criteria,
including their availability for interview. No emphasis
was placed on selecting a random sample of nurses who
would represent the whole of nursing knowledge, but
attention was paid to ensuring some variety of specialty,
32


age, and degree of nursing education among the subjects.
The nurses were selected as being individuals who are
routinely biomedicine's direct contact with patients and
families.
The demographics of the three groups are presented
in Table 2.1.
Table 2.1
Summary of Demographics
Nurses Latinas Ancrlos
Age distribution:
<20 0 0 2
20-29 9 7 9
30-39 8 7 4
40-49 4 4 7
50 + 1 4 0
Number of Birth Children: 0 11 0 7
1 6 2 3
2 4 10 7
3 1 5 3
4 0 3 2
5 0 2 0
Number in Household:
1 7 0 2
2 7 3 14
3 4 7 6
4 3 10 0
5 1 0 0
6 0 2 0
Marital Status:
Mar 12 20 16
Sgl 6 0 2
Div 3 1 3
Wid 0 1 0
Coh 1 0 1
33


Table 2.1 (Continued
Household Income:
-$15,000 0 2 7
$15-25k 0 9 3
$25-45k 11 9 4
$45,000+ 11 2 8
Educational Levels:
Some H.S. 0 0 1
H.S. Grad 0 6 11
Some Coll 12 5 4
Coll Grad 7 5 5
Grad Degree 3 0 0
Vo-Tech School 0 6 1
Ability to Speak Spanish Bilingual 0 6 1
Some Fluency 4 10 3
Low Ability 4 1 3
No Spanish 14 5 15
Type of Clinic Last Used Private/HMO 18 13 18
Emergency Dept. 1 1 2
Walk-In Clinic 3 2 0
Public Health 0 6 2
Totals:
22 22 22
Registered Nurses
The sample of 22 registered nurses was part of the
inpatient care staff of a large, non-profit teaching
hospital in a downtown neighborhood. All of the nurses
are involved in the care of seriously ill patients; none
are limited to outpatient care.
34


Roles. Two are pediatric nurses, and nine are
medical/surgical caregivers, of which six are regularly
assigned to intensive care. Three nurses are emergency
care certified, and work primarily in the emergency
wards. Four more work primarily with cancer patients,
two with neurological patients, and two are generalized
nurses whose primary role is on-site training of student
nurses. Three of the nurses also work in the local
metropolitan public health system clinics, mostly part-
time (one was full-time in the city clinics, and only
part-time at the hospital).
Demographics. The nurses range in age from 23 years
old to 55 years old, with an average of 33 years. Their
experience as nurses ranges from just under a year to
over 25 years, with an average of 11 years of active
patient care.
Twelve nurses are married, two are single, and three
are divorced; one lives in a marriage-like relationship.
Half of the women have no children; the remaining eleven
have an average of 1.4 children. Seven live alone, but
the average household among these nurses holds 2.3
people.
All of the nurses have at least some college
education, and seven are college graduates. Three more
35


have Master of Science in Nursing degrees. All report
household incomes of at least $25,000, and eleven report
household incomes of more than $45,000.
Ethnic Assignment and Social Status. The nurses
have a relatively high degree of access to educational
resources, and relative fiscal security. None of the
nurses reported being in the lowest two categories of
household income. All of them own vehicles, and
generally express their ability to utilize private
physicians with whom they are personally comfortable (the
exception to this is those nurses who are insured through
Health Maintenance Organizations, who in some cases
expressed feelings that visiting an HMO physician is not
the same as seeing a private physician). None had
utilized a public health resource on their most recent
physician visit.
Within the major categories of census ethnic groups,
the nurses fit into a predominantly anglo national
mainstream. Twenty consider themselves culturally
different from latino patients. None consider themselves
fully fluent in both Spanish and English, and only four
report that they are at least "somewhat comfortable"
using Spanish. None speak any Spanish with their own
families. Semi-structured interviews suggest that those
36


nurses who report some language capability in Spanish
refer to medical Spanish, useful in their careers.
Compared by matched-pairs to the anglo sample set, nurses
speak no more Spanish than anglos (Z-score= -.1529, 2
tailed p= .8785).
The Latina Sample
Twenty-two latinas were interviewed over a series of
Wednesday night meetings at the Westside Protestant
Church, a bilingual congregation situated in a semi-
urban, predominantly hispanic neighborhood. The near-600
members of the congregation are an active and growing
feature of the community, and have participated in
community health and neighborhood revitalization projects
since taking occupation of the site nine years ago
(Westside was previously housed in a smaller facility in
the same area).
The Westside Protestant Church is fully bilingual in
its offerings. Church materials are printed in both
Spanish and in English. Almost all the meetings of the
congregation are held in both Spanish and English.
Sunday morning services are held in both Spanish and
English separately, but in order to unify the
37


congregation, smaller group meetings and scriptural
discussions are held with both languages together,
facilitated by bilingual deacons and clergy.
Westside is a rapidly growing congregation,
following a contemporary trend of conversion of United
States latinos to Protestant denominations. Most of the
congregation is composed of people who dwell in the
immediate, area; Westside is a neighborhood organization,
and is a focal point of many processes in the community.
The use of the term "latina" is chosen entirely for
the convenience of the writer and reader; the respondents
use the term occasionally, but more frequently call
themselves "hispanic" or "Mexican" (but were never heard
to use the Spanish form of the adjective, "Mexicana").
Some of the older women were heard identifying other
women with the label "Spanish lady," as in, "have you met
Janet? She's another Spanish lady here" (no respondents
were from Spain). Various other labels for people of
Spanish-speaking descent are used widely.
Demographics. The women range in age from twenty to
sixty-five years of age, with a mean age of 37 years.
This is not significantly different from the age set of
the nurses or of the anglo women.
38


Five of the Westside women say they were born
outside the United States-- all in Mexico-- but all the
women were reportedly raised within the Western United
States. Eleven were born to parents who themselves were
U.S. born. The remaining six were all born in the U.S.
as well, but reported having at least one parent born in
Mexico.
All of the Westside women-- the entire latina
sample-- had been married at least once. One 55-year-old
woman was widowed, and one was divorced. This high
marital rate differs significantly from the marital rates
of the nurses (Chi-square= 9.2, one-tailed p<.005), and
from the marital rates of the anglo women (Chi-square=
1.66, p<.01) .
The Westside women also more frequently had children
than the nurses-- all of the latinas have children. This
difference is statistically significant (Chi-square=
9.23, p<.005). The higher number of births per
respondent compares significantly to the other groups.
Number of individuals in the household differs from that
of the nurses (Z-score= -2.47, p= .0136), and from that
of the anglo women (Z-score= -3.375, p= .0007).
The latinas are significantly less-educated than the
nurses (Z-score= -2.864, 2 tailed p= .0042). Six of the
women have gone through vocational or technical
39


certification programs, and five women have graduated
from a baccalaureate-level degree program.
Ethnic Assignment and Social Status. The Westside
community is economically depressed, and the reported
household incomes of the sample group reflect the women's
low socioeconomic status compared to the nurses. Eleven
of the women report household incomes in the bottom two
categories available-- less than $25,000. This differs
significantly from the income levels of the nurses, of
which none had reported household incomes in the bottom
two categories (Z-score= -3.258, 2 tailed p= .001). Two
of the women reported total household incomes of greater
than $45,000 annually. Two of the women reported that no
one in their household owned an automobile (this
departure from nurses is not statistically significant).
The Westside latinas also have significantly larger
households and numbers of birth children, than do either
the nurses or the anglos. Their lower income designation
makes them poorer than nurses but the latinas' budgets
are more often spread over larger households. With more
mouths to feed, their low income is reduced further.
The latinas reported having used a public health
clinic, on their last physician visit, in six of the
twenty-two cases. One other was seen in an emergency
40


room, and another two visited walk-in care convenience
clinics. The remaining thirteen all saw a private
physician (or HMO equivalent) on their most recent
physician visit. This use of public health clinics is
significantly different from that of the nurses.
Comparing numbers of public health uses to all other
types of physician visits, the latinas used public health
resources at higher frequency (Chi-square= 8.25, 1-tailed
p<.005). Compared to the anglo group, the difference is
also directional and significant, though to one lower
order of magnitude in confidence (Chi-square= 3.66, 1-
tailed pc.05). Although two of the latinas respond that
no one in their household owns a car, all reported having
traveled to their last physician visit in a private car.
The latinas differ from the anglo women, and the
(also predominantly anglo) nurses, in that their use of
Spanish is greater. Six of the latinas report themselves
as fully fluent and bilingual in both English and
Spanish, and ten more designate themselves as at least
"somewhat comfortable" in speaking Spanish. One woman
reports herself as "not very comfortable" speaking
Spanish, and five more report that they do not speak
Spanish at all. The degree to which these women speak
Spanish sets them apart as significantly different from
both the anglos and the nurses. By Wilcoxon matched-
41


pairs analysis, the latinas are in a different category
of fluency in Spanish, with high confidence compared to
anglos (Z-score= -2.9733, p= .0029), and to nurses (Z-
score= -3.266, p= .0011).
The latinas' language use may be more Spanish-
oriented than the survey responses suggest. On one
occasion, a respondent was observed to describe herself
with the category, "I don't speak Spanish." Moments
later, this same woman turned to a toddler playing
noisily at her feet: "Sal de aquil Busca a Papa!" Other
latinas may have answered questions regarding their
Spanish ability with similar modesty.
The Anglo Sample
Twenty-two anglo women were interviewed at the
Eastside Protestant Congregation during weekly meetings
of the congregation's women's group, and during meetings
for the church's single mothers ministry.
Demographics. The mean age of the women is 31.7
years, ranging from 19 years to 47 years of age. This
age grouping is not significantly different from that of
the nurses.
42


Ten of the twenty-two respondents reported household
incomes of less than $25,000; of these, seven designated
their household resources as less than $15,000 dollars
(the seven women who reported the lowest income are all
under the mean age, ranging from 19 to 28 years of age,
with a mean of 22.7 years). This low income level is
significantly different from that of the nurses (Z-score=
-2.978, 2-tailed p= .0029).
Seven of the anglo women have no children. Those
who are mothers have an average of 2.7 children. All but
two of the women live with family and friends, with an
average household population of 2.18 individuals. This
household size is not significantly different from that
of nurses (p= .925).
Sixteen of the anglo women are married. Of the
remaining six, four are divorced, but one is living in a
marriage-like relationship with her ex-husband. Two are
single. The number of anglo women who are married is
higher than that of the nurses. But within each group,
the number of married individuals versus all others, is
perhaps directional, but not significantly different
(Chi-square= 1.69, .10> p >.05).
All but one twenty-one year-old woman had completed
at least high school or a G.E.D. certification. One
woman reported having completed a secretarial school
43


program, in medical records coding. Nine of the women
had attended at least some college, and of these, five
had earned a bachelor's level degree. None had earned
graduate degrees.
In fifteen of twenty-two cases, the nurses declare a
higher level of education than the anglos by matched
pairs analysis. This difference is statistically
significant (Z-score= -2.864, 2 tailed p= .0042).
Ethnic Assignment and Social Status. Eastside
Congregation is located in an industrial urban fringe,
which in recent, years has seen an increase in the size of
local minority populations. Eastside members note that
their congregation was once the norm for the local
neighborhood, in which a long-empty church building
nearby has recently re-opened, housing a charismatic,
evangelical congregation composed mostly of local
minority members. The Eastside Congregation is not
shrinking, and many of its members are young families.
But members do report that Eastside is not growing at the
rate of churches which are identified as latino.
Eastside Congregation engages in a number of joint
projects with some other local churches, and their credo
of the unity of their faith is credible. However, they
are aware of the changing demography of the neighborhood.
44


The anglo women's use of health care resources does
not differ significantly from that of the nurses,
although two of the anglo women had used public health
clinics on their last visits to physicians. Eighteen of
the remaining women had used a private physician or an
HMO physician on their last visit, and two had most
recently been seen in emergency rooms.
Methods Utilized
Three data-sets for this exploration were collected
from each of the three sample groups. An example of the
survey instrument is found in Appendix A. At the same
time the survey was administered, the informants each
participated in a pile-sort exercise and a semi-
structured interview.
Multi-Dimensional Scaling
Multi-Dimensional Scaling uses aggregate pile-sort
data to build a map of the similarities (relatedness)
among items from a given list. The procedure is non-
metric .
45


Respondents are given a stack of cards, each labeled
with a disease term. They are asked to sort the cards
into piles which show how these items are similar or
different. The respondents were all given the uniform
instruction:
Please place these cards into five to eight piles,
showing how they are similar or different. Use
whatever reasons you want to, to decide how they go
together.
If further clarification was needed, the respondent was
given an example:
If these were cars, you might want to put all of the
trucks together, or maybe all the Chevys together
and all the Hondas together, and so on.
In all cases, this explanation was sufficient. All
respondents were able to find satisfactory arrangements
for their piles.
The disease terms were selected by asking six nurses
to list the first 40 medical ailments that came to mind.
From these lists, the most, frequently mentioned items
were selected (rare diseases, or disease terms mentioned
by only one or a few informants, were regarded as less
important than diseases that all informants listed). The
list was culled to 21 items for Multi-Dimensional
Scaling.
Anthropac (Borgatti, 1993) software was used to
convert the respondent's piles into aggregate data
46


matrices. Items frequently placed together by
respondents were assigned high proximity scores (from
zero for no co-occurrences, to 22 co-occurrences where
all informants placed the items together). These
frequencies of co-occurrence became the proximity
ratings, used to assemble a best-fit map of the items,
via the Minissa-X algorithm. An example of a correlation
matrix may be found in Appendix B. Each illness term was
assigned an abbreviation in order to reduce clutter on
the MDS maps; a key to the MDS labels is found in
Appendix C.
Disease Severity Surveys
Using a Likert scale, all respondents were asked to
assign a rank for typical severities of some different
diseases. The outcomes were ranked on the categories:
1) Always or Almost Always Fatal
2) Sometimes Fatal
3) Rarely Fatal
4) Never or Almost Never Fatal
The illness terms selected were acceptable to most
respondents, and most were able to complete the survey
without further explanation. The exceptions were a few
nurses who felt that some terms needed to be more
47


specific ("is this Lymphoma Hodgkin's or non-Hodgkin's
type?"); they were asked to consider the class of
diseases subsumed by that label. A few of the lay women
indicated that the terms were too precise; one woman
wanted to know why the list included both emphysema and
asthma, when these "are almost the same thing." These
cases were exceptional; most respondents understood the
question and were able to complete the survey.
Technical Term Assessment. All respondents were
asked to assess the severity of the terms "Myocardial
Infarction" and "Heart Attack." These two terms have the
same clinical definition. It was suspected that the
percent agreement between the two terms for each group
would depend on their familiarity with biomedicine.
Specifically, the two terms should "tie" in severity more
often by matched pairs analysis when assessed by nurses
than when assessed by lay people.
Semi-Structured Interviews
Each respondent was interviewed using a semi-
structured format, either during or immediately following
their completion of the survey and pile-sort exercises.
In some cases the interview was used to complete a
missing value in the demographic survey.
48


All respondents were asked to describe some of their
pile-sort categories and discuss their criteria for
sorting the disease terms, using their own piles.
Each respondent was asked about the type of doctor
they last saw, to clarify the answers elicited by the
demographic survey. Each was asked to compare the
doctors they usually use now, from the doctors they saw
as a child, and whether they were satisfied with the care
and attention they receive from their current source of
health care.. They were also asked about alternative
sources of health care they might have used in recent
years.
To help qualify data regarding the women's language
and information source preferences, each was asked what
kind of television they usually watch. This was to
determine whether there were a likely difference in types
of mass media exposure, between the latinas and anglo
women.
49


