Citation
The complex patient's experience with depression

Material Information

Title:
The complex patient's experience with depression the role of competing illness in depression symptom improvement
Creator:
Moralez, Ernesto Andres ( author )
Language:
English
Physical Description:
1 electronic file. : ;

Subjects

Subjects / Keywords:
Depression, Mental ( lcsh )
Depression, Mental ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Background Growing evidence shows the co existence of physical chronic illnesses and mental health disorders like depression can confound and hinder physical symptom improvement and depression treatment. Objectives The purpose of this mixed methods study was examine the potential relationship between coexisting illness severity and change in depression over time and to identify contextual factors that influence beliefs attitudes and treatment preferences for depression and coexisting illness care for primary care patients experiencing a new episode of depression. Participants and Setting 168 primary care patients experiencing a new episode of depression recruited from a primary care network in Denver Colorado. Nineteen participants were interviewed for the qualitative portion of the study. Measures Patient Health Questionnaire 9 PHQ 9 to measure depression and the Cumulative Illness Rating Scale CIRS to measure illness severity. ( , )
Review:
Results Hierarchical Linear Modeling analyses reveled CIRS was not a statistically significant predictor of change in depression over time give the estimated variance components for the PHQ 9 reliabilities were low Intercept 0.42 Time Slope 0.054 . Participants described a variety of causes for their depression including coexisting illnesses as well as describing coexisting illness as a competing demand to seeking mental health care. Additionally in spite of a new diagnosis of depression participants for the most part continued to focus on physical ailments and controlling existing chronic disease. Conclusions Although this study did not show a statistical relationship between chronic illness severity and change in depression qualitative findings did suggest that illness severity has some influence on depression symptoms which should be further tested. The interviews offer some challenges to the potential assumptions of depression interventions in primary care settings that may see depression as a separate condition from patients chronic illnesses and need isolated treatment. It can be theorized that the patients interviewed for this study see their depression as the result of many of the difficulties associated with illness complexity. Though this study was unable to identify a relationship between chronic illness severity and depression symptom improvement additional efforts are needed to improve the understanding of chronic illness burden on depression outcomes.
Thesis:
Thesis (Ph.D.)--University of Colorado Denver.
Bibliography:
Includes bibliographic references.
General Note:
Department of Health and Behavioral Sciences
Statement of Responsibility:
by Ernesto Andres Moralez.

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:
930294689 ( OCLC )
ocn930294689

Downloads

This item has the following downloads:


Full Text
THE COMPLEX PATIENTS EXPERIENCE WITH DEPRESSION:
THE ROLE OF COMPETING ILLNESS IN DEPRESSION
SYMPTOM IMPROVEMENT
By
ERNESTO ANDRES MORALEZ
B.H.C.S., New Mexico State University 2007
M.P.H., New Mexico State University 2009
A thesis submitted to the Faculty of the Graduate School of the
University of Colorado in
partial fulfillment of the requirements for the degree of
Doctor of Philosophy
Health and Behavioral Sciences
2015


This thesis for the Doctor of Philosophy degree by
Ernesto Andres Moralez
has been approved for the
Health and Behavioral Science Program
by
Karen Lutfey, Chair
Debbi Main, Advisor
Elizabeth A. Bayliss
Robert D. Keeley
Jini Puma


Moralez, Ernesto, Andres (Ph.D., Health and Behavioral Sciences)
Depression Treatment for Complex Patients in Primary Care: The Role of Illness Severity in
Depression Symptom Improvement
Thesis directed by Professor Debbi Main
ABSTRACT
Background: Growing evidence shows the co-existence of physical chronic illnesses and mental
health disorders like depression can confound and hinder physical symptom improvement and
depression treatment.
Objectives: The purpose of this mixed methods study was examine the potential relationship
between coexisting illness severity and change in depression over time; and to identify contextual
factors that influence beliefs, attitudes, and treatment preferences for depression and coexisting
illness care for primary care patients experiencing a new episode of depression.
Participants and Setting: 168 primary care patients experiencing a new episode of depression
recruited from a primary care network in Denver, Colorado. Nineteen participants were
interviewed for the qualitative portion of the study.
Measures: Patient Health Questionnaire-9 (PHQ-9) to measure depression, and the Cumulative
Illness Rating Scale (CIRS) to measure illness severity.
Results: Hierarchical Linear Modeling analyses reveled CIRS was not a statistically significant
predictor of change in depression overtime give the estimated variance components for the PHQ-
9 reliabilities were low (Intercept = 0.42; Time Slope = 0.054).
Participants described a variety of causes for their depression including coexisting illnesses as
well as describing coexisting illness as a competing demand to seeking mental health care.
Additionally, in spite of a new diagnosis of depression, participants for the most part continued to
focus on physical ailments and controlling existing chronic disease.
m


Conclusions: Although this study did not show a statistical relationship between chronic illness
severity and change in depression, qualitative findings did suggest that illness severity has some
influence on depression symptoms which should be further tested. The interviews offer some
challenges to the potential assumptions of depression interventions in primary care settings that
may see depression as a separate condition from patients chronic illnesses and need isolated
treatment. It can be theorized that the patients interviewed for this study see their depression as
the result of many of the difficulties associated with illness complexity. Though this study was
unable to identify a relationship between chronic illness severity and depression symptom
improvement, additional efforts are needed to improve the understanding of chronic illness
burden on depression outcomes.
The form and content of this abstract are approved. I recommend its publication.
Approved: Debbi Main
IV


ACKNOWLEDGEMENTS
First of all I want to thank my father, Alejandro Moralez, for being my hero and my biggest
fan. I owe all my success to your success as a dad.
I would like to thank my dissertation committee for their support and patience. Dr. Keeley for
bringing me onto his project and being a friend and supporter throughout this endeavor. Dr.
Bayliss for being so positive and supportive of my ideas. Dr. Lutfey for being the right kind of
motivator. All the teachers, professors, and researchers that have contributed to my professional
development; Jeanine Voller, Mark Hailing, Beti Thompson, Mary OConnell, Satya Rao,
Yvonne Kellar-Guenther (thank you for being not just a mentor but a wonderful friend), Jean
Scandlyn, Erin Wright, Howard Waitzkin, David Tracer, and Arthur Kleinman.
My family. Thank you Alex for being proud of your little brother. Thank you Isabelle, Bobby,
and Lilly for always having my back and supporting this long endeavor.
Countless friends (who tolerated my incessant self-doubt and were always there to keep me on
track): Andrianna Martinez, Chris Cantrell, Leighton Kaufman (always there to ask hows is
coming along), Mateo Banegas (would have done it without you doing it first); Joe Beling, Tres
Morley, and Dave Troyer; Christine, Ryan, Whitney Jones, Debbie, Melanie, Sarah, and all the
HBS doctoral students I met along the way. Abby Litch and Jessica Halliday (special thanks for
everything); Carlos and Sonia, Ray, Jerry, Pete, Robert, Tomas and Angela, and all my friends in
Albuquerque who will never let me forget what matters most; the Kaufmans, Jennifer Cantrell
(always there to listen and get me out of trouble); Mike Anderson for help uploading 2.0. and
seeing me through at the finish.
A very special thank you to Natalie Slevin for knowing this day would come, even when I was
sure it wouldnt.
A very special thank you to Jini Puma! for always supporting, helping, and smiling. Could not
have done it without you
A very special thank you to Ashley Loldes and the Loldes family for being unconditionally
supportive and giving me the belief in myself to keep going.
A very special thank you to Gilbert Escarcida for always being there to offer a certain kind of
empathy and understanding (and a place to live).
A very special thank you to Indy for being my best friend.
A very special thank you to Matthew Engel for doing so much for me.
Special people and places: Commonground Golf Course for being my sanctuary, Sugar Bake
Shop, Tattered Cover, and Hooked on Colfax for the coffee and free Wi-Fi.
Mentors along the way: Basil Walter, Max Contreras, Jon Stewart, Robert F. Kennedy, and
Ernesto Galarza.
v


DEDICATION
To my mother, Blanca Velasco, who I know is proud of her son and even more proud that I have
spent so much time studying a disease that needs attention and care. I love you very much and I
hope to continue to make you proud.
I would not be who I am today without the love from my grandmother Lucy Caballero.
To my grandfather, Catarino Moralez Jr., who didnt make it to see me finish but called me
doctor anyway...
To Debbi Main, for rescuing me when I had no idea what I was doing, or where I was going. You
took me on and made sure I found something I was passionate about and that I was developing
the fortitude to finish. Thank you Debbi, I truly would not have done this without you.
vi


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION......................................................................1
Study Purpose and Rationale.......................................................3
Research Questions..............................................................3
Specific Aims...................................................................4
Overview of the Research Methods..................................................4
Study Setting and Participants..................................................5
Significance of the Study.......................................................6
Terminology.....................................................................9
Structure of the Dissertation....................................................11
II. REVIEW 01 HU] LITERATURE..........................................................12
Chronic Illness and Depression among Medically-Indigent
Populations in Primary Care......................................................16
Medical Illness Comorbidity and Depression.......................................20
Prevalence of Depression Comorbid with Specific Medical Illnesses..............21
Studies comparing depression treatment outcomes for patients with and without
coexisting illnesses...........................................................23
Theoretical Approach.............................................................33
Competing Demands..............................................................34
The Illness Narrative..........................................................36
Summary of the Theoretical Frameworks..........................................38
III. METHODS..........................................................................39
Description of the Original Randomized Controlled Trial..........................39
Study Setting..................................................................42
Study Participants.............................................................42
Measures.......................................................................43
Description of the Dissertation Study............................................45
Measures.......................................................................45
Data Collection Procedure......................................................47
Data Analysis..................................................................48
Qualitative Primary Data Research Design.......................................50
Data Collection Procedure......................................................50
Data Analysis..................................................................52
Qualitative Validity...........................................................54
IV. RESULTS: QUANTITATIVE ANALYSIS...................................................56
Quantitative Patient Population..................................................56
Research Question 1..............................................................58
Depression and Illness Severity................................................58
Results from the Hierarchical Linear Modeling..................................62
Response to Research Question 1..................................................65
vii


V. RESULTS: QUALITATIVE ANALYSIS....................................................66
Qualitative Patient Population..................................................66
Research Question 2.............................................................69
Perceived Causes of Depression................................................69
Perceived Physical Symptoms of Depression.....................................76
Perceived Emotional Symptoms of Depression....................................77
Living as a Complex Patient...................................................79
Response to Research Question 2...............................................85
Research Question 3...............................................................86
Illness Priority..............................................................86
Depression Treatment Preferences..............................................88
Competing Demands/ Barriers to Treatment and Improving Symptoms...............91
Response to Research Question 3...............................................95
VI. DISCUSSION......................................................................99
Change in Depression Over Time in Complex Patients.............................100
Key Findings from the Qualitative Analyses.....................................101
Patient Description of Depression............................................101
Perceived Relationship between Depression and Chronic Illness................103
Illness Priority and Competing Demands.......................................104
Assessing Methodology..........................................................106
Conclusion.....................................................................106
Limitations....................................................................108
Limitations of the Quantitative Analyses.....................................109
Limitations of the Qualitative Analyses......................................109
Dissemination of Results.......................................................110
REFERENCES..........................................................................Ill
APPENDIX............................................................................124
A. Glossary of Terms..........................................................124
B. Patient Health Questionnaire-2.............................................128
C. Patient Health Questionnaire-9.............................................129
D. Demographic Variables Questionnaire........................................130
E. Cumulative Illness Rating Scale (CIRS).....................................133
F. Qualitative Interview Guide................................................134
E. Diagram of the relationships between levels of coding in the qualitative analysis.137
viii


LIST OF TABLES
TABLE
II. 1 Diagnostic Criteria for Depression/Depressive Disorder.....................13
II. 2 Summaries of the Existing Literature Since 2002 on the Effect of Coexisting
Illness on Depression ...................................................29
III. 1 Description of the Primary and Dissertation Studies .......................41
IV. 1 Patient Demographics ......................................................57
IV. 1 PHQ-9 Mean Scores at all Four Times Points ................................59
IV.3 Total CIRS by Age..........................................................60
IV.4 Main Variable Correlations.................................................61
IV.5 Results of Regressing Change in Depression on Illness Severity.............62
IV.6 Results From the Random Coefficient Estimates .............................63
IV.7 Results From Depression Over Time Model ...................................64
IV. 8 Results From the Depresion Change by CIRS Total ...........................65
V. l Interview Participant Characteristics .....................................68
V.2 Total Counts of the Codes Identified ......................................96


LIST OF FIGURES
FIGURE
1.1 Order of the Data Collection Procedure.....................................5
1.2 Structure of the Disseration...............................................8
II. 1 Factors Contributing to Depression Outcomes in Primary Care...............19
11.2 The Competing Demands of Psychosocial Care ...............................36
III. 1 Iterative Process Model...................................................54
111.2 Description of the Qualitative Analysis ..................................55
IV. 1 Total CIRS Score Reported by Percent of Sample ...........................59
IV.2 Change In Depression Over Time Stratified by CIRS ........................61
x


CHAPTER I
INTRODUCTION
The Centers for Disease Control and Prevention (CDC) estimate that by 2020, depression
will be the second leading cause of disease burden in the U.S. behind cardiovascular disease
(Centers for Disease Control and Prevention, 2013). Depression is a debilitating illness typically
associated with a loss of pleasure in activities and perceptible changes in mood and behavior such
as sleeping patterns, appetite, and concentration, as well as feelings of hopelessness (Gilbody,
2011; National Institute for Health and Clinical Excellence (NIHCE), 2009). Both the CDC
(2010) and the NIHCE (2009) confirm that depression is more prevalent among people with
chronic illnesses and the coexistence of chronic physical health problems can cause and
complicate depression. The NICHE report concludes that depression is approximately two to
three times more common in patients with a chronic physical health problem then in people who
have good physical health (2009).
Growing evidence shows the co-existence of physical chronic illnesses and mental health
disorders like depression can confound and hinder physical symptom improvement and
depression treatment (Gonzales, Esbitt, Schneider, Osborne, & Kupperman, 2011; Katon &
Schulberg, 1992; Koike, Unutzer, & Wells, 2002; Nutting, Rost, Smith, Wemer, & Elliot, 2000;
Pagoto, Schneider, Appelhans, Curtin, & Hajduk, 2011; Rost, et al., 2000; Rutledge, Reis, Linke,
Greenberg, & Mills, 2006). Depression has a bi-directional relationship with physical illness in
that it adversely affects the severity of physical illness, but similarly the presence of physical
illness exacerbates the severity of depressive symptoms (Clarke, 2009), complicating the
management of both physical and mental health conditions (Morris, et al., 2012). Reports from
the World Health Organization indicate that of any chronic disease, depression produces the
greatest diminution of overall personal health and, as a coexisting condition, is more debilitating
1


than any other disease combination contributor (Moussavi, et al., 2007). The prevalence of
depression is reported anywhere from three to five times higher among complex patients than the
general population (Bair, Robinson, Katon, & Kroenke, 2003; Teh, Reynolds, & Cleary, 2008).
Additionally, physical illness can disguise the symptoms of depression since some symptoms are
common to both physical and mental disorders, complicating the assessment of depression
(NICHE, 2009).
Furthermore, diagnostic decisions, depression treatment adherence, and symptom
improvement are further complicated by the economic, socio-cultural, and environmental factors
often experienced by individuals with lower socioeconomic status (SES) (Alegria, et al., 2008;
Everson, Maty, Lynch, & Kaplan, 2002; Lorant, et al., 2003). For example, studies show higher
odds of persisting depression (Lorant, et al., 2003), and lower rates of service use for depression
(Alonso, et al., 2004) among low SES populations. Given most patients (particularly low SES
populations) with poor mental health and coexisting illnesses are cared for in primary care
settings (Amow, et al., 2006), developing and targeting more effective interventions for primary
care settings is critical for reducing disparities in the diagnosis and treatment of depression,
especially for low-income primary care patients with limited access to mental healthcare.
Providing adequate behavioral health treatment in primary care is challenging, especially
for medically indigent populations suffering from poorer mental and physical health. Even when
depression recognition is adequate for low-SES populations, gaps remain in the quality of the
depression care and in the understanding of how the patients competing life demands interplay
with depression care and improved symptomatology. Research related to improving behavioral
health treatments and interventions for medically indigent complex patients are not only a
research priority, but also a clinical imperative. Fortunately, there is evidence that treating
depression in patients with chronic physical health problems can potentially increase life
expectancy and their overall quality of life (NICHE, 2009); however, we lack understanding
about how the severity of chronic illnesses moderates the effect of depression treatment on
2


symptom improvement (Teh, et al., 2008), particularly for patients who are medically indigent
and economically disadvantaged.
The purpose of this two-phase sequential mixed methods study was two-fold: examine
the potential relationship between coexisting illness severity and change in depression over time;
and to identify contextual factors that influence beliefs, attitudes, and treatment preferences for
depression and coexisting illness care for primary care patients experiencing a new episode of
depression. This study will build upon the findings of a randomized control trial (RCT) that
trained primary care providers in a communication strategy to improve clinical dialogue around
depression (Keeley & Brody, 2012). The setting for the study was an integrated safety-net health
care network called Denver Health and Hospital Authority (DHHA) in Denver, Colorado that
primarily treats economically-indigent patients. Though existing research has identified the
adverse effect of chronic illness on depression symptom improvement, gaps remain in
understanding how the severity of chronic illnesses-as well as patients competing demands-
interplay with depression care and improved symptomatology.
Study Purpose and Rationale
Research Questions
Given the lack of evidence concerning the potential relationship between physical illness
severity and depression symptom improvement, as well as the patient narrative around living with
depression and coexisting illnesses, this dissertation addressed the following research questions:
1. Does the severity of coexisting medical illnesses impact depression symptom
improvement for primary care patients experiencing a new episode of depression?
2. How do complex depression patients describe their lived experience with concurrent
illnesses?
3. How does a diagnosis of a new episode of depression impact the management and
prioritization of concurrent illnesses for complex patients?
3


Influenced by the model of competing demands-particularly patient-level demands-
(Jaen, Stange, & Nutting, 1994), I hypothesized that patients with high competing illness severity
would experience significantly less depression symptom improvement overtime.
Specific Aims
Specifically, this dissertation research addressed the following aims:
Aim 1 Determine whether severity of coexisting illnesses were
associated with improvement of depressive symptom outcomes
for moderately to severely depressed primary care patients with
coexisting illnesses (complex patients)
Aim 2 Identify how patients perceive the bi-directional relationship
between depression and physical illness
Aim 3 Describe the effects of illness complexity on patients quality of
life including, but not limited to, physical and social functioning.
Aim 4 Describe the contextual factors (competing demands) that
influence complex patients beliefs, attitudes, and treatment
preferences in receiving care for their depression and coexisting
illnesses in a primary care setting.
Overview of the Research Methods
This mixed-methods study used a combination of quantitative analysis of secondary data,
as well as in-depth qualitative interview data to address the study aims (FIGURE 1.1). For the
quantitative analysis, I used Hierarchical Linear Modeling (HLM) to examine the association
between the number and severity of coexisting medical illnesses and depression symptom
improvement, over time, for complex patients with a major depressive disorder (MDD) in a
clinical trial study. For the qualitative analysis, I conducted a thematic analysis of semi-
structured interviews. I selected a mixed-method design because it provided a way to place
statistical findings in context, allowing for a rich explanation of related phenomena (Johnson &
Onwuegbuzie, 2004).
4


Data Collection Procedures
Sequential Implementation
Equal Priority (weighting)
FIGURE 1.1: Order of the Data Collection Procedure
Study Setting and Participants
This secondary analysis used data from 168 patients screened for depression from eight
primary care health clinics in the Denver Health system that provide health care services to low-
income and underinsured people (medically indigent) in Denver, Colorado. Patients were enrolled
in Medicare, Medicaid or the Colorado Indigent Care Program (CICP) or both, and of various
race/ethnicities, and at least 18 years of age. For the qualitative interviews, and given the intent to
understand the potential effect of comorbid illnesses on depression symptom improvement,
purposeful sampling techniques were used to identify patients with low, medium and high illness
severity, and with increased, neutral and decreased depression symptom improvement at 6-12-,
and 36-weeks after baseline.
5


Significance of the Study
A review of the literature on treating depression in primary care highlights that although
the impact of coexisting illness on depression symptom improvement has been examined, most
studies have focused on either the impacts of antidepressant medications (Koike, et al., 2002;
Kurdyak & Gnam, 2004; Lin, et al., 2012), isolating only one co-existing chronic illness for
comparison (e.g., cancer, diabetes), or included a sample consisting of only older adults, or all or
some of the above (Iosifescu, 2007) (all relevant studies are reviewed in Chapter II).
Additionally, a disproportionate number of the studies used a simple count of morbidities, which
may not reflect severity of symptoms or the chronicity of specific illnesses, which are limitations
when assigning equal weight to all illnesses. A more nuanced approach was to assess the severity
of each illness (Fortin, Bravo, Hudon, Vanasse, & Lapointe, 2005; Huntley, Johnson, Purdy,
Valderas, & Salisbury, 2012). Thus, a primary aim of this dissertation was to test whether
severity of coexisting illnesses are associated with improvement of depressive symptom
outcomes among moderately to severely depressed medically-indigent primary care patients ages
>18 participating in a non-pharmaceutical treatment focused on improving clinical dialogue
around depression delivered by primary care providers (PCPs). Improving our understanding the
pathways through which coexisting illnesses influence depression symptom improvement in
primary care is imperative. From a clinical perspective, complex patients (patients with multiple
chronic medical conditions occurring simultaneously) represent the major users of health care
services in the U.S. accounting for more than two-thirds of health care spending (Tinetti, Fried, &
Boyd, 2012) Furthermore, depression as a coexisting illness with the more prevalent chronic
disorders (e.g., coronary heart disease, diabetes) is associated with poor self-care, and increased
complications (Katon 2010), which can lead to more severe illness severity and earlier death.
Thus, the identification of effective interventions to treat patients with multiple conditions,
particularly, the coexistence of physical and psychological disorders, is timely and important.
6


A second aim of this dissertation was to collect in-depth qualitative data from semi-
structured interviews to identify specific contextual factors (e.g., competing demands) that
influence patients beliefs, attitudes, and treatment preferences for their depression and coexisting
illness care. Though competing demands at the provider-level have been previously explored
(Jaen et. al., 1994; Nutting et.al., 2008; Rost et. al., 2000; Stange et.al., 1994), less understood are
the competing demands at the patient-level (e.g., financial strain, poor living conditions, co-
occurring illnesses), and what those demands mean for chronic illness care and depression
symptom improvement. A third aim was to explore how patients perceive the relationship
between depression and chronic physical illness, specifically which, between depression and
chronic illness, patients prioritize as crucial for immediate care. Though coexisting depression
and chronic illness is common, it is not satisfactory to conclude that they have a linear
relationship, but in fact a more complex, bi-directional one, where depression influences chronic
illness severity and treatment, or vice versa. A conceptual model including descriptions of the
three phases, along with sampling strategies, brief descriptions of the quantitative and qualitative
methods, and how they associate with each of the aims is included below (FIGURE 1.2).
7


Description of Original Study (Keeley, 2012)
Previously recruited from providers current list of patients.
Recruited by telephone.
All patients completed the Patient Health Questionnaire-2 during initial telephone
call.
Informed consent and baseline data collected prior to visit in clinic waiting areas.
PHASE II
Test for the potential effect of
coexisting illness severity on
depression symptom
improvement
PHASE I (Secondary Data)
n=168
Score each patients coexisting
illness severity using the
Cumulative Illness Rating Scale
(CIRS)
FIGURE 1.2: Structure of the Dissertation
8


Terminology
Given the number of terms used in this study and the complexity of some of those terms,
definitions of the various terms along with citations are included to improve understanding.
These citations include existing studies published in peer-reviewed academic journals, the
Diagnostic and Statistical Manual IV (DSM-IV) published by the American Psychiatric
Association, the Centers for Disease Control and Prevention (CDC), and the National Institutes of
Mental Health (NIMH). A glossary of terms (Appendix A) is included with the definitions of
terms central to this study.
The DSM-IV offers a series of symptoms associated with depressive disorder including
depressed mood most of the day (e.g., feeling sad or empty), markedly diminished pleasure or
interest in most daily activities including those activities necessary to maintain manageable living
conditions (e.g., employment, family and other social interactions), change in sleep patterns, lack
of energy and capacity to concentrate (e.g., reading, watching television), and recurrent
thoughts of hurting oneself or suicidal ideation with or without a specific plan (American
Psychiatric Association, 2000). For the purposes of this study, the term depression is used to
refer to a mental health condition that meets diagnostic criteria such as those published in the
DSM-IV.
Though major depressive disorder (MDD) is typically reported as a dichotomous variable
(YES or NO), this study aims to measure improvement in depressive symptoms necessitating a
continuous measure instrument. While acknowledging that MDD can be difficult to quantify or
interpret uniformly across populations of different socioeconomic status, cultural traditions, and
communities, there are several instruments for detecting MDD the aimed at accurately assessing
an individuals mental health status. The study assumed that distinctions exist in the severity of
depression (Aim 1) critical to fully understanding not only the current depressive state of each
patient, but to better assess the effectiveness of depression interventions at multiple levels of
depressive states. Specifically, depression was given diagnostic scores listed as moderate (10-


14), moderately severe (15-19), and severe (> 20) depression in accordance with a 9-item
depression module called the Patient Health Questionnaire-9 (PHQ-9) (Kroenke, Spitzer, &
Williams, 2001). The PHQ-9 is a questionnaire that assesses depressive symptom criteria in
accordance with the symptoms described in the DSM-IV including "little interest in doing
things, feeling down or depressed, trouble concentrating, and thoughts that you would be
better off dead or hurting yourself in some way(Kroenke, et al., 2001). As a measure of
depression severity, the PHQ-9 scores range from 0-27 with each of the nine items having
possible individual scores of 0 (not at all) to 3 (nearly every day). A copy of the PHQ-9 is
included in the appendix.
I used the term comorbidity in this research as an encompassing term to refer to both
physical and mental health conditions and is defined as the presence of more than one distinct
(health) condition in an individual (Valderas, Starfield, Sibbald, Salisbury, & Roland, 2009).
Comorbidity can often be either the cause or the consequence of an index disease and can affect
disease detection, therapy, and desired outcomes or changes to behavior (de Groot, Beckerman,
Lankhorst, & Bouter, 2003). Additionally, evidence suggests that comorbid illnesses affect the
effectiveness of treatments and interventions. For the purposes of my research, comorbidity was
counted, weighted, and reported by severity of the illnesses using the Cumulative Illness Rating
Scale (CIRS) which rates comorbidity on a five-point severity scale based on 13 anatomical
domains (e.g., cardio-vascular respiratory system, neurological etc.), making it a methodological
superior approach to measuring comorbidity to using strictly a simple count.
This study used complex patient as a term to refer to those patients with multiple chronic
conditions that often require unique services and care management strategies (Bayliss, Ellis, &
Steiner, 2005; Fortin, Soubhi, Hudon, Bayliss, & van den Akker, 2007). Complex patients can
experience adverse health outcomes, higher healthcare costs, as well as a decreased quality of
life, and psychological distress (Bayliss, Edwards, Steiner, & Main, 2008; Bayliss, et al., 2005;
Fortin, Soubhi, et al., 2007; Newcomer, Steiner, & Bayliss, 2011).
10


This study used several descriptors to refer to social issues and stressors experienced by
the patient populations. The term socially and economically disadvantaged is used as an
encompassing term to describe individuals and groups experience conditions of poverty,
inadequate resources, and deprivation, as well as limited political capital and unequal access to
opportunities, social rewards, and social status (Centers for Disease Control and Prevention,
2011c; U.S. Department of Health and Human Services, 2001). This study included patients that
are socially and economically disadvantaged with limited access to health services along with
transportation issues, language barriers, or other issues hindering adequate medical care and are
referred throughout this research as medically-indigent.
Lastly, this study referred to providers as Primary Care Providers (PCPs), which
includes nurse practitioners, family doctors, physicians assistants, and internists. Specific titles
are delineated when necessary.
Structure of the Dissertation
This dissertation follows a six-chapter format and includes a Table of Contents, List of
Tables and Figures, Glossary of Terms, Appendices, and a List of References. Following this
introduction (Chapter 1), Chapter 2 discusses related literature supporting the studys aims, as
well as health and social theories supporting the studys hypotheses, methodology, and
interpretation of the results. Chapter 3 outlines the research methods and design used for data
collection (both quantitative and qualitative respectively), the analysis of the findings, and
methods to verify the findings. The results are presented in Chapters 4 (Quantitative) and 5
(Qualitative). Chapter 6 summarizes and concludes the study by discussing the findings relative
to existing literature, the identified theoretical concepts (both introduced in Chapter 2),
implications of this work, study limitations, and how the findings contribute to the current body
of knowledge, as well as suggestions for future research.
11


CHAPTER II
REVIEW OF THE LITERATURE
In the United States, mental disorders are common, and often serious, significantly
impacting the population. In 2010, the Substance Abuse and Mental Health Services
Administration (SAMHSA) estimated 45.9 million adults (ages 18 and older) had a mental
illness1, representing 20% of the total U.S. adult population (Substance Abuse and Mental Health
Services Administration (SAMHSA), 2012). Globally, mental illnesses account for more than
11% of disease burden (Ustun, Ayuso-Mateos, Chatteiji, Mathers, & Murray, 2004), with that
number increasing to 15% for developed nations, including the United States (Murray & Lopez,
1996). Depression disorders2 in particular can result in serious impairment and societal costs
(Katon, 2009; Substance Abuse and Mental Health Services Administration (SAMHSA), 2012).
The economic toll depression has in the United States is estimated at over $83 billion annually, of
which $52 billion are attributed to workplace costs including absenteeism or productivity
impairment (Greenberg & Bimbaum, 2005). The intrapersonal toll of depression includes
continuous feelings of diminished interest in previously enjoyed activities, recurring feelings of
guilt, chronic fatigue, difficulties concentrating, sleep disorders, weight gain or loss, and recurrent
thoughts of inflicting pain on oneself and/or suicide (Centers for Disease Control and Prevention,
2011a). Both the World Health Organization International Classification of Diseases (ICD-10)
and the American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders
(DSM-IV) list criteria for depression with some general agreements on the etiology of the
disorder (TABLE II. 1).
1
Mental Illness is defined as currently or at any time in the past year having a diagnosable mental, behavioral, or
emotional disorder (excluding substance abuse) of sufficient duration to meet diagnostic criteria specified within the
Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; American Psychiatric Association).
2 Major depressive disorder will be referred to as depression in this chapter.
12


TABLE II. 1: Diagnostic Criteria for Depression/Depressive Disorder from the World Health
Organization ICD-10 (WHO, 1990) and the American Psychiatric Association DSM-IV (APA,
2000)
Clinical
significance
Duration of
symptoms
Severity
ICD 10 Depressive disorder
Some difficulty in continuing with ordinary work
and social activities, but will probably not cease
to function completely in mild depressive
episode; considerable distress or agitation, and
unlikely to continue with social, work, or
domestic activities, except to a very limited
extent in severe depressive episode
A duration of at least 2 weeks is usually required
for diagnosis for depressive episodes of all three
grades of severity
Depressed mood, loss of interest and enjoyment,
and reduced energy leading to increased fatigue
and diminished activity in typical depressive
episodes; other common symptoms are:
DSM-IV Major depressive
disorder
Symptoms cause clinically
significant stress or impairment in
social, occupational, or other
important areas of functioning
A major duration of the day, nearly
every day, for at least 2 weeks
Five or more of the following
symptoms; at least one symptom is
either depressed mood or loss of
interest or pleasure:
(1) Reduced self-esteem/self-confidence
(2) Reduced concentration and attention
(3) Recurring feelings of guilt and unworthiness
(4) Bleak and pessimistic views of the future
(5) Suicidality/self-harm
(6) Not enough or too much sleep
(7) Change in appetite
(8) Recognizable changes in speech
For mild depressive episode, 2 of most typical
symptoms of depression and of the other
symptoms are required
For moderate depressive episode, 2 of 3 of most
typical symptoms of depression and at least 3 of
the other symptoms are required.
For severe depression episode, all 3 of the
typical symptoms noted for mild and moderate
depressive episodes are present and at least 4
other symptoms of severe intensity are required
(1) Depressed mood
(2) Loss of interest
(3) Significant weight loss or gain
or decrease or increase in
appetite
(4) Insomnia or hypersomnia
(5) Psychomotor agitation
(6) Fatigue or loss of energy
(7) Feelings of worthlessness or
excessive or inappropriate guilt
(8) Diminished ability to think or
concentrate
(9) Recurring Suicidality/self-harm
Adapted from Gilbody (2011)
In the US, chronic conditions, which require ongoing treatment and self-management, are
the most commonly seen in health care settings (Grumbach, 2003; Ridgeway, et al., 2014) and are
responsible for nearly 70% of all mortality affecting approximately 133 million adults (National
Center for Chronic Disease Prevention and Health Promotion, 2012). The rate of people living
13


with chronic conditions is increasing faster than originally predicted, supporting the critical need
for more research around improving chronic care. In 2010, approximately 75 million people in
the U.S. fit the definition of complex patients, defined as having multiple (two or more)
coexisting chronic illnesses, conditions of long duration and characterized by ongoing
medical attention and/or limit activities of daily living (Eton, et al., 2015; Parekh & Barton,
2010; Warshaw, 2006; World Health Organization, 2013). Chronic illnesses are costly,
accounting for nearly 65-80% of total national health care expenditures in the U.S., (Anderson,
2010; Guthrie, 2014; Parekh & Barton, 2010; Robert Wood Johnson Foundation, 2010; Ward,
Schiller, & Goodman, 2014), and contribute an undue individual medical burden by decreasing
quality of life, increasing levels of psychological distress, and social deprivation (Creed, et al.,
2002; Fortin, Bravo, Hudon, Fapointe, Almirall, et al., 2006; Fortin, Bravo, Hudon, Fapointe,
Dubois, et al., 2006; Fortin, Soubhi, et al., 2007).
The bi-directional effects of depression and chronic illness are well documented with
depression contributing to the development and progression of various physical illnesses, and
physical illnesses increasing the risk of depression (Steptoe, 2007). Depression is prevalent
among patients with competing chronic illnesses and adversely impacts self-care (e.g., patient
initiated behaviors like exercise, diet, medication adherence) of chronic disease resulting in
increased resource utilization (e.g., expensive secondary care referrals) (Gilbody, 2011; Guthrie,
2014). Depression in people with coexisting illnesses impairs functioning by (1) amplifying
reactions to somatic symptoms; (2) reducing motivation to care for physical ailments; and (3)
limiting the energy, and cognitive capacity to cope with physical illness while increasing
subjective senses of shame or social stigma (Creed & Dickens, 2006; Creed, et al., 2002;
Martucci, et al., 1999). The prevalence of depression coexisting with other chronic conditions is
staggering, a problem compounded by the fact that depression is rarely detected and treated
adequately in people who have physical illness (Creed & Dickens, 2006, p. 3). Though
evidence-based treatments for depression are available and have been shown to improve
14


depressive symptoms for the chronically ill, many, if not most, complex patients continue to
suffer from untreated depression.
Further complicating the deleterious effects of coexisting depression and chronic physical
illness are economic disadvantages, which are associated with the initial onset of depression
symptoms and worse prognosis (Gilman, Kawachi, Fitzmaurice, & Buka, 2002; Gilman, Trinh, et
al., 2013). Although mental illness is a ubiquitous problem affecting all races and ethnicities
(U.S. Department of Health and Human Services, 2001), economic and social disadvantages
magnify the consequences of mental illnesses, and limit access to adequate care, leading to
mistrust of the medical system (Alegria, et al., 2002; Alegria, et al., 2008; Smedley, Stith, &
Nelson, 2003; U.S. Department of Health and Human Services, 2001). One way that mental
illness compounds a persons economic status is the additional health costs that accrue, including
estimates showing between a 33 to 170 percent increase in monthly medical costs for complex
patients with a psychological disorder (e.g., depression and/or anxiety) (Guthrie, 2014; Melek &
Norris, 2008). Being economically disadvantaged is also linked to reduced access to mental
health treatment, and poorer clinical depression outcomes. (Gilman, Bruce, et al., 2013; Gilman,
Fitzmaurice, et al., 2013). Potentially, the relationship between being economically
disadvantaged and poorer mental health is mediated by chronic exposure to adverse life situations
(e.g., unemployment, low wages, housing conditions), and little or no social support (Clarke,
2009; Clarke & Currie, 2009; Wilhelm, Mitchell, Slade, Brownhill, & Andrews, 2003). The
World Federation for Mental Health (2012) identified additional risk factors for depression such
as low education level and exposure to violence, which is more prevalent in economically
disadvantaged populations.
The problems faced by patients with depression and chronic illness have highlighted the
need for research to address issues around (1) access to adequate mental health care, (2) equity of
access to distribute mental health care, and (3) the effectiveness of mental health services
including better collective management of mental illnesses and eliminating inappropriate
15


allocation and use of psychotropic medications (Gilbody, 2011). Concurrently, the U.S.
Department of Health and Human Services identified areas to address including cultural, social,
and economic factors contributing to mental health care disparities including stigma, poverty, and
the overall financing of services (U.S. Department of Health and Human Services, 2001), along
with integrating mental health services into primary care (Galson, 2009). Consistent with these
recommendations, this dissertation research explored the potential effect coexisting illness
severity has on depression symptom improvement in economically disadvantaged primary care
patients using secondary data from a completed RCT on depression treatment (Keeley & Brody,
2007).
The following background sections include a review what is known about the connection
between chronic health conditions and depression, delivery of behavioral health services in
primary care, challenges in caring for medically- and economically-indigent complex patients
with depression, the complications faced by primary care providers caring for medically- and
economically-indigent complex patients, as well as the theoretical assumptions guiding the
studys methodology.
Chronic Illness and Depression among Medically-Indigent
Populations in Primary Care
Given the increased susceptibility to mental disorders, lack of access to mental health
resources, and numerous barriers to care, developing and implementing strategies to better
manage depressive disorders for economically- and medically-indigent2 individuals is a public
health imperative. One setting that can have a significant impact on mental health disparities seen
in populations of low socioeconomic status (SES) is primary care. However, the barriers to
adequate detection and treatment of depression in primary care can be challenging for PCPs
caring for socially and economically disadvantaged; disadvantages typically associated with
2
Some of the literature on medically-indigent populations only includes those individuals with no health insurance or
means to pay for health care. For the purposes of this study, medically-indigent will include those individuals
enrolled in safety-net programs for indigent populations including but not limited to MEDICAID.
16


poorer access to care and limited health care resources including little or no health insurance (i.e.,
medically-indigent) (Bovbjerg & Kopit, 1986).
Health outcomes follow SES gradients related to socioeconomic status (SES) from those
living in poverty to those individuals with relatively high SES with the most affluent having
better health outcomes compared to the less advantaged (Glazier, Agha, Moineddin, & Sibley,
2009; Macintyre, 1994; Pickett & Pearl, 2001). The relationship between chronic disease
prevalence and SES is clear, as is the correlation between SES and the risk factors for chronic
disease (Adler & Ostrove, 1999). Low SES is associated with increased risk for multiple health
problems leading to an increase risk in premature morbidity and mortality (Inglis, Ball, &
Crawford, 2005; Lantz, et al., 1998; Marmot, Kogevinas, & Elston, 1987; Pampel, Krueger, &
Denney, 2010; Stringhini, et al., 2010). Economically disadvantaged populations not only suffer
disproportionally from adverse health outcomes and behaviors (Stringhini, et al., 2010; Walsh,
Seen, & Carey, 2013) but concurrently face personal and contextual barriers to successful
treatment of those diseases (Mansyur, Pavlik, Hyman, Taylor, & Goodrick, 2013).
Existing research suggests psychosocial factors have an impact on mental disorders (e.g.,
depression) even after controlling for genetic factors (Piccinelli & Wilkinson, 2000; Stansfeld &
Rasul, 2006). Individuals in the lower socioeconomic strata (i.e., poverty) experience
significantly higher rates of psychological stress and mental disorders including depression
(Akhtar-Danesh & Landeen, 2007; Hudson, Neighbors, Geronimus, & Jackson, 2012; Kohn,
Dohrenwend, & Mirotznik, 1998; Piccinelli & Wilkinson, 2000; Saraceno, Levav, & Kohn,
2005). In 2010, The Office of Minority Health reported adults living below the poverty level (as
determined using the U.S. Census Bureau computations) were three times more likely to have
serious psychological distress as compared to adults over twice the poverty level (The Office of
Minority Health, 2013).
The predominance of mental disorders, especially depressive disorders, among people of
lower SES has been hypothesized to be explained by financial disadvantages, high rates of
17


unemployment (Mossakowski, 2008), limited educational opportunities and low standard of
living (Stansfeld & Rasul, 2006), all demonstrated to have proximal effects on risk for
depression. Exacerbating these disparities for less affluent populations are lower access and
availability of mental health services (U.S. Department of Health and Human Services, 2001) for
certain racial/ethnic groups. Recent reports demonstrate the reduced utilization rates and
adherence to mental health treatment for black and Latino groups in rates, noting that the needs of
minority racial/ethnic groups remain unmet (Chow, Jaffee, & Snowden, 2003; U.S. Department
of Health and Human Services, 2001; US Department of Health and Human Services, 1999).
Although the relationship between racial/ethnic populations, low SES, and mental health
utilization is complex, several factors are clear:
Poorer communities inhabited by historically marginalized racial/ethnic groups do not
have the community resources to recognize and treat mental health (Chow, et al., 2003;
U.S. Department of Health and Human Services, 2001);
Patients may expect mistreatment due to perceived discrimination and prejudice (Wang,
et al., 2005);
Patients may not trust the medical system (Katon, 2003); and
Provider-level factors including competing demands, insufficient reimbursement
protocols for mental health care, and inadequate training and experience treating mental
health disorders are exacerbated in poorer communities (Klinkman, 1997; Nutting, et al.,
2000; Wang, et al., 2005).
Figure II. 1 illustrates the relationships between those levels and how they potentially hinder
depression symptom improvement, including the potential relationship posed by patients not
receiving adequate care on the overall health care system (e.g., increased cost, increased
emergency room visits).
18


Svstem-level Patient-level
Lack of mental health services for medically- indigent populations Competing demands from other chronic illness(es)
Overburdened social service staff Health service utilization
Insufficient reimbursement protocols for mental health care Psychological distress
Provider-level Sociodemographics
Feeling unprepared to adequately Perceived discrimination
discuss and/or treat mental illness (clinical inertia) Allostatic load
Inadequate detection and treatment of mental illness
Competing demands caring for patients' other illness(es) f
Burnout
Direct Effect
Feedback Effect
Poorer Outcomes
FIGURE II. 1: Factors Contributing to Depression Outcomes in Primary Care
Addressing system-, provider-, and patient-level factors contributing to health outcomes
of medically-indigent populations, including the Medicaid populations, becomes even more
important with recent legislation projects expanding Medicaid coverage to 15 million more
people by 2019 (National Association of Community Health Centers, 2010).
19


Medical Illness Comorbidity and Depression
A substantial amount of evidence demonstrates that individuals with psychological
disorders (e.g., depression) are disproportionately affected by chronic disease, often contributing
to poorer outcomes for both chronic disease management and mental health improvement
(Benton, Staab, & Evans, 2007). Concomitant physical and mental illnesses are associated with
poor treatment (both physical and psychological) response, lost work productivity, occupational
disability, lower reported quality of life, increased participation in health compromising
behaviors, and increased healthcare costs (Benton, et al., 2007; Gonzales, et al., 2011; Iosifescu,
2007; Katon, 2003; Pagoto, 2011; Steptoe, 2007). Though research suggests the existence of a bi-
directional relationship between depression and chronic illness, understanding the contextual and
latent factors that make up the relationship is imperative to improve both physical and
psychological health outcomes.
Recently, research on the relationship between depression and physical illness has
increased, along with interventions designed to improve depression recognition and treatment for
PCPs. However, these studies did not aim to understand the potential bi-directional relationship
between chronic illness and depression, focusing rather on depression as a cause and/or
consequence of a chronic illness, as well as depression being associated with poor outcomes of
medical illness and increased mortality (Iosifescu, 2007). Examples of this unidirectional
relationship include depression caused by obesity (Clark, Cargill, Medeiros, & Pera, 1996; Linde,
et al., 2004; Pagoto, et al., 2007), type 2 diabetes (Anderson, Freedland, Clouse, & Lustman,
2001; Black, Markides, & Ray, 2003; de Groot, Anderson, Freedland, Clouse, & Lustman, 2001;
Katon, et al., 2005; Zhang, et al., 2005), cardiovascular disease, a disease that depression
independently predicts (Barth, Schumacher, & Herrmann-Lingen, 2004; Kronish, Rieckmann,
Schwartz, Schwartz, & Davidson, 2009; van Melle, et al., 2004), and cancer (Spiegel & Giese-
Davis, 2003). Since understanding the association between depression and the overall burden of
comorbid medical illness is critical, an equally important area of research pertains to the course of
20


depression in complex patients. Though Benton and colleagues (2007) as well as Pagoto (2011)
have extensively studied the relationships between psychiatric illness and physical illness, this
dissertation includes a review of some of those findings with concentration on relevant illnesses
seen in the population recruited for this study.
Prevalence of Depression Comorbid with Specific Medical Illnesses
A large body of literature describes the prevalence of depression among complex patients
in primary care. Most of the studies look at the relationship between depression and one
particular chronic disease, primarily the most common illnesses seen in primary care (i.e.,
obesity, type II diabetes, and chronic pain). The studies identified as relevant to this dissertation
research are described below to demonstrate the coexistence of depressive disorders and chronic
medical illness.
The prevalence of obesity in patients with diagnosed psychiatric illnesses has rapidly
increased compared to the general population (Allison, et al., 2009). Obese individuals not only
suffer disproportionately from psychiatric illnesses, but obesity appears to be a risk factor for
various mental problems like depression (Pagoto, et al., 2011). The association between
depressive disorders and obesity is shown in both epidemiological and clinical studies. Simon
and colleagues using a U.S. representative sample showed a significantly higher percentage of
obese adults met the criteria for depression compared to their non-obese counterparts (Simon, et
al., 2006). One important finding from studies on obesity and depression is the relationship
between increasing degrees of obesity with increasing severity of depression. Psychological
factors could potentially explain the association between obesity and depression. Given the
association of depression and obesity, assessments of depression for obese individuals in clinical
settings (e.g., primary care) should be part of standard care, especially given depression could
contribute to further weight gain, as well as patient decision making around weight loss
treatments, and patients confidence for weight loss success and management. Though
21


antidepressant medication treatment options are available for obese patients with depressive
symptoms, psychotropic medication has been linked as a residual contributor to the growing
obesity epidemic in the U.S., given weight gain is a side effect of antidepressants, which can lead
to medication noncompliance (McAllister, et al., 2009), and further psychological frustration felt
by the patient.
Type 2 diabetes (T2DM) continues to be a major public health problem estimated to
affect more than 435 million adults worldwide, and a 20% increase is projected in developed
nations over the next 20 years (Gonzales, et al., 2011). Optimal treatment for T2DM is
significantly dependent on patient behavior (lifestyle factors like diet and exercise) and treatment
adherence to blood glucose monitoring, and prescribed medications related to their counterparts
without depression. Individuals with T2DM commonly show symptoms of depression and
distress, which has been associated with poorer treatment adherence and increased mortality
(Black, et al., 2003; Gonzalez, et al., 2008). Furthermore, depression and diabetes mellitus seem
to have a bi-directional relationship, with depression symptoms preceding the development of
diabetes (Gonzales, et al., 2011; Knol, et al., 2006) and with development of T2DM followed by
an increase in depressive symptoms (Bogner, Morales, de Vries, & Cappola, 2012). Though
available literature does not offer definitive evidence of biological pathways mediating the
relationship between diabetes and depression, it does show that the health compromising
behaviors (HCBs) often associated with the development of T2DM are prevalent in those
individuals either clinically diagnosed with depression or exhibiting depressive symptoms.
These HCBs include tobacco use, poor diet, lack of physical activity, and medication non-
adherence (Gonzales, et al., 2011; McClave, et al., 2009). Additionally, while pharmacological
interventions for depression have been recommended for patients with diabetes (including
T2DM) (Goodnick, 2001; Lustman, et al., 2000), antidepressant medications frequently lead to
undesired side effects including weight gain and hyperglycemia both serious complications for
patients with diabetes (Gonzales, et al., 2011).
22


Chronic pain continues to be one of the most commonly seen medical conditions in
primary care, and ranks as the 3rd most prevalent complaint for all primary care visits (Upshur,
Luckmann, & Savageau, 2006; Van Dorsten & Weisberg, 2011). Adding to the burden of pain
are the feelings of depression commonly associated with chronic pain (Fishbain, Cutler,
Rosomoff, & Rosomoff, 1997). The symptoms of depression (e.g., loss of pleasure in everyday
activities, sadness, hopelessness, and fatigue) are often the residual effect of persistent pain, and
of debilitating episodes experienced regularly by patients with chronic pain (Van Puymbroeck,
Zautra, & Harakas, 2006). The strong association between pain and depression also contributes
to adverse immune functioning critical to physical and mental illness improvement. Literature on
illnesses coexisting with depression identifies pain as the most frequently seen concomitant
complaint (Van Dorsten & Weisberg, 2011), with some rates as high as 30-56% of patients with
depression. Concurrently, patients with depression experience symptoms of pain more than
patients with cardiac disease, cancer, diabetes, and neurological disorders (Campbell, Clauw, &
Keefe, 2003; Van Dorsten & Weisberg, 2011; Van Puymbroeck, et al., 2006; Von Korff &
Simon, 1996).
Studies comparing depression treatment outcomes for patients with
and without coexisting illnesses
Some of the studies on the relationship of coexisting illness with depression symptom
improvement focused on the effects of comorbidity on antidepressant medication tolerance and
adherence (Iosifescu, et al., 2003; Koike, et al., 2002; Oslin, et al., 2002; Papakostas, et al., 2003;
Perlis, et al., 2004). Of those, two studies had an open-label design using one single
antidepressant (Iosifescu, et al., 2003; Papakostas, et al., 2003), one looked at a single
antidepressant (Perlis, et al., 2004), and three looked at different antidepressant medications to
test for rates of prescription use and adherence (Koike, et al., 2002; Oslin, et al., 2002; Simon,
Von Korff, & Lin, 2005).
23


Six additional studies examined the differences in response to depression care
interventions between individuals with depression and coexisting chronic illnesses (Bogner, et al.,
2005; Duhoux & et. al., 2009; Harpole, et al., 2005; Koike, et al., 2002; Morris, et al., 2012; Vera,
et al., 2010). Teh and colleagues (2008) evaluated the effect of chronic medical conditions on
depression diagnosis and care, including outcomes related to depression diagnosis, patient
satisfaction, patient/provider communication and continuity of care. The instruments used to
measure depression in these studies included the Composite International Diagnostic Interview
(Short-Form, and standard versions) (Kurdyak & Gnam, 2004; Teh, et al., 2008), Hamilton
Rating Scale for Depression (Bogner, et al., 2005; Iosifescu, et al., 2003; Morris, et al., 2012;
Papakostas, et al., 2003; Perlis, et al., 2004), Centers for Epidemiological Studies Depression
scale (CES-D) (Koike, et al., 2002), mental component score of the Short-Form 12 (Harpole, et
al., 2005), Structured Clinical Interview for DSM-IV/Hopkins Symptom Checklist (Simon, et al.,
2005; Vera, et al., 2010), Geriatric Depression Scale (Oslin, et al., 2002), and the Patient Health
Questionnaire-2 and -9 (Katon, et al., 2010; Vera, et al., 2010). The strategies/instruments to
measure medical illness varied by study and included noting the presence of medical illness
(Koike, et al., 2002), survey data completed by patients (count of medical illness) (Harpole, et al.,
2005; Kurdyak & Gnam, 2004; Morris, et al., 2012; Oslin, et al., 2002; Teh, et al., 2008; Vera, et
al., 2010), the Cumulative Illness Rating Scale (CIRS) which classifies comorbidities by organ
system along with the severity of each illness (Iosifescu, 2007; Papakostas, et al., 2003; Perlis, et
al., 2004), the Charlson Comorbidity Index (Bogner, et al., 2005), and computerized chart
reviews of existing medical illness (Katon, et al., 2010; Simon, et al., 2005).
Differences among the study designs and interventions make comparisons of their
findings difficult. For example, four of the studies found that coexisting illness had little or no
effect on depression remission and symptom improvement (Bogner, et al., 2005; Harpole, et al.,
2005; Papakostas, et al., 2003; Perlis, et al., 2004). Of those studies, two (Papakostas, et al., 2003;
Perlis, et al., 2004) included only patients who had treatment-resistant depression and had small
24


sample sizes (n=101 and n=92) limiting their ability to detect a difference. Two of the studies
(Iosifescu, et al., 2003; Koike, et al., 2002) showed that comorbid medical disorders had an
impact on depression symptom improvement for both antidepressant and behavioral
interventions. Iosifescu et al. (2003) investigated the role of comorbid medical illness on severity
of depression and antidepressant outcomes in depressed patients, hypothesizing that patients with
depression would experience more severe symptoms of depression which would lead to lower
rates of treatment response (>50% reduction in depression score) and remission (score <7 for
depression score at the end of the trial) compared to patients with depression but no medical
comorbidity. The sample consisted of 384 outpatients meeting DSM-III-R criteria for depression,
enrolled in an 8-week open treatment with fluoxetine. Using the CIRS to measure comorbidity,
the authors concluded that the total burden of medical illness (CIRS overall score) and the
number of organ systems affected response to fluoxetine treatment and clinical remission.
Additionally, patients with higher burden of medical comorbidity had significantly higher
depression scores at the end of the 8-week treatment, showing that medical comorbidity can
have a significant negative impact on outcome of acute treatment of depression (p. 2125).
Koike, Unutzer, and Wells (2002) examined two quality improvement programs for
depression, specifically a comparative analysis of treatment rates and outcomes for depressed
participants with and without coexisting medical conditions. The authors hypothesized that 1)
depressed patients with coexisting medical illness(es) would have worse outcomes compared to
depressed patients without a coexisting illness; 2) treatment rates would not vary significantly
between groups; and 3) a quality improvement program for depression would benefit both groups
improving treatment rates and health-related outcomes. The sample consisted of 1,336 patients
with depression randomized into one of three groups; usual care, quality improvement program
with medication, or quality improvement program with therapy. The findings showed that though
depressed patients with coexisting medical illness had similar rates of treatment, they had worse
depression outcomes compared to patients without coexisting medical illness. However,
25


improving the quality of treatment did show some improvements for depressed patients with
coexisting medical illness, offering evidence of the importance of making greater efforts to
include the medically ill when implementing quality improvement programs for depression.
Two studies (Kurdyak & Gnam, 2004; Teh, et al., 2008) examined the effects of chronic
medical conditions on the quality of depression care for persons with coexisting illness and are
described below. Teh and colleagues (2008) analyzed data from the National Survey of Alcohol,
Drug, and Mental Health Problems, a survey of community-based individuals, to describe the
use and quality of depression care for people with one or more chronic medical condition
(CMC) (p.529) as well as the potential relationship between CMCs and aspects of the patient-
provider relationship. The sample consisted of 1,309 patients with depression. The findings
indicated that patients with a CMC and depression were more likely to have their depression
recognized by a health care provider compared to patients without a CMC, but the existence of a
CMC did not impact depression treatment. A second finding from the study showed the patient-
provider relationship potentially mediates depression recognition among patients with chronic
conditions, suggesting improving the patient-provider relationship could increase the probability
of having depression recognized by a health care provider. Kurdyak and Gnam (2004) compared
utilization of mental health services as well as the quality of medication management delivered in
health care settings between depressed patients with and without CMCs. The sample consisted of
278 individuals with the diagnosis of major depression. The data showed that depressed persons
with CMCs are more likely to receive guideline-level antidepressant treatment compared to
depressed persons without a competing CMC.
Katon et al. (2010) conducted a randomized controlled trial consisting of primary care
patients diagnosed with diabetes, coronary heart disease, or both and concomitant depression.
Patients all reported at least one measure of poor disease control including high blood pressure,
high levels of low-density lipoprotein (LDL) cholesterol, and glycated hemoglobin level of
>8.5%. The primary purpose of the RCT was to determine if medical outcomes and depressive
26


symptoms would improve using a care-management intervention delivered collaboratively by
nurses and PCPs in a primary care setting. After baseline biomarker measures were collected,
nurse practitioners initiated depression treatment, which included self-care strategies,
pharmacotherapy to control depression, and motivational coaching to help patients set treatment
goals and handle barriers impeding treatment. At 12-months, the intervention group had
significantly greater overall improvement than their control group counterparts with respect to
glycated hemoglobin levels, LDL cholesterol, systolic blood pressure, and depression outcomes,
as well as significant between-group differences on three of the four disease-control measures
(Katon, et al., 2010, p. p. 2616) Additionally, intervention patients were more satisfied with their
care, and reported higher scores of quality of life, contributing to the evidence that training health
care providers in communication strategies (i.e., motivational coaching) can potentially affect
depression and disease outcomes, especially around creating dialogue around goal setting, and
competing demands.
Common suggestions from the existing research include the need for further research on
how best to recognize and treat depressed patients with coexisting illness. Some of the studies
suggest a collaborative approach to treatment including antidepressants, behavioral therapy, and
quality improvement programs being effective in this population. Though the previous studies
emphasized the importance of treating complex patients with depression, it is important to
reiterate that those reasons are amplified for medically-indigent populations given issues around
access to care, lower social support around mental health, and daily hassles competing with self-
care of medical conditions. Important to this dissertation research is to examine the hypothesis
that competing demands experienced by the patient, including their competing illnesses, may not
affect depression treatment (Ani, et al., 2009; Vyas & Sambamoorthi, 2011). A summary of the
studies described is included below (TABLE II.2).
In conclusion, studies show the potential impact of coexisting medical illness on
depression and depression treatments (both behavioral and psychotropic). Although the described
27


studies support the rationale and study design of the proposed analysis, they also help identify
important areas this research will address. For example, the proposed study included the
following suggestions made by previous research including methods and gaps to be addressed:
Complex patients >18 years of age compared to those studies focusing on older adults
(Bogner, et al., 2005; Harpole, et al., 2005; Oslin, et al., 2002),
Medical chart reviews to compile a list of coexisting illness compared to self-report
(Koike, et al., 2002; Morris, et al., 2012; Teh, et al., 2008),
With the use of the CIRS, most medical illnesses included compared to specific illnesses
identified pre-study. Additionally, the severity of those illnesses compared to a simple
count (Bogner, et al., 2005; Harpole, et al., 2005; Katon, et al., 2010; Koike, et al., 2002;
Kurdyak & Gnam, 2004; Morris, et al., 2012; Simon, et al., 2005; Teh, et al., 2008; Vera,
et al., 2010)
Whereas some of the mentioned studies only examined the potential effect of
comorbidity on medication adherence, (Iosifescu, et al., 2003; Koike, et al., 2002; Morris,
et al., 2012; Papakostas, et al., 2003; Perlis, et al., 2004), and given the potentially
harmful effects of antidepressant medications, these may not be viable options for all
complex patients, the proposed study is a secondary analysis evaluating a communication
intervention versus a pharmacologic regimen alone.
Attention to the patient-provider relationship and to the continuity of care as it pertains to
depression improvement for complex patients.
Primarily, this study aimed to address the gaps of (1) including the medically ill (especially
giving notice to the severity of their health conditions) in an intervention to improve the quality of
depression care (Koike, et al., 2002), and (2) to identify illness severity (opposed to a count of
illness) as a potential causal mechanism hindering depression symptom improvement (Benton, et
al., 2007). Additionally, this study explored patients perceptions of coexisting illness and the
influence on clinical encounters and communication, and on depression care in order to gain
perspective on the competing demands and treatment preferences that matter to patients (Bayliss,
2012).
28


TABLE II.2: Summaries of the Existing Literature since 2002 on the Effect of
Coexisting Illness on Depression
Author(s) (Year) Objective(s) Population Outcomes
Oslin, et al. (2002) To examine the relationship between specific medical illnesses and the outcomes of treatment for late-life depression. 671 older adult patients receiving inpatient treatment for depression from one of 71 psychiatric treatment facilities in the U.S. Physical disability and the total number of medical illnesses were significantly related to change in depressive symptoms. Certain somatic disorders play a role in the treatment response of late-life depression suggesting that the effect of specific illnesses on depression may be mediated by the presence of functional disability.
Koike, Unutzer, & Wells (2002) Compare treatment and outcomes for depressed primary care patients with and without comorbid medical conditions and assess the impact of quality improvement programs with medication and with therapy for these patients. Depression was measured using the CES-D Scale and chronic medical conditions using self-report. 1,356 depressed primary care patients from six managed care organizations (46 primary care clinics) from five U.S. states. At 6- and 12- month follow-up, he likelihood of having a probable depressive disorder was higher but the rates of use of antidepressant medication and specialty counseling were similar, for depressed patients with comorbid medical disorders than for depressed patients without a coexisting illness. Quality improvement programs (some were an additional cost to the patient) resulted in greater use of antidepressant medications and psychotherapy and lower rates of probable depressive disorder compared to usual care.
Papakostas, Petersen, Iosifescu, Roffi, Alpert, et al. (2003) To test whether the presence of comorbid medical conditions can predict clinical response in patients with treatment-resistant major depressive disorder treated with open-label nortriptyline (NT). Tested whether comorbidity (using the CIRS-G) predicted clinical response or depression severity at endpoint. 92 patients with treatment-resistant major depressive disorder starting a 6- week trial of NT. The results failed to confirm the relationship between comorbid medical conditions and poor outcome in the treatment of major depressive disorder for patients with treatment-resistant depression.
Iosifescu, et al. (2003) Investigate the impact of medical comorbidity on the acute phase of antidepressant treatment in patients with major depressive disorder. 384 adult depressed (determined using the Hamilton Rating Scale) outpatients enrolled in an 8-week open treatment with fluoxetine, 20 mg/day. The total burden of medical illness (using the CIRS-G instrument) and the number of organ systems affected by medical illness had a significantly negative predictive value for clinical outcome in the acute phase of treatment in major depressive disorder.
Harpole, et al. (2004) RCT to determine if the presence of multiple coexisting medical illnesses (average was 3.8 conditions) affects patient response to a multidisciplinary depression treatment program. 1,801 depressed older adults (> 60 years of age) from 18 primary care clinics from eight health care organizations in five U.S. states. The presence of multiple coexisting medical conditions did not affect patient response to a multidisciplinary depression treatment program (a trained nurse or psychologist working directly with the patient to determine a treatment option for depression that included either antidepressants or psychotherapy).
29


Author (s) (Year) Objective(s) Population Outcomes
Kurdyak, & Gnam (2004) To examine the difference in quality of care for depression between depressed persons with and without a chronic medical condition. 287 adults ages 18 to 64 with the diagnosis of major depression evaluated using the Composite International Diagnostic Interview. Depressed persons with comorbid medical conditions are more likely to receive guideline-level care for depression than are depressed persons without comorbid medical illnesses. However, the association did not persist once high utilizing patients were excluded.
Perlis, Iosifescu, Alpert, Nierenberg, Rosenbaum, & Fava(2004) To examine the moderating effect of general medical illnesses on treatment outcome in a controlled trial with patients whose major depressive disorder failed to respond to an 8-week trial of fluoxetine (treats depression, obsessive compulsive disorder, and other mood disorders; commonly known as Prozac). 386 outpatients (mean age 39.9 years) who met the criteria for major depressive disorder using the Hamilton Depression Rating Scale and had coexisting illnesses as determined by the Cumulative Illness Rating Scale. Using logistic regression analysis, CIRS score was not associated with likelihood of depression remission. Coexisting medical illness(es) does not appear to be associated with significantly poorer outcomes among patients whose major depressive disorder failed initially to respond to an initial trial of fluoxetine.
Simon, VonKorff, & Lin (2005) A longitudinal study of depressed primary care patients with and without specific co-morbid chronic medical conditions (ischemic heart disease, diabetes, chronic obstructive lung disease) to assess differences in baseline characteristics, course of depressive symptoms following initiation of antidepressant treatment, and course of functional impairment and disability. 204 primary care patients identified using health plan administrative data to identify those patients initiating antidepressant treatment. Depression severity in patients with diabetes at baseline was not affected by comorbidities but was in patients with ischemic heart disease. All groups were not significantly different in terms of social and emotional functioning, but those patients with coexisting illnesses reported greater physical impairment. Improvement in depression during treatment was strongly associated with change in disability.
Bogner, et al. (2005) To describe the influence of specific medical conditions on clinical remission of major depression in a clinical trial evaluating a care management intervention among older (> 60 years of age) primary care patients. 324 older adults and were randomly assigned to either usual care or to an intervention consisting of depression care managers offering algorithm-based depression care. Usual care showed mixed results of remission depending on the illness (e.g., patients with myocardial infarction reported faster remission compared to patients without; but slower for patients with chronic pulmonary disease). Intervention patients showed no significant associations between treatment and remission. Results suggest that the association of medical comorbidity and treatment outcomes for major depression for older adults may be determined by the intensity of treatment for depression.
30


Author(s)
(Year)
Objective(s)
Outcomes
Teh, Reynolds, &
Cleary (2008)
Ani et al.
(2009)
Katon et al. (2010)
Population
Determine of the effect of chronic
medical conditions (CMCs) on the
use of and quality of depression
care and to understand if the
patient/provider relationship
mediates the relationship between
CMCs and depression care quality.
Severity of depression was
measured by the presence of
suicidal ideation and CMCs were
identified by self-report.
Cross-sectional study using survey
data to compare guideline-
concordant treatment, and follow-up
care between primary care patients
with chronic medical conditions and
depression, and depression alone.
1309 adults in the U.S.
with Major Depressive
Disorder (MDD)
identified from the
National Survey of
Alcohol, Drug, and
Mental Health
Problems.
6.1% Medicaid
16.1% uninsured
60.5% had private
insurance (employer or
individual).
315 primary care
patients recruited from
3 separate public
primary care clinics
with depression at
baseline using the
PHQ-9. Comorbidities
were measured using
the Charlson
Comorbidity Index.
An RCT designed to determine
whether a primary care-based, care-
management intervention (self-care
and pharmacotherapy) for complex
patients, delivered in collaboration
by nurses and primary care
providers, could improve medical
outcomes (i.e., diabetes, coronary
heart disease, or both) and
depressive symptoms. A portion of
the intervention consisted of
training nurses in motivational
coaching to help patients set goals
and solve problems around
medication adherence and self-
control around diabetes
management.
214 primary care
patients identified to
have a diagnosis of
diabetes, coronary heart
disease or both
according to the ICD-9
and a PHQ-9 score of
>10 of which ~75%
were White, and ~12%
were unemployed or
disabled.
Depressed people with at least one
comorbid CMC were more likely to
have their depression recognized than
those without a CMC, though were no
more likely to receive adequate
depression care or patient satisfaction.
Additionally, aspects of the
patient/provider relationship including
trust, and continuity of care may help
explain the increased rate of depression
recognition among patients with CMCs.
No significant difference in the
likelihood of depression diagnosis,
guideline-concordant treatment, or
follow-up care in individuals with
depression alone compared to those
with concomitant chronic illnesses was
found. Severity of depression did
contribute to being diagnosed with
depression. The authors concluded that
physician depression care in primary
care settings is not influenced by
competing demands for care for other
coexisting chronic illnesses.
Those patients in the intervention group
had greater overall one-year
improvement across measurements of
diabetes control and depressive
symptoms. Other improvements for
intervention patients included higher
ratings of diabetes, coronary heart
disease, and depression care compared
to control patients as well as overall
rating of quality of life.
31


Al\w(S) Objective(s)
Population
Outcome(s)
Veraetal. (2010)
Vyas, &
Sambamoorthi
(2011)
To examine if using a collaborative
care model (including cognitive-
behavioral therapy and anti-
depressant medication) would
improve clinical and fiinctional
outcomes compared to usual care
(care managers encouraging
depressed patients to talk to their
provider about their mental health)
for complex primary care patients
with depression in Puerto Rico
(specifically to replicate other
findings outside the U.S.)
Adults with depression and at least
one chronic physical condition were
clustered into body systems (e.g.,
cardiometabolic, respiratory) to
compare treatment for depression
among individuals with multiple
chronic illnesses to a single chronic
physical condition.
179 adult (>18 years of
age), Spanish-speaking
primary care patients
with depression (PHQ-
9 and the Hopkins
Symptom Checklist-20)
and at least one other
chronic illness (e.g.,
diabetes, heart disease,
hypertension).
1,376 adults with
depression and at least
one chronic condition;
45% had at least two
conditions, 80%
identified as white, and
less 10% were
uninsured.
Depressed patients with coexisting
medical illnesses offered an evidence-
based antidepressant treatment or
cognitive behavioral therapy reported
significant reduction in depressed
symptoms and improved fiinctioning
within a collaborative care model that
included psychiatric consultation, case
management, and patient-provider
education compared to their usual care
counterparts. Furthermore, those
patients with depression and chronic
illness use of depression care
significantly increased when given the
opportunity to receive care in primary
care settings.
It was concluded that presence of
multimorbidity in the fully adjusted
model including total number of
outpatient visits was not associated with
depression treatment. The authors
concluded that competing demands
from other illnesses did not affect
depression treatment.
Morris et al.
(2012)
To determine if differences exist in
overall antidepressant treatment
outcomes based on the number of
general medical illnesses (using
self-report) in terms of depression
symptom severity, medication
tolerability, and psychosocial
fiinctioning.
Adult (18-75 years of
age) primary care and
psychiatric patients
enrolled in a RCT
comparing single
medication and
multiple medications to
treat depression.
Number of coexisting illnesses had little
or no effect on antidepressant treatment
response. Those patients with 3 or
more coexisting illnesses reported
higher rates of social and occupational
fiinctioning (using the Work and Social
Adjustment Scale). The study showed
complex patients can be effectively
treated with antidepressants regardless
of the existence and total number of
coexisting medