Assessing health literacy and health information needs of persons at risk for cardiovascular disease

Material Information

Assessing health literacy and health information needs of persons at risk for cardiovascular disease
Gaglio, Bridget Ann
Place of Publication:
Denver, Colo.
University of Colorado Denver
Publication Date:
Physical Description:
xviii, 300 leaves : ; 28 cm.

Thesis/Dissertation Information

Doctorate ( Doctor of Philosophy)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Health and Behavioral Sciences, CU Denver
Degree Disciplines:
Health and Behavioral Sciences
Committee Chair:
Bull, Sheana S.
Committee Members:
Bayliss, Elizabeth A.
Glasgow, Russell E.
McKenzie, Thomas
Tracer, David P.


Subjects / Keywords:
Cardiovascular system -- Diseases -- Patients -- Health and hygiene ( lcsh )
Health education -- United States ( lcsh )
Patient education -- United States ( lcsh )
Literacy -- United States ( lcsh )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Thesis (Ph. D.)--University of Colorado Denver, 2010.
Includes bibliographical references (leaves 271-300).
General Note:
Department of Health and Behavioral Sciences
Statement of Responsibility:
by Bridget Ann Gaglio.

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:
656256564 ( OCLC )

Full Text
Bridget Ann Gaglio
B.S. University of Houston, 1998
M.P.H. University of Texas Houston, School of Public Health, 2000
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Health and Behavioral Sciences

2010 by Bridget Ann Gaglio
All rights reserved.

This thesis for the Doctor of Philosophy
degree by
Bridget Ann Gaglio
has been approved
Sheana S. Bull
David P. Tracer
Zl\0l 2|l}

Gaglio, Bridget A. (Ph.D., Health and Behavioral Science)
Assessing Health Literacy and Health Information Needs of Persons at Risk
for Cardiovascular Disease
Thesis directed by Associate Professor Sheana S. Bull
Individuals with limited health literacy have poorer health outcomes,
under-utilize preventive services, and encounter challenges in understanding
health information. The overall aim of this study was to investigate the
experiences of, and preferences, for receiving health information from
individuals with varying health literacy and numeracy abilities who are at risk
for CVD. Patient-level quantitative and qualitative data were collected from
adults, age 40 years and older, who received care at either Kaiser
Permanente Colorado (KPCO) or Denver Health and Hospital Authority
(DHHA), and had two or more risk factors for cardiovascular disease. At
KPCO a list of 32,041 members meeting the study criteria was generated
from the electronic medical records system. A randomly selected sample
(N=493) from the patient list were invited by letter to participate in the study.
Seventy-five members completed the study survey. At DHHA a two-tiered
recruitment strategy was used consisting of a recruitment letter mailed to 173
individuals followed by in-clinic recruitment. Seventy-five patients from DHHA
completed the study survey. All 150 participants provided data on health
literacy ability, numeracy ability, sociodemographics, health status, patient
activation, illness perception, and information related to their experiences with
accessing, using, and understanding health information. We conducted in-
depth interviews with 28 participants to further explore experiences and
preferences for health information. In-depth interviews were also conducted
with 9 health educators to understand how health information is currently
shared with patients/members. The majority of participants had adequate
functional health literacy, while 65% had low numeracy skills. Regardless of
health literacy or numeracy ability, participants preferred to receive health

information during a face-to-face conversation with their physician, primarily
to receive immediate answers to their health questions. Individuals with
limited health literacy did not prefer to receive information via written or
impersonal modalities such as pamphlets and the internet. Numeracy
abilities appear to be of equal or greater importance than health literacy
abilities for comprehension of health information. While patients
overwhelmingly prefer to receive health information during face-to-face
encounters with their health care providers, having multiple modalities
available in plain, clear language is key to patients understanding and
applying information to health behavior.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.

The future belongs to those who believe in the beauty of their dreams -
Eleanor Roosevelt
I dedicate this dissertation to my father, Joseph Gaglio, my stepmother,
Latresa Fuller Gaglio and to my grandparents, Louis and Josephine Gaglio
and Angelo and Margarite Nasella, for emphasizing the importance of
education and dreams.

I would like to acknowledge the financial support of the Institute for Health
Research at Kaiser Permanente Colorado, Colorado Health Outcomes, and
my academic program Health and Behavioral Sciences. I would like to thank
a number of people who have been integral to the completion of this
dissertation. First of all, special thanks to the participants of this project. I
appreciate all of you sharing your experiences with me. Thanks to my
committee members Sheana S. Bull, Elizabeth A. Bayliss, Russell E.
Glasgow, Thomas MacKenzie, and David P. Tracer for their sage wisdom
and advice. I would especially like to thank Cohort 10 (Aimee Ferraro,
Colleen Julian, Stacy Salomnsen-Sautel, and Megan Wilson) for making this
entire journey worth taking and enjoyable. I would like to thank my aunt, Dr.
Connie Marie Gaglio, for always reminding me that I am the master of my
own destiny and supporting me in any experience I have decided to pursue. I
would like to acknowledge the experiences I have had, both personally and
professionally, during the course of this dissertation. These experiences
along with this dissertation have forced me to grow and evolve in ways I
never imagined. I would like to acknowledge a few key friends; Amy
Opperman for volunteering to make recruitment calls so I could take a much
needed vacation and Holly Whitesides for assisting with coding qualitative
data. I would like to thank all the wonderful women at the Adult Care clinic at
Denver Healths Eastside clinic for all their assistance with recruiting
participants. I would like to thank the LUCHAR group at COHO and Denver
Health and Hospital Authority for allowing me to present challenges incurred
during this project to them and for the suggestion of ideas on how to
overcome them. I would like to thank Dr. Rebecca Henratty from Denver
Healths Eastside clinic for all her help and support during participant
recruitment. Finally, I want to thank my friends and family who have always
been there to listen to me struggle, celebrate, and everything in-between
during this whole process. This work would not have been possible without
all of you. Thank you.

Figures.................................................. xv
Tables................................................... xvi
1. INTRODUCTION AND BACKGROUND................................ 1
Structure of the Literature Review...................... 3
History of Health Literacy.............................. 3
Trends in the Literature................................ 6
Definitions of Health Literacy.......................... 7
Assessments Used to Evaluate Health Literacy....... 10
Current Areas of Emphases in Health Literacy
Literature............................................. 17
Printed Materials............................... 17
The Internet.................................... 19
The Medical Encounter........................... 21
Health Outcomes and Disease Knowledge....... 22
Screening and Preventive Behaviors.............. 26
Economic Implications........................... 27
Emerging Areas of Study................................ 28

System Barriers................................. 28
Oral Exchange................................... 30
Numeracy........................................ 31
Cardiovascular Disease (CVD) in Relation to Health
Literacy............................................... 34
Summary of Existing Research and Gaps.................. 36
Overview of the Dissertation........................... 36
Research Questions and Specific Aims............ 36
Structure of the Dissertation................... 39
2. THEORETICAL PERSPECTIVES.................................. 41
Individual Level Theories.............................. 45
Health Belief Model............................. 45
Social Cognitive Theory......................... 48
Structural Level Theories.............................. 52
Foucault Knowledge/Power...................... 52
Qualitative Theory..................................... 57
Phenomenology................................... 57
Association of Sociodemographics and Health
Characteristics and Health Literacy and Numeracy 63

Study Population................................. 63
Study Instruments................................ 63
Data Collection........................................ 72
Analytic Plan.......................................... 73
Sample Size Calculations......................... 73
Overview of Analyses............................. 74
Patients Attitudes and Preferences Toward Current
Methods of Obtaining and Receiving CVD Prevention
Information........................................... 77
Study Population................................. 77
Data Collection........................................ 77
Patient Interviews............................... 77
Current Methods Used by Health Care Professionals to
Distribute and Provide CVD Health Information.......... 79
Health Educator Interviews....................... 79
Analytic Plan.......................................... 80
Compiling Results of Mixed-Methods Analysis............ 81
AND HEALTH INFORMATION STUDY............................... 82
Introduction........................................... 82
Methods................................................ 86

Obtaining IRB Approval from Multiple Sites.... 86
IRB Approvals...................................... 87
Research Settings.................................. 88
Eligibility Criteria for Participation............. 90
Participant Identification KPCO.................. 91
Passive Recruitment Sampling Frame KPCO.... 93
Recruitment Methods DHHA......................... 94
Consent of Study Participants...................... 96
Analyses........................................... 96
Results................................................... 97
Characteristics of Participants and Non-
Potential Effect of Health Literacy on
Discussion.............................................. 102
Limitations............................................. 105
Conclusions............................................. 107
Introduction........................................ 108
Methods............................................. 110

Analyses................................................ 111
Descriptive Analyses.............................. 112
Bivariate Analyses................................ 112
Multivariate Analyses............................. 113
Results................................................. 113
Overall Sample Characteristics.................... 113
CVD Risk Factors and Health Status................ 116
Health Literacy and Numeracy...................... 120
CVD Knowledge/Perception.......................... 131
Discussion.............................................. 135
Limitations............................................. 139
Conclusions............................................. 140
AND NUMERACY ABILITY?.......................................... 141
Introduction............................................ 141
Methods................................................. 144
Quantitative Methods.............................. 144
Qualitative Methods............................... 145
Analyses................................................ 147

Qualitative..................................... 148
Quantitative Results................................... 149
Computer Access and Use of Internet for Health
Information..................................... 149
Information Seeking Preferences................. 158
Qualitative Results.................................... 158
Open-ended Questions from Survey................ 159
Patient I nterviews............................. 166
Health Educator Interviews...................... 170
Discussion............................................. 173
Limitations............................................ 180
Conclusions........................................... 181
DIRECTIONS................................................ 182
Summary of Study Findings.............................. 182
Results in the Context of Theoretical Perspectives... 186
Health Belief Model............................. 186
Social Cognitive Theory......................... 187
Foucault Knowledge and Power.................. 192

Revised Theoretical Model
Study Contributions........................ 197
Study Implications......................... 199
Future Research............................ 200
A: STUDY INSTRUMENTS............................ 202
B: KPCO CONSENT FORM............................ 248
C. DHHA CONSENT AND HIPPA FORMS................. 254
D. KPCO RECRUITMENT MATERIALS................... 264
E. DHHA RECRUITMENT MATERIALS................... 267
BIBLIOGRAPHY........................................ 271

2.1 Overview of theoretical perspectives and constructs....... 44
3.1 Overview of research design for study..................... 62
4.1 CONSORT Diagrams KPCO compared to DHHA 98
4.2 CONSORT Diagrams DHHA recruitment methods................... 99
7.1 Revised theoretical model.................................... 196

1.1 Comparison of original assessments created to assess
adult health literacy abilities........................... 15
3.1 Summary of measures used in the data collection process
and associated theoretical constructs..................... 70
3.2 Overview of analyses by research question conducted for
this dissertation......................................... 75
4.1 Representativeness of KPCO membership compared to the
Denver metropolitan population............................ 89
4.2 Representativeness of DHHA Eastside clinic membership
compared to the Denver metropolitan population............ 90
4.3 Summary of health literacy abilities for all study participants
and by health care organization............................... 102
5.1 Demographic characteristics of study participants by health
care organization............................................. 115
5.2 Cardiovascular risk factors and health status characteristics
of study participants by health care organization............. 119
5.3 Health literacy and numeracy abilities of study
participants.................................................. 120
5.4 Demographic characteristics of study participants by health
literacy and health numeracy level............................ 122
5.5 Cardiovascular risk factors and health status characteristics
of study participants by health literacy and health numeracy
level......................................................... 123

5.6 Associations between participant characteristics and health
literacy...................................................... 125
5.7 Association between participant characteristics and
numeracy...................................................... 126
5.8 Multivariate logistic regression assessing respondent
characteristics significantly associated with adequate
functional health literacy.................................... 128
5.9 Multivariate logistic regression assessing respondent
characteristics significantly associated with high
numeracy...................................................... 130
5.10 Study samples beliefs about causes of cardiovascular
disease by health literacy level.............................. 132
5.11 Study samples beliefs about causes of cardiovascular
disease by health numeracy level.............................. 134
6.1 Characteristics of respondents who use the internet for
health information............................................ 151
6.2 Perceived benefits of internet use on participants
understanding of health care issues - stratified by health
literacy and numeracy......................................... 153
6.3 Access to a computer as function of health literacy: multiple
logistic regression........................................... 155
6.4 Use of internet for health information as a function of health
literacy: multiple logistic regression........................ 157
6.5 Summary of study participants most frequent qualitative
data responses by health literacy and numeracy abilities.... 161
7.1 Matrix of key study findings by components of health
literacy definition........................................... 185

Results from correlations for predictor variables used in
health literacy and numeracy analyses.....................

Chronic illnesses affect a large proportion of the American population.
In 2005, 133 million people, almost half of all Americans lived with at least
one chronic condition (Centers for Disease Control and Prevention, 2009).
As of September, 1.2 million people had died from a chronic disease in 2009
(Centers for Disease Control and Prevention, 2009). Cardiovascular disease
(CVD) is the leading cause of chronic disease-related death with roughly
630,000 deaths in 2006 (Centers for Disease Control and Prevention, 2008c).
In addition to the high prevalence of chronic illness in this country, people with
chronic conditions are taking on more responsibility for their care in the form
of self-management (Jamison, 1998). Self-management is the ability of a
patient to deal with all that a chronic illness entails, including symptoms,
lifestyle changes, and day-to-day treatment (Barlow, Wright, Sheasby, Turner,
& Hainsworth, 2002). Having multiple chronic morbidities makes self-care
more complex and challenging (Bayliss, Ellis, & Steiner, 2007; Gazmararian,
Williams, Peel, & Baker, 2003; Williams, Baker, Parker, & Nurss, 1988).
One factor of substantive importance in chronic disease prevention,
self-management interactions with the health care system, and health
outcomes is health literacy-the degree to which individuals have the
capacity to obtain, process, and understand basic health information and
services needed to make appropriate health decisions (Ratzan & Parker,
2005). Persons with high health literacy are better able to both prevent
and/or manage disease relative to persons with limited health literacy (Baker
et al., 2002; Gazmararian, et al., 2003; Nielsen-Bohlman, Panzer & Kindig,

2004; Schillinger, Barton, Karter, Wang, & Alder, 2006; Schillinger et al.,
2002; Scott, Gazmararian, Williams, & Baker, 2002; Williams, Baker, Parker,
et al., 1998). While we know how health literacy is associated with health
outcomes, we know little about preferences for receiving health information
among individuals with varying health literacy levels. CVD exemplifies a
condition for which self-management may be dependent on health literacy
and we know little about preferences for receiving health information in this
population. Additionally, we know little about how to best disseminate CVD
prevention information to low health literacy populations within and among
health care delivery systems. A separate skill to also consider and is a
broader part of health literacy is numeracy the degree to which individuals
have the capacity to assess, process, interpret, communicate and act on
numerical health information.
This dissertation focuses on the assessment of health literacy and
numeracy skills and the experience of and preferences for receiving CVD
prevention information using a mixed methods approach in a sample of
patients with multiple CVD risk factors recruited from two health care
systems. The patient data are supplemented with information obtained by
education and care providers on their preferences and plans for relaying CVD
prevention information to high-risk patients. Recommendations for matching
patient/provider needs and preferences are made. This chapter provides
background information on health literacy, research to date, in the context of
the burden of cardiovascular disease. Lastly, the specific aims and the
structure of this dissertation are presented.

Structure of the Literature Review
The literature review is structured in such a way as to provide an
overview of health literacy, in terms of the development of the field, relevance
and importance to health and health outcomes, key current developments
and future directions. First, I discuss the history of the field of health literacy.
This includes a presentation of the various terminology and definitions in
current use. Second, I present the assessments used to evaluate individual
health literacy abilities. Third, I discuss the current areas of emphasis in
health literacy: printed health-related materials; use of the internet for health
information; the medical encounter; the impact of health literacy on health
outcomes, disease knowledge, and screening/preventive behaviors; and the
economic implications of health literacy. Fourth, I discuss the emerging areas
of emphasis in the field including system barriers for individuals with limited
health literacy, the oral exchange of information with patients, and health
numeracy. Lastly, health literacy issues in the context of CVO are discussed
since this dissertation focuses on individuals with multiple risk factors for the
History of Health Literacy
Interest in health literacy can be traced back to 1992 when the first
National Adult Literacy Survey (NALS) was conducted in the United States.
Approximately 25,000 adults ages sixteen and older were interviewed for the
study. The survey's focus was on functional literacy, defined as literacy skills
used in everyday activities. The NALS focused on an individual's abilities a)

to read and understand text, b) to locate and use information in documents,
and c) to apply arithmetic operations and use numerical information contained
in printed materials (Baker, 2006; Kirsch, Jungeblut, Jenkins, & Kolstad,
1993). The results from the study were bleak based on the studys findings;
it was estimated that roughly 90 million US adults 47% of the population -
had limited literacy skills (40-44 million Americans were functionally illiterate
at the time of survey) (Kirsch et al., 1993). A follow up assessment, National
Assessment of Adult Literacy (NAAL), was conducted almost ten years later
and the results were released in 2006. The NAAL focused on the three areas
of emphasis in the previous study but also included a health literacy
component. The health literacy component of the NAAL survey is the first-
ever national assessment designed specifically to measure adults ability to
use literacy skills to read and understand health-related information. Health-
related materials used in the assessment included medication information,
medical instructions, health insurance forms, and prevention and wellness
information (National Assessment of Adult Literacy, 2005a).
The NAAL surveyed 19,000 U.S. adults. The literacy items on the
NAAL were guided by the definition of health literacy used by the Institute of
Medicine and Healthy People 2010 (stated earlier on page 19 of this
dissertation) (Health and Human Services, 2000; Nielsen-Bohlman et al.,
2004). A little over half (53%) of the adults had an intermediate level of health
literacy. However, about 36% (which the authors extrapolated to about 93
million Americans based on the findings of the study) were at basic or below
basic levels (Kutner, Greenberg, Jin, & Paulsen, 2006). Factors related to
literacy include sociodemographic factors such as age, race/ethnicity,
language background, income level, and education level (National

Assessment of Adult Literacy, 2005b). The average literacy scores of
Hispanics fell on average 15 points from 1992 to 2003. Average literacy
scores among whites did not change significantly. Among blacks and
Asians/Pacific Islanders, average literacy scores increased (National
Assessment of Adult Literacy, 2005a). While Hispanics represent 12 percent
of the total NAAL survey population, 39 percent of the Hispanic adults were
classified as below basic literacy. In contrast, while 73% of the study
population was white, only 37% were classified as below basic (National
Assessment of Adult Literacy 2005a). Also, individuals 65 years of age and
older comprised 15% of the total study population while 26% were classified
as having below basic literacy levels.
The original 1992 NAAL study has influenced the vast majority of health
literacy studies conducted. Study results in the current literature have
identified the same demographic characteristics associated with low literacy
in the 1992 study to be similarly associated with limited health literacy:
education level (Kim, Love, Quistberg & Shea, 2004; Nielsen-Bohlman et al.,
2004; Van Servellen, Brown, Lombardi, & Herrera, 2003; Williams et al.,
1995), age (Chew, Bradley, Flum, Cornia, & Koepsell, 2004; Gazmararian et
al., 1999; Kim et al., 2004; Nielsen-bohlman et al., 2004; Schillinger et al.,
2002; Williams, Baker, Honig, Lee, & Nowlan, 1998; Williams, Baker, Parker
et al., 1998; Williams et al., 1995), ethnicity (Gazmararian et al., 1999;
Nielsen-Bohlman et al., 2004; Schillinger et al., 2002), and income level
(Gazmararian et al., 1999; Kalichman, Ramachandran, & Catz, 1999; Kim et
al., 2004; Nielsen-Bohlman et al., 2004). A few studies also listed
acculturation and English not being the primary language as additional
predisposing factors. These studies argue that the level of acculturation is a

predictor of utilization of health care services in the host country and may
have a significant impact on individuals abilities to learn and adapt to medical
regimens (Guerra, Krumholz, & Shea, 2005; Van Servellen et al., 2003).
Paasche-Orlow, Parker, Gazmararian, Nielsen-Bohlman, & Rudd (2005)
reviewed 85 publications regarding health literacy. While the methodologies
and data collection instruments varied, overall ethnicity/race (Hispanics and
African Americans) was associated with limited health literacy (p=0.003), as
was not completing high school (p=0.02), and older age (p=0.004) (Paasche-
Orlow et al., 2005). While the review of the literature did find statistically
significant associations between ethnicity/race and limited health literacy, the
association identified may be even greater in the general population as a
majority of the studies exclude participants who are not native English
Trends in the Literature
The health literacy literature has grown rapidly over the last two
decades. In the 1980s there were seven articles in the health literature that
focused on general literacy and health issues. In the 1990s, especially after
the results from the NALS were released, there was an increased focus on
health literacy and its impact and effect on disease knowledge and health
outcomes. In the first half of the decade over 100 articles on health literacy
were published. These articles mainly focused on patient medical forms and
health education materials. The second half of the decade saw a continued
growth in the number of articles published. These articles focused on the
development of health literacy assessments and the differences in disease
knowledge of patients with varying health literacy levels (Rudd, Moeykens, &

Colton, 1999). The current decade continues to see an increased number of
publications. While a majority of the articles continue to place an emphasis
on literacy levels related to health education materials, there is a noticeable
shift to documenting interventions designed to test various mechanisms and
technologies for communicating health information more effectively to
patients, describing an expansion in the number of health literacy
assessments available, increased recognition of numeric demands on the
patient, and a focus on the oral exchange of health information.
Definitions of Health Literacy
Health literacy and numeracy are becoming increasingly recognized as
areas of focus and emphasis in health services and public health research.
One challenging aspect of health literacy and numeracy is the lack of
standardized terminology. This can be partly explained by the relatively new
appearance of health literacy and numeracy on the health research agenda
and by the different perceptions of skills associated with literacy and
The definitions presented in this chapter reflect the terminology used in
the most frequently cited manuscripts on health literacy and numeracy, by the
Institute of Medicine, monographs, reports, and in funding announcements of
major federal funding agencies (i.e. Centers for Disease Control and
Prevention [CDC] and National Institutes of Health [NIH]). There is debate in
the field as to which definitions are the most encompassing in identifying all
the key components associated with health literacy and numeracy. To

identify where the field is currently in reaching consensus about terms, I have
created a list of definitions encountered during the course of conducting the
literature review for this dissertation. The definitions are presented in the
sections/pages that follow. For each definition, I have included the most
relevant reference so that the literature may be consulted for a more in-depth
discussion and description of its application.
In order for a field to prosper and thrive, a common language is
essential. As is often the case when numerous individuals with different
perspectives and backgrounds converge in the development of a field, health
literacy and numeracy research is characterized by inconsistent terminology.
The purpose of this section is not to advocate or argue for the superiority of
one term over another, but to facilitate discussion by compiling all the
commonly used terms in this field of research. A common language should
help to advance the scientific progress of the field by not only working from a
common language but to have the language help guide the advance of more
robust measures for assessment and comparison of findings, as well as
indentifying gaps in knowledge. Given the increased interest and financial
support from the NIH and CDC related to health literacy, it is anticipated that
the terminology for the field will be refined and expanded.
The American Medical Association (AMA) defines health literacy as a
constellation of skills, including the ability to perform basic reading and
numerical tasks required to function in the health care environment
(American Medical Association, 1999, p.553). This definition implies that
health literacy is the ability to apply basic reading, writing, and numeracy skills
to health-related materials and activities within a health care setting.

However, this definition fails to address the capacity of the individual to
identify, understand, and apply the information.
A broader definition is offered by Healthy People 2010 (Health and
Human Services, 2000) (first defined by Ratzan and Parker), the degree to
which individuals have the capacity to obtain, process, and understand basic
health information and services needed to make appropriate health decisions
(Nielsen-Bohlman et al., p.32). Health literacy in this context suggests that
individuals use a more complex level of thinking in order to act or make
informed decisions related to health care. Unlike the previous definition
offered for the concept of health literacy it means more than being able to
read pamphlets and make appointments. But rather incorporates the
processes of obtaining and applying health information to ones personal
health situation(s).
Finally, Nutbeam (1998) on behalf of the World Health Organization
and Zarcadoolas, Pleasant, & Greer (2005) offer definitions of health literacy
that are even broader in scope. The WHO definition builds upon the previous
definitions offered in that it moves beyond providing information to initiating
empowerment. The definition offered by Nutbeam (1998) in Health Promotion
Glossary created for the WHO is health literacy represents the cognitive and
social skills which determine the motivation and ability of individuals to gain
access to, understand and use information in ways which promote and
maintain good health (Nutbeam, 1998, p. 357). Zarcadoolas, et al. (2005, p.
196), believe the definition of health literacy to be the wide range of skills,
and competencies that people develop over their life course to seek out,
comprehend, evaluate, and use health information an concepts to make
informed choices, reduce health risks, and increase quality of life. These

definitions, while very broad in scope, imply the constant progress (movement
along a continuum) towards achievement of knowledge, personal skills,
confidence to take action to improve personal health by changing personal
lifestyles and living conditions, and to participate in the discourse related to
health, medicine, and scientific knowledge.
For the purposes of this study, I used the Ratzan and Parker (2005)
definition of health literacy for three reasons. First, it focuses on the individual
and the emphasis on the ability to interact with health care providers. Second,
it is the one that is generally agreed upon and cited most often. Third, this
definition allows a theoretical exploration of the individual experience
(phenomenology) in relation to obtaining health information (level of agency
over information seeking and obtaining), processing of health information
(perceived susceptibility, severity, threat, barriers, and benefits), and
understanding of health information in relation to personal and environmental
factors and behavior change.
Assessments Used to Evaluate Health Literacy
Assessments to test individuals understanding of printed health
information came about in the 1990s. Although health literacy is a complex
and multifaceted construct, the instruments available for assessing health
literacy abilities are predominantly composed of questions that evaluate
literacy skills using printed health-related materials. None of the instruments
to date test beyond the reading and comprehension of printed information.
See Table 1.1 for a summary of these instruments. The Rapid Estimate of
Adult Literacy (REALM) and the Test of Functional Health Literacy in adults

(TOFHLA) (Davis et al., 1993; Parker, Baker, Williams, & Nurss, 1995) were
the first two assessments developed and they vary considerably.
The REALM is a 66-item word recognition assessment. It consists of
common medical terms that patients might run across in the course of their
health care. This test does not assess comprehension. A patient receives
the list of words on a sheet of paper and is asked to read them aloud. S/he
receives one point for each word pronounced correctly to create the raw
score. The raw score is then converted in to a grade range estimate. The
REALM is not suitable to translate into Spanish as the language usually has
one sound associated with one letter (unlike English). If someone can
recognize letters in Spanish and the sound associated with that letter it would
be easier for lower-literate individuals to complete the test, thus
overestimating their literacy abilities (Davis, Michielutte, Askov, Williams, &
Weiss, 1998).
The TOFHLA, on the other hand, assesses reading comprehension
(50-items) and numeracy (17 items). The reading comprehension portion of
the test was developed using commonly used hospital texts such as
instructions for preparation for an upper gastrointestinal series, the patient
rights and responsibilities section of a Medicare application, and an
informed consent document. In addition, the questions in the reading
comprehension section utilize the modified Cloze procedure (Parker et al.,
1995). This procedure omits every 5th to 7th word from a sentence while
providing 4 options for the reader to select the correct response. The
numeracy section tests an individuals ability to understand directions for
taking a medication, keeping an appointment, ability to understand
instructions for monitoring blood glucose, and obtaining financial assistance.

The results of the TOFHLA are categorized as either being inadequate,
marginal, or adequate. The TOFHLA, while more comprehensive in nature,
does require more time to administer and score than the REALM assessment.
The TOFHLA is translated and validated in Spanish.
One criticism of these two instruments is the time required for
administration. Time constraints make it difficult to effectively administer
either of these instruments in the course of a routine encounter with the
healthcare system. In response to this criticism, both the REALM and the
TOFHLA underwent revisions to condense the assessments into shorter
formats that would take less time to administer in the busy health care setting.
The REALM-R and the S-TOFHLA were the products of these refinements
(Baker, Williams, Parker, Gazmararian, & Nurss, 1999; Bass, Wilson, &
Griffith, 2003). The S-TOFHLA is the most common assessment currently
used. It takes about 12 minutes to administer and consists of 4 numeracy
items and 36 reading comprehension items. It is also available and has been
validated for use in Spanish.
With the increased attention on health literacy, efforts continue to
create shorter screening instruments which might be used in a clinical
encounter and/or by a variety of clinical staff (Bass et al.,2003; Chew,
Bradley, & Boyko, 2004; Seligman et al., 2005; Weiss et al., 2005). While the
formal assessments listed above continue to be the gold standard for health
literacy assessment, more informal assessment methods are being
evaluated. One informal means to assess a patients literacy level is to ask
three questions (Chew et al., 2004). The three questions are: 1) how often do
you have problems learning about your medical condition because of difficulty
understanding written information, 2) how confident are you filling out medical

forms by yourself, an 3) how often do you have someone (like a family
member, friend, hospital/clinic worker, or caregiver) help you read hospital
materials. A few other studies investigated the ability to predict health literacy
abilities of patients with the use of one screening question; how confident are
you filling out medical forms by yourself? (Chew et al., 2008; Wallace,
Rogers, Roskos, Holiday, & Weiss, 2006). Based on initial assessment of
these three questions compared to the REALM and the S-TOFHLA, the
question how confident are you filling out medical forms by yourself? had
the largest area under the Receiver Operating Characteristic Curve (AUROC)
of 0.74 (95% Cl: 0.69-0.79) and 0.84 (95% Cl: 0.79-0.89) based on the S-
TOFHLA and REALM, respectively. AUROCs were lower for detecting
"inadequate or marginal" health literacy than for detecting inadequate health
literacy for each of the 3 questions (Chew et al., 2008). Both of these study
results showed promise in detecting limited health literacy abilities with this
one question. A third assessment created with the interest of time in mind is
the Newest Vital Sign (NVS) (Osborn et al., 2007; Weiss et al., 2005). The
NVS is a six-item literacy assessment structured around the activity of
reading and demonstrating how to interpret information from a nutritional food
label. The NVS demonstrated high sensitivity for detecting limited literacy
and moderate specificity (AUROC curve 0.71-0.73). The NVS was less
effective than the S-TOFHLA for predicting health outcomes (Osborn et al.,
2007). The assessment is brief and is available in English and Spanish.
As the focus and attention on health literacy continues to increase, so
does the overall number of assessments available to evaluate patients
abilities. In addition to shorter, more informal means of assessment, the
focus has also shifted to specific areas of content, such as dentistry (Gong et

al., 2007; Lee, Rozier, Lee, Bender, & Ruiz, 2007; Richman et al., 2007),
combining knowledge assessment with health literacy (diabetes and
electronic health resources) (Norman & Skinner, 2006; Rothman et al., 2005;
Yost et al., 2009) and tailoring of the assessments to various languages and
cultures (Baron-Epel, Balin, Daniely, & Eidelman, 2007; Ibrahim et al., 2008;
Lee, Bender, Ruiz, & Cho, 2006; Lee, Kang, Lee, & Hyun, 2009). The
greatest limitation regardless of which assessment is used to assess health
literacy abilities is that none are robust enough to comprehensively assess
health literacy they serve as approximations.
Currently the field is largely only focused (with a few exceptions) on
measuring the skills of the patient; not the person who is communicating the
information. The focus of change is not on the system. In addition, emphasis
has been on the individual encounter with the physician or health care
system. While this is important, this encounter is also limited in frequency
and duration. There is a need to shift the focus to a broader context, one that
might include home/family, community, prevention, social support, and public
health information management (i.e. the recent swine flu outbreak and the
information associated with it and how it was communicated to the general
public). The current literature on health literacy has maintained an almost
exclusive focus on patient literacy skills. I maintain that an important
consideration to facilitate improved care and self-management is to shift our
focus on health literacy to the study of provider and system level facilitators
and barriers to delivery of understandable and actionable health information.

Table 1.1. Comparison of original assessments created to assess adult health literacy abilities.
Full Name of Assessment Test of Functional Health Literacy in Adults Rapid Estimate of Adult Literacy in Medicine Short-Test of Functional Health Literacy in Adults Rapid Estimate of Adult Literacy in Medicine Revised Wide Range Achievement Test Revised
Administration Time 22 minutes 1 to 3 minutes 12 minutes 2 minutes 10 minutes
Content 3 prose reading passages (50-item reading comprehension section) and 17 numeracy items (how to read rx bottles, apptcard, etc) 66-word list of common medical and lay terms that are routinely used in primary care patient education materials 2 prose passages (36- item reading comprehension) and 4 numeracy items 8 item word list of medical terms -osteoporosis -allergic -jaundice -anemia -fatigue -directed -colitis -constipation Letter reading (naming 15 letters of the alphabet) and word reading (pronouncing 42 words) not specific to medicine

Table 1.1 (Cont).
Administration Method Patient reads the reading comprehension section and answers questions. For the numeracy section patients are given cue cards or bottles to read and answer questions about the items. Patient is given the long list of words and is asked to read them aloud while the examiner keeps score on a separate sheet of paper. Similar to TOFHLA. Patients read the reading comprehension section and answer questions. Then for the numeracy section they are given cue cards or prescription bottles. Patient is given list of words and is asked to read them aloud. Does not test comprehension. Patient is given list of words and asked to pronounce words and letters. Patient must continue to read aloud until they miss 10 consecutive items. Does not test comprehension & words are not from health context.
Formats available Valid in both English & Spanish. English Valid in both English & Spanish English English
References Parser, Baker, et al. J Gen Intern Med 1995;10(10):537-41. Davis, Long, et al. Fam Med 1993;25(6):391 -395. Baker, Williams, et al. PatEduc Couns. 1999;38 (1):33-42. Gazmararian et al. JAMA 1999. 281(6): 545-551. Bass, Wilson, et al. J Gen Intern Med 2003:18:1036- 1038. Davis, Michielutte, et al. Health Educ & Behav 1998;25(5):613- 624.

Current Areas of Emphases in Health Literacy Literature
Research to date on health literacy has largely focused on the
materials used to communicate health information to patients. These
educational materials include: handouts or brochures, graphic illustrations,
audiotapes and compact discs, DVDs, cartoon books, and interactive
applications such as computer kiosks, internet programs, and health
education classes. Although many different modes of communication are
available for health information, there has been a recent trend towards
technology-based modalities (DeWalt, Pignone et al., 2004; DeWalt &
Pignone, 2005: Gerber et al., 2005; Gustafson, Rovinson, Ansley, Adler, &
Brennan, 1999; Kim etal., 2001; Santo, Laizner, & Shohet, 2005).
Simultaneously, there is an overabundance of studies that have assessed
reading and literacy levels of materials currently available and a lack of
studies that have assessed what modes of communication are most favorable
to the patients (Cotugna, Vickery, & Carpenter-Haefele, 2005; Freda, 2005;
Kirksey, Harper, Thompson, & Pringle, 2004; Rudd, Moeykens, & Colton,
2000; Santo et al., 2005).
Printed Materials
Most medical and public health literature focused on health literacy
report on reviewing the readability levels of print materials used in health care
settings and for health promotion purposes. Some studies have assessed the
readability of materials targeted at specific diseases, such as CVD, cancer,
and diabetes, while others take a broader approach, examining a specific

type of material, such as patient package inserts or informed consents for
research studies (Cotugna et al., 2005; Coyne et al., 2003; Davis, Crouch,
Wills, Miller, & Abdehou, 1990; Doak & Doak, 1987; Freda, 2005; Goldstein,
Frasier, Curtis, Reid, & Kreher, 1996; Hunter, 2005; Kirksey et al., 2004;
Leichter, Nieman, Moore, Collins, Rhodes, 1981; Meade & Byrd, 1989;
Meade, Diekmann, & Thornhill, 1992; Overland, Hoskins, McGill, & Yue,
1993; Rudd, Comings, & Hyde, 2003; Smith & Adams, 1978; Spadero, 1983;
TenHave et al., 1997). Despite the many kinds of health-related materials that
have been analyzed for readability, a clear trend emerges from the literature:
there is a profound mismatch between the literacy skills of the public and the
literacy demands of the materials.
A study by Davis et al. (1990) assessed reading comprehension and
the readability of patient education materials. They found that 60% of the
participants were reading at least three grade levels below the grade they last
attended. They also found that of the 150 patient education materials
assessed, only nine were written below a 9th grade level. In addition, a few
studies have noted that patients who read at higher levels often prefer
information to be written at a lower grade level and presented in plain
language (Koo, Krass, & Aslani, 2003; Parker, 2000; Safeer & Keenan, 2005).
Current efforts have been focused on revising written educational materials
(Davis, Bocchini et al., 1996; Davis, Fredrickson et al., 1998; Eaton &
Holloway, 1980; Holmes-Rovner et al., 2005; Hwang, Tram, & Knarr, 2005;
Michielutte, Bahnson, Dignan, & Schroeder, 1992: Nurss et al, 1997;
Overland et al., 1993). While few studies have shown that simpler written
materials can improve knowledge, no published studies have shown a
beneficial effect on patients health outcomes from the simplified materials

(American Medical Association, 1999). Even though much effort has gone
into revising written materials, it is still unclear which is the best way to
communicate health information to patients (Pignone, DeWalt, Sheridan,
Berkman, & Lohr, 2005). Lastly, consistently absent from the literature were
information on the makeup and layout of the printed materials. Without
providing the actual documents used in the assessment it is challenging to
know exactly what is being measured.
The Internet
In addition to print materials commonly used in health care settings,
people are increasingly turning to electronic media to obtain health
information. The internet is pervasive, use continues to grow, and has great
potential for health promotion. It is estimated, as of December 2009, that 74%
of American adults (18 and older) use the internet, with over half of them
seeking health care/medical information (Rainie, 2010). According to Pew
Internet, searching for disease-specific information for medical purposes is
one of the primary reasons that individuals use internet technology (Madden
& Rainie, 2004).
The internet is available 24 hours a day, 7 days a week. This
translates to no waiting time for obtaining information and getting information
when it is convenient for the individual. In addition, individuals can search
anonymously. The internet can assist health consumers in decisions about
treatment, when to seek medical care, locating social support, and helping
prepare them to actively participate in their care (Rainie & Fox, 2000). The
internet also has the potential to affect health-related behaviors. One study

reported positive behavioral actions following internet health-related site
visits. These actions included asked a doctor about treatment (54%), made
treatment decisions (45%), and altered exercise/eating habits (46%)
(CyberCitizen, 2001). Given the popularity of the internet, technological
advances are progressing with the advent of e-health. Given the popularity of
the internet, one has to ask, does increased agency in seeking and obtaining
health information translate to increased comprehension or confusion?
Unfortunately, there is a negative side to the ever-increasing popularity
and reliance on the internet for health information similar to that of printed
health education materials: the readability levels and information
management demands are by and large too high for the general population
(Birru et al., 2004; Zarcadoolas, Blanco, Boyer, & Pleasant, 2002).
Furthermore, sites can be difficult to navigate, erroneous information is
abundant, and English dominates the entire forum. A study conducted by
Berland et al. (2001) evaluated health information on four medical conditions
available through English and Spanish websites. They found that the English
language websites were, on average, written at the college level, whereas the
Spanish websites had an average reading level of 10th grade. While wider
access to computer technologies is becoming more and more available
(United States Department of Commerce, 2000), thus narrowing the digital
divide, the online health information currently available is difficult for many
people to understand. The majority of the literature related to the internet and
health focuses on the number of Web resources and the quality of health
information available. More work is needed regarding the information
available on the internet and literacy level of the users.

Even though scant research has been published on the experiences
individuals with low health literacy have with these new technologies, several
barriers based on internet use have been identified that could exacerbate the
current inequalities in health and health care (Birru & Steinman, 2004;
Eysenback & Jadad, 2001). Barriers identified include: lack of access the
digital divide, lack of knowledge on how to use the internet effectively, lack of
knowledge on how to decipher credible from non-credible sources, and
inability to comprehend information provided (Birru et al, 2004; Eng et al.,
1998). It is important to consider the experiences of all health information
seekers, especially diverse, underserved populations in the formation of new
technology, so as to ensure that those who are at highest risk of preventable
or treatable health problems have adequate access to the necessary and
appropriate information.
The Medical Encounter
Having limited literacy skills may directly affect access to care. Baker
et al. (1996) conducted a focus group study to understand the difficulties that
patients with poor literacy have when interacting with the health care system.
The study found that individuals with poor literacy had problems navigating
the physical health care environment, completing necessary forms for care,
following medication instructions, and understanding information received
during the physician encounter.
In addition, individuals with limited literacy skills may be less informed
regarding the key components of the care plan. A study by Schillinger,
Bindman, Wang, Stewart, and Piette (2004) explored the association of health

literacy and the quality of patient-physician communication among diabetes
patients. Individuals with inadequate health literacy reported worse
understanding regarding physicians explanation of their condition,
explanations of processes of care, and considerations of patients desire
and/or ability to adhere to treatment plans.
A second study conducted by Schillinger et al. (2003) focused on
measuring the extent to which primary care physicians in a public hospital
assessed patient recall and comprehension of new concepts introduced in the
course of an outpatient, diabetic visit. Physicians rarely assessed patient
comprehension and recall during the course of the study. Changes in
medication regimens were the most prevalent concepts introduced in the
visits. While medication management is key to diabetes care, physicians in
the study only assessed comprehension 13% of the time. This study also
found that physician assessment of patient recall and comprehension was
associated with improved glycemic control. Patients with low health literacy
were more likely to have poor control of their diabetes, yet when physicians
assessed comprehension it was associated with good glycemic control
independent of patients health literacy level.
Health Outcomes and Disease Knowledge
The impact of limited literacy on the health care experience is being
documented in a growing body of literature including the United States
national health objectives Healthy People 2010. Health literacy was added
as a main objective (#11-2) to the health communication goals for Healthy
People 2010 (United States Department of Health and Human Services,

2000). In addition, health literacy indirectly impacts almost 40 other
objectives listed in the document. Health literacy has many individual and
population correlates such as age, language, race and ethnicity (culture),
education, socioeconomic status, and environment (access to care and
technology) (Pawlak, 2005). Health literacy can potentially be viewed as a
health disparities issue derived from these characteristics.
Research over the last several years has consistently shown that
people with limited health literacy skills have higher health care costs
(Howard, Gazmararian, & Parker, 2005; Parker, Ratzan, & Lurie, 2003; Weiss
et al., 1994; Weiss & Palmer, 2004), use health care services more frequently
(Baker et al, 2002; Baker, Parker, Williams, & Clark, 1998; Baker, Parker,
Williams, Clark, & Nurss, 1997; Barragan et al., 2005; Kalichman et al., 2000),
have poorer understanding of chronic disease management or ones
condition (Endres, Sharp, Haney, & Dooley, 2004; Gazmararian et al., 2003;
Kalichman et al., 1999; Kennen et al., 2005; Mancuso & Rincon, 2006; Pandit
et al., 2009; Schwartzberg, 2002; Williams, Baker, Honig et al., 1988;
Williams, Baker, Parker et al., 1998), have poorer health outcomes (Baker et
al., 1996; Mancuso & Rincon, 2006; Nielsen-Bohlman et al., 2004; Wolf,
Gazmararian, & Baker, 2005), under-use preventive health services (Bennett,
Chen, Soroui, & White, 2009; Guerra, Dominguez & Shea, 2005; Howard et
al., 2005; Scott et al., 2002), have poorer self-management skills (Chew,
Bradley, Flum et al., 2004; Kalicham et al., 1999; Schillinger et al., 2002;
Williams, Baker, Honig et al., 1998), and tend to be less knowledgeable about
health-related topics than people with adequate health literacy skills, (DeWalt,
Berkman, Sheridan, Lohr, & Pignone, 2004; Kalichman etal., 2000; Wolf et
al., 2005) even after adjusting for confounding variables such as

sociodemographics. A majority of these studies have focused on three
primary areas: chronic disease management and outcomes, patient-physician
encounters, and health education (DeWalt, Berkman et al., 2004; DeWalt,
Pignone et al., 2004; Pignone et al., 2005; Safeer & Keenan, 2005; Schillinger
et al., 2003). While many of these studies focused on the quality of patient
education and chronic disease management, few have looked at better ways
to communicate health information to patients with low health literacy.
Unfortunately, health-related decisions are not always easy. The
United States health care system is complex and very disjointed. Patients
must be able to access information, communicate with their healthcare
professionals about their illness/health, complete and sign forms, understand
treatment options presented to them, and follow through with treatment plans
given to them (Mika, Kelly, Price, Franquiz, & Villarreal, 2005). Patients are
constantly forced to make complex decisions on multiple levels. Having
sufficient and comprehensible information is necessary for patients to make
health-related choices that best suit them and their conditions (Janz,
Champion, & Strecher, 2002; Kim et al., 2004; Lorig & Holman, 2003;
Montano & Kasprzyk, 2002). The AMA Council of Scientific Affairs report,
raised concern that patients with the greatest health care needs may have the
least ability to read and comprehend important medical information (American
Medical Association, 1999). With the average doctors visit lasting only 16
minutes, it is crucial that patients medical needs are met in this limited time
frame (Blumenthal et al., 1999). For this to happen, patients need to be able
to read, comprehend, and understand information they receive in the health
care setting.

Baker and colleagues (1998 and 2002) conducted two studies
assessing functional health literacy and the risk of hospital admission. The
1998 study reported that individuals with inadequate functional health literacy
had a 52% higher risk of hospitalization (Baker et al., 1998). Given that this
study was confined to one hospital, they conducted a multi-site study in 2002.
This study was a prospective study conducted with 3,260 Medicare managed
care enrollees in four major US cities. Enrollees were contacted for
participation in the study three months after enrollment in the managed care
program. Health literacy was identified as being an independent risk factor
for hospital admission among elderly managed care enrollees. Hospitalization
data were obtained from claims from the managed care organization. A total
of 29.5% of the participants were hospitalized at least once during the follow-
up period. Individuals with inadequate health literacy were more likely than
individuals with adequate health literacy to be hospitalized (34.9% vs. 26.7%)
(Baker et al., 2002). Those with inadequate health literacy were also more
likely to be hospitalized two or more times than were those with adequate
health literacy (19.9% vs. 14.0%) (Baker et al., 2002). Further examination
of this Medicare population also found that there was no difference in
outpatient physician services utilization between the literacy groups.
However, individuals with low health literacy had higher utilization rates for
the emergency department. The authors questioned whether some of the
emergency department visits were a substitution for an office visit in this
population (Baker et al., 2004).
Furthermore, a study on health literacy and patient outcomes for
persons with Type 2 diabetes showed that low health literacy was
independently associated with poor glycemic control (Schillinger et al., 2002).

Twenty percent of individuals identified with low health literacy had tight
glycemic control versus 33% of patients with adequate health literacy. In
addition, low health literacy was also associated with higher prevalence rates
of self-reported diabetes complications, such as retinopathy. Lastly, a study
conducted by Williams, Baker, Parker et al. (1998), focused on patients with
type 2 diabetes from two public hospitals found that only half of the patients
with low health literacy knew the symptoms of hypoglycemia compared with
94% of patients with adequate health literacy. This same study assessed
health literacy and patients knowledge on hypertension. Results showed that
patients with poor literacy skills were less likely to answer knowledge
questions correctly compared to more literate patients (Williams, Baker,
Parker et al., 1998). One third of participants with inadequate health literacy
did not know that losing weight lowers blood pressure. In addition, 60% of
participants with low health literacy did not know that exercise lowers blood
pressure. Moreover, 45% of the low health literacy group did not know that a
blood pressure reading of 160/100 mm was high.
Screening and Preventive Behaviors
Von Wagner, Knight, Steptoe, and Wardle (2007) assessed the
relationship between functional health literacy and health-promoting
behaviors in a national sample of adults in the United Kingdom. Using the
TOFHLA for assessment of health literacy abilities, they found that a one
point increase on the scale increased the likelihood of eating five servings of
fruits and vegetables a day, being a non-smoker, and having good/excellent
self-rated health.

Miller, Brownlee, McCoy, and Pignone (2007) and Peterson, Dwyer,
Mulvaney, Dietrich, and Rothman (2007) both conducted studies that
assessed the association of health literacy with colorectal cancer screening
behavior. The Miller study recruited individuals from a large university-
affiliated community-based internal medicine practice. The Peterson study
recruited low-income individuals from a community health clinic in Nashville,
TN. The REALM was used to assess health literacy abilities in both studies in
addition to structured interviews to capture cancer screening behaviors and
colorectal cancer knowledge in terms of risks, screening, causes, treatments,
and cures. Individuals with low health literacy were less likely to be up to date
on colorectal cancer screenings. Both studies also found that individuals with
low health literacy had decreased knowledge about colorectal cancer and
reported more barriers to completing recommended screenings and tests.
Davis and colleagues assessed the association of health literacy and
mammography knowledge among low income women in Louisiana (Davis,
Arnold, et al., 1996). Using the REALM to assess health literacy abilities they
found that low health literacy was associated with less mammography
knowledge. Findings from this study could explain why low-income women
tend to be lower utilizers of mammography.
Economic Implications
Although the research that quantifies the financial burden of low health
literacy is limited, it is estimated that average annual health care costs of
persons with low health literacy may be four times greater than the general

population (Pawlak, 2005). Additional costs incurred by these individuals are
due to unnecessary visits, hospitalizations, and difficulty in following treatment
plans (Kleinbeck, 2005). Howard, et al., (2005) conducted a study that looked
at the medical charges of over 3000 Medicare recipients. Using economic
data from the Baker et al., (2002) study, they predicted that inpatient
spending for persons with limited health literacy was higher than persons with
adequate health literacy after controlling for health status (Howard et al.,
2005; Nielsen-Bohlman etal., 2004)(Howard, Gazmararian, & Parker, 2005;
Nielsen-Bohlman, Panzer, & Kindig, 2004). The results showed that patients
with limited health literacy incurred disproportionately higher medical costs
and utilized health services inefficiently (Howard et al., 2005). Given that the
causal relationships between literacy and health care utilization and costs
have yet to be identified, it is not possible to establish a valid cost figure for
the impact of low health literacy (Nielsen-Bohlman et al., 2004).
Emerging Areas of Study
System Barriers
The current United States healthcare system is organized for highly
educated individuals. There is a need to shift the current focus of
investigation in the field of health literacy from the patient to healthcare
system. As previously discussed, the majority of health information available
is inadequate for the majority of the healthcare system users. In addition, it
has been well documented that patients with limited health literacy have
challenges with navigating the healthcare system as well as understanding

common healthcare documents such as discharge instructions, consent
forms, insurance forms and applications, medication instructions, explanation
of health services and benefits, and advance directives (Baker, et al., 1996;
Ott & Hardie, 1997; Powers, 1988; Sudore et al., 2007; Sudore et al., 2006;
Williams, Counselman, & Caggiano, 1996).
While research on health literacy has identified several areas that are
problematic for patients, these areas consistently fall into two broader
categories that are slowly becoming the new priorities for study in this area.
These include making the healthcare system more patient-centered and
simplifying and standardizing access and utilization processes (Institute of
Medicine, 2009; Paasche-Orlow, Schillinger, Greene, & Wagner, 2006).
Patient-centered care includes tailoring health messages to patients (with the
emphasis being on limited literacy), prioritizing of collaborative goal-setting,
following up with patients after information is originally given to confirm initial
and continued understanding, and broadening the range of available self-
management strategies. Additionally, there is a great need to simplify and
standardize how patients access and use the healthcare system. This is
relevant across the entire continuum of use; from the point of entry in applying
for insurance or public assistance programs, to benefits and rights within the
system, to self-management of illness, to end-of-life decisions and options.
Using plain and clear communication accompanied by standardizing
processes across the system could potentially increase satisfaction with the
system and possibly increase understanding and retention of information for
individuals over time.

Oral Exchange
Another emerging area of emphasis in health literacy is focusing on the
oral exchange of information. This entails assessing what patients hear,
whether they are able to comprehend oral instructions and/or information, and
whether they have the oral skills to communicate or describe symptoms,
questions, or concerns. Most of the research in this area has centered on the
patient-physician encounter. Effective communication with patients cannot
simply rely on telling patients something or giving them a handout. The
physician needs to go beyond the relaying of information to making sure a
patient has understood what s/he has been told (Schartzber, Cowett,
VanGeest, & Wolf, 2007). The present literature has focused on a handful of
communication techniques to address understanding during the medical
encounter. These include: closing the loop, the teach-back method, teach to
goal, and Ask me 3 (Kripalani & Weiss, 2006; Oates & Paasche-Orlow,
2009; Pfizer, 2009; Schillinger et al., 2003). These methods can work in
tandem and have several similar key components: 1) an emphasis on
avoiding jargon, 2) limiting the amount of information discussed to two or
three items during an encounter, 3) being clear and specific regarding steps
or actions patient needs to take, 4) using more than one communication
modality to convey new information, 5) creating an environment that
encourages patients to ask questions, and 6) confirming comprehension.
Confirmation of comprehension is the desired endpoint. This can be done by
using the teach-back method, an example of which is the phrase tell me what
you will tell your family about what we discussed today. After the patient
responds, then feedback should be focused on areas that were not mastered
(Oates & Paasche-Orlow, 2009). The clinician would then need to go one

more step by "closing the loop which is to reassess comprehension and
provide additional feedback until mastery of the information is obtained
(Schillinger et al., 2003).
Health numeracy is defined as the degree to which individuals have
the capacity to assess, process, interpret, communicate, and act on numerical
quantitative, graphical, biostatistical, and probabilistic health information
needed to make effective health decisions (United States Department of
Health and Human Services, 2000). This definition not only emphasizes that
health numeracy is about processing and interpreting information but also
incorporates communicating and acting on numeric concepts pertaining to
health. To date, this is the most frequently cited definition; however, as with
health literacy, we lack general agreement on a standard definition (Peters,
Hibbard, Slovic, & Dieckmann, 2007; Rothman, Montori, CHerrington, &
Pignone, 2008).
In 1992, the NALS found that 26% of adults surveyed did not posses
basic quantitative skills or had extremely limited skills. Another 32% of the
respondents had basic skills consisting of performing simple one-step
arithmetic problems (Kirsch et al., 1993). Although health literacy and
numeracy are correlated, it has been found that individuals with adequate
health literacy may have poor numeracy abilities (Lipkus, Samsa, & Rimer,
2001). Several health-related tasks, such as understanding risk/benefit
information, understanding food labels, management of weight, portion-size
estimations, or interpreting blood sugar and/or blood pressure levels, all

depend on numeracy (Cavanaugh et al., 2088; Donelle, Arocha, & Hoffman-
Goetz, 2008; Huizinga, Beech, Cavanaugh, Elasy& Rothman, 2008; Huizinga
et al., 2009; Rothman et al., 2006; Schwartz, Woloshin, Black & Welch,
1997). Patients who are required to self-manage an illness and who have
poor numeracy abilities are at risk for having poorer health outcomes. One
way to help patients understand numerical information is to change the way
current numerical information is communicated. Risk information
communicated to patients can be confusing when presented as a single event
probability, a conditional probability, or as a relative risk. To minimize the
amount of confusion, the same information can be presented in a frequency
statement (i.e. 3 out of 10 patients have a reaction to this medication), natural
frequencies (i.e. eight out of every 1000 women have breast cancer; of these
eight women with breast cancer seven will have a positive result on
mammography), or absolute risk (i.e. in every 1000 women who undergo
screening one will be saved from dying of breast cancer) either independently
or in conjunction with relative risk data (Gigerenzer & Edwards, 2003).
The assessments currently available to evaluate numeracy abilities
pale in comparison to the number currently available to evaluate health
literacy abilities. The original assessments for numeracy were objective in
nature and focused on simple quantitative skills, such as calculating
percentages or proportions (Lipkus, et al., 2001; Schwartz et al., 1997).
Similar to the evolution of assessments focusing strictly on health literacy,
assessments on health numeracy are still evolving. The most recent
assessment has moved beyond measuring the ability to perform
mathematical operations and focuses on whether people can use numbers to
make comparisons for decision making. The Medical Data Interpretation test

assesses the ability to compare disease risk and risk reduction (Schwartz,
Woloshin, & Welch, 2005). The test is comprised of twenty questions focused
on direct-to-consumer prescription drug advertising, news media reports, and
statements a clinician might make to a patient. Additional numeracy
assessments include the numeracy component of the TOFHLA, the three-
item numeracy measure by Schwartz et al., (2005) and the 11-item numeracy
measure by Lipkus et al., (2001). Furthermore, there is a trend to develop
subjective numeracy assessments as opposed to objective assessments as
discussed above. Subjective numeracy consists of self-report of perceived
ability to perform numerical tasks and identification of preference for
numerical versus prose information (Fagerlin et al., 2007; Zikmund-Fisher,
Smith, Ubel, & Fagerlin, 2007). While the field continues to evolve, more
assessments of specific disease numeracy (i.e. diabetes numeracy) are being
developed along those related to general numeracy (Apter et al., 2006).
Numeracy will be an important area of focus as the field of health
literacy continues to evolve. As patients are asked to participate in the self-
management of their health conditions, they will be faced with the need to
understand numerical health information. For example, a health care provider
might tell a patient that s/he needs to lose 10% of their body weight. Would a
patient be able to understand what this means? A better approach could be
to have the provider calculate the exact number of pounds a patient needs to
lose and convey this number to him or her. Future directions for research
focused on numeracy include: the need to develop a unified and
encompassing definition of numeracy, the need to develop better ways to
assess numeracy skills, and identification of better ways to display and orally
communicate numerical and risk information to patients.

Cardiovascular Disease (CVD) in Relation to Health Literacy
This dissertation focuses on individuals with two or more risk factors
for cardiovascular disease (i.e. diabetes, smoking, hypertension,
hypercholesterimia, and BMI > 30). CVD is the leading cause of death and
disability in the United States. Nearly 2,600 Americans die each day from
CVD. This equates to 1 death every 34 seconds (American Heart
Association, 2004). The most recent statistics show that mortality from CVD
in adults (ages 18 and older) claimed roughly 630,000 lives in 2006 (Centers
for Disease Control and Prevention, 2008c). It is estimated for 2005 that
direct and indirect costs of CVD (not including stroke) were $336 billion
(Thom et al., 2006). In the past, CVD has largely affected adults 65 years of
age and older and was traditionally perceived to be an illness of males.
However, increases in mortality due to heart disease have been increasingly
noticed in younger populations and women over the last several years
(Brownson, Remington, & Davis, 1998; Centers for Disease Control and
Prevention, 2005). In addition, CVD is unequally distributed in America; it is
highest in those who are already disadvantaged by social structures of
inequality. For example, in 2001, the number of premature deaths (<65 years
of age) from CVD was greatest among American Indians or Alaska Natives at
36%, blacks at 31.5% compared to only 14.7% for whites (American Heart
Association, 2004). In addition, premature deaths, given a diagnosis of CVD,
were higher for Hispanics than non-Hispanics (23.5% versus 16.5%). CVD
has several well-known, modifiable risk factors including: high blood pressure
(systolic pressure of 140mm Hg or higher, or diastolic pressure of 90 mm Hg
or higher), elevated lipids (total blood cholesterol of 200mg/dL or higher),

cigarette smoking, physical inactivity, diabetes, and obesity (American Heart
Association, 2004; Brownson et al., 1998; Dawber, 1980; Marmot & Elliott,
2005). All of these modifiable risk factors depend on the ability to follow
health instructions, understand treatment plans, follow medication schedules,
etcetera. As documented in the previous sections of this chapter, individuals
with low health literacy and/or have limited numeracy skills have difficulty
understanding and managing health related tasks.
Primary prevention of CVD involves controlling or minimizing these
major risk factors. The prevalence of two or more risk factors for adults age
18 or older increased from 23.6% in 1991 to 27.9% in 1999 (American Heart
Association, 2004). The multi-factored etiology of CVD has resulted in a wide
variety of interventions. To date, the majority of these interventions have
placed emphasis on health education aimed at the individual. This can be
problematic for some patients as previously discussed, most patient
education materials are written at too high a level for most individuals to
understand. Moreover, many studies to date have shown that individuals with
diabetes and/or high blood pressure (two risk factors for CVD) and low health
literacy have higher rates of disease complications and decreased knowledge
about disease (Williams, Baker, Parker, et al.,1998). Furthermore, focusing
solely on the individual provides a limited amount of success in behavior
change. To be more effective in modifying behaviors, an understanding of
the context in which the individual operates (i.e. through phenomenology) and
identification of the structural issues that inhibit individual level change (i.e.
lack of agency and control of health information) are needed.

Summary of Existing Research and Gaps
Recent studies have supported the finding that lower health literacy
levels are clearly associated with poorer health outcomes. These studies
have also highlighted the fact that typical patient educational materials and
care practices are inadequate to overcome the barriers experienced by
individuals with limited health literacy. It is necessary to make improvements
in practice such as an increased emphasis on rigor in design and preparation
of materials and on clear oral communication skills. In addition, a shift in the
locus of responsibility from the patient to a shared responsibility with the
provider and the healthcare system needs to be made when studying the
communication of health information. Lastly, more comprehensive and
broadly applicable measures are needed to better assess health literacy and
numeracy abilities. Based on the current state of the field, I designed this
study to address patient needs in terms of health literacy and health
information preferences in an attempt to begin to shift the responsibility for all
involved in the healthcare process.
Overview of the Dissertation
Research Questions and Specific Aims
This dissertation focuses on the assessment of health literacy and
health numeracy skills of persons at risk for CVD. In addition, preferences for
receiving CVD prevention information were explored. Both quantitative and
qualitative methods were used. The following research questions and
specific aims are addressed in the dissertation:

1. What is the association between sociodemographic and health
characteristics and health literacy and numeracy abilities of adult patients with
multiple risk factors for cardiovascular disease in two different health care
Aim 1: Use electronic medical records and disease registries from Kaiser
Permanente Colorado (KPCO not for profit health maintenance
organization) and Denver Health and Hospital Authority (DHHA a public
safety net institution) to identify 500 English-speaking adults who have
multiple risk factors for cardiovascular disease (confirmed diagnoses of at
least two of the following: diabetes mellitus, hypertension,
hypercholesterolemia, smoker, and BMI >30) but no known diagnosis of
coronary heart disease (CHD).
Aim 2: Administer the Test of Functional Health Literacy in Adults short
form and numeracy survey to 150 adults (75 from KPCO and 75 from
DHHA) with multiple risk factors for CVD.
Aim 3: Survey study participants regarding sociodemographic and health
characteristics, disease knowledge, access to and ease of use of
information outlets for prevention information, and self-efficacy regarding
obtaining and successfully using prevention information.

2. What are patients attitudes and preferences toward current methods of
obtaining and receiving CVD prevention information? Do they vary by health
literacy and numeracy?
Aim 4: Conduct in-depth, semi-structured interviews with a sub-set of
adults from the study sample to further explore participants experiences
related to receiving and understanding health information, preferred
modalities for health communication, and experiences with accessing
health information.
3. What are a) the current methods used by health care professionals to
distribute and provide CVD health information and b) what are their
preferences for dissemination of health information?
Aim 5: Identify health education specialists who provide information on
CVD prevention and CVD in general at both KPCO and DHHA.
Aim 6: Conduct in-depth, semi-structured interviews with a sample of
health education specialists at KPCO and DHHA regarding their opinions
on effective methods for communicating CVD health information, current
methods used in communicating health information, methods preferred for
delivering health information, potential benefits and barriers to the use of
technologically-oriented health communication strategies, and methods
used to gauge patients understanding of information received.

4. Integrating the above information, are there recommendations for ways to
enhance CVD health information to make it more appealing and useful for
patients with low health literacy?
Aim 7: Analyze the data from Aims 1-6 using statistical analyses,
analytical induction, and triangulation to assess similarities and variations
of themes, interpretations, and categories between patient, health
educator responses, and between health care systems.
Aim 8: Develop recommendations for each health care organization and
share results via written reports, oral presentations, bullet point
conclusions and recommendations.
Structure of the Dissertation
The preceding sections of this chapter have outlined the history of the
field of health literacy and its impact on health outcomes. While the field is
still growing and evolving rapidly, the focus needs to shift from documenting
the readability of patient education materials to the development of more
comprehensive measures to assess health literacy abilities. In addition, the
burden of understanding must shift from the patient to the healthcare system
and providers with the focus on communication of information. The summary
of the current state of the field and the discussion of research gaps led to the
four research questions and eight specific aims that are addressed in
Chapters 4 through 6.

Chapter 2 discusses theoretical perspectives in conjunction with the
definition of health literacy to understand the structural and individual factors
that may impede a person from becoming an active participant in their health
care. Individual level theories discussed include the Health Belief Model and
Social Cognitive Theory. The structural level theory used to guide this
dissertation is Michel Foucaults work on knowledge and power. Lastly,
phenomenology was used to understand the patients lived experience as
well as to guide the qualitative interviews for this dissertation. Chapter 3
describes the design of the study as well as the quantitative and qualitative
methods used. Chapter 4 describes the recruitment methods used for this
study and the impact using various methods had on the project. Chapter 5
describes the health literacy and numeracy abilities of the study participants
and summarizes sociodemographic, health characteristics, and CVD
knowledge. Chapter 6 provides a summary of results and explores patient
preferences for receiving health information, and compares these preferences
to current health education methods at both heath care organizations.
Finally, Chapter 7 summarizes the key findings and conclusions from the
previous chapters and discusses how they addressed the previously
described research gaps and what their contribution is to the current field of
health literacy. The dissertation concludes with the presentation of
recommendations for future research and implications of these

Traditional theories of health behavior have focused on effective health
education efforts that may translate into desired or recommended behavior
change(s). Research over the last decade has identified health literacy as an
important issue to consider when developing health education programs and
materials. For health education to truly be effective, it should be designed
with an understanding of the individual and his/her health, social
characteristics, beliefs, norms, current skills, and the environment (Glanz,
Rimer, & Viswanath, 2008; Institute of Medicine, 2002).
Current theories of individual behavior change consider a multitude of
demographic characteristics; including education and income. These data
are important to collect; however, they do not provide enough information to
create effective health education and ultimately health-related behavior
change. The concepts and constructs of health literacy and health numeracy
are not explicitly considered in any of the individual behavior change
theoretical models. While individual educational attainment is routinely
collected; health literacy is more nuanced and focused. Health literacy and
numeracy are assessed with increasing frequency, yet this information is
typically lumped in with demographic characteristics when in fact it should be
more prominent in the explanation of health behavior change. Another
limitation of current behavior change theories is they often assume an
adequate level of health literacy and numeracy abilities of individuals. When
these theories are applied by health education specialists, interventions tend
to focus on increasing an individuals self-efficacy level or providing

information on barriers and/or risk, with little focus on whether or not a person
understands the information provided by the intervention.
Income level, usually assessed as the annual household income, is
also important, however; assessing ones income level without collecting
additional information is very limiting in helping to explain health behavior
change. Additional information that is related to socioeconomic status and
income such computer access, internet access, and internet use are
additional data that would be useful in understanding whether or not an
individual engages in recommended behavior changes.
The last decade has seen massive changes in the way people obtain
and receive information, especially health information. The popularity of the
internet has grown in the last two decades and seeking health information is
one of the most popular uses (CyberCitizen, 2001). While communication
technologies show great promise, considerable obstacles in using technology
to improve health have been identified and should be considered in health
behavior change theories. The digital divide, while thought to be narrowing,
continues to be a concern of access and equity for diverse populations (Pew
Internet, 2009). In addition, content-barriers (such as literacy barriers,
language barriers, quality of content, and lack of cultural suitability) continue
to be problematic (Institute of Medicine, 2002).
Another consideration missing from these theories is that there is no
focus or discussion on how individuals prefers to receive health information.
Health education covers a broad continuum ranging from disease prevention,
to detection, to treatment, and to a lifetime of self-management or long-term
care. In addition, health education is conducted in a variety of settings and
through diverse mechanisms. Contemporary health education efforts

emphasize the need to understand the target audience in order to be
effective. Characterization of the target audiences typically includes:
sociodemographic characteristics, life cycle stage, and disease or at-risk
status (Glanz et al., 2008). The current theoretical models assume that if
information is targeted and shared, regardless of format and/or literacy level,
an individual will engage with the information and apply it as directed. Given
that health education is routinely conducted with limited resources, a one-size
fits all approach is typically used. While this way of conducting health
education is efficient and contains cost, it may not reach individuals who need
and could benefit most from the information.
The current investigation used several theoretical perspectives to
create a conceptual model that integrates structural and individual factors that
may impact an individuals ability to locate, understand, and apply health
information related to the prevention of cardiovascular disease (see Figure
2.1). Included in this study are theories of individual behavior, i.e. the Health
Belief Model (HBM) and Social Cognitive Theory (SCT); theory of post-
structural-functionalism, i.e. Foucaults theory of power and knowledge. In
addition, phenomenology theory was used to guide the qualitative portion of
the study as well as to gain a better understanding of the patient experience
with accessing and using health information.


Individual Level Theories
Health Belief Model (HBM)
The HBM was first developed in the 1950s by individuals from the United
States Public Health Service to explain poor participation by the public in
health screening programs (Janz et al., 2002). In the 1970s, Kirscht and
Becker elaborated on the model to include an individuals response to
symptoms and diagnosis of illness (Janz et al., 2002). The HBM has been
used widely to predict and/or understand why people accept preventive health
services and why they do or do not adhere to other kinds of health care
recommendations and regimens. Since its creation, the HBM has been
applied to a broad range of health behaviors. The most prevalent areas of
application for the HBM are health-promotion (i.e. diet, exercise, cancer
screenings) and health decisions based on risk (i.e. smoking and condom
use) (Abood, Black, & Feral, 2003; Byrd, Peterson, Chavez, & Heckert, 2004;
Chew, Palmer, & Kim, 1998; Doshi, Patrick, Sallis, & Calfas, 2003; Hay et al.,
2003; Hounton, Carabin, & Henderson, 2005; Janz et al., 2002; Janz, Wren,
Schottenfeld, & Guire, 2003; Juniper, Oman, Hamm, & Kerby, 2004; Kaufert,
Rabkin, Syrotuik, Boyko, & Shane, 1986; Lollis, Johnson, & Antoni, 1997).
The HBM follows an individual agency perspective the capacity of
individuals to act independently and to make their own choices. The HBM
assumes an individual will take a health-related action if s/he believes that a
negative health condition can be avoided, has a positive expectation that by
taking a recommended action, s/he will avoid a negative health condition, and
believes that s/he can successfully take a recommended health action (Janz
et al., 2002).

Six main constructs comprise the revised HBM: perceived
susceptibility, perceived severity, perceived benefits, perceived barriers, cues
to action, and self-efficacy. Perceived susceptibility refers to an individuals
perception of the risk of becoming ill (or re-contracting an illness). Perceived
severity refers to an individuals perceptions about the seriousness of the
illness, if contracted. It not only includes the medical consequences (such as
disability or pain) but also the social consequences (such as burdens to family
and work life). Perceived threat is the combination of susceptibility and
severity. Perceived benefits refer to an individuals beliefs) that taking an
advised action will produce an effective outcome, i.e. risk is decreased,
money will be saved, or conditions/diseases are avoided. Perceived barriers
refer to an individuals belief of the potential negative aspects of the
requested action might prohibit him/her from performing the action.
Perceived benefits and barriers as described in this model can also influence
the use of health information. With respect to the current study, benefits of
locating and using health information can range from a better understanding
of the risks for developing CVO to better awareness of CVD in general;
barriers can range from comprehension difficulties due to limited health
literacy, lack of access, or perceived relevance of health information.
Cues to action refer to mechanisms that trigger an individual to act, i.e.
a media campaign focused on increasing colon cancer screening/awareness
or symptoms experienced by the individual. Self-efficacy refers to an
individuals sense of confidence that he or she can successfully perform the
requested action (Janz et al., 2002). Hence, high perceived self-efficacy in
persons at risk for chronic illness might reduce feelings of apprehension and
helplessness that could inhibit modifications to diet and exercise. However,

low perceived self-efficacy could be detrimental to patient health if s/he does
not have confidence in talking to health care providers or is unsure when to
seek help.
In relation to this study, perceived threat (susceptibility in combination
with severity) and self-efficacy stand out as key components possibly affected
by limited health literacy. Individuals with limited health literacy are at greater
risk of having a false perception of threat due to the challenges associated
with understanding the health information they have received (Baker, et al.,
2002; Gazmararian et al., 2003; Williams, Baker, Honig, et al., 1998; Williams,
Baker, Parker, et al., 1998). For example, a person with limited health literacy
may be diagnosed with diabetes and is then given a multitude of complex
instructions that must be followed. Unfortunately, this person may not fully
understand the short and long-term consequences associated with not
checking their blood sugar level several times per day, the need for diet
modification, and the effects of exercise on blood sugar levels. This individual
only follows the instructions that are easy to understand and perceives that
the severity of the disease has been decreased when it reality it has not.
In addition, self-efficacy, while situation and task-specific, is influenced
by the knowledge an individual has in relation to the situation and/or task.
Basic knowledge and skills are important. For instance, if a patient with
limited health literacy does not understand that weight loss occurs over time
(not an overnight event based on one physical activity encounter) and steps
on the scale the following morning to see no change in weight, this person
might give up on exercise after only a few attempts. Constructs from the
HBM were used in the survey portion of this study to explore the capacity to
understand portion of the health literacy definition by capturing basic CVD

knowledge, beliefs, barriers, perceived severity and susceptibility to CVD, and
self-efficacy of individuals with varying degrees of health literacy. Once an
individual receives information, do they understand it? How does it impact
their perceptions in relation to perceived susceptibility, or perceived
benefits/barriers to perform a requested action?
Trying to get individuals to change or adopt a desired health behavior
by knowing how they rate their severity, susceptibility, benefits, barriers, and
self-efficacy is just the first of many steps to understanding why they adopt (or
do not adopt) a health behavior. Missing from the HBM is the answer to the
questions why do these individuals have a particular perceived susceptibility
or severity? Do they understand the information they have been given? What
is the context and agency these individuals are operating within and how
does that affect their self-efficacy? By solely focusing on the individual the
HBM fails to identify and understand the circumstances in which the individual
operates. Without a true understanding of an individuals context, educational
information and materials have limited effectiveness in promoting behavior
change. Due to the main limitation of the HBM (focus on the individual), this
study also incorporated an additional individual level theory as well as a post-
modern theory and a phenomenological perspective to identify the context of
the individuals and its effect on the individual and his/her health behaviors.
Social Cognitive Theory (SCT)
Albert Bandura (1977, 1986, 1997) developed his theory of social
learning in the 1960s which later evolved into what is currently known as
Social Cognitive Theory (SCT). Similar to the HBM, SCT follows a mostly

agency perspective but takes into account some structural effects. SCT
assumes human beings act with intention and forethought and are capable of
adjusting behavior based on social and environmental feedback as well as
personal reflection (Glanz et al., 2008). Reciprocal determinism between
personal factors, environmental factors, and behavior is a key construct of
SCT (Bandura, 1986; Bandura, 1997). An individual receives cues from the
environment, processes these cues cognitively, and then changes behavior
(either positively or negatively) due to this interaction of cognition and
environment. People and their behaviors are shaped by their environment,
while at the same time people can also influence their environment through
behavior and expectations.
Observational learning, self-regulation, and self-efficacy are key
components influencing behavior in this model (Bandura, 1986; Bandura,
1997). Observational learning occurs when a person watches the actions of
another person and is able to act in a similar manner to what was observed.
Self-regulation is the deliberate monitoring of ones own behavior and
evaluation of how this behavior affects ones health. If the desired effect is
realized, the individual reinforces the effect by continuing the behavior.
Lastly, as in HBM, self-efficacy, which originated in SCT and has since been
added to many theories including HBM, is the belief in ones capabilities to
produce desired results by ones actions (Bandura, 1997).
While SCT tends to focus on the individual (personal factors), it does
incorporate a contextual component (environmental factors) in an attempt to
understand why an individual engages in a particular health behavior. The
social cognitive perspective on health behavior change is that individual
behavior change can be facilitated by modifying individuals personal factors

(such as knowledge, skills, self-efficacy, and outcome expectations) and by
modifying environmental factors (i.e. social and institutional) to encourage
healthful behavior.
Within the framework of SCT, the concepts of self-efficacy, outcome
expectations (or perceived benefits in HBM terms), and reciprocal
determinism stand out as key components affected by limited health literacy.
Individuals with limited health literacy are at greater risk of having decreased
knowledge, skills, and self-efficacy (personal factors) as discussed previously
in terms of HBM (Baker, et al., 2002; Gazmararian et al., 2003; Williams,
Baker, Honig et al., 1998; Williams, Baker, Parker et al., 1998). In SCT,
knowledge is a necessary precondition for behavior change. Before behavior
change can occur, individuals must have knowledge about both their risk
factors (conditions that put them at risk for developing disease) and the ways
in which their risk factors can be reduced (substitute behaviors). Without
adequate knowledge, people are unlikely to acquire skills and/or have the
self-efficacy needed to engage in behavior change.
While knowledge is important, the influences of environmental factors
must also be considered. What information is the individual receiving from
his/her health care organization (institutional)? What other sources are
individuals accessing to obtain health information (family/friends, electronic or
print media, etc)? Are they taking a more passive role and waiting for the
information to come to them? How the individual interacts (or not) with these
resources also affects behavior change. In terms of promoting health literacy,
health care professionals could apply constructs from SCT to assist their
patients in better understanding health information. For example, an
individual with limited health literacy most likely will have low self-efficacy

regarding making recommended diet changes necessary to decrease weight
and BMI because of difficulty understanding the information given about the
need for diet changes. Considering the concept of reciprocal interaction
between personal and environmental factors, health professionals/educators
may want to ensure that diet instructions are presented and shared within a
supportive environment (such as one-on-one lessons, small classes, or with
family). Verbal encouragement, praise, and advice, along with opportunities
for repeated accomplishments, would serve as external reinforcements and
thus build the individuals self-efficacy beliefs (Poor, Drummond, & Richter,
Constructs from SCT were used in the survey portion of this study to
explore both the capacity to understand and the capacity to obtain portions
of the health literacy definition. Personal factors such as knowledge and skills
were used to investigate understanding. Self-efficacy (personal factor) along
with abilities to extract CVD related health information from healthcare
professionals and other sources of information along with environmental
factors such as appraisal of the health information environment and resources
available related to CVD health information were also explored.
Trying to get individuals to change or adopt a desired health behavior
by focusing on personal factors in addition to some structural effects in the
way of environmental factors in connection with constructs from the HBM
continues to add to the understanding of why an individual would adopt a
health behavior. However, since SCT only partially explains environmental
effects on an individual, this study also incorporated a post-modem theory to
identify the context of the individual and its effect on the individual and health

Structural Level Theories
Foucault Knowledga/Power
Michel Foucaults (1973, 1975) work has an emphasis on knowledge
and power, how they are intertwined, and how the relationship between the
two are used to regulate bodies'. Foucaults work regarding biomedicine
provides a useful conceptual starting point for understanding key changes to
medicine and clinician-patient relationships. Foucault was also concerned
with processes whereby the human is turned into the subject of scientific
investigation and control. In The Birth of the Clinic: An Archaeology of
Medical Perception, Foucault outlines how medical standards of normalcy
spread into the human sciences individualizing the self, producing
subjectivity, and at the same time, provides the basis for social control
(Foucault, 1973). He illustrates how the medical system is set up to not see
patients as individuals but as pathologies. In addition, he demonstrates how
this knowledge is then turned into power by prescribing to the patient how
they should react in relation to the illness. Foucault uses the doctor-patient
relationship to illustrate how power flows through social relationships.
Moreover, Foucault outlined the various changes in the medical field
especially the transition from the observational gaze to the clinical gaze.
This transition saw the ability of the physicians move from indirectly observing
a body through the patients descriptions of what they were experiencing to a
more direct observation through the use of new medical technology (at the
time, reference was made to the stethoscope) and the physical exam
(Foucault, 1973). The gaze was seen as additional knowledge. The gaze

supplemented what physicians had learned through textbooks (Ritzer &
Goodman, 2004). Foucault attributes the gaze to the change in Western
medicine (White, 2002). The patient had become the object of scientific
Disciplines such as medicine began to use knowledge to define who
has or does not have the authority to make decisions (or diagnoses), in what
way information should be gathered, and by whom. During this change in the
practice of medicine, the rules and character of the discourse had also
changed in relation to how diseases were named and categorized (Foucault,
1973). Medical professionals began to define categories (labeling) of people -
deviant, sick, mentally ill, etc. Normalizing judgments were also part of the
new way of categorizing diseases and the continual analysis of whether one
deviates in any way from normality. These judgments tended to be
dichotomous in nature instead of evaluated along a continuum. An individual
is considered compliant or non-compliant, well or ill. These judgments put
limits on behavior as to what was considered acceptable. Medicine became
(i.e. physicians) an agent of social control.
When Foucault originally wrote about professional power in the
medical field and the control of information flow to the patient it was in
reference to the patient-clinician encounter. He believed that the physician
(expert) could not possibly communicate all the knowledge relevant to a
patients condition during the course of the medical encounter because the
sciences have a high degree of specialized knowledge. Traditionally, doctors
had a monopoly on the medical knowledge and a monopoly concerning the
definition of diagnosis and treatment decisions. He wrote about the uneven
relationship (hierarchical) between provider and patient in the health care

system. The patients were seen as passive agents in the system. Is this still
the case with the onset of chronic illness? Current health education
philosophy says the more information the chronically-ill patient has the better
s/he is able to manage her/his illness, evaluate options available, and
participate in the decision process (i.e. self-management). The advent of the
internet eliminated many physical barriers related to accessing information,
including health information. Now, patients can search the web to find an
overwhelming amount of information on any given topic in other words, they
have the power to receive knowledge traditionally available only to
physicians. So, one could argue that the patients possibility to obtain
accurate medical information about his/her disease and possible treatments
perhaps gives the patients the opportunity for greater participation in the
decision making process in the health care system now making them more
active agents in their care. Does agency over seeking and obtaining
information translate in to understanding and knowledge?
According to Foucault, the individual is not an autonomous agent, but
rather a social construct (Gordon, 1980). He argued that society dictates the
values and norms by which we live. He also argued that power relies on the
constant supervision and control of individuals in concurrence with what is
considered normal. Even when an individual appears to be self-regulating
their behavior to live in concordance with the concepts of normality, they are
in reality regulating themselves within a large discipline or system of power.
Foucault used these arguments to emphasize that individuals can only come
into being as a construct of power and knowledge and that no individual can
truly be an autonomous agent. While individuals cannot have autonomy they
can have agency. Foucault argues that while even individuals do have

agency, it is limited to a degree. Agency is the combination of the
experiences a person has, the way they use reason, how they make
decisions, and the beliefs they hold. While there are differences between
people in all these actions, they are still occurring within the social context
define by a discipline or system of power.
Foucaults discussions of power were focused on various disciplines.
One discipline of specific focus was medicine, mainiy the patient-provider
relationship and the structure and knowledge the medical community created.
While limited health literacy and numeracy are associated with the field of
medicine, it is enmeshed in social structures. Structures of a society create
the individual. In terms of health literacy and numeracy, we must
acknowledge that while the interactions with a physician can perpetuate the
cycle of limited health literacy, health literacy and numeracy are fundamental
manifestations of structural inequalities, mainly at the level of the educational
system, the economic system, and the health care system. These current
systems are set up in such a way that they continually create and recreate the
conditions that foster limited health literacy.
The educational system in the U.S. routinely suffers from budgetary
cutbacks. Due to budgetary constraints, health education has been removed
or offered in limited capacity to students. In addition, students continue to
progress to the next grade level when in fact they should be held back.
Results from the NAAL showed that over three-quarters of adults with less
than a high school degree were below basic health literacy level or at basic
health literacy level; that percentage decreased as education level increased.
While health literacy increased with higher educational attainment, 44 percent
of high school graduates and 12 percent of college graduates had below

basic or basic health literacy (National Assessment of Adult Literacy, 2005a).
Given that education is associated with literacy and health literacy, an
individual who receives a poor education, is considerably more likely to have
limited health literacy and numeracy skills.
In the United States, health insurance is tied to employment. Given
the present economic environment of this country, a national unemployment
rate of almost 10%, millions of Americans are currently without a job and
without health insurance. Results from the NAAL identified that uninsured
adults and those enrolled in Medicare and Medicaid were more likely to be at
the below basic or basic level of health literacy than those adults who
received insurance from an employer. Whereas about one-fourth of those
with employment-based health insurance were in the below basic or basic
health literacy group, well over half of uninsured persons, Medicare
beneficiaries, and Medicaid beneficiaries were in these groups (National
Assessment of Adult Literacy, 2005a). Not having health insurance also limits
the frequency and consistency with which an individual interacts with the
health care system.
The lack of stable coverage and consistent healthcare access are significantly
associated with low health literacy, as both problems keep people from
learning to use health care appropriately while trying to meet their own health
The health care system in the U.S. is complex. Unneeded complexity
remains commonplace and is problematic for individuals with limited health
literacy. At present there is no consistent way to interact with the system.
Each clinic, hospital, and health insurance plan has their own way to
determine eligibility, provide services, establish which benefits will be

covered, and payment processes. There is also no consistent way in which
people learn to interact with the health care system. There is an urgent need
to address the gap between the policies and procedures of the system and
the skills people have to understand and use this information to interact with
the system and to make health-related decisions.
For the purposes of this study, Foucaults knowledge and power was
examined in the context of health literacy via survey and in-depth interviews.
We examined the amount of information participants receive from a health
professional during a health encounter, the ability of participants to obtain
information independently of the health encounter to inform health-related
decisions, and the ability of participants to understand the information
obtained to maintain health and overcome illness.
Qualitative Theory
Qualitative research and methodology has a long and distinguished
history in the social sciences. In sociology, the work of the "Chicago School"
in the 1920s and 1930s established the importance of qualitative research for
the study of human group life (Patton, 2002). During the same period,
Anthropology introduced the fieldwork method (ethnography), in which the
observer went to a foreign setting to study and participate in (participant
observation) the customs and habits of another society and culture (Agar,
1996; Trotter & Schensul, 1998). Soon, qualitative research was employed in
other social science disciplines such as linguistics, education, social work,

communications, and the management and organizational sciences (Patton,
Qualitative methods are a detailed description of situations, events,
people, and behaviors. It includes what people say about their experiences,
attitudes, beliefs, and thoughts through recordings, documents, transcripts,
records, and narrative histories (Berg, 2004; Patton, 2002). Qualitative data
sources include observation and participant observation, interviews, texts,
and the researcher's notes of impressions and reactions. Qualitative methods
are typically open-ended and do not impose outsider predetermined,
categories such as the response choices that comprise typical questionnaires
or tests (an etic approach) (Agar, 1996; Berg, 2004; Patton, 2002). Instead
qualitative methods begin with specific observations and build towards
general observations and explorations as identified by patterns illustrated by
the people being studied (an emic approach) (Agar, 1996; Berg, 2004; Patton,
Phenomenology focuses on exploring how people make sense of
experiences and then transform these experiences into awareness (Patton,
2002; Trotter & Schensul, 1998). A phenomenological approach is best used
with narrative research, where an investigator uses the words of individuals to
understand, for example how patients experience a phenomenon like a health
event or problem (Patton, 2002). This theoretical perspective is designed to
capture the lived experience. Phenomenology was used to obtain a better
understanding of the individual experience with locating and using health
information. In order to understand the lived experience one must understand
how individuals describe, feel, perceive, make sense of, and talk about the
phenomenon of interest in the case of this study, phenomenology was used

to explore the full health literacy definition the capacity to obtain, process,
and understand basic health information and services needed to make
appropriate health decisions" (Patton, 2002). How do individuals prefer to
obtain health information and are they able to use information to make
decisions and does it translate into behavior change?
In the context of this dissertation, individuals of various health literacy
and numeracy abilities look for and attempt to apply information gained from
health information. The emphasis is placed on how people experience this
phenomenon, obtaining and using health information, instead of on the
phenomena itself. The thematic field may include different aspects of this
situation as described by individuals which can include personal factors,
impact of economic, educational, and/or health care systems, knowledge,
environmental influence, social support, perceived severity or susceptibility to
a health condition, or general ideas about health information. Each
individuals recount of his/her experience will depend on the parts of the
phenomena of which they are aware. Discussing the lived experience is
necessary for gaining an understanding of all the potential factors that impact
whether or not someone understands health information and whether this
varies by health literacy or numeracy ability. With this information we can
then begin to identify better ways to provide individuals health information
they can truly use.
When exploring the experiences of patients with limited health literacy,
one cannot just look at the patient, one must look at all the inputs and factors
at all levels outside the individual (Paasche-Orlow & Wolf, 2007); thus
combining all of the previously stated theoretical approaches with the lived
experiences of individuals related to receiving and using health information

we can begin to understand how people locate health information, their
preferences for health information, and whether or not they understand the
information received.

This chapter is an overview of the study design and methods used in
this dissertation to provide a context for the three results chapters that follow.
Study population, data sources, study instruments, data collection techniques,
and analytic plans are presented. In this research project, I used both
quantitative and qualitative methods to explore and understand patients and
health educators preferences and attitudes toward the delivery of
cardiovascular disease prevention health information. In addition, patient
preferences for health information were compared with health literacy and
numeracy to assess if there were any differences based on abilities or skills. I
implemented a sequential, two-phase design, starting with a patient-level
quantitative survey followed by semi-structured, in-depth interviews with
patients and health educators (see Figure 3.1) (Tashakkori &Teddlie, 1998;
Tashakkori & Teddlie, 2003).

QUAN qual
Dk7-0 Interviews
with subset
* Interviews

Association of Sociodemographics and Health Characteristics
and Health Literacy and Numeracy Abilities
Study Population
The study population consisted of 150 individuals, 75 from Kaiser
Permanente Colorado (KPCO) and 75 from Denver Health and Hospital
Authority (DHHA). All study participants were 40 years of age or older and
had multiple risk factors for cardiovascular disease (CVD) (at least two of the
following: diabetes mellitus, hypertension, hypercholesterolemia, smoker, and
a BMI >30) at the start of the study period and had completed a primary care
clinic visit in the past 12 months. All study participants needed to be able to
read and understand English. Exclusion criteria included any diagnosis of
mental imparity and the inability to read or understand English. Visual acuity
was the last requirement, which was assessed prior to consent. Individuals
with vision of 20/100 or better were eligible. Details on recruitment methods
can be found in Chapter 4.
Study Instalments
Health Literacy
The first instrument consisted of the short version of the Test of
Functional Health Literacy in Adults (S-TOFHLA). The S-TOFHLA is an
instrument that has been validated in multiple patient populations and is one
of the most popular assessments used for evaluating health literacy (Baker et
al., 1999; Gazmararian, et al., 1999). The S-TOFHLA is an abbreviated

version of the full TOFHLA. The full TOFHLA has a 50-item reading
comprehension test and assesses prose literacy (ability to read and
understand prose passages) and document literacy (ability to understand
appointment slips and Medicare paperwork) and takes up to 22 minutes to
administer. Each reading passage in the TOFHLA has every 5th to 7th word
deleted throughout the passage (referred to as a modified Cloze method)
(Parker et alM 1995). Respondents must then select the appropriate missing
word from a list of four options. TOFHLA is self-administered. The S-
TOFHLA consists of two reading comprehension passages with a total of 36
Cloze items. The S-TOFHLA requires a maximum of 12 minutes to
administer. The internal reliability of the S-TOFHLA has been demonstrated
to be very good (Cronbachs alpha 0.97) (Baker et al., 1999).
Scoring for the S-TOFHLA consists of assigning one point to each
correctly answered question. The maximum total for the S-TOFHLA raw
score is 36. The cutoffs for each level of health literacy are: 0-16: inadequate
functional health literacy, 17-22: marginal functional health literacy, and 23-
36: adequate functional health literacy.
Numeracy was assessed independently of the S-TOFHLA using
questions from the Medical Data Interpretation Test (Schwartz et al., 2005).
The S-TOFHLA does have a numeracy component but due to time
constraints another measure was selected. The original Medical Data
Interpretation Test is an 18-item assessment used to measure the ability to
interpret medical statistics (Schwartz et al., 2005). Questions were created
around real-world health information that a person routinely encounters.

Topics include news media coverage reports, direct-to-consumer prescription
drug advertisements, and statements health professionals typically use in
conversations with patients regarding medical risk. In addition to the 18-item
assessment, the authors have also tested and validated a three-item measure
of numeracy (L.M. Schwartz, personal communication, November 2005;
Schwartz et al, 1997). The three-item measure was used for this research
study to reduce the amount of time participants spent completing the surveys.
No composite score is calculated for the 3 items. Individuals are classified
based on the number of questions answered correctly.
A compilation of several survey instruments were used to collect data
on individual demographics, health status, cardiovascular disease knowledge,
obtaining and using health information, information seeking preferences, and
patient activation (See Table 3.1). Study instruments can be found in
Appendix A of this dissertation.
Sociodemographic questions were asked regarding income, education,
occupation, race/ethnicity, country of birth, and primary language spoken at
home. These questions were used to collect data on personal factors related
to Social Cognitive Theory from all study participants.
Overall Health Status
Overall health status was assessed using the general health question
from the CDC health-related quality of life scale (HRQOL-4) used in the
Behavioral Risk Factor Surveillance System (BRFSS) and the National Health
and Nutrition Examination Survey (NHANES) (Centers for Disease Control

and Prevention, 2008a, 2000b). In addition, the World Health Organization
Well-being Index (WHO-5) was used (Henkel et al., 2003). This scale was
chosen to be included in the survey over the additional CDC HRQOL
questions because they are written at a much lower literacy level. Comparing
the WHO-5 to other depression screening tools, the WHO-5 had the greatest
sensitivity (93%) and negative predictive value (98%) (Henkel et al., 2003).
The scale consists of 5 questions with six possible responses assigned a
score from 0 to 5. A raw score is calculated by totaling the responses to the 5
questions. The raw score ranges from 0 to 25 were 0 corresponds to the
worst possible quality of life and 25 corresponds to the best possible quality of
life. A score below 13 represents poor well-being and is recommended
further testing for depression be conducted. Age, body mass index (asked in
terms of height and weight), diagnosis of high cholesterol, diagnosis of high
blood pressure, diagnosis of diabetes, and smoking status were all self-
reported and collected during the screening process to determine eligibility for
the study. The organization from which the individual receives care was also
Disease Knowledge, Attitudes, and Beliefs
Two subscales (9 questions) from the 13-item Patient Activation
Measure (PAM) were used to assess confidence, and beliefs in relation to
CVD prevention (Hibbard, Stockard, & Mahoney, copyright 2003-2005;
Hibbard, Mahoney, Stockard & Tusler, 2005). The questions from this
assessment were used to explore the capacity to understand portions of the
health literacy definition in terms of perceived susceptibility from the Health
Belief Model and self-efficacy from Social Cognitive Theory. This instrument

was developed based on previous research on self-care, self-management,
doctor-patient communication, and using comparative information to inform
health care choices. The PAM uses statements with the respondent
indicating degree of agreement (four categories ranging from disagree
strongly to agree strongly).
In addition, 17 questions from the Revised Illness Perception
Questionnaire (IPQ-R) were used to assess understanding of cardiovascular
disease in relation to consequences, personal control, treatment control,
overall understanding, and emotional representations (Moss-Morris et al.,
2002). The questions from the IPQ-R were also used to explore the capacity
to understand portion of the health literacy definition in terms of perceived
severity, perceived benefits and barriers, and perceived susceptibility from the
Health Belief Model as well as self-efficacy and self regulation from Social
Cognitive Theory. The IPQ-R questions are rated on a 5-point Likert scale
(ranging from strongly disagree to strongly agree) and were adapted
specifically to cardiovascular disease (Moss-Morris et al., 2002). No
summary score is calculated from the causal items from the IPQ-R. Mean
scores for each item are calculated and presented.
Perceived Effects of Internet on Health Information Understanding
While few instruments are available to assess experiences with
accessing, using, and understanding health information, I used questions
from two validated surveys. The first set of questions (4 total) were derived
from the National Survey of the Internet and e-mail for Health Care
Information Survey (non-chronic illness version) (Baker, Wagner, Singer, &
Bundorf, 2003). These questions were used to explore the capacity to obtain

portion of the health literacy definition in terms of reciprocal determinism
(environment and behavior) from Social Cognitive Theory. The questions
were asked of respondents who reported that they had ever used the internet
or e-mail for health purposes. The questions focused on the effects of
internet or e-mail use on their knowledge about healthcare issues and their
use of health care. Responses are on a 4-point Likert scale ranging from
strongly agree to strongly disagree" that the internet and/or e-mail had
various effects. No summary score is calculated from the four internet and e-
mail questions. Mean scores for each item are calculated and presented.
Information Seeking Preferences
The next set of questions were derived from the Autonomy Preference
Index (API) eight question scale on information seeking (Ende, Kazis, Ash, &
Moskowitz, 1989). These questions were used to explore the capacity to
obtain and the capacity to process portions of the health literacy definition in
terms of reciprocal determinism (individual cognition interacting with the
environment) from Social Cognitive Theory. Responses for each question
were measured on a 5-point Likert scale ranging from strongly disagree to
strongly agree. The questions focus on when information should be given to
a patient (i.e. at what stage of treatment information should be shared) and
information sharing to increase knowledge regarding treatment and
medication use. Preferences for information seeking were expressed on a
scale adjusted to range from 0 to 100, where 0 refers to strong disagreement
with statements favoring patients being informed, 50 to a neutral reaction to
statements, and 100 refers to strong agreement.

Experiences Accessing and Using Health Information
In addition to the quantitative survey items listed above regarding
experiences accessing and using health information, I included a series of
open-ended questions. These questions focused on asking study participants:
their preferred sources of health information,
the extent to which they trust these sources of information and
how well they understand the information,
what kinds of health information would they like to have and
which sources do they use the most, and
what is their most (and least) preferred way to receive health
information and why.
These open-ended questions were asked to explore preferences for health
information and the lived experience with obtaining and using health
information from a phenomenological perspective and are listed in Appendix
A in the study instrument.

Table 3.1. Summary of measures used in the data collection process and
associated theoretical constructs.
Variables Collection Instrument & Method Domains Theoretical Constructs
Health Literacy Self-administered survey -S-TOFHLA (Baker, 1999; Gazmararian et al., 1999) Health Literacy Structural, SCT, HBM
Numeracy Self-administered survey- Medical Data Interpretation Test (3 questions) (Schwartz et al., 2005) Numeracy Structural, SCT, HBM
Organization individual receives care Recruitment lists KPCO or DHHA Organization individual receives care SCT personal factors
Demographics Oral Survey-gender, marital status, number of household members, education, income, occupation, race, ethnicity, country of birth, language spoken at home, number of comorbid conditions Individual patient characteristics SCT personal factors
Overall Health Status Oral Survey-WHO-5 Well-Being Index (Henkel et al, 2003; Tarlov et al, 1989) Individual level rating of health status SCT- personal factors

Table 3.1 (Cont).
Variables Collection Instrument & Method Domains Theoretical Constructs
Disease Knowledge, Attitudes, and Beliefs Oral Survey The Revised Illness Perception Questionnaire (IPQ-R) and Patient Activation Measure (PAM-non chronic illness version) knowledge and attitudes specific questions (Hibbard et al., copyright 2003-2005; Hibbard et al., 2005; Hibbard et al., 2004; Moss-Morris et al., 2002) Timeline (acute/chronic) Treatment control Personal control Consequences Causes Illness coherence Emotional representations HBM Perceived severity HBM Benefits/Barriers HBM Perceived Susceptibility SCT- personal factors (knowledge, skills, and self-efficacy)
Perceived effects of internet ore- mail use on health care understanding Oral Survey National Survey on use of the internet and e-mail for health care information (4 questions non- chronic illness patients) (Baker et al., 2003) Effects of internet or e-mail use on knowledge about health care issues/ use of health care SCT environmental factors and behavior
Information- seeking preferences Oral Survey Autonomy Preference Index (8 questions) (Ende et al., 1989) Patients preference for being informed SCT personal factors
Qualitative Data Open-ended questions and in-depth interviews Experiences with locating and using health information Phenomenology

Data Collection
Upon completion of the consent process (discussed in further detail in
Chapter 4) participants were given instructions on how to complete the S-
TOFHLA survey and then were left alone for seven minutes (the standard
time allotment for survey) to complete as much of the survey as s/he could in
the prescribed time. Once the participant completed the S-TOFHLA or time
expired (whichever came first), s/he was instructed to place the survey in an
envelope so as not to bias the investigator as to the level of health literacy of
the participant. Upon completion of the S-TOFHLA, the participant was given
the three written numeracy-related questions. The participants were
instructed on how to complete this short survey and left to work on it as long
as they desired. Again, the participant was instructed to place the survey in
an envelope upon completion prior to returning the materials to the
investigator. After the health literacy and numeracy assessments were
completed, the main study survey was administered verbally by the study
investigator to each of the participants.
To assure feasibility of administering this survey in a busy clinic
setting, a small pilot with a sample of seven volunteers of both sexes and
racial/ethnic background that meet study criteria were asked to complete the
survey and provide the investigator with feedback, particularly on items not
previously standardized. Adjustments were made after the pilot and prior to
implementation. It was originally proposed to use the full IPQ-R scale
consisting of 47 questions. However, after the pilot, it was decided due to
time constraints, to only ask 17 questions from a sub-scale focusing on
causes of cardiovascular disease. The revised survey took approximately 20
to 30 minutes for study participants to complete.

Upon completion of all surveys, study participants were given a $20 gift
card to a local grocery store as a thank you for their time and participation.
Individuals needing assistance with transportation were also given bus tokens
for travel to and from the study appointment. Surveys were coded and scored
at the end of each day and entered in a password protected database.
Analytic Plan
All statistical analyses were conducted using SPSS (SPSS, 2009).
Sample size calculations
Based on previous CVD-focused studies comparing 1) the effect of
decision aids on patients knowledge of options and outcomes for heart
disease compared with usual care; and 2) mean knowledge score for heart
disease and hypertension by health literacy level, the effect size was
estimated at r = 0.30 (Gazmararian et al., 2003; OConnor et al., 1999). One
hundred and fifty participants will result in a power of 0.90 to detect a
difference of -0.30 between the null hypothesis correlation (between health
literacy and disease knowledge) of 0.00 and the alternative hypothesis of
correlation of 0.30 using a two-sided hypothesis test with significance level
set at 0.05.. In addition, the rule of thumb for survey research and related
regression equations is 10 subjects for every variable measured (Pedhazur &
Schmelkin, 1991). Based on this guideline, a sample size of n=150 was
considered to be adequate and was obtained.

Overview of Analyses
This section provides an overview of the quantitative analyses used for
this dissertation. Specific analyses for each research question can be found
in Chapters 4, 5, and 6. All data were cleaned and visually inspected to
ensure they met assumptions prior to analyses. Missing data for individual
respondents were minimal, due to the survey being conducted in person, and
were generally limited to refusal to answer a particular demographic item,
such as income level. Because missing self-reported demographic data were
rare, they were in most cases ignored and the remaining data were analyzed.
Skipped answers for the standardized surveys were scored according to the
instructions for that specific survey. Total Ns are reported for all analyses.
Significance values were set at p<0.05, values are reported as mean +
SD, and proportions as %. Distributions, frequencies, cross-tabulations,
means, and standard deviations of patient characteristics including
demographics, health literacy level, risk level for developing CVD, disease
and risk factor knowledge were calculated and any necessary transformations
needed were performed. Participation rate in the surveys and characteristics
of respondents versus non-respondents were also documented. Correlations
and associations were determined between the dependent and independent
variables listed above using chi-square tests, independent t-tests, and
correlation coefficients. Appropriate analyses, depending on type of variable
and whether the assumptions were met, were conducted to assess
associations/correlations between dependent and independent variables.
Since the dependent variables were dichotomous, independent variables
found to be significant in bivariate analyses were further analyzed using
logistic regression.

Table 3.2. Overview of analyses by research question conducted for this
Study Aims 1, 2, and 3
Were there differences in participation rates based on the various recruitment methods used for this study? Descriptive analyses and CONSORT diagram Comparison of active and passive recruitment methods Chapter 4
What are the sociodemographics of adults with multiple risk factors for CVD as a function of health literacy and numeracy abilities? Chi square test, t- test, binary logistic regression Demographic characteristics associated with limited health literacy and low numeracy Chapter 5
What are the health characteristics of adults with multiple risk factors for CVD as a function of health literacy and numeracy abilities? Chi square test, t- test, binary logistic regression Health characteristics associated with limited health literacy and low numeracy Chapter 5
What do participants believe are the causes of CVD? Descriptive statistics (mean, SD) and t-test Mean scores related to CVD knowledge by health literacy and numeracy ability Chapters
What are participants' information seeking preferences related to health and health information? Descriptive statistics (mean, SD) and t-test Mean score for information seeking preferences and demographic characteristics associated with higher information seeking preferences scores. Chapter 6

Table 3.2 (Cont.).
Study Aim 4
What are participants preferences for receiving health information? Coding for themes patient interviews Participants' most and least preferred methods for receiving health information by health literacy and numeracy abilities Chapter 6
Study Aims 5 and 6
What are the current methods used to communicate health information to patients at DHHA and KPCO? Coding for themes health educator interviews Description of current methods used at DH and KPCO to communicate health information to members Chapter 6
Study Aims 7 and 8
Comparison of patient preferences' to current methods used by DHHA and KPCO. Recommendations for ways to enhance how health information is given to patients. Comparison of themes from qualitative interviews with patients and health educators Recommendations for meeting patient needs for health information Chapter 6

Patients Attitudes and Preferences Toward Current Methods
of Obtaining and Receiving CVD Prevention Information
Study Population
The study population for the qualitative portion of the study consisted
of 150 study participants as previously described in the quantitative data
section of this chapter. Twenty-eight individuals (14 from each health care
organization) completed a follow-up, in-depth interview. Nine health
educators also participated in an in-depth interview (5 from DHHA and 4 from
Data Collection
Patient Inten/iews
Convenience sampling was used for the patient-level interviews. The
sampling method was based on study participants who were identified as
having expressed in the initial study visit a willingness to participate in a
follow-up interview. Interested individuals were called approximately two
weeks after their initial study visit to remind them about the study and to
extend an invitation to participate in an in-depth, semi-structured interview.
Interview topics included participants knowledge, attitudes, beliefs, and
experiences related to receiving and understanding health information,
preferred modalities for health communication, experiences with accessing
health information, and benefits and barriers to various modalities based on
experience. Discussing the lived experience (phenomenological perspective)
is necessary for gaining an understanding of all the potential factors that

impact whether or not someone understands health information and whether
this varies by health literacy or numeracy ability. The researcher also used
clarification probes" (based on Kleinmans explanatory model queries) to
encourage interviewees to elaborate on the domain being discussed
(Kleinman, 1998; Patton, 2002). Questions were asked followed by a
discussion allowing respondents to elaborate as little or as much as they
wanted to. Initial questions included:
Tell me about the last time you tried to find out something related to
health. Where did you look? Who did you ask?
When you did get health information (from patient solicited sources in
previous question), what was helpful? Not helpful? What did you
easily understand? What did you have difficulty understanding? How
did it make you feel to understand (or not)?
What do you feel is your personal risk for getting heart disease?
In your own words how would you describe heart disease?
Where have you gone to look for information on heart disease or
related to heart disease?
Tell me about the last time you talked with your health care provider.
How did you feel? What did you talk about? Was the information you
received helpful? Why? Why not?
If you could choose any way of receiving health information, what
would it be? Why?
It was originally proposed to have participants complete a talk aloud
experience (VanSomeren, Barnard, & Sandberg, 1994; Zarcadoolas et al.,
2002) consisting of them thinking out loud as they navigated one of two e-

Health products. Unfortunately this part of the study had to be dropped due
to lack of wireless internet connectivity in the clinics where the research visits
were conducted.
Current Methods Used by Health Care Professionals to
Distribute and Provide CVD Health Information
Health Educator Interviews
Snowball sampling was used to identify the health educators. An e-
mail was sent to professional contacts at each health care organization with a
short description of the study and a solicitation for names of individuals who
might be interested in participating in an interview. During each interview
participants were asked to recommend additional individuals for potential
participation in the study.
In-depth, semi-structured interviews were conducted by the study
investigator. Interview topics included role in health care organization,
description of interactions with patients when communicating health
information, methods preferred for delivering health information, and
preferences for future methods. Questions were asked followed by a
discussion allowing respondents to elaborate as little or as much as they
wanted to. Initial questions included :
When communicating health information to patients either at risk for
CVD or diagnosed with CVD, what methods do you find most
effective? Not so effective?

Describe for me how you normally/routinely communicate CVD health
information to patients. (A request will be made by the interviewer for
copies of class materials/handouts to assess literacy level.)
How do you know if a patient understood the information you gave
them? Tell me about a time when a patient did not understand? How
did you know? What did you do?
What modalities for health communication would you like to see your
organization embrace in the future? Why? Which ones would you
hope are not utilized? Why?
Analytic Plan
All open-ended question responses and all in-depth interviews were
transcribed verbatim by the study investigator. The narrative data were
entered into ATLAS.ti and coded according to the constant comparative
method (Berg, 2004; Patton, 2002). Data analyses included identifying
emerging themes, interpretations, and categories related to patient
preferences for receiving and using health information. Data analysis
occurred throughout the data collection and analysis phase of the study. The
content of the qualitative data was coded for recurrent themes by the study
investigator as well as a masters-level research assistant. The emerging
themes were discussed between the two coders which led to the
development of a thematic framework and coding structure. Thematic
analysis continued until data saturation was reached and no new themes

Compiling Results of Mixed-Methods Analysis
The information gained from the patient surveys and interviews related
to research questions 1 and 2 were compared to the information gained from
the health educator interviews related to research question 3 to assess how
the needs of patients with multiple risk factors for CVD are currently being
met. In addition, areas for potential improvement in communicating health
information were also identified. Comparisons were made based on patient
characteristics, as well as overall themes, categories, and interpretations of
the qualitative data. Specific analyses and results can be found in Chapter 6
of this dissertation. Analysis of these data consisted of using the results from
the statistical analyses conducted in research question 1, results from the
analytical induction from the qualitative analyses in research questions 2 and
3, and overall triangulation to assess similarities and variations of themes,
interpretations, and categories between patient and health educator
responses (Patton, 2002; Tashakkori & Teddlie, 1998). Once the synthesis
was completed, practical recommendations were developed and can be
found in Chapter 7 of this dissertation. Findings will be shared with both
health care organizations via written reports and oral presentations.

Recruitment of a diverse participant sample is necessary to improve
the generalizability of research findings. While the goal for research studies
is to enroll individuals who will commit to the study for the entire duration and
carry out all components as instructed, the individuals enrolled should also be
representative of the population to which the intervention is directed. Since
the National Institutes of Health (NIH) Revitalization Act of 1993, researchers
are now obligated to recruit diverse study samples for projects receiving
federal funds (Freedman et al., 1995; National Institutes of Health, 1993).
How this is achieved can depend on the recruitment methods and efforts of
the research staff.
Currently efforts to diversify research samples have had mixed results.
Studies that have experienced challenges in recruiting diverse samples report
barriers such as language, lack of knowledge about research studies offered,
site accessibility, lack of child care, lack of transportation, lack of trust, and
lack of time are the main reasons for non participation by minority and low
income populations (Brown, Long, Weitz, & Milliken, 2000; Eakin et al., 2007;
Shavers, Lynch, & Burmeister, 2002; Wendler et al., 2006). Recruitment
strategies that have been shown to be effective in increasing sample diversity
include building relationships and trust with the community of interest
(Gauthier & Clarke, 1999), using multiple recruitment methods (Patrick,
Pruchno, & Rose, 1998), having culturally competent processes and research