CHAPTER 3
FINDINGS AND RESULTS
Multi-Dimensional Scaling
The exploratory MDS maps generated from pile-sort
data do not reveal differences in the taxonomies applied
by latinas, anglos, and nurses to the biomedical terms
used. The maps may suggest a higher degree of consensus
and certainty in the nurses' taxonomy for the terms, but
this is speculation regarding the non-metric MDS maps.
Abbreviations of the illness terms were used in preparing
the graphical maps; a key to the mapped labels is
presented as Appendix D.
The "map stress" or "graphical strain" expresses the
degree of mathematical inaccuracy the graphical map must
allow, to calculate a fit for the data points (from no
stress at 0, to unity at a total distortion of the data).
Romney and Weller (1988) suggest that map stress of
between 0.10 and 0.15 is generally considered acceptable.
MDS Mapping: The Nurses
The nurses' responses to pilesort exercises yield a
50


well organized set of categories and axes with which to
group the illness terms. The final map is presented as
Figure 3.1. The map stress of .105 is deemed acceptable,
following Weller and Romney (1988).
Clustering. The clusters may be seen in five major
components, mapped in a cruciate pattern. At upper left,
the cancers are all listed together: Stomach, Breast,
Liver, Colon, and Leukemia.
At lower left, cardiovascular diseases are found in
a unified group. Hypertension (High Blood Pressure),
Coronary Artery Disease, Myocardial Infarction, and Heart
Attack come together with Aneurism and Stroke (Cerebral
Vascular Accident). These diseases are unified by the
quality of relating to heart and vascular disease.
Although related anatomically to cranial structures, the
Stroke category (resulting from blocked or bleeding brain
vasculature) is placed with vascular disorders.
At upper right are illnesses contracted by lifestyle
or by voluntary exposure to risk factors (as shall be
further examined under interview results). AIDS,
Alcoholism, and Hepatitis (all types together) are
clustered together. There is one curiously absent item
in this cluster: Emphysema. Emphysema comes primarily
from chronic inhalation of smoke and particulate
51


material, almost always from habitual tobacco use, but
here is less associated.with lifestyle and risk factor.
Emphysema is more closely related to the cluster at lower
right. As is discussed under interview findings, below,
this area of the map features a group of diseases
apparently organized by two principles: respiratory
ailments and infectiousness. Emphysema appears related
to Meningitis, Influenza, and Pneumonia. Diabetes and
Kidney Failure are grouped together roughly at the center
of the other four clusters.
Axes. The axis of greatest importance in this map
appears to be the horizontal axis from left to right,
across the map. At left, the cardiovascular and
malignant illnesses may be seen as originating in
processes internal to the patient's body. .While
certainly mediated by behavioral and environmental
factors, both cancers and cardiovascular diseases are
internal processes of declining or improper biological
function.
This contrasts with the right side of the map at the
other end of the axis. The disease terms at right all
result (primarily) from external forces which impinge on
the body. The infectious diseases, clearly, are brought
on not by failed body function, but by the insult of
52


GsCA
BrCA LiCA
LnCA
luk
AIDS
ale
hep
CAD HTN
rf
diab
emp
flu
MIt
anuCVA
pnu
mng
Figure 3.1
MDS Result for Registered Nurses
53


pathogenic organisms. At the same time, AIDS, Hepatitis,
and Alcoholism depend on individual behavior patterns and
risk factors (semi-structured interviews suggest that the
Diabetes/Kidney Failure cluster is a part of the left end
of this axis, graphically pulled toward the
infectious/behavioral end of the axis by two factors: the
perception that diabetics' prognoses are mediated in part
by their ability to control the illness through their own
behaviors, and the nurses' awareness of the increased
frequency of diabetes in populations at risk for these
other somewhat behavioral diseases, including minorities
of low socioeconomic status).
MDS Mapping: The Latino Women
The latino women's MDS mapping results differ in
some ways from the map produced by nurses, and is
presented as Figure 3.2. Map stress is calculated at
.133, indicating an acceptable fit to the aggregate
matrix, following Romney and Weller (1988) .
Clusters. At upper right, the latino women's MDS
results show a cluster of cardiovascular illnesses.
Myocardial Infarction and Heart Attack lie very near one
another, and immediately below these lie Coronary Artery
54


Disease, High Blood Pressure, and Stroke. Aneurism lies
in this cluster as well, set off to the lower left.
At the upper right portion of the map, cancers lie
together. Breast, Lung, Liver, Colon, and Stomach
Cancers virtually overlay one another, and Leukemia lays
nearby to the left. The organization of cancers
together, of course, is in large part to be expected.
The terms all include the word "Cancer," and are clearly
related by this trait. However, the latino women largely
included Leukemia in the cancers, although its name does
not give away its biological category.
At lower right, Pneumonia, Emphysema, Influenza,
Meningitis, and Hepatitis come together in a group.
Hepatitis lies-here and not with the "behavioral"
diseases of AIDS and Alcoholism.
AIDS and alcoholism lie together in the lower left
portion of this map. At far left, the categories of
Diabetes and Kidney Failure lie close together.
Axes. Some similar axial distributions of the
illness terms to those of the nurses exist. Infectious
diseases occur at the lower portion of the map, mostly
off to the right side. Absent is the central location of
Diabetes and Kidney Failure. The two are associated, but
not placed between internal and external etiologies.
55


MI
Hrt
diab
rf
CAD HTN
CVA
anu
BrCA
luk
LiCA
LnCA
AIDS
ale

emp
hep
mng
flu
pnu
Figure 3.2
MDS Result for Latinas
56


In sum, the latinas' MDS results display apparent
similarity in the clusters of taxonomy to those of
nurses, but are not fully congruent with the axes of
biomedicine as constructed from the nurses' results.
MDS Mapping: The Anglos
The map generated by anglo women does not appear to
be clearly unified along the cruciate pattern observed
among the nurses. Two-dimensional model stress is
indexed at .117, acceptable under the standards put forth
by Weller and Romney (1988). The anglo map is presented
as Figure 3.3.
Clusters. The clustering of illness terms among the
anglo sample does not follow the same cruciate pattern as
the nurses produced, although the same cluster groups are
apparent.
At upper left are vascular and cardiac ailments.
Stroke, Aneurism, and Myocardial Infarction lay together
with Coronary Artery Disease and Heart Attack.
Hypertension lays near this group but further toward the
center of the map.
At top center, we again see the co-occurrence of
57


infectious and respiratory ailments. Pneumonia and
Influenza lay together with Meningitis, and Emphysema is
not far (though this item strays toward the center of the
map). Lung Cancer lies nearby, to the right margin of
this cluster.
Further right, the other cancers are grouped
together. Breast, Liver, Stomach, Breast and Colon
Cancer cluster together, all at nearly the same
coordinates (the outlying Lung Cancer is positioned
between the cancers and the respiratory cluster described
above). Leukemia lies outward from the cluster also, to
the lower extreme.
A grouping of AIDS, Hepatitis, and Alcoholism
appears at the lower border of the anglo women's map.
Kidney Failure and Diabetes are at lower right.
Axes. The axes of separation in the anglo map seem
to be similar to those of the nurses' map. Increasing in
power to lower right, we may see an axis of causation--
the further to the lower right, the more the illness
items appear to relate to lifestyle and external risk
factors.
58


pnu
MID mng
CVA anu emp flu LnCA Br LiC
HTN luk
rf
diab hep
ale AIDS
Figure 3.3
MDS Result for Anglo Women
59


Comparative MDS Appraisal
The MDS maps do not appear to show the lay groups
using a distinct folk model to organize these disease
terms. The MDS results for the lay women appear to
follow the nurses' more authoritative model, yet lack
expertise in its detailed principles. The disease terms
used are not native knowledge-- they do not derive from
the people who have been asked to sort them, and they not
necessarily part of the lay women's usual vocabulary.
People have axes of discrimination and cognitive models
for the terms that they themselves use, but these lay
women do not routinely use all of the biomedical terms
they have been asked to sort.
The lay groups are able to recognize, mostly, what
goes together and does not go together according to the
biomedical model, as gathered from the nurses. AIDS,
diabetes, cardiovascular problems, and cancers are all
much discussed in the public arena and mass media.
This may be described as the ability of the lay-
people to recognize basic stock categories of biomedical
illnesses, and their simultaneous inability to fully
utilize the biomedical taxonomy. For instance, both lay
groups are able to recognize that Coronary Artery Disease
goes along with Heart Attack, often by thinking of what
went together in people that they have known (evidenced
60


by interview data, below). And when confronted with
technical terms such as "Myocardial Infarction," they are
able to recognize the "-cardial" portion of the term as
heart-oriented. This converts the exercise into a task
of sorting by meanings other than the biomedical meanings
which the nurses explained as the basis for their
pilesorts, yet produce essentially the same clusters in
the lay groups' MDS map results.
The nurses describe using a biomedical taxonomy
which sorts illnesses into clusters by organ system and
by axes of disease etiology. By this model, it appears
that diseases which result proximally from internal body
system malfunction, aging, or degradation, occur as an
opposite pole from illnesses caused by exposure to insult
from outside the body. Two related medical problems--
diabetes and its associated renal failure-- occur as
problems which are mostly internal medicine issues, but
are impacted and mediated by lifestyle, thus drawing them
graphically closer to the outside-insult end of the axis.
The anglos and latinas both cluster diseases into
categories similar to the clusters that the nurses use--
to this extent, they are able to appreciate some of the
same categories of illness as co-occurrent and similar.
But they do not seem to utilize, at least as thoroughly,
the same axes of discrimination. Lacking knowledge of
61


the internal logic of the nurse-derived list of diseases,
their responses turn to their personal experiences and
relationships with the terms, instead of to the
biomedical taxonomy that produced them.
Were the lay groups asked to pilesort terms that
came from a free list exercise among latino women, or
among low socioeconomic status anglo women, it may be
that the MDS results would reveal a very different sort
of clustering taxonomy. In some situations, ethnomedical
categories could become relevant to understanding how the
groups think about illness in general. But in coming
into biomedical relationships, patients and families are
faced not with their own taxonomies, but with those
offered (and imposed) by biomedicine. In this case, the
respondents show how they deal with organizing terms that
they might be confronted with in biomedical situations.
Rather than dealing with these terms with a coherent
ethnoscience, they organize the terms by virtue of
experience with family and acquaintances, and basic
principles not reserved only for recipients of biomedical
training: cancers go together; high blood pressure has
something to do with the heart; AIDS and alcoholism have
to do with how you live your life.
This finding does not indicate whether less-
disadvantaged lay people would use a more expert version
62


of the biomedical model. Lawyers and accountants, for
instance, are also excluded from biomedical training, but
they control more highly-valued cultural capital. This
may afford them greater participation in biomedical
relationships, and consequently greater access to the
biomedical knowledge.
Survey Findings
A survey conducted regarding the severity of several
health problems found few significant differences between
the groups. Matched pairs analysis was used to compare
the nurses' responses to each lay group, and to compare
the responses between the lay groups.
A table of the significant differences found is
presented as Table 3.1. For three groups of respondents
asked to assess 36 different disease terms, the data
consists of 108 individual data points (latina versus
nurses, latina versus anglos, and nurses versus anglos
for each item). At the p <.05 level of confidence, we
should not be surprised to see at least six significant
differences arise by chance alone.
63


Table 3.1
p-Values of Significant Differences
In Severity Assessments Between Groups
DISEASE TERM Lat/Ang RN/Lat RN/Ang
HEART ATTACK CORONARY ARTERY DISEASE . 0077
DIABETES TUBERCULOSIS (TB) EMPHYSEMA . 0032 . 0151
ANEURISM ARTERIOSCLEROSIS STROKE LEUKEMIA LYMPHOMA KIDNEY FAILURE HEADACHE STOMACH CANCER CERVICAL CANCER COLON CANCER HERPES . 0106 . 0106
CHICKENPOX STOMACH ULCER LUNG CANCER MENINGITIS HIGH BLOOD PRESSURE .0164
HEPATITIS . 0281
SYPHILIS GONORRHEA LIVER CANCER BREAST CANCER UTERINE CANCER AIDS (SIDA) MYOCARDIAL INFARCTION ALCOHOLISM THE COLD INFLUENZA . 0022 .0071
MEASLES PNEUMONIA ASTHMA OBESITY .0414 .0468
64


There is no indication that the differences found
emerge from latina cultural ideas about illness. The
data contains eleven significant differences in the
entire matrix. Two are differences between latina and
anglo assessments, four are between latinas' and nurses'
answers, and the remaining five are between anglos and
nurses.
Technical Term Assessment Findings
By matched pairs analysis, the nurses showed the
highest number of "ties" between "Myocardial Infarction"
and "Heart Attack." Nurses tied on these items in 18 of
22 cases, for an 81% rate of agreement. The latinas tied
the two terms in 14 of 22 cases (64%), and anglos in 13
of 22 cases (59%).
When the items are compared by matched pairs within
the groups, a directional difference emerges between
nurses' assessments of "Myocardial Infarction" and "Heart
Attack." The difference is not statistically significant
(p= .06), but may indicate a directional trend in nurses'
use of the term "Myocardial Infarction" (or simply
"M.I.") against "Heart Attack."
65


Semistructured Interview Findings
Semistructured interviews did not reveal meaningful
differences in the fundamental disease models employed by
the three groups, except in levels of expertise in the
biomedical model. The interviews indicated that the lay
groups generally place confidence in biomedical authority
on health and disease, but do not have detailed
understanding of the biomedical model.
Interviews with the nurses suggest a high degree of
consensus regarding how the disease terms should be
organized. Etiology and organ system seem to be the most
important aspects of the disease taxonomy; nurses would
routinely describe their pile-sort organization by
pointing to their piles: "That's all cancers, those are
vascular and heart, and these are contagious..."
Nurses reserve a special category for diseases
caused by risk factors invited by personal lifestyle and
activities. AIDS, hepatitis, and alcohol addiction all
are ascribed first to the actions of the patient, who
nurses may see as responsible for his own health problem.
One 26 year-old nurse said of these ailments, "...and
these are things you do to yourself."
Latinas did not mention cultural difference as a
problem with their current physicians and caregivers.
The latinas did relate their dissatisfaction with health
66


care resources to their difficulties finding insurance
coverage and gaining access to health care more desirable
than the public health clinics.
Discussion with interviewees regarding their pile-
sort results do not indicate a humoral or other folk
model as an organizing principle, among the latinas or
among the anglo women. Both the latinas and the anglo
women refer to a biomedical taxonomy to organize the
terms, but their interviews reveal a limited
understanding of the biomedical terms. They are less
familiar with all of the terms, and less confident about
using them. Explaining her placement of "Myocardial
Infarction," one 28 year-old anglo woman said, "I don't
know what this is, but it sounds pretty bad. I think
that cardiac means your heart." Of pneumonia, she
reported that "it goes with emphysema because of people
get it in the lungs [sic] when they're old. And
influenza you can also catch from other people."
Both lay groups augment their limited biomedical
understanding by appealing to personal knowledge of
people who have had these health problems. The lay
groups place disease terms together in piles by virtue of
who they've known in connection with the disease. They
appear to echo the nurses' categories of disease, but
with a less explicit understanding of biosciences, and a
67


greater reliance on their personal understandings of
diseases and individuals who have been afflicted:
Anglo, 27 years old: "I think this is what my aunt
with diabetes got" (referring to "Kidney Failure").
Anglo, 31 years old:"I don't know what this is, but
they talked about it when my dad had his heart
attack" (holding card labeled "Arteriosclerosis") -.
Latina, 30 years old: "My cousin drinks too much,
and my mother says that's why he has hepatitis"
(placing "Hepatitis" with "Alcoholism").
Latina, 23 years old: "My baby was premature and had
these, and the neighbor boy had the flu real bad the
same way" (pointing to a pile including "Pneumonia"
and "Meningitis").
Latina, 65 years old: "These are things that happen
with people who aren't living right" (placing
"Alcoholism" and "AIDS" together).
Although the latina woman quoted above placed "Hepatitis"
and "Alcoholism" together, the overall latina MDS map
shows "Hepatitis" closer to other infectious diseases
than to behavioral-risk factor diseases (infectious
Hepatitis-A is more common in Latin American immigrant
populations than in the United States mainstream, and
this may inform the context of the latinas' placement of
Hepatitis).
68


Among both latinas and anglos, the respondents
expressed concern about, their level of access to
biomedical care. Both groups of women report putting
considerable effort into using private doctors instead of
public health clinics. Their underutilization is not a
willing resistance to a model they do not believe in; it
is a symptom of their lack of access to services. One 30
year-old latina, a mother of two children, said,
I think that the reason that Hispanics and Mexicans
don't go to the doctor is that they don't have good
insurance. I mean, I have a good job here, I work
in the church doing bookkeeping, but they don't have
insurance for me. I don't like going down there
[Public Health Clinic], and I have to wait with my
kids in the little room with these people, all dirty
and stuff. They treat us like we're those people
and we're not.
69


CHAPTER 4
DISCUSSION AND DIRECTIONS
This exploratory examination of three groups of
women raises questions and points out areas for further
inquiry. The cultural orientation and the political-
economic orientation address somewhat different levels of
analysis in biomedical relationships and
underutilization. Cultural orientation and the primacy
of negotiated explanatory models hold relevance for the
individual clinical relationship, while the political
economic perspective speaks to the needs of communities
and larger populations.
The Merits of a Cultural Orientation
A cultural perspective on latino underuse of
biomedicine has its greatest value at the micro-level of
biomedicine, the individual patient-healer relationship.
Where different cultural models exist between patients
and caregivers, they exist in the form of the cognitive
schemata that individuals carry into the clinics and
70


hospitals. But because the individuals' cognitive schema
and explanatory models are rooted in a larger cultural
worldview, their actions and needs can be called
cultural, rather than personal.
The cultural orientation in understanding latinos'
problematic use of biomedicine is relevant to the
clinician who deals with individual patients, and with
the impact of differing schema. This is rather a
different issue from the general ignorance of lay
populations of specialized biomedical knowledge.
Negotiating explanatory models is not simply a matter of
providing biomedical information to the patient, filling
a cognitive void. Patients of some backgrounds may
already have a set of standards and schematics for making
sense of illness experiences-- different schemata
regarding the workings of the world and body, and
extending also into the patient's social suurroundings.
Among the social influences on the individual's
clinical relationship is the cultural ambiguity of
informed consent and clinical decision-making. A
biomedicine better attuned to questions of individualism,
collectivismo, community, and meanings, is a biomedicine
in greater accord with the needs of its patients.
Under a culturological agenda, the clinical role
becomes one of negotiation over authority (or toward
71


understanding in a "culturally sensitive" medicine), but
this is not to suggest that all cognitive models find
voice in ideal biomedical relationships. An individual's
mental status, for instance, may confound and preempt
appreciation of culturally-different explanatory models.
Culture shapes individuals' beliefs and experiences
of health and sickness. It is with the patient and
caregiver that cultural incongruence and shared meanings
emerge. To come to a resolution of meanings in the
therapeutic relationship is to come to know the
explanatory model of the patient, and to negotiate and
participate in the meanings of the sickness, just as in
the rigors of the treatment. The ethos is as fleeting,
perhaps, as the art is long.
It is simpler for anthropology than for biomedicine
to accept non-biomedical cultural models' legitimacy
beyond individuals' discourse with caregivers. But
patients will live and die within their explanatory
models with or without the approval of biomedicine.
Where the stated goal of biomedicine is the healing of
the patient, sensitivity to the cultural meanings of
illness is central to the relationship of patient and
healer.
72


The Merits of a Political- Economic Orientation
The data at hand do not show that these groups of
women use different cultural models for health and
illness, but suggest that the lay women control only a
novitiate understanding of biomedical knowledge.
Regarding the health of communities and of marginalized
populations of United States, the political-economic
orientation is of greater value.-
The political-economic orientation suggests that the
two groups of lay women do not see the world in radically
different, culturally-programmed ways. They understand
biomedical terms and ailments, and their own biomedical
relationships, from relatively similar positions of
economic and educational marginalization.
Lay people in general are not privileged to
specialized biomedical knowledge, and all the more so for
lay people who are marginalized by their devalued
cultural capital and socioeconomic status. As Singer er
al. write,
[W]e must consider the social origins of illness and
class (as well as racial and gender) control of
health institutions and knowledge (1990:185).
Exclusion from biomedical knowledge relates to a social
hegemony expressed in biomedicine's relationship with lay
populations.
73


Where biomedicine is positioned as the gold standard
of clinical understanding, limited knowledge gives rise
to misinformation. Ironically, patients' own
misunderstandings can create a false sense of
communication in the biomedical relationship, further
confounding troublesome clinical encounters. The women
examined in this exploration, for example, created
similar clusters of the same illness terms, but this does
not mean that the lay groups and the nurses came to these
similar categories in similar ways.
The political-economic orientation advocates greater
sensitivity to poor lay people's marginalization from
more-valued cultural capital. Providing information to
patients cannot be limited to providing raw information,
but must also address the lack of the overarching
principles into which this information is supposed to
fit.
A focus on culturally different meanings of illness
can become a romanticized distraction from attention to
the social stratification embedded in biomedical praxis:
Characteristic of this circumscribed explanation is
its inattention to the manifold ways doctor-patient
interactions are structured by a wider field of
class and other power relations embedded within, but
not always visible form, the narrow confines of the
clinical setting (Singer et al. 1990:179).
74


Where marginalized class status is the basis for
biomedical exclusion, devalued cultural capital is the
cause of biomedical underutilization. A critical social-
science view of biomedicine must be:
concerned with synthesizing the macrolevel
understanding of political-economy with the
microlevel sensitivity and awareness of [cultural]
anthropology (Singer et al. 1990:179).
Appropriate strategies for expanding access to
biomedicine must address not only the financial logistics
of access, but the complex issues of cultural
marginalization as a function of socioeconomic position.
75


Appendix A- Example of Survey Instrument
YOUR AGE: NUMBER OF CHILDREN:
TOTAL NUMBER OF PEOPLE IN YOUR HOME:
You are (circle one):
1) MARRIED
2) SINGLE
3) DIVORCED
4) WIDOWED
5) LIVING IN A MARRIAGE-LIKE RELATIONSHIP
Total Household Income, from all sources, before taxes:
(circle one)
1) less than $15,000
2) between $15,000 and $25,000
3) between $25,000 and $45,000
4) over $45,000
What is
the one
1)
2)
3)
4)
5)
6)
your highest level of education? Please circle
category that best describes you:
Some high school
High school diploma OR G.E.D. certificate
Some College
Bachelor's Degree
Graduate Degree
Vocational or Technical Certification in:
76


Do you, or a member of your household, own a car?
1) YES
2) NO
How comfortable are you speaking Spanish?
1) VERY- I'm bilingual
2) SOMEWHAT COMFORTABLE
3) NOT VERY COMFORTABLE
4) I DON'T SPEAK SPANISH
What language do you speak with your children, if you
have any?
1) SPANISH
2) ENGLISH
3) BOTH
4) OTHER
What language do you usually speak with your spouse, if
you have one?
1) SPANISH
2) ENGLISH
3) BOTH
4) OTHER
How many hours do you watch television in a typical day?
In your home, how many hours is a television turned on in
a typical day? _____________
How far did you have to travel the last time you or your
family went to see a nurse or doctor?___________________
77


What kind of medical facility did you use the last time
you or your family went to see a nurse or doctor?
1) PRIVATE DOCTOR'S OFFICE
2) EMERGENCY ROOM
3) CONVENIENCE CARE CLINIC
4) PUBLIC HEALTH CLINIC (County Health Dept, etc)
5) OTHER______________________
Did you get there by (circle one):
1) PRIVATE CAR
2) BUS
3) WALKING
4) OTHER______________________
78


PLEASE INDICATE HOW LIKELY ONE IS TO SURVIVE, OR HOW
LIKELY ONE IS TO DIE FROM, THE FOLLOWING DISEASES AND
HEALTH PROBLEMS, ON THE SCALE BELOW:
1) Always or Almost Always Fatal
2) Sometimes Fatal
3) Rarely Fatal
4) Never or ALmost Never Fatal
TERMS:
Heart Attack
Coronary Artery Disease
Diabetes
Tuberculosis (TB)
Emphysema
Aneurism
Arteriosclerosis
Stroke
Leukemia
Lymphoma
Kidney Failure
Headache
Stomach Cancer
Cervical Cancer
Colon Cancer
Herpes
Chickenpox
Stomach Ulcer
Lung Cancer
Meningitis
High Blood Pressure
Hepatitis
Syphilis
Gonorrhea
Liver Cancer
Breast Cancer
Uterine Cancer
AIDS (SIDA)
Myocardial Infarction
Alcoholism
The Cold
Influenza
Measles
Pneumonia
Asthma
Obesity
79


Appendix B- An Example Correlation Matrix File Contents
For Multi-Dimensional Scaling
DL N = 21 SIMILARITIES, LOWERHALF
LABELS:
HrAtk,CAD,diab,emp,anu,CVA,luk,RF
hep,LiCA,BrCA,AIDS,MI,ale,flu,pnu
DATA:
21
2 3
0 1 4
18 17 3 0
16 15 2 0 16
1 1 1 0 1 1
3 4 16 4 2 4 1
1 1 0 0 1 1 17 0
1 1 0 0 1 1 17 0 21
1 1 0 3 1 1 15 0 18 18
1 0 4 3 2 3 1 4 0 0
18 19 5 3 14 14 1 7 1 1
0 0 6 2 0 0 1 9 2 1
1 1 0 0 1 1 16 0 21 20
1 1 0 0 1 1 17 0 20 21
0 1 1 2 0 0 1 1 0 0
22 21 2 0 18 16 1 3 1 1
0 1 6 3 0 0 0 7 2 1
0 0 0 6 0 0 0 0 0 0
1 0 i : 16 1 1 0 1 0 0 2
GsCA,CoCA,LnCA,mng,HTN,
0
1 0
0 4 1
19 0 1 2
18 0 1 0 19
1 3 1 8 1 0
1 1 18 0 1 1 0
0 0 2 13 2 0 5 0
1 8 0 2 0 0 2 0 0
7 0 1 0 0 1 1 1 11
80


Appendix C- Guide for Semi-structured Interviewing
1) Last time you saw a doctor, how would you rate the
medical attention you received?
2) Is there anyone besides a doctor that you and your
family see when you're sick?
3) Have you.ever not gone to a doctor when you thought
that you or your family should? Why?
4) Do the doctors you've usually seen explain enough
about a sickness to you? Do you ever remember them
asking you where you think it came from?
5) Was it easy for you to get to the doctor the last time
you went?
6) Is there anything people can do to prevent serious
illness?
7) How are (MDS TERMS IN SAME PILE) similar?
8) How did you decide what else should go in this pile?
[CANCERS] [CARDIO] [BEHAVIOR] [INFECT] [OTHER]
9) What kind of TV do you watch? How about your kids?
(Span / Eng)
10) Are the doctors you see now similar to the doctors
your family used when you were a child?
81


Appendix D- Labels for MDS Mapping of Terms
Illness Term MDS Label
Heart Attack HrAtk
Coronary Artery Disease CAD
Diabetes diab
Emphysema emp
Aneurism anu
Stroke CVA
Leukemia luk
Kidney Failure RF
Stomach Cancer GsCA
Colon Cancer CoCA
Lung Cancer LnCA
Meningitis mng
High Blood Pressure HTN
Hepatitis hep
Liver Cancer LiCA
Breast Cancer BrCA
AIDS (SIDA) AIDS
Myocardial Infarction MI
Alcoholism ale
Influenza flu
Pneumonia pnu
82


REFERENCES
Acosta, A., Hamel, V.
1995 C.S.A.P. Implementation Guide: Hispanic/Latino Natural
Support Systems. United States Department of Health and Human
Services Publication No (SMA) 95-3033. Washington, DC: U.S.
Government Printing Office.
Baer, Hans
1989 The American Dominative Medical System as a Reflection of
Social Relations in the Larger Society. Social Science and
Medicine. 28 (11) :1103-12.
Bernard, H. Russell
1988 Research Methods in Cultural Anthropology. Newbury Park,
CA: Sage.
Bourdieu, Pierre
1977 Reproduction in Education, Society, and Culture. London:
Sage Press.
Borgatti, Stephen
1992 Anthropac (computer program). Columbia, SC: Analytic
Technologies.
Casson, Ronald
1983 Schemata in Cognitive Anthropology. Annual Review of
Anthropology 12:429-62.
Chen, A.
1991 Non-compliance in Community Psychiatry: Review of Clinical
Interventions. Hospital and Community Psychiatry 42:282-287.
Crimm, Allan and Greenberg, Raymond
1981 Reflections on the Doctor-Patient Relationship:
Paternalism and Autonomy. In Ethical Dimensions of
Clinical Medicine. Dyer and Robbins eds. Pp 104-111.
Springfield IL: Charles C Thomas.
Csordas, A., Kleinman, Arthur
1990 The Therapeutic Process. In Medical Anthropology: Handbook
of Theory and Method. Johnson and Sargent, eds. Pp 11-42. NY:
Greenwood Press.
D'Andrade, Roy
1979 Culture and Cognitive Science. Memorandum to Cognitive
Science Committee, Sloan Foundation.
Dietrich, K.
1989 Ethnic Differences in the Demand for Physician and
Hospital Utilization Among Older Adults in Major American
Cities: Conspicuous Evidence of Considerable Inequalities. The
Milbank Quarterly 67 (3-4) :412-50.
83


Elder, J., Mayer, J., Campbell, N., Castro, F., Harmon, M., Coe, K.
1994 Breast Self Examination: Knowledge and Practices of
Hispanic Women in Two Southwestern Metropolitan Areas. Journal
of Community Health 19(G):433-449.
Foster, George McClelland
1994 Hippocrates' Latin American Legacy: Humoral Medicine in
the New World. Langhorne, PA: Gordon and Breach.
Geertz, Clifford
1973 The Interpretation of Cultures. NY: Basic Books.
Ginzberg, E.
1991 Access to Health Care for Hispanics. The Journal of the
American Medical Association 265 (2) :238-242 .
Goodenough, Ward
1970 General and Particular. In Description and Comparison in
Cultural Anthropology. Pp 98-130. Chicago: Aldine Publishing.
Guarnaccia, P., Angel, R.
1989 Mind, Body, and Culture: Somaticization Among Hispanics.
Social Science and Medicine 28 (12) :1229-1239.
Harris, Marvin
1979 Cultural Materialism: The Struggle for a Science of
Culture. NY: Random House.
1988 Theoretical Principles of Cultural Materialism. In High
Points in Anthropology. Bohannon and Glazer, eds. Pp 379-403.
NY: McGraw-Hill.
Jordan, Brigitte
1983 Birth in Four Cultures. Montreal: Eden Press.
Kay, Margarita
1977 Health and Illness in a Mexican Barrio. In Ethnic
Medicine in the Southwest. E.H. Spicer, ed. Pp 97-167. Tucson:
University of Arizona Press.
Keefe, S., Padilla, A.
1979 The Mexican-American Family as an Emotional Support
System. Human Organization 38:144-152.
Kleinman, Arthur
1981 Patients and Healers in the Context of Culture: An
Exploration of the Borderland Between Anthropology, Medicine,
and Psychiatry. Berkeley, CA: University of California Press.
1988 The Illness Narratives: Suffering, Healing, and the Human
Condition. NY: Basic Books.
Kleinman, Arthur, Good, Bryon, Guarnaccia, Peter
1990 A Critical Review of Epidemiological Studies of Puerto
Rican Mental Health. American Journal of Psychiatry
147 (11) :1449-57.
84


Kleinman, Arthur, and Margaret Lock
1996 The analysis of social suffering. Daedalus. vl25 nl pXI-
XXI.
Lantz, Paula, Dupuis, Lawrence, Reding, Douglas, Krauska, Michelle,
and Lappe, Karen
1994 Peer discussions of cancer among Hispanic migrant farm
workers. Public Health Reports 109(4):512-521.
Leacock, M.
1982 Marxism and Anthropology. In The Left Academy: Scholarship
on American Campuses. NY: McGraw-Hill.
Marx, Karl
1970 [1859] A Contribution of the Critique of Political-
Economy. NY: International Publishers.
McManus, M., Newachek, P., Lieu, T.
1993 Race, Ethnicity, and Access to Ambulatory Care Among US
Adolescents. The American Journal of Public Health 83(7):960-6.
Perez-Stable, E., Sabogal, F., Hiatt, S., McPhee, S.
1992 Misconceptions About Cancer Among Latinos and Anglos. The
Journal of the American Medical Association 268(22) :3219-3224.
Pettitt, L., Cumberland, W., Fielding, J.
1994 Immunization Status of Children of Employees in a Large
Corporation. The Journal of the American Medical Associationv
271 (7) :525-31.
Randall, Terry
1991 Key to Organ Donation May Be Cultural Awareness. The
Journal of the American Medical Association 265(2) :176-177.
Rankin-Hill, L., and M. Bates
1994 Control, Culture, and Chronic Pain. Social Science and
Medicine 39 (5) : 629-46.
Ray, Laura, F. Trevino, and A. Estrada
1990 Health Care Utilization Barriers Among Mexican Americans:
Evidence from HHANES 1982-84. The American Journal of Public
Health 80 (12) -.21-22 .
Ray, Laura, F. Trevino, and J. Higginbotham
1990 Utilization of Curanderos by Mexican Americans: Prevalance
and Predictors. The American Journal of Public Health
80 (12) :32-36.
Romney, A. Kimball, Susan Weller, and William Batchelder
1986 Culture as Consensus: A Theory of Culture and Informant
Accuracy. American Anthropologist 88:313-338.
Saunders, L.
1954 Cultural Differences and Medical Care: The Case of the
Spanish-speaking People of the Southwest. New York: Russell
Sage.
85


Scheper-Hughes, Nancy
1983 Curanderismo in Taos County, New Mexico- A Possible Case
of Anthropological Romanticism? Western Journal of Medicine.
December 139 (6) :875-884.
Schlesinger, H., B. Burns, C. Patrick, and D. Padgett
1994 Ethnicity and the Use of Outpatient Mental Health Services
in a National Insured Population. The American Journal of
Public Health 84(2) -.222-221.
Shelton, D., M. Garcia, G. Marks,and J. Solis
1990 Acculturation, Access to Care, and Use of Preventative
Services by Hispanics: Findings from HHANES 1982-84. The
American Journal of Public Health 80(12):11-20.
Seidel, Henry, Jane Ball, Joyce Dains, and William Benedict
1995 Mosby's Guide to Physical Examination. St. Louis, MO:
Mosby.
Singer, Merrill
1989 The Coming of Age of Medical Anthropology. Social Science
and Medicine 28(11):1103-12.
1994 Community-centered Praxis: Toward an Alternative Non-
dominative Applied Anthropology. Human Organization 53 (4) :336-
45..
Singer, Merrill, Hans Baer, and Ellen Lazarus
1990 Reinventing Medical Anthropology: Toward a Critical
Realignment. Social Science and Medicine 30(2):179-188.
Smith, Harmon L.
1981 A Valid Consent: Informed, Voluntary, Competent. In
Ethical Dimensions of Clinical Medicine. Pp 70-81.
Springfield, IL: Charles C Thomas.
Snowden, L. and F. Cheung
1990 Community Mental Health and Ethnic Minority Populations.
Community Mental Health Journal 26(3) :277-293.
Solis, J., M. Garcia, and G. Marks
1990 Health Risk Behaviors of Hispanics in the US: Findings
from HHANES 1982-84. The American Journal of Public Health
80 (12) :20-27.
Spradley, James
1972 Foundations of Cultural Knowledge. In Culture and
Cognition: Rules, Maps, and Plans. J. Spradley, ed. Pp 3-41.
London: Chandler Publishing Company.
Stroup-Benham, C., R. Valdez, and E. Moyer
1991 Health Insurance Coverage and Utilization of Health
Services by Mexican Americans, Mainland Puerto Ricans, and
Cuban Americans. The Journal of the American Medical
Association 265 (2) :233-238.
86


Taussig, I.M.
1987 Comparative Responses of Mexican-Americans and Anglo-
Americans to Early Goal-Setting in a Public Mental Health
Clinic. Journal of Counseling Psychology 34:214-217.
Valle, R., and W. Vega
1982 Hispanic Natural Support Systems: Mental Health
Perspectives. Sacramento, CA: State of California Department of
Mental Health.
Vega, W.
1989 What is a Culturally-Sensitive and -Specific Prevention
Program for Hispanic High-Risk Youth? Rockville, MD: U.S.
Department of Health and Human Services, Office of Substance
Abuse Prevention.
Weller, Susan; Romney, A. Kimball
1988 Systematic Data Collection. Beverly Hills, CA: Sage.
Winn, Deborah
1991 HIV Transmission and General Knowledge of HIV and AIDS.
The American Journal of Public Health 81(12):1591-1596.
Young, Zonana, Schepler
1986 Medical Non-compliance in Schizophrenia: Codification and
Update. Bulletin of the American Academy of Psychiatry and the
Law 14:105-122.
87


Full Text

PAGE 1

EXPLORING CULTURAL PERCEPTIONS OF ILLNESS IN LAY AND EXPERT SOCIAL GROUPS by Jeremy David Graham B.A., University of Colorado at Denver, 1993 A thesis submitted to the University of Colorado at Denver in partial fulfillment of the requirements for the degree of Master of Arts Anthropology 1996

PAGE 2

This thesis for the Master of Arts degree by Jeremy David Graham has been approved by Duane Quiatt \ Brett Date

PAGE 3

Graham, Jeremy David (M.A., Anthropology) Exploring Cultural Perceptions of Illness in Lay and Expert Social Groups Thesis directed by Assistant Professor Kitty K. Corbett. ABSTRACT This thesis explores the problematic underutilization of biomedicine by United States latinos. A model of culture is described, under which the relative merits of a cultural orientation versus a political-economic orientation are explored. Data were collected from registered nurses, lay anglo women, and lay hispanic women in a metropolitan area. The data suggest that both lay groups differ from the nurses in their level of biomedical understanding, but is inconclusive regarding the presence of culturally-different explanatory models. The implications for biomedical relationships of unshared meanings and devalued cultural capital are discussed. This Abstract accurately represents the content of the candidate's thesis. I recommend its publication. Signed Kitty iii

PAGE 4

ACKNOWLEDGEMENT I extend thanks to the individuals who generously agreed to be interviewed for this exploration, and to the community members who provided introductions. I am deeply indebted to my advisors for their patience,. their considerable editing, and especially for encouraging me to explore the project freely. Valuable statistical advice was provided by Dr. Tammy Stone. I thank especially G.L.G., L.L.G., A.B.D., C.M.G, E.M.M., and my family and friends without whom this thesis could not have been completed.

PAGE 5

CONTENTS CHAPTER 1 INTRODUCTION .................................. 1 An Overview of Culture ...................... 1 The Culturological Orientation .............. 8 The Political-Economic Orientation ......... 16 Issues: Underutilization of Biomedicine .... 21 Issues: Unshared Meanings .................. 26 2. METHODS AND SAMPLES USED ..................... 31 The Samples ................................ 32 Registered Nurses ..................... 34 The Latina Sample ..................... 37 The Anglo Sample ...................... 42 Methods Utilized ........................... 45 Multi-Dimensional Scaling ............. 45 Disease Severity Surveys .............. 47 Semi-Structured Interviews ............ 48 v

PAGE 6

3. FINDINGS AND RESULTS ......................... 50 Multi-Dimensional Scaling .................. 50 MDS Mapping: The Nurses ............... 50 MDS Mapping: The Latino Women ......... 54 MDS Mapping: The Anglos ............... 57 Comparative MDS Appraisal ............. 60 Survey Findings ............................ 63 Technical Term Assessment Findings .... 65 Semi-Structured Interview Findings ......... 66 4. DISCUSSION AND DIRECTION ..................... 70 The Merits of A Cultural Orientation ....... 70 The Merits of A Political-Economic Orientation ................................ 73 APPENDIX A. EXAMPLE OF SURVEY INSTRUMENT ............ 76 B. AN EXAMPLE CORRELATION MATRIX FILE FOR MULTI-DIMENSIONAL SCALING ........... 80 C. GUIDE FOR SEMISTRUCTURED INTERVIEWING ... 81 D. LABELS FOR MDS MAPPING OF TERMS ......... 82 REFERENCES ...................................... 83 vi

PAGE 7

CHAPTER 1 INTRODUCTION. In the problematic utilization of health care services by some Latino populations in the United States, what are the relative merit.s of a cultural relevance approach versus a political economic approach? This thesis explores some indices of whether cultural background and socioeconomic status impact lay people's understandings of some biomedical illness terms. An Overview of Culture What is meant in this paper by culture, and by cultural difference? For the purposes of this exploratory work, culture is conceived as the belief system that governs how individuals make sense of and cognize events and objects in the world of experience. This aspect of culture involves what Spradley described as "social rules," which are "instructions for constructing, combining, interpreting, and otherwise 1

PAGE 8

dealing with symbols11 (1972:29). Symbols are the mental labels applied to objects in the world. Assemblies of cultural rules form cognitive maps and models, which govern the ways that an individual assesses reality. Spradley's focus emphasizes the ways in which these rules are codified in language structure, but the concept is equally native _to the areas of systems programming and psychology,. Casson uses the psychological term 11Schema,11 to describe these cognitive knowledge systems. Schemata, he writes, are conceptual abstractions that mediate between stimuli received by the sense organs and behavioral responses ... Schemata are autonomous and automatic-once set in motion they proceed to their conclusion-and they are generally unconscious, nonpurposive, and irreflexive (1983:430). Where schemata are individuals' mental structures for cognizing the world, some aspects of schemata are idiosyncratic, and other aspects are common to all humans. Some aspects of schemata are common to members of some groups but not others: Schemata differ in their distribution in populations ... cultural schemata are neither unique to individuals nor shared by all humans, but rather shared by members of particular societies (1983:440). 2

PAGE 9

Members of given social groups share many of the same cognitive schema and cultural ideation because they share similar experiences and life trajectories. When people live much the same, they experience much of the same stimuli in the formation of their cognitive schemata. The concept is emphasized in Marx's assertion that the material existence of a human shapes his consciousness, making class and status the major underlying cause of cultural difference in a mass society. Focusing on this "cognitive aspect of culture," Ward Goodenough writes, People who deal recurringly and frequently with one another develop expectations regarding the manner of conducting these dealings. They make some of their expectations explicit and formulate some of them as rules of conduct. The do not consciously formulate others but react to a person's failure to abide by them as a breach of appropriate behavior, saying that he behaves or talks oddly or in a mixed-up fashion (1970:98). Ideation orders social behavior by providing a set of automatic expectations regarding how people interact with each other and the world. Not everyone in a social group steers by the exact same cognitive map, but as Goodenough continues, [T]he variance in their individual expectations must be small enough so that they are able to accomplish their purposes with and through one another 3

PAGE 10

reasonably well most of the time ... In this respect a people's culture is like a people's language (1972: 100) Cognitive maps include taxonomies with which people organize their world are built to make the world predictable, so that human individuals can synchronize their interactions with each other, and with the. needs of life. The knowledge structure within the individual is the component of culture that is explored by most methodology, from unstructured interviewing to cognitive testing. However, the schema takes shape in the world through social interaction. The determinism of knowledge structures has been challenged and redefined in the last two decades, based on anthropology which views culture as negotiated and transitional. Geertz (1973) discussed culture as a "web of meanings" within which any one fact or experience draws its meaning from context. Culture becomes a text to be read and interpreted. But where literal text may inspire human action only indirectly, culture gives human ideation immediate physical reality, through behavior and activity. However, Geertz's sense of shared meanings has value in understanding how social ideas constrain human relationships. 4

PAGE 11

Bordieu's notion of habitus recodifies "webs of meaning" in a manner useful to exploring cultural differences between social groups. Habitus is: A system of lasting, transposable dispositions which, integrating past experiences, functions at every moment as a matrix of perceptions ... and actions ... [it is] a subjective but not individual system of internalized structures, schemes of conception, and. action common to all members of the same group or class (1977:85). The habitus concept integrates the notions of culture as traditional knowledge, shared meanings, and dynamic mechanisms for living. It is the environment within which the rise and fall of cultural ideas take place, and is the currency by which experiences and new ideas are evaluated and appraised. Moreover, Bourdieu's habitus is conceived not only in terms of different societies, but of the shared life chances" Weber described for groups within a society divided by social class, party, and status. Habitus.offers a materialist theory of culture, which credits cultural ideas with material outcomes. The material conditions determine consciousness, but consciousness impacts material conditions. The shared life chances of common-group members shape their shared meanings, and their shared meanings reproduce and enforce their common life chances. 5

PAGE 12

Bourdieu's notion of "cultural capital" is a useful tool in understanding how habitus impacts class relations in biomedicine. Built on scientific discourse and the ideal of objectivity, biomedicine devalues the explanatory models of lay people, whether they are based on distinct folk models or on vague shadows of biomedicine. Cultural capital relates to social class reproduction--how the rich beget rich and the poor beget poor--as a series of processes, which contribute to overall denial of most social mobility. Cultural capital, as delineated by McLeod, involves the general cultural background, "knowledge," disposition, and skills passed on from one generation to the next (1987:12). It incorporates common use of sentence structure and vocal accent, dress, and posture, as well as familiarity with museums, arts, sports, and other canonized social activities. Bourdieu suggests that children of higher socioeconomic classes inherit different "capital" than do children of lower classes. Authoritative institutions reward a performance specific to a valorized, upperclass, set of social technologies. Schools therefore reward those who enter with more prized cultural capital with higher academic credentials, which then serve as the credentials which grant one access to higher-class 6

PAGE 13

prosperity in the future. The strata of academic achievement are the strata of social class, covered over by a veneer of scholastic achievement. Similar processes occur in biomedical relationships. Non-biomedical explanatory models are actively (if quietly) devalued in clinical relationships. Like an upper class setting the value of different sets of cultural capital in schools, the biomedical establishment serves as an upper class determining and rewarding only select cultural capital in clinics. Biomedical relationships are hindered by (and perhaps exploit) precisely the same processes of cultural capital and devaluation that act in education. Biomedicine may in fact play a covert role in the reproduction of social class through this mechanism. Valued cultural capital is turned into greater power to access health resources (including simple awareness of health services), which directly and indirectly allow greater control of material and economic resources. This control of material aspects of life translates back into access to higher social class. The role of the biomedical model as a social control mechanism has yet to be fully explored. The problem of non-participation in biomedical relationships hangs between economic inaccessibility on 7

PAGE 14

one end, and a lack of shared meanings on the other, a question of how and whether culture and class impact utilization. Do United States latinos lack shared meanings with biomedicine by virtue of their different cultural traditions, or simply by lack of access to cultural capital, the authoritative knowledge valued by the biomedical_enterprise? The former notion suggests that an exploration should focus on cultural understanding and appropriateness, while the latter suggests that the problem be viewed as an aspect of class marginalization. The Culturological Orientation A cultural appropriateness approach favors the reconciliation of conflicting "explanatory models" as a key to understanding some group differences in health care utilization. Incongruous explanatory models can impede shared understandings, and thereby limit the patient-caregiver relationship. The basic idea of a cultural approach can be stripped of some implications and reduced to a central theme: that patients go to health resources that are consistent with their belief systems about the way the world works; by implication, 8

PAGE 15

they tend to avoid using care resources that deny their worldview and explanatory models. Explanatory models are defined by Kleinman as "the notions about an episode of sickness and its treatment that are employed by all those engaged in the clinical process." The relationship between the explanatory models held by_patients and caregivers is a "central component of health care11 (1981:105). Kleinman separates the idea of pathological disease as a somatic event from the semiotic progression and relationships within the cultural construction of an illness as a personal and social experience. Where the ethical landscape appears differently to different parties involved in a clinical relationship, communication depends on recognition of meanings of actions for all involved (1988) Kleinman (1988) defines five major areas where explanatory models seek to make sense of an illness event: 1) etiology 2) how symptoms appear 3) pathophysiology 4) predicting course and outcome of sickness 5) defining sensible treatment Of these, agreement on the likely outcome of an illness, 9

PAGE 16

and feeling comfortable with the treatment regimen, are major elements in cultural compatibility. Phrasing the patient-caregiver exchange in terms reasonable to the patient's explanatory model can be crucial. The formal organization of these elements is more precise within the biomedical model than for most lay persons, and lay people's may not attend to all the categories: Lay explanatory models disclose the significance of a given health problem for the patient and his family, along with their treatment goals ... Vagueness, multiplicity of meanings, frequent changes, and lack of sharp boundaries between ideas and experiences are characteristic of lay explanatory models (1981:106-7). Within the cultural approach, the patient-caregiver relationship is a central component of health care, because people go to healers because of the illness that they experience, not the pathological disease behind it. Using the term 11symbolic healing11 can have 11the unfortunate connotation that there are other forms of healing which are not symbolic11 (Csordas and Kleinman 1990:11). Some level of sharing of meanings is necessary to fulfill the need for the therapeutic relationship. Explanatory models are not the same as a person's general views and beliefs about health and illness. As Kleinman defines the term, an explanatory model is a set of ideas and expectations related to a particular illness 10

PAGE 17

event. However, the information to build an explanatory model comes largely from the backgrounded, general set of beliefs an individual uses to interpret such events: culture, varying across ethnicity and class. Since cultural schemata give rise to explanatory models, people of different cultural backgrounds can experience diseases in different ways, and assign them different meanings depending partly on social group membership. Crimm and Greenberg (1981) have suggested that biomedicine must increasingly temper paternalistic beneficence with a cross-cultural vision of autonomy. Guarnaccia and Angel (1989) have examined somaticization among United States latinos, suggesting that latinos may translate psychological stresses into physical pain and symptoms in ways that differ from anglo patterns. Lock and Kleinman (1996) suggest that some patterns of pain experience may relate more to social expression than to pragmatic biomedical problems or psychological processes. Rankin-Hill and Bates (1994) related patients' beliefs regarding control of pain and illness to ethnic background and to forms of expression of chronic pain. The pain experiences revealed were different for groups in Puerto Rico from groups in New England. The authors recommend the development of culturally appropriate and .11

PAGE 18

relevant programs which help patients feel in control of their illness experience. Kleinman, Good and Guarnaccia (1990) reviewed literature on the mental health status of latinos in the United States. Among many latino subgroups (especially Puerto Ricans), the ailments nervios and ataques de nervios are poorly understood. Primarily psychosomatic symptoms are expressed mostly as culturally-appropriate times, and occurrences more upon levels of acculturation than upon clinical symptoms. The reviewers suggest that the benefit of biomedicine to these patients is limited and resisted because current methodologies do not attend to cultural factors. Latinos in the United States are a large category containing a great deal of variation, but some workers have attempted to describe some major unifying cultural traits useful to shaping biomedical services directed at latinos. Acosta and Hamel (1995) provide a simplified index of cultural traits associated with latino identity. Among the central characteristics they propose are: 1) Collectivismo-a sense of the community as the common overarching good; willingness to forego individual gain to maintain a role in the community. 2) Orgullo-a strong sense of local pride, specific to both ethic identity and the immediate community. 12

PAGE 19

3) Espiritualismo-commonly-held concept of life issues (personal success, illness) as representations of spiritual forces. 4) Machismo-culturally 110fficial11 male authority, value of family males as providers. 5) Marianismo-ideal value surrounding women's roles, favoring submissiveness, nurturing, devotion to marital and extended familial networks, and children. Acosta and Hamel see major cultural features as social terrain to which biomedical services must be made to fit. They propose their generalizations as guidelines to locating 11Natural Support Systems11 within ethnic communities, particularly in the areas of community mental health and disease prevention. Their predictions are highly generalized and do not reveal the workings of any particular latino individual or subgroup, but do point out some potentially important community values. Valle and Vega (1982) identified similar traits as parts of social support networks utilized in latino communities. Other cultural-appropriateness approaches in biomedicine have used similar typologies. The primary text used to train medical students in physical examination illustrate the typological mindset of cultural romanticism. The text states that: 13

PAGE 20

To assume homogeneity in the beliefs, attitudes, and behaviors ... is to court error and to miss the mark in the effort to understand the individual. The stereotype, a fixed image of any group that rejects the potential of originality or individuality within the group, is itself to be rejected. People can and do respond differently to the same stimuli (Seidel et al 1995:34). The value of this passage is belied by the reductionism of the following text; Seidel et al go on to present a table of cultural traits associated with various ethnic backgrounds. Latinos are described as people who 11have little regard for the past11 among whom 11males are considered big and strong.11 They live with a 11relaxed concept of time--may be late for appointments11 (1995:39-49) Boulette (1978) has warned against confidence in stereotypical cultural determinism, especially regarding the values of destino, machismo, and espiritualismo, which fails to understand the agency of the individual. Typologies can themselves impede cultural understanding. A cultural relevance approach, she suggests, may often be better served by a negotiated model of exchange between and patient, than by a typological model. Typologies are also problematic because some latino cultural traditions have faded in recent decades. In some subgroups of Untied States latinos, traditional folk-models of illness have less impact on people's ideas 14

PAGE 21

than they have had in the past. Saunders (1954) focused on the creation of guidelines for field health personnel, suggesting that ideas of personal and spiritual fate (destino) made "faith and fatalism ... the first ingredients in Spanish-American folk medicine" (1954:152). Scheper-Hughes and Stewart (1983) measured a marked decline_in the importance of curanderismo in rural New Mexico, compared to older ethnographies, and describe a cultural landscape shaped by mainstream biomedical values: We would have to conclude that curanderismo in Northern.New Mexico today has moved from being a primary and important source of medical care to an alternative and very occasional source in cases of chronic pain and illness or for those puzzling psychosomatic and stress-related afflictions for which biomedicine has little respect and virtually no cure and therefore for which it has none of the creative names and faces that the folk medical system readily supplies (1983:884). Ray, Trevino, and Higginbotham (1990) also report that use of traditional folk medicine (curanderismo) by United States latinos relates to dissatisfaction with biomedicine (they found also that curandero use did not relate to access to biomedical care or to income) George Foster has suggested that although traditional humoral folk medicine occurs among United States latinos, humoral healing practices are often empty 15

PAGE 22

cultural artifacts, "shorn of theory" and lacking a generally accepted humoral theory of the body (1994) Looking at latinos even outside the United Stats, Weller has described a decline of the humoral model in Guatemala, concluding that: [A] simple rule may be neither possible nor desirable. Service providers need to be aware that [a folk model] may exist, but that if a particular patient/client does not subscribe to it, that actual interpretation can be highly idiosyncratic and need not interfere with the delivery of health care (1983: 256). The Political-Economic Orientation The political-economic concept of health care underutilization emphasizes the relative inaccessibility of biomedical understanding to poor and marginalized. This emphasis focuses not on the conflict of meanings in the biomedical relationship, but upon lack of control of information. Cultural difference may at times cover over the political-economic issues in biomedicine: A focus on 11culture11-hence on traditions, values, and attitudes--rather than on class relations can be and is widely used to mask and distort the brutal realities of power and exploitation (Leacock 1982: 257) From the political economy perspective, latina underutilization of health care is a case of 16

PAGE 23

marginalization by a dominant social class, whose meanings (the biomedical model) are outside the economic grasp of marginalized groups in general. Baer (1989) has described United States biomedicine as a "dominative" institution which mirrors class relationships in ways that cover over class exclusion; such a system falsely places the responsibility for utilizing available resources with the marginalized patient. A critical social science in biomedicine [C]annot take biomedicine at face value as the scientific medical system, as some are inclined to do, but neither can it stop with merely declaring biomedicine to be yet another ethnomedicine, as others would have it (Singer et al. 1990:183). Rather, individual biomedical relationships and the health care utilization practices of social groups result from underlying class-stratified control of biomedical knowledge and resources. In building a methodological perspective, Singer suggests that the empowerment of communities and social classes is crucial to their members' individual power in biomedical realtionships (1994). In practice, Singer's concept translates into a 11community-centered praxis," an applied anthropology agenda which advocates communities' rights to shape their own relations with larger social systems (the individual patient's biomedical relationships are antecedent to the class position of his 17

PAGE 24

community, and are not addressed directly by the political-economic orientation) Biomedicine is the dominant model of health and illness in the United States. It involves a great deal of specialized knowledge and interpretive frameworks which are understood as objective truths by members of biomedical institutions. Full competence in the biomedical model is reserved for the few individuals who are charged with its use. The biomedical model within the mainstream of United States populations is similarly regarded as privy to objective truths about how illnesses will proceed and be reconciled, and how pathologies relate to the health of the individual. Mainstream anglo ideation about body praxis posits bioscience as the arbiter of truth, and generally validates biomedical knowledge; they generally rely on_ the biomedical model, but have less-expert understanding of the model than biomedical professionals. Coherent belief in a non-biomedical model is the exception in mainstream anglo groups. A political economy of medicine suggests that the degree of agreement between the explanatory models of caregivers and of patients (and their families) varies by degree of economic power the patient holds. Individuals from relatively higher social classes have more access to 18

PAGE 25

legitimized cultural capital, and are able to trade more powerfully in the currency of dominant knowledge. In the case of biomedicine, simply being in control of dominant cultural capital does not bring biomedical understanding. Biomedicine's highly specialized form of knowledge places all lay people in one category; neither poor latinas nor anglo businessmen are generally privy to biomedical training. The difference is that the latter participates from the position of more highly-valued cultural capital. Lay people in general have a limited inventory of biomedical understandings, but poorer lay people have less access to valued cultural capital, with which to gain more understanding or to organize further information into agreement with the dominant model. By this view, the issue of latina underutilization stems primarily from their marginalization from the dominant explanatory model of health care. Where their models do not reflect biomedical understandings, the situation is not necessarily to be interpreted as the presence of an alternative explanatory model, but as a lack of access to (and control of) the dominant discourse. Studies have indicated relatively low levels of biomedical knowledge in latina populations. Elder et al. (1994) examined the use of Breast-Self-Examination (BSE) 19

PAGE 26

among latinas in two southwestern cities. Even in populations where a high percentage (approximately 80%) of latinas had been taught BSE by health professionals, and a high number (62-63%) reported performing a BSE in the past month, only 0.7 percent were found adequately proficient at BSE. English language ability and acculturation were the most significant predictors, suggesting that control of dominant cultural capital affords a better grasp of biomedical knowledge to lay people. Lantz et al. conducted focus groups among migrant farm workers in the American Midwest, finding that lack of knowledge and information regarding the causes of cancer, its prevention, and its early detection and treatment was evident among participants (1994:512). Lantz et al. ascribe latino fatalism regarding cancer to the advanced disease states with which they frequently seek medical assistance. Perez-Stable et al. (1992) compared latino impressions and misconceptions regarding cancer to those of angles. Latinos were more likely to have a "fatalistic" attitude regarding cancer, and to have less knowledge than their anglo counterparts of the risks and symptoms of cancers. All subjects were members of the same pre-paid health plan. Latinos ascribed some 20

PAGE 27

individual cases of cancer to a form of punishment by God. Latinos also suspected a number of worldly phenomena as of cancer, including microwave appliances, breast feeding, spicy foods, and pork. Similarly limited understandings of infectious disease have been identified, especially regarding AIDS and HIV transmission. Winn (1991) asserts that minorites are less able. to gather information regarding AIDS and are also more susceptible to myth and misinformation. Individuals who were over 50 years of age, or did not complete high school held the most misconcepti.ons regarding HIV infection. Issues: Underutilization of Biomedicine Latinos in the United States do not utilize health care resources at the same rates as the mainstream anglo population. It is unclear from current literature the degree to which underutilization is due to ethnic identity, as opposed to the low socioeconomic status that characterizes many United States latinos. Pettitt et al. (1994) examined the children of employees of a large corporation, and found that children were most likely to have received all of their immunizations if their parents had higher incomes or were more educated. Parents with 21

PAGE 28

the most comprehensive insurance options were also more likely to have their children fully immunized. Latino (and African-American) children were the least likely to be fully immunized at all levels of parental insurance or pay level. Intracultural variability within the category "latino" is great, and not all subsets of latinos show the same degree of utilization (Aday 1984) The degree of utilization of latino subgroups by both ethnic assignment (e.g. Puerto Ricans utilize services more than ethnic Mexicans), and by socioeconomic status. Overall, latinos in the United States have relatively low averadge income and educational levels. The median latino family income in 1988 was $21,800, compared to $33,900 for angles (Ginzberg 1991) Ray et al. (1990) suggest that "cultural barriers" beyond low socioeconomic status act as contributing factors to the underutilization of medical services by Mexican Americans. However, access factors play the primary role in their analysis (culturally, Ray et al recommend reduction of waiting time in clinics and provision of transportation as responses to the problem) Underutilization is compounded by disproportionately prevalent risk factors in some latino communities. A review by McManus et al. (1993) showed that heart 22

PAGE 29

disease, cancer, and perinatal health problems are more likely to kill latinos than anglos. Minority children in general were more frequently exposed to heavy metals and other environmental toxins, and more frequently suffer from asthma than their anglo counterparts. Individuals who smoke, drink, and have a poor diet are the least likely to use health care services, with Mexican-American and Puerto Rican males the most likely of all latino groups to consume large quantities of.alcohol (Solis 1990). Among elderly people in urban areas, latinos utilize physicians and hospitals at significantly lower rates than angles. At the same time, urban latinos are at higher risk for diabetes, hypertension, and several other major diseases than are their anglo counterparts (Dietrich et al 1989). Taken together, the underutilization of inpatient care by aging urban latinos is substantial. A major factor in latino underutilization is lack of regular primary care insurance coverage. Roberts and Lee (1980) found that those latinos lacking regular primary care coverage were also the least likely to utilize other more specialized medical care. This is especially true of those latinos at the lowest levels of socioeconomic status (Aday 1980) Current Population Survey (CPS) data showed that 10 percent of angles were without insurance 23

PAGE 30

in the mid-1980s, compared with 32 percent of latinos. Within the latino category, Mexican-Americans were the least insured subgroup, with 37% uninsured. Those without insurance were the least likely to see a physician; nearly 40 percent of Mexican-Americans lacked a regular primary doctor. Stroup-Benham (1991) reasons that a high number of United States latinos .live with undiagnosed and untreated medical problems. Schlesinger et al. (1994) reviewed 1.2 million insurance claims, examining two different insurance packages, of differing degrees of coverage. Latinos used the services significantly less than angles. Among women, greater education was related to more frequent use of services. The number of latinos lacking primary care coverage increased in the last decade at nearly three times the rate of anglos, and their lack of coverage is expected remain high into the next century (Andersen 1986, Trevino 1983). Where lack of primary coverage increasingly precludes use of many other medical services, low coverage is a significant factor in overall patterns of latino health care underutilization. Taussig (1987) analyzed appointment records to count the number of broken, kept, and canceled appointments for therapy in public mental health clinics. A number of 24

PAGE 31

indices, such as primary home language and ethnic selfidentification, were built into scores for acculturation, to define Mexican-American and Anglo-American samples. Taussig found no significant difference between appointment'keeping and breaking between the two ethnic groups. However, socioeconomic status related significantly to all broken appointments, regardless of ethnic group assignment. Snowden and Cheung (1990) describe United States latinos as having among the lowest rates of use of psychiatric resources. Young et al. (1986) examined factors in non-compliance with psychiatric medication regimens. Their review of 21 studies, without attention to ethnic background, found the greatest predictive force in two factors: living alone, and low socioeconomic status. Shelton et al. (1990) determined that the most important factors in latino health care utilization were logistical access to services and acculturation. Ethnic identity was a relatively weak predictor of health resource usage compared to access to services and the ability to speak English. 25

PAGE 32

Issues: Unshared Meanings Divergent explanatory models impact the patientcaregiver relationship in a number of ways. One of the most relevant contemporary issues is in the area of informed consent as a hallmark of medical ethics and legal issues. Informed consent draws its medicolegal validity from the positivist legal tradition of the rational, positing a natural law by which any fully informed person would come to her decisions. Contemporary clinical ideas of informed consent do not adequately appreciate the impact of unshared meanings. Informed consent relies on a basic premise of positivistic, rational science: information is simply information, which has meaning on its own. Such implies that the meaning of information (the knowable) is independent of the person dealing with it (the knower) Informed consent means telling patients all the raw information that a reasonable person would need in order to make a health decision, aware of the possible outcomes of each option. The assumption of an ideal person" means that any person who does not accept or interpret the information in the biomedical manner is 26

PAGE 33

less than reasonable, swayed by superstition or irrational fear, or is impeded in their clarity by psychological stressors of illness. Participation in becoming informed may be seen as compliance with, and acceptance of, the meanings assigned to information by biomedicine itself, compliance which can deny the social meanings inherent. to the persona of the patient But humans act on information by virtue of its meanings when translated into the lives of persons. Its meaning cannot be divorced from the web of habitus through which it is read. It depends upon the matrix of experience and meaning with which the information is assimilated into an individual's meaningful life possibilities. By virtue of being sick, a person might temporarily assign unexpected meanings to information; by virtue of habitus and psycho-cultural determinants, that person might also develop lifelong understandings which depart from theobjective intended meanings (which are themselves shaped thoroughly by cultural expectations and medical socialization) Supposedly objective data cannot properly inform when divorced from semiotic context. With the populations examined here, numerous difficulties emerge. One difficulty is the problem of the individual. Informed consent is about informing an individual, who 27

PAGE 34

makes his own decisions. In the event that this one person cannot make decisions for himself, biomedicine turns to a rather prescribed set of medico-legal rules for determining another individual, who has the authority to make decisions for the patient. But the individual, or a proxy, is not always a valid agent of _decision processes in all social contexts. Decision making may be shared among numerous individuals. Within latino cultural traditions, where an individual's identity is often understood in terms of family and community, familialismo and collectivismo are potentially inconsistent with the informed consent of the individual patient. An individual can appear to be the agent of decision, especially when viewed from a biomedical model built on this very assumption. Caregivers can be distracted from the needs of the actual social units of decision. Conflicts arise in the care of the patient which compromise the patient-caregiver relationship. Notwithstanding a medicolegal foundation of informed consent as a centerpiece of bioethics, a more universally valid goal of participation, consent, and clinical relationships is needed. It is useful to call this goal valid consent, rather than informed consent. The patient in a consent covenant must be knowledgeable, voluntary, 28

PAGE 35

and competent. The implications for cultural appropriateness are that being knowledgeable and competent to interpret medical information assumes acculturation to the biomedical model, which is not always the case (Smith, 1981) The matter of cultural compatibility carries significant weight in some issues, but discussion of cultural appropriateness too often centers on somehow validating the ways in which the Other sees the world. At worst, this strategy asserts 11folk11 models as equal to biomedical perspectives--a strategy doomed to fail within biomedicine. To a biomedicine rooted in scientific discourse, this is a direct assault on a sovereign claim to truth, a threat to a monopoly on valid explanatory models. This has not helped make 11culture11 palatable to biomedicine. This contentiousness need not exist, for cultural appropriateness which validates other systems of meanings, which elevates these systems to an equal position of authority on illness, misses the mark itself. The issue at hand is not that biomedicine must relinquish its authoritative stance on the meanings of illness; biomedical respect for culture need have little to do with political recognition (suspending for a moment, the normalizing force of biomedicine as discussed by 29

PAGE 36

Foucault, Scheper-Hughes, and others). The central issue is this: that people will live inside .their (nonscientific) systems of meaning, regardless of whether these systems are recognized as truths. To communicate effectively across or among explanatory systems, acknowledging and negotiating points from the nonbiomedical systems is essential, regardless of any question of whose systems more closely approximate "truth." 30

PAGE 37

CHAPTER 2 METHODS AND SAMPLES USED The data in this exploration are in three sets: pile-sort data, subjected to Multi-Dimensional Scaling and cluster analysis; survey data regarding the severity of some biomedical disease terms; and interviews about respondents' beliefs about illnesses and biomedical relationships. Each of the three data types were collected with all respondents. Samples were composed of twenty-two latinas, twenty-two female registered nurses, and twenty-two anglo women. These forms of data collection were chosen because of the willingness of the subjects to participate in this form of research; one of the goals was to determine how these techniques might be utilized in relatively rapid anthropological studies. All of these data sets were collected in one sitting with each informant. Unlike pure survey research, these techniques can be applied to only a few subjects at any one time. Most of the collection involved one informant at a time, but some were surveyed in small groups (up to four individuals) 31

PAGE 38

with the individuals interviewed one at a time during the sitting. The Samples Women church members were targeted as "purposive samples" for three main reasons. First, they are members of communities to whom others turn for health care decisions, and they are often charged with the care of children and aging family members. Second, their interest in being participants was higher than members of other organizations; leaders of the women's ministries at the churches readily provided introduction and legitimization of the inquiry. Third, relatively cohesive social groups are less variable than random samples: "when intracultural variability is high, it makes little sense to report aggregate data ... results are not only inaccurate, but misleading" (Weller, 1983 :256). The nurses were selected by some related criteria, including their availability for interview. No emphasis was placed on selecting a random sample of nurses who would represent the whole of nursing knowledge, but attention was paid to ensuring some variety of specialty, 32

PAGE 39

age, and degree of nursing among the subjects. The nurses were selected as being individuals who are routinely biomedicine's direct contact with patients and families. The de.mographics of the three groups are presented in Table 2 .1. Table 2.1 Summary of Demographics Nurses Latinas Anglos Age distribution: <20 0 0 2 20-29 9 7 9 30-39 8 7 4 40-49 4 4 7 50+ 1 4 0 Number of Birth Children: 0 11 0 7 1 6 2 3 2 4 10 7 3 1 5 3 4 0 3 2 5 0 2 0 Number in Household: 1 7 0 2 2 7 3 14 3 4 7 6 4 3 10 0 5 1 0 0 6 0 2 0 Marital Status: Mar 12 20 16 Sgl 6 0 2 Div 3 1 3 Wid 0 1 0 Coh 1 0 1 33

PAGE 40

Table 2.1 (Continued Household Income: -$15,000 0 2 7 $15-25k 0 9 3 $25-45k 11 9 4 $45,000+ 11 2 8 Educational Levels: Some H.S. 0 0 1 H.S. Grad 0 6 11 Some Coll 12 5 4 Coll Grad 7 5 5 Grad Degree 3 0 0 Vo-Tech School 0 6 1 Ability to Speak Spanish: Bilingual 0 6 1 Some Fluency 4 10 3 Low Ability 4 1 3 No Spanish 14 5 15 Type of Clinic Last Used: Private/HMO 18 13 18 Emergency Dept. 1 1 2 Walk-In Clinic 3 2 0 Public Health 0 6 2 Totals: 22 22 22 Registered Nurses The sample of 22 registered nurses was part of the inpatient care staff of a large, non-profit teaching hospital in a downtown neighborhood. All of the nurses are involved in the care of seriously ill patients; none are limited to outpatient care. 34

PAGE 41

Roles. Two are pediatric nurses, and nine are medical/surgical caregivers, of which six are regularly assigned to intensive care. Three nurses are emergency care certified, and work primarily in the emergency wards. Four more work primarily with cancer patients, two with neurological patients, and two are generalized nurses whose role is on-site training of student nurses. Three of the nurses also work in the local metropolitan public health system clinics, mostly parttime (one was full-time in the city clinics, and only part-time at the hospital) Demographics. The nurses range in age from 23 years old to 55 years old, with an average of 33 years. Their experience as nurses ranges from just under a year to over 25 years, with an average of 11 years of active patient care. Twelve nurses are married, two are single, and three are divorced; one lives in a marriage-like relationship. Half of the women have no children; the remaining eleven have an average of 1.4 children. Seven live alone, but the average household among these nurses holds 2.3 people. All of the nurses have at least some college education, and seven are college graduates. Three more 35

PAGE 42

have Master of Science in Nursing degrees. All report household incomes of at least $25,000, and eleven report household incomes of more than $45,000. Ethnic Assignment and Social Status. The nurses have a relatively high degree of access to educational resources, and _relative fiscal security. None of the nurses reported being in the lowest two categories of household income. All of them own vehicles, and generally express their ability to utilize private physicians with whom they are personally comfortable (the exception to this is those nurses who are insured through Health Maintenance Organizations, who in some cases expressed feelings that visiting an HMO physician is not the same as seeing a private physician) None had utilized a public health resource on their most recent physician visit. Within the major categories of census ethnic groups, the nurses fit into a predominantly anglo national mainstream. Twenty consider themselves culturally different from latino patients. None consider themselves fully fluent in both Spanish and English, and only four report that they are at least "somewhat comfortable" using Spanish. None speak any Spanish with their own families. Semi-structured interviews suggest that those 36

PAGE 43

nurses who report some language capability in Spanish refer to medical Spanish, useful in their careers. Compared by matched-pairs to the anglo sample set, nurses speak no more Spanish than angles (Z-score= -.1529, 2 tailed p= .8785). The Latina Sample Twenty-two latinas were interviewed over a series of Wednesday night meetings at the Westside Protestant Church, a bilingual congregation situated in a semiurban, predominantly hispanic neighborhood. The near-600 members of the congregation are an active and growing feature of the community, and have participated in community health and neighborhood revitalization projects since taking occupation of the site nine years ago (Westside was previously housed in a smaller facility in the same area) The Westside Protestant Church is fully bilingual in its offerings. Church materials are printed in both Spanish and in English. Almost all the meetings of the congregation are held in both Spanish and English. Sunday morning services are held in both Spanish and English separately, but in order to unify the 37

PAGE 44

congregation, smaller group meetings and scriptural discussions are held with both languages together, facilitated by bilingual deacons and clergy. Westside is a rapidly growing congregation, following a contemporary trend of conversion of United States latinos to Protestant denominations. Most of the congregation is composed of people who dwell in the immediate area; Westside is a neighborhood organization, and is a focal point of many processes in the community. The use of the term "latina" is chosen entirely for the convenience of the writer and reader; the respondents use the term occasionally, but more frequently call themselves "hispanic" or "Mexican" (but were never heard to use the Spanish form of the adjective, "Mexicana"). Some of the older women were heard identifying other women with the label "Spanish lady," as in, "have you met Janet? She's another Spanish lady here" (no respondents were from Spain) Various other labels for people of Spanish-speaking descent are used widely. Demographics. The women range in age from twenty to sixty-five years of age, with a mean age of 37 years. This is not significantly different from the age set of the nurses or of the anglo women. 38

PAGE 45

Five of the Westside women say they were born outside the United States-all in Mexico--but all the women were reportedly raised within the Western United States. Eleven were born to parents who themselves were U.S. born. The remaining six were all born in the U.S. as well, but reported having at least one parent born in Mexico. All of the Westside women--the entire latina sample--had been married at least once. One 55-year-old woman was widowed, and one was divorced. This high marital rate differs significantly from the marital rates of the nurses (Chi-square= 9.2, one-tailed p<.005), and from the marital rates of the anglo women (Chi-square= 1 6 6 1 p<. 01) The Westside women also more frequently had children than the nurses--all of the latinas have children. This difference is statistically significant (Chi-square= 9.23, p<.005). The higher number of births per respondent compares significantly to the other groups. Number of individuals in the household differs from that of the nurses (Z-score= -2.47, p= .0136), and from that of the anglo women (Z-score= -3.375, p= .0007). The latinas are significantly less-educated than the nurses (Z-score= -2.864, 2 tailed p= .0042). Six of the women have gone through vocational or technical 39

PAGE 46

certification programs, and tive women have graduated from a baccalaureate-level degree program. Ethnic Assignment and Social Status. The Westside community is economically depressed, and the reported household incomes of the sample group reflect the women's low socioeconomic status compared to the nurses. Eleven of the women report household incomes in the bottom two categories available--less than $25,000. This differs significantly from the income levels of the nurses, of which none had reported household incomes in the bottom two categories (Z-score= -3.258, 2 tailed p= .001). Two of the women reported total household incomes of greater than $45,000 annually. Two of the women reported that no one in their household owned an automobile (this departure from nurses is not statistically significant) The Westside latinas also have significantly larger households and numbers of birth children, than do either the nurses or the anglos. Their lower income designation makes them poorer than nurses but the latinas' budgets are more often spread over larger households. With more mouths to feed, their low income is reduced further. The latinas reported having used a public health clinic, on their last physician visit, in six of the twenty-two cases. One other was seen in an emergency 40

PAGE 47

room, and another two visited walk-in care convenience clinics. The remaining thirteen all saw a private physician (or HMO equivalent) on their most recent physician visit. This use of public health clinics is significantly different from that of the nurses. Comparing numbers of public health uses to all other types of physician visits, the latinas used public health resources at higher frequency (Chi-square= 8.25, 1-tailed p<.OOS). Compared to the anglo group, the difference is also directional and significant, though to one lower order of magnitude in confidence (Chi-square= 3.66, 1-tailed p<.OS). Although two of the latinas respond that no one in their household owns a car, all reported having traveled to their last physician visit in a private car. The latinas differ from the anglo women, and the (also predominantly anglo) nurses, in that their use of Spanish is greater. Six of the latinas report themselves as fully fluent and bilingual in both English and Spanish, and ten more designate themselves as at least "somewhat comfortable" in speaking Spanish. One woman reports herself as "not very comfortable" speaking Spanish, and five more report that they do not speak Spanish at all. The degree to which these women speak Spanish sets them apart as significantly different from both the anglos and the nurses. By Wilcoxon matched-41

PAGE 48

pairs analysis, the latinas are in a different category of fluency in Spanish, with high confidence compared to anglos (Z-score= -2.9733, p= .0029), and to nurses (Zscore= -3.266, p= .0011). The latinas' language use may be more Spanishoriented than the survey responses suggest. On one occasion, a respondent was observed to describe herself with the category, "I don't speak Spanish." Moments later, this same woman turned to a toddler playing noisily at her feet: "Sal de aqui! Busca a Papa!" Other latinas may have answered questions regarding their Spanish ability with similar modesty. The Anglo Sample Twenty-two anglo women were interviewed at the Eastside Protestant Congregation during weekly meetings of the congregation's women's group, and during meetings for the church's single mothers ministry. Demographics. The mean age of the women is 31.7 years, ranging from 19 years to 47 years of age. This age grouping is not significantly different from that of the nurses. 42

PAGE 49

Ten of the twenty-two respondents reported household incomes of less than $25,000; of these, seven designated their household resources as less than $15,000 dollars (the seven women who reported the lowest income are all under the mean age, ranging from 19 to 28 years of age, with a mean of 22.7 years). This low income level is significantly from that of the nurses (Z-score= -2.978, 2-tailed p= .0029). Seven of the anglo women have no children. Those who are mothers have an average of 2.7 children. All but two of the women live with family and friends, with an average household population of 2.18 individuals. This household size is not significantly different from that of nurses (p= .925). Sixteen of the anglo women are married. Of the remaining six, four are divorced, but one is living in a marriage-like relationship with her ex-husband. Two are single. The number of anglo women who are married is higher than that of the nurses. But within each group, the number of married individuals versus all others, is perhaps directional, but not significantly different (Chi-square= 1.69, .10> p >.05). All but one twenty-one year-old woman had completed at least high school or a G.E.D. certification. One reported having completed a secretarial school 43

PAGE 50

program, in medical records coding. Nine of the women had attended at least some college, and of these, five had earned a bachelor's level degree. None had earned graduate degrees. In fifteen of twenty-two cases, the nurses declare a higher level of education than the angles by matched pairs This difference is statistically significant (Z-score= -2.864, 2 tailed p= .0042). Ethnic Assignment and Social Status. Eastside Congregation is located in an industrial urban fringe, which in recent. years has seen an increase in the size of local minority populations. Eastside members note that their congregation was once the norm for the local neighborhood, in which a long-empty church building nearby has recently re-opened, housing a charismatic, evangelical congregation composed mostly of local minority members. The Eastside Congregation is not shrinking, and many of its members are young families. But members do report that Eastside is not growing at the rate of churches which are identified as latina. Eastside Congregation engages in a number of joint projects with some other local churches, and their credo of the unity of their faith is credible. However, they are aware of the changing demography.of the neighborhood. 44

PAGE 51

The anglo women's use of health care resources does not differ significantly from that of the nurses, although two of the anglo women had used public health clinics on their last visits to physicians. Eighteen of the remaining women had used a private physician or an HMO physician on t.heir last visit, and two had most recently been seen in emergency rooms. Methods Utilized Three data-sets for this exploration were collected from each of the three sample groups. An example of the survey instrument is found in Appendix A. At the same time the survey was administered, the informants each participated in a pile-sort exercise and a semistructured interview. Multi-Dimensional Scaling Multi-Dimensional Scaling uses aggregate pile-sort data to build a map of the similarities (relatedness) among items from a given list. The procedure is nonmetric. 45

PAGE 52

Respondents are given a stack of cards, each labeled with a disease term. They are asked to sort the cards into piles which show how these items are similar or different. The respondents were all given the uniform instruction: Please place these cards into five to eight showing how they are similar or different. whatever reasons you want to, to decide how together. piles, Use they go If further clarification was needed, the respondent was given an example: If these were cars, you might want to put all of the trucks together, or maybe all the Chevys together and all the Hondas together, and so on. In all cases, this explanation was sufficient. All respondents were able to find satisfactory arrangements for their piles. The disease terms were selected by asking six nurses to list the first 40 medical ailments that came to mind. From these lists, the most frequently mentioned items were selected (rare diseases, or disease terms mentioned by only one or a few informants, were regarded as less important than diseases that all informants listed) The list was culled to 21 items for Multi-Dimensional Scaling. Anthropac (Borgatti, 1993) software was used to convert the respondent's piles into aggregate data 46

PAGE 53

matrices. Items frequently placed together by respondents were assigned high proximity scores (from zero for no co-occurrences, to 22 co-occurrences where all informants placed the items together) These frequencies of co-occurrence became the proximity ratings, used to assemble a best-fit map of the items, via the Minissa-X algorithm. An example of a correlation matrix may be found in Appendix B. Each illness term was assigned an abbreviation in order to reduce clutter on the MDS maps; a key to the MDS labels is found in Appendix C. Disease Severity Surveys Using a Likert scale, all respondents were asked to assign a rank for typical severities of some different diseases. The outcomes were ranked on the categories: 1) Always or Almost Always Fatal 2) Sometimes Fatal 3) Rarely Fatal 4) Never or Almost Never Fatal The illness terms selected were acceptable to most respondents, and most were able to complete the survey without further explanation. The exceptions were a few nurses who felt that some terms needed to be more 47

PAGE 54

specific ("is this Lymphoma Hodgkin's or non-Hodgkin's type?11); they were asked to consider the class of diseases subsumed by that label. A few of the lay women indicated that the terms were too precise; one woman wanted to know why the list included both emphysema and asthma, when these "are almost the same thing." These cases were exceptional; most respondents understood the question and were able to complete the survey. Technical Term Assessment. All respondents were asked to assess the severity of the terms "Myocardial Infarction" and "Heart Attack." These two terms have the same clinical definition. It was suspected that the percent agreement between the two terms for each group would depend on their familiarity with biomedicine. Specifically, the two terms should "tie" in severity more often by matched pairs analysis when assessed by nurses than when assessed by lay people. Semi-Structured Interviews Each respondent was interviewed using a semistructured format, either during or immediately following their completion of the survey and pile-sort exercises. In some cases the interview was used to complete a missing value in the demographic survey. 48

PAGE 55

All respondents were asked to describe some of their pile-sort categories and discuss their criteria for sorting the disease terms, using their own piles. Each respondent was asked about the type of doctor they last saw, to clarify the answers elicited by the demographic survey. Each was asked to compare the doctors they usually use now, from the doctors they saw as a child, and whether they were satisfied with the care and attention they receive from their current source of health care. They were also asked about alternative sources of health care they might have used in recent years. To help qualify data regarding the women's language and information source preferences, each was asked what kind of television they usually watch. This was to determine whether there were a likely difference in types of mass media exposure, between the latinas and anglo women. 49

PAGE 56

CHAPTER 3 FINDINGS AND RESULTS Multi-Dimensional Scaling The exploratory MDS maps generated from pile-sort data do not reveal differences in the taxonomies applied by latinas, anglos, and nurses to the biomedical terms used. The maps may suggest a higher degree of consensus and certainty in the nurses' taxonomy for the terms, but this is speculation regarding the non-metric MDS maps. Abbreviations of the illness terms were used in preparing the graphical maps; a key to the mapped labels is presented as Appendix D. The "map stress" or "graphical strain" expresses the degree of mathematical inaccuracy the graphical map must allow, to calculate a fit for the data points (from no stress at 0, to unity at a total distortion of the data). Romney and Weller (1988) suggest that map stress of between 0.10 and 0.15 is generally considered acceptable. MDS Mapping: The Nurses The nurses' responses to pilesort exercises yield a 50

PAGE 57

well organized set of categories and axes with which to group the illness terms. The final map is presented as Figure 3.1. The map stress of .105 is deemed acceptable, following Weller and Romney (1988) Clustering. The clusters may be seen in five major components, mapped in a cruciate pattern. At upper left, the cancers are all listed together: Stomach, Breast, Liver, Colon, and Leukemia. At lower left, cardiovascular diseases are found in a unified group. Hypertension (High Blood Pressure), Coronary Artery Disease, Myocardial Infarction, and Heart Attack come together with Aneurism and Stroke (Cerebral Vascular Accident) These diseases are unified by the quality of relating to heart and vascular disease. Although related anatomically to cranial structures, the Stroke category (resulting from blocked or bleeding brain vasculature) is placed with vascular disorders. At upper right are illnesses contracted by lifestyle or by voluntary exposure to risk factors (as shall be further examined under interview results) AIDS, Alcoholism, and Hepatitis (all types together) are clustered together. There is one curiously absent item in this cluster: Emphysema. Emphysema comes primarily from chronic inhalation of smoke and particulate 51

PAGE 58

material, almost always from habitual tobacco use, but here is less associated.with lifestyle and risk factor. Emphysema is more closely related to the cluster at lower right. As is discussed under interview findings, below, this area of the map features a group of diseases apparently organized by two principles: respiratory ailments and infectiousness. Emphysema appears related to Meningitis, Influenza, and Pneumonia. Diabetes and Kidney Failure are grouped together roughly at the center of the other four clusters. Axes. The axis of greatest importance in this map appears to be the horizontal axis from left to right, across the map. At left, the cardiovascular and malignant illnesses may be seen as originating in processes internal to the patient's body .. While certainly mediated by behavioral and environmental factors, both cancers and cardiovascular diseases are internal processes of declining or improper biological function. This contrasts with the right side of the map at the other end of the axis. The disease terms at right all result (primarily) from external forces which impinge on the body. The infectious diseases, clearly, are brought on not by failed body function, but by the insult of 52

PAGE 59

GsCA BrCA LiCA luk LnCA CAD HTN Mit anuCVA Figure 3.1 rf diab MDS Result for Registered Nurses 53 AIDS ale hep emp flu pnu mng

PAGE 60

pathogenic organisms. At the same time, AIDS, Hepatitis, and Alcoholism depend on individual behavior patterns and risk factors (semi-structured interviews suggest that the Diabetes/Kidney Failure cluster is a part of the left end of this axis, graphically pulled toward the infectious/behavioral end of the axis by two factors: the perception that diabetics' prognoses are mediated in part . by their ability to control the illness through their own behaviors, and the nurses' awareness of the increased frequency of diabetes in populations at risk for these other somewhat behavioral diseases, including minorities of low socioeconomic status) MDS Mapping: The Latino Women The latino women's MDS mapping results differ in some ways from the map produced by nurses, and is presented as Figure 3.2. Map stress is calculated at .133, indicating an acceptable fit to the-aggregate matrix, following Romney and Weller (1988). Clusters. At upper right, the latina women's MDS results show a cluster of cardiovascular illnesses. Myocardial Infarction and Heart Attack lie very near one another, and immediately below these lie Coronary Artery 54

PAGE 61

Disease, High Blood Pressure, and Stroke. Aneurism lies in this cluster as well, set off to the lower left. At the upper right portion of the map, cancers lie together. Breast, Lung, Liver, Colon, and Stomach Cancers virtually overlay one another, and Leukemia lays nearby to the left. The organization of cancers together, of course, is in large part to be expected. The terms all include the word "Cancer," and are clearly related by this trait. However, the latino women larg_ely included Leukemia in the cancers, although its name does not give away its biological category. At lower right, Pneumonia, Emphysema, Influenza, Meningitis, and Hepatitis come together in a group. Hepatitis lies.here and not with the "behavioral" diseases of AIDS and Alcoholism. AIDS and alcoholism lie together in the lower left portion of this map. At far left, the categories of Diabetes and Kidney Failure lie close together. Axes. Some similar axial distributions of the illness terms to those of the nurses exist. Infectious diseases occur at the lower portion of the map, mostly off to the right side. Absent is the central location of Diabetes and Kidney Failure. The two are associated, but not placed between internal and external etiologies. 55

PAGE 62

anu diab rf AIDS ale Figure 3. 2 MDS Result for Latinas MI Hrt CAD HTN CVA 56 luk emp mng hep BrCA LiCA LnCA flu pnu

PAGE 63

In sum, the latinas' MDS results display apparent similarity in the clusters of taxonomy to those of nurses, but are not fully congruent with the axes of biomedicine as constructed from the nurses' results. MDS Mapping: The Anglos The map generated by anglo women does not appear to be clearly unified along the cruciate pattern observed among the nurses. Two-dimensional model stress is indexed at .117, acceptable under the standards put forth by Weller and Romney (1988) The anglo map is presented as Figure 3.3. Clusters. The clustering of illness terms among the anglo sample does not follow the same cruciate pattern as the nurses produced, although the same cluster groups are apparent. At upper left are vascular and cardiac ailments. Stroke, Aneurism, and Myocardial Infarction lay together with Coronary Artery Disease and Heart Attack. Hypertension lays near this group but further toward the center of the map. At top center, we again see the co-occurrence of 57

PAGE 64

infectious and respiratory ailments. Pneumonia and Influenza lay together with Meningitis, and Emphysema is not far (though this item strays toward the center of the map). Lung Cancer lies nearby, to the right margin of this Further right, the other cancers are grouped together. Breast, Liver, Stomach, Breast and Colon Cancer cluster together, all at nearly the same coordinates (the outlying Lung Cancer is positioned between the cancers and the respiratory cluster described above) Leukemia lies outward from the cluster also, to the lower extreme. A grouping of AIDS, Hepatitis, and Alcoholism appears at the lower border of the anglo women's map. Kidney Failure and Diabetes are at lower right. Axes. The axes of separation in the anglo map seem to be similar to those of the nurses' map. Increasing in power to lower right, we may see an axis of causation-the further to the lower right, the more the illness items appear to relate to lifestyle and external risk factors. 58

PAGE 65

MID CVA anu emp HTN rf diab Figure 3.3 MDS Result for Anglo Women 59 pnu liUlg flu hep ale LnCA AIDS luk Br LiC

PAGE 66

Comparative MDS Appraisal The MDS maps do not appear to show the lay groups using a distinct folk model to organize these disease terms. The MDS results for the lay .women appear to follow the nurses' more authoritative model, yet lack expertise in its detailed principles. The disease terms used are not native knowledge--they do not derive from the people who have been asked to sort them, and they not necessarily part of the lay women's usual vocabulary. People have axes of discrimination and cognitive models for the terms that they themselves use, but these lay women do not routinely use all of the biomedical terms they have been asked to sort. The lay groups are able to recognize, mostly, what goes together and does not go together according to the biomedical model, as gathered from the nurses. AIDS, diabetes, cardiovascular problems, and cancers are all much discussed in the public arena and mass media. This may be described as the ability of the laypeople to recognize basic stock categories of biomedical illnesses, and their simultaneous inability to fully utilize the biomedical taxonomy. For instance, both lay groups are able to recognize that Coronary Artery Disease goes along with Heart Attack, often by thinking of what went together in people that they have known (evidenced 60

PAGE 67

by interview data, below) And when confronted with technical terms such as "Myocardial Infarction," they are able to recognize the "-cardial" portion of the term as heart-oriented. This converts the exercise into a task of sorting by meanings other than the biomedical meanings which the nurses explained as the basis for their pilesorts, yet _produce essentially the same clusters in the lay groups' MDS map results. The nurses describe using a biomedical taxonomy which sorts illnesses into clusters by organ system and by axes of disease etiology. By this model, it appears that diseases which result proximally from internal body system malfunction, aging, or degradation, occur as an opposite pole from illnesses caused by exposure to insult from outside the body. Two related medical problems-diabetes and its associated renal failure--occur as problems which are mostly internal medicine issues, but are impacted and mediated by lifestyle, thus drawing them graphically closer to the end of the axis. The anglos and latinas both cluster diseases into categories similar to the clusters that the nurses use-to this extent, they are able to appreciate some of the same categories of illness as co-occurrent and similar. But they do not seem to utilize, at least as thoroughly, the same axes of discrimination. Lacking knowledge of 61

PAGE 68

the internal logic of the nurse-derived list of diseases, their responses turn to their personal experiences and relationships with the terms, instead of to the biomedical taxonomy that produced them. Were the lay groups asked to pilesort terms that came from a free list exercise among latino women, or among low socioeconomic status anglo women, it may be that the MDS results would reveal a very different sort of clustering taxonomy. In some situations, ethnomedical categories could become relevant to understanding how the groups think about illness in general. But in coming into biomedical relationships, patients and families .are faced not with their own taxonomies, but with those offered (and imposed) by biomedicine. In this case, the respondents show how they deal with organizing terms that they might be confronted with in biomedical situations. Rather than dealing with these terms with a coherent ethnoscience, they organize the terms by virtue of experience with family and acquaintances, and basic principles not reserved only for recipients of biomedical training: cancers go together; high blood pressure has something to do with the heart; AIDS and alcoholism have to do with how you live your life. This finding does not indicate whether lessdisadvantaged lay people would use a more expe.rt version 62

PAGE 69

of the biomedical model. Lawyers and accountants, for instance, are also excluded from biomedical training, but they control more highly-valued cultural capital. This may afford them greater participation in biomedical relationships, and consequently greater access to the biomedical knowledge. Survey Findings A survey conducted regarding the severity of several health problems found few significant differences between the groups. Matched pairs analysis was used to compare the nurses' responses to each lay group, and to compare the responses between the lay groups. A table of the significant differences found is presented as Table 3.1. For three groups of respondents asked to assess 36 different disease terms, the data consists of 108 individual data points (latina versus nurses, latina versus anglos, and nurses versus anglos for each item). At the p <.05 level of confidence, we should not be surprised to see at least six significant differences arise by chance alone. 63

PAGE 70

Table 3.1 p-Values of Significant Differences In Severity Assessments Between Groups DISEASE TERM Lat/Ang RN/Lat RN/Ang HEART ATTACK .0077 CORONARY ARTERY DISEASE DIABETES TUBERCULOSIS (TB) .0032 .0151 EMPHYSEMA ANEURISM. ARTERIOSCLEROSIS STROKE LEUKEMIA LYMPHOMA KIDNEY FAILURE HEADACHE STOMACH CANCER .0106 CERVICAL CANCER COLON CANCER HERPES .0106 CHICKENPOX STOMACH ULCER LUNG CANCER. .0164 MENINGITIS HIGH BLOOD PRESSURE HEPATITIS .0281 SYPHILIS GONORRHEA LIVER CANCER .0022 .0071 BREAST CANCER UTERINE CANCER AIDS (SIDA) MYOCARDIAL INFARCTION ALCOHOLISM THE COLD INFLUENZA MEASLES PNEUMONIA ASTHMA .0414 .0468 OBESITY 64

PAGE 71

There is no indication that the differences found emerge from latina cultural ideas about illness. The data contains eleven significant differences in the entire matrix. Two are differences between latina and anglo assessments, four are between latinas' and nurses' answers, and the remaining five are between angles and nurses. Technical Term Assessment Findings By matched pairs analysis, the nurses showed the highest number of "ties" between "Myocardial Infarction" and "Heart Attack." Nurses tied on these items in 18 of 22 cases, for an 81% rate of agreement. The latinas tied the two terms in 14 of 22 cases (64%), and angles in 13 of 22 cases (59%) When the items are compared by matched pairs within the groups, a directional difference emerges between nurses' assessments of "Myocardial Infarction" and "Heart Attack." The difference is not statistically significant (p= .06), but may indicate a directional trend in nurses' use of the term "Myocardial Infarction" (or simply "M.I.") against "Heart Attack." 65

PAGE 72

Semistructured Interview Findings Semistructured interviews did not reveal meaningful differences in the fundamental disease models employed by the three groups, except in levels of expertise in the biomedical model. The interviews indicated that the lay groups generally place confidence in biomedical authority on health and disease, but do not have detailed understanding of the biomedical model. Interviews with the nurses suggest a high degree of consensus regarding how the disease terms should be organized. Etiology and organ system seem to be the most important aspects of the disease taxonomy; nurses would routinely describe their pile-sort organization by pointing to their piles: "That's all cancers, those are vascular and heart, and these are contagious ... Nurses reserve a special category for diseases caused by risk factors invited by personal lifestyle and activities. AIDS, hepatitis, and alcohol addiction all are ascribed first to the actions of the patient, who nurses may see as responsible for his own health problem. One 26 year-old nurse said of these ailments, .. and these are things you do to yourself." Latinas did not mention cultural difference as a problem with their current physicians and caregivers. The latinas did relate their dissatisfaction with health 66

PAGE 73

care resources to their difficulties finding insurance coverage and gaining access to health care more desirable than the public health clinics. Discussion with interviewees regarding their pilesort results do not indicate a humoral or other folk model as an organizing principle, among the latinas or among the anglo Both the latinas and the anglo women refer to a biomedical taxonomy to organize the terms, but their interviews reveal a limited understanding of the biomedical terms. They are less familiar with all of the terms, and less confident about using them. Explaining her placement of 11Myocardial Infarction," one 28 anglo woman said, "I don't know what this is, but it sounds pretty bad. I think that cardiac means your heart." Of pneumonia, she reported that "it goes with emphysema because of people get it in the lungs [sic], when they're old. And influenza you can also catch from other people." Both lay groups augment their limited biomedical understanding by appealing to personal knowledge of people who have had these health problems. The lay groups place disease terms together in piles by virtue of who they've known in connection with the disease. They appear to echo the nurses' categories of disease, but with a less explicit understanding of biosciences, and a 67

PAGE 74

greater reliance on their personal understandings of diseases and individuals who have been afflicted: Anglo, 27 years old: "I think this is what my aunt with diabetes got" (referring to "Kidney Failure"). Anglo, 31 years old:"I don't know what this is, but they talked about it when my dad had his heart attack" (holding card labeled 11Arteriosclerosis11); Latina, 30 years old: 11My cousin drinks too much, and my mother says that's why he has hepatitis11 (placing 11Hepatitis11 with 11Alcoholism11). Latina, 23 years old: 11My baby was premature and had these, and the neighbor boy had the flu real bad the same way11 (pointing to a pile including 11Pneumonia11 and 11Meningitis"). Latina, 65 years old: 11These are things that happen with people who aren't living right11 (placing "Alcoholism11 and "AIDS11 together) Although the latina woman quoted above placed 11Hepatitis11 and 11Alcoholism11 together, the overall latina MDS map shows 11Hepatitis11 closer to other infectious diseases than to behavioral-risk factor diseases (infectious Hepatitis-A is more common in Latin American immigrant populations than in the United States mainstream, and this may inform the context of the latinas' placement of Hepatitis) 68

PAGE 75

Among both latinas and angles, the respondents expressed concern about. their level of access to biomedical care. Both groups of women report putting considerable effort into using private doctors instead of public health clinics. Their underutilization is not a willing resistance to a model they do not believe in; it is a symptom of their lack of access to services. One 30 year-old latina, a mother of two children, said, I think that the reason that Hispanics and Mexicans don't go to the doctor is that they don't have good insurance. I mean, I have a good job here, I work in the church doing bookkeeping, but they don't have insurance for me. I don't like going down there [Public Health Clinic] and I have to wait with my kids in the little room with these people, all dirty and stuff. They treat us like we're those people and we're not. 69

PAGE 76

CHAPTER 4 DISCUSSION AND DIRECTIONS This exploratory examination of three groups of women raises questions and points out areas for further inquiry. The cultural orientation and the politicaleconomic orientation address somewhat different levels of analysis in biomedical relationships and underutilization. Cultural orientation and the primacy of negotiated explanatory models hold relevance for the individual clinical relationship, while the political economic perspective speaks to the needs of communities and larger populations. The Merits of a Cultural Orientation A cultural perspective on latino underuse of biomedicine has its greatest value at the micro-level of biomedicine, the individual patient-healer relationship. Where different cultural models exist between patients and caregivers, they exist in the form of the cognitive schemata that individuals carry into the clinics and 70

PAGE 77

hospitals. But because the individuals' cognitive schema and explanatory models are rooted in a larger cultural worldview, their actions and needs can be called cultural, rather than personal. The cultural orientation in understanding latinos' problematic use of biomedicine is relevant to the clinician who deals with individual patients, and with the impact of differing schema. This is rather a different issue from the general ignorance of lay populations of specialized biomedical knowledge. Negotiating explanatory models is not simply a matter of providing biomedical information to the patient, filling a cognitive void. Patients of some backgrounds may already have a set of standards and schematics for making sense of illness experiences--different schemata regarding the workings of the world and body, and extending also into the patient's social suurroundings. Among the social influences on the individual's clinical relationship is the cultural ambiguity of informed consent and clinical decision-making. A biomedicine better attuned to questions of individualism, collectivismo, community, and meanings, is a biomedicine in greater accord with the needs of its patients. Under a culturological agenda, the clinical role becomes one of negotiation over authority (or toward 71

PAGE 78

understanding in a "culturally sensitive" medicine), but this is not to suggest that all cognitive models find voice in ideal biomedical relationships. An individual's mental status, for instance, may confound and preempt appreciation of culturally-different explanatory models. Culture shapes individuals' beliefs and experiences of health and sickness. It is with the patient and caregiver that cultural incongruence and shared meanings emerge. To come to a resolution of meanings in the therapeutic relationship is to come to know the explanatory model of the patient, and to negotiate and participate in the meanings of the sickness, just as in the rigors of the treatment. The ethos is as fleeting, perhaps, as the art is long. It is simpler for anthropology than for biomedicine to accept non-biomedical cultural models' legitimacy beyond individuals' discourse with caregivers. But patients will live and die within their explanatory models with or without the approval of biomedicine. Where the stated goal of biomedicine is the healing of the patient, sensitivity to the cultural meanings of illness is central to the relationship of patient and healer. 72

PAGE 79

The Merits of a Political-Economic Orientation The data at hand do not show that these groups of women use different cultural models for health and illness, but suggest that the lay women control only a novitiate understanding of biomedical knowledge. Regarding the health of communities and of marginalized populations of United States, the political-economic orientation is of greater value. The political-economic orientation suggests that the two groups of lay women do not see the world in radically different, culturally-programmed ways. They understand biomedical terms and ailments, and their own biomedical relationships, from relatively similar positions of economic and educational marginalization. Lay people in general are not privileged to specialized biomedical knowledge, and all the more so for lay people who are marginalized by their devalued cultural capital and socioeconomic status. Ai Singer er al. write, [W]e must consider the social origins of illness and class {as well as racial and gender) control of health institutions and knowledge {1990:185). Exclusion from biomedical knowledge relates to a social hegemony expressed in biomedicine's relationship with lay populations. 73

PAGE 80

Where biomedicine is positioned as the gold standard of clinical understanding, limited knowledge gives rise to misinformation. Ironically, patients' own misunderstandings can create false sense of communication in the biomedical relationship, further confounding troublesome clinical encounters. The women examined in this exploration, for example, created similar clusters of the same illness terms, but this does not mean that the lay groups and the nurses came to these similar categories in similarways. The political-economic orientation advocates greater sensitivity to poor lay people's marginalization from more-valued cultural capital. Providing information to patients cannot be limited to providing raw information, but must also address the lack of the overarching principles into which this information is supposed to fit. A focus on culturally different meanings of illness can become a romanticized distraction from attention to the social stratification embedded in biomedical praxis: Characteristic of this circumscribed explanation is its inattention to the manifold ways doctor-patient interactions are structured by a wider field of class and other power relations embedded within, but not always visible form, the narrow confines of the clinical setting (Singer et al. 1990:179). 74

PAGE 81

Where marginalized class status is the basis for biomedical exclusion, devalued cultural capital is the cause of biomedical underutilization. A critical social-science view of biomedicine must be: concerned with synthesizing the macrolevel understanding of political-economy with the microlevel sensitivity and awareness of [cultural] anthropology (Singer et al. 1990:179). Appropriate strategies for expanding access to biomedicine must address not only the financial logistics of access, but the complex issues of cultural marginalization as a function of socioeconomic position. 75

PAGE 82

Appendix A-Example of Survey Instrument YOUR AGE: NUMBER OF CHILDREN: TOTAL NUMBER OF PEOPLE IN YOUR HOME: You are (circle one) : 1) MARRIED 2) SINGLE 3) DIVORCED 4) WIDOWED 5) LIVING IN A MARRIAGE-LIKE RELATIONSHIP Total Household Income, from all sources, before taxes: (circle one) 1) less than $15,000 2) between $15,000 and $25,000 3) between $25,000 and $45,000 4) over $45,000 What is your highest level of education? Please circle category thatbest describes you: the one 1) 2) 3) 4) 5) 6) Some high school High school diploma OR G.E.D. certificate Some College Bachelor's Degree Graduate Degree Vocational or Technical Certification in: 76

PAGE 83

Do you, or a member of your household, own a car? 1) YES 2) NO How comfortable are you speaking Spanish? 1) VERY-I'm bilingual 2) SOMEWHAT COMFORTABLE 3) NOT VERY COMFORTABLE 4) I DON'T SPEAK SPANISH What language do you speak with your children, if you have any? 1) SPANISH 2) ENGLISH 3) BOTH 4) OTHER What language do you usually speak with your spouse, if you have one? 1) SPANISH 2) ENGLISH 3) BOTH 4) OTHER How many hours do you watch television in a typical day? In your home, how many hours is a television turned on in a typical day? How far did you have to travel the last time you or your family went to see a nurse or doctor? ------------------77

PAGE 84

What kind of medical facility did you use the last time you or your family went to see a nurse or doctor? 1) PRIVATE DOCTOR'S OFFICE 2) EMERGENCY ROOM 3) CONVENIENCE CARE CLINIC 4) PUBLIC HEALTH CLINIC (County Health Dept, etc) 5) OTHER ____________________ __ Did you get there by (circle one) : 1) PRIVATE CAR 2) BUS 3) WALKING 4) OTHER 78

PAGE 85

PLEASE INDICATE HOW LIKELY ONE IS TO SURVIVE, OR HOW LIKELY ONE IS TO DIE FROM, THE FOLLOWING DISEASES AND HEALTH PROBLEMS, ON THE SCALE BELOW: 1) Always or Almost 2) Sometimes Fatal 3) Rarely Fatal 4) Never or ALmost TERMS: Heart Attack Coronary Artery Disease Diabetes Tuberculosis (TB) Emphysema Aneurism Arteriosclerosis Stroke Leukemia Lymphoma Kidney Failure Headache Stomach Cancer Cervical Cancer Colon Cancer Herpes Chickenpox Stomach Ulcer Lung Cancer Meningitis High Blood Pressure Hepatitis Syphilis Gonorrhea Liver Cancer Breast Cancer Uterine Cancer AIDS (SIDA) Myocardial Infarction Alcoholism The Cold Influenza Measles Pneumonia Asthma Obesity Always Fatal Never Fatal 79

PAGE 86

Appendix B-An Example Correlation Matrix File Contents For Multi-Dimensional Scaling DL N = 21 SIMILARITIES, LOWERHALF LABELS: HrAtk,CAD,diab,emp,anu,CVA,luk,RF,GsCA,CoCA,LnCA,mng,HTN, DATA: 21 2 3 0 1 4 18 17 3 0 16 15 2 0 16 1 1 1 0 1 1 3 4 16 4 2 4 1 1 1 0 0 1 1 17 0 1 1 0 0 1 1 17 0 21 1 1 0 3 1 1 15 0 18 18 1 0 4 3 2 3 1 4 0 0 0 18 19 5 3 14 14 1 7 1 1 1 0 0 0 6 2 0 0 1 9 2 1 0 4 1 1 1 0 0 1 1 16 0 21 20 19 0 1 2 1 1 0 0 1 1 17 0 20 21 18 0 1 0 19 0 1 1 2 0 0 1 1 0 0 1 3 1 8 1 0 22 21 2 0 18 16 1 3 1 1 1 1 18 0 1 1 0 0 1 6 3 0 0 0 7 2 1 0 0 2 13 2 0 5 0 0 0 0 6 0 0 0 0 0 0 1 8 0 2 0 0 2 0 0 1 0 1 16 1 1 0 1 0 0 2 7 0 1 0 0 1 1 1 11 80

PAGE 87

Appendix C-Guide for Semi-structured Interviewing 1) Last time you saw a doctor, how would you rate the medical attention you received? 2) Is there anyone besides a doctor that you and your family see when you're sick? 3) Have you.ever not gone to a doctor when you thought that you or your family should? Why? 4) Do the doctors you've usually seen explain enough about a sickness to you? Do you ever remember them asking you where you think it came from? 5) Was it easy for you to get to the doctor the last time you went? 6) Is there anything people can do to prevent serious illness? 7) How are (MDS TERMS IN SAME PILE) similar? 8) How did you decide what else should go in this pile? [CANCERS] [CARDIO] [BEHAVIOR] [INFECT] [OTHER] 9) What kind of TV do you watch? How about your kids? (Span I Eng) 10) Are the doctors you see now similar to the doctors your family used when you were a child? 81

PAGE 88

Appendix D-Labels for MDS Mapping of Terms Illness Term MDS Label Heart Attack HrAtk Coronary Artery Disease CAD Diabetes diab Emphysema emp Aneurism anu Stroke CVA Leukemia luk Kidney Failure RF Stomach Cancer GsCA Colon Cancer CoCA Lung Cancer LnCA Meningitis mng High Blood Pressure HTN Hepatitis hep Liver Cancer LiCA Breast Cancer BrCA AIDS (SIDA) AIDS Myocardial Infarction MI Alcoholism ale Influenza flu Pneumonia pnu 82

PAGE 89

REFERENCES Acosta, A., Hamel, v. 1995 C.S.A.P. Implementation Guide: Hispanic/Latina Natural Support Systems. United States Department of Health and Human Services Publication No (SMA) 95-3033. Washington, DC: U.S. Government Printing Office. Baer, Hans 1989 The American Dominative Medical System as a Reflection of Social Relations in the Larger Society. Social Science and Medicine. 28(11) :1103-12. Bernard, H. Russell 1988 Research Methods in Cultural Anthropology. Newbury Park, CA: Sage. Bourdieu, Pierre 1977 Reproduction in Education, Society, and Culture. London: Sage Press. Borgatti, Stephen 1992 Anthropac (computer program). Columbia, SC: Analytic Technologies. Casson, Ronald 1983 Schemata in Cognitive Anthropology. Annual Review of Anthropology 12:429-62. Chen, A. 1991 Non-compliance in Community Psychiatry: Review of Clinical Interventions. Hospital and Community Psychiatry 42:282-287. Crimm, Allan and Greenberg, Raymond 1981 Reflections on the Doctor-Patient Relationship: Paternalism and Autonomy. In Ethical Dimensions of Clinical Medicine. Dyer and Robbins eds. Pp 104-111. Springfield IL: Charles C Thomas. Csordas, A., Kleinman, Arthur 1990 The Therapeutic Process. In Medical Anthropology: Handbook of Theory and Method. Johnson and Sargent, eds. Pp 11-42. NY: Greenwood Press. D'Andrade, Roy 1979 culture and Cognitive Science. Memorandum to Cognitive Science Committee, Sloan Foundation. Dietrich, K. 1989 Ethnic Differences in the Demand for Physician and Hospital Utilization Among Older Adults in Major American Cities: Conspicuous Evidence of Considerable Inequalities. The Milbank Quarterly 67(3-4) :412-50. 83

PAGE 90

Elder, J., Mayer, J., Campbell, N., Castro, F., Harmon, M., Coe, K. 1994 Breast Self Examination: Knowledge and Practices of Hispanic Women in Two Southwestern Metropolitan Areas. Journal of Community Health 19(6) :433-449. Foster, George McClelland 1994 Hippocrates' Latin American Legacy: Humoral Medicine in the New World. Langhorne, PA: Gordon and Breach. Geertz, Clifford 1973 The Interpretation of Cultures. NY: Basic Books. Ginzberg, E. 1991 Access to Health Care for Hispanics. The Journal of the American Medical Association 265(2) :238-242. Goodenough, Ward 1970 General and Particular. In Description and Comparison in Cultural Anthropology. Pp 98-130. Chicago: Aldine Publishing. Guarnaccia, P., Angel, R. 1989 Mind, Body, and Culture: Somaticization Among Hispanics. Social Science and Medicine 28(12) :1229-1239. Harris, Marvin 1979 Cultural Materialism: The Struggle for a Science of Culture. NY: Random House. 1988 Theoretical Principles of Cultural Materialism. In High Points in Anthropology. Bohannon and Glazer, eds. Pp 379-403. NY: McGraw-Hill. Jordan, Brigitte 1983 Birth in Four Cultures. Montreal: Eden Press. Kay, Margarita 1977 Health and Illness in a Mexican Barrio. In Ethnic Medicine in the Southwest. E.H. Spicer, ed. Pp 97-167. Tucson: University of Arizona Press. Keefe, S., Padilla, A. 1979 The Mexican-American Family as an Emotional Support System. Human Organization 38:144-152. Kleinman, Arthur 1981 Patients and Healers in the Context of Culture: An Exploration of the Borderland Between Anthropology, Medicine, and Psychiatry. Berkeley, CA: University of California Press. 1988 The Illness Narratives: Suffering, Healing, and the Human Condition. NY: Basic Books. Kleinman, Arthur, Good, Bryon, Guarnaccia, Peter 1990 A Critical Review of Epidemiological Studies of Puerto Rican Mental Health. American Journal of Psychiatry 147 (11) :1449-57. 84

PAGE 91

Kleinman, Arthur, and Margaret Lock 1996 The analysis of social suffering. Daedalus. v125 n1 pXI XXI. Lantz, Paula, Dupuis, Lawrence, Reding, Douglas, Krauska, Michelle, and Lappe, Karen 1994 Peer discussions of cancer among Hispanic migrant farm workers. Public Health Reports 109(4) :512-521. Leacock, M. 1982 Marxism and Anthropology. In The Left Academy: Scholarship on American Campuses. NY: McGraw-Hill. Marx, Karl 1970 [1859] A Contribution of the Critique of PoliticalEconomy. NY-: International Publishers. McManus, M. Newachek, P. I Lieu I T. 1993 Race, Ethnicity, and Access to Ambulatory Care Among US Adolescents. The American Journal of Public Health 83(7) :960-6. Perez-Stable, E., Sabogal, F., Hiatt, S., McPhee, S. 1992 Misconceptions About Cancer Among Latinos and Anglos. The Journal of the American Medical Association 268(22) :3219-3224. Pettitt, L., Cumberland, W., Fielding, J. 1994 Immunization Status of Children of Employees in a Large Corporation. The Journal of the American Medical Associationv 271 (7) :525-31. Randall, Terry 1991 Key to Organ Donation May Be Cultural Awareness. The Journal of the American Medical Association 265(2) :176-177. Rankin-Hill, L., and M. Bates 1994 Culture, and Chronic Pain. Social Science and Medicine 39(5) :629-46. Ray, Laura, F. Trevino, and A. Estrada 1990 Health Care Utilization Barriers Among Mexican Americans: Evidence from HHANES 1982-84. The American Journal of Public Health 80(12) :27-32. Ray, Laura, F. Trevino, and J. Higginbotham 1990 Utilization of Curanderos by Mexican Americans: Prevalance and Predictors. The American Journal of Public Health 80 (12) :32-36. Romney, A. Kimball, Susan Weller, and William Batchelder 1986 Culture as Consensus: A Theory of Culture and Informant Accuracy. American Anthropologist 88:313-338. Saunders, L. 1954 cultural Differences and Medical Care: The Case of the Spanish-speaking People of the Southwest. New York: Russell Sage. 85

PAGE 92

Scheper-Hughes, Nancy 1983 Curanderismo in Taos County, New Mexico-A Possible Case of Anthropological Romanticism? Western Journal of Medicine. December 139(6) :875-884. Schlesinger, H., B. Burns, C. Patrick, and D. Padgett 1994 Ethnicity and the Use of Outpatient Mental Health Services in a National Insured Population. The American Journal of Public Health 84(2) :222-227. Shelton, D., M. Garcia, G. Marks,and J. Solis 1990 Acculturation, Access to Care, and Use of Preventative Services by Hispanics: Findings from HHANES 1982-84. The American Journal of Public Health 80(12) :11-20. Seidel, Henry, Jane Ball, Joyce Dains, and William Benedict 1995 Mosby's Guide to Physical Examination. St. Louis, MO: Mosby. Singer, Merrill 1989 The Coming of Age of Medical Anthropology. Social Science and Medicine 28(11) :1103-12. 1994 Community-centered Praxis: Toward an Alternative Nondominative Applied Anthropology. Human Organization 53(4) :336-45 .. Singer, Merrill, Hans Baer, and Ellen Lazarus 1990 Reinventing Medical Anthropology: Toward a Critical Realignment. Social Science and Medicine 30(2) :179-188. Smith, Harmon L. 1981 A Valid Consent: Informed, Voluntary, Competent. In Ethical Dimensions of Clinical Medicine. Pp 70-81. Springfield, IL: Charles C Thomas. Snowden, L., and F. Cheung 1990 Community Mental Health and Ethnic Minority Populations. Community Mental Health Journal 26(3) :277-293. Solis, J., M. Garcia, and G. Marks 1990 Health Risk Behaviors of Hispanics in the US: Findings from HHANES 1982-84. The American Journal of Public Health 80 (12) :20-27. Spradley, James 1972 Foundations of Cultural Knowledge. In Culture and Cognition: Rules, Maps, and Plans. J. Spradley, ed. Pp 3-41. London: Chandler Publishing Company. Stroup-Benham, C., R. Valdez, and E. Moyer 1991 Health Insurance Coverage and Utilization of Health Services by Mexican Americans, Mainland Puerto Ricans, and Cuban Americans. The Journal of the American Medical Association 265(2) :233-238. 86

PAGE 93

Taussig, I.M. 1987 Comparative Responses of and AngloAmericans to Early Goal-Setting in a Public Mental Health Clinic. Journal of Counseling Psychology 34:214-217. Valle, R., and W. Vega 1982 Hispanic Natural Support Systems: Mental Health Perspectives. Sacramento, CA: State of California Department of Mental Health. Vega, w. 1989 What is a Culturally-Sensitive and -Specific Prevention Program for Hispanic High-Risk Youth? Rockville, MD: U.S. Department of Health and Human Services, Office of Substance Abuse Prevention. Weller, Susan; Romney, A. Kimball 1988 Systematic Data Collection. Beverly Hills, CA: Sage. Winn, Deborah 1991 HIV Transmission and General Knowledge of HIV and AIDS. The American Journal of Public Health 81(12) :1591-1596. Young, Zenana, Schepler 1986 Medical Non-compliance in Schizophrenia: Codification and Update. Bulletin of the American Academy of Psychiatry and the Law 14:105-122. 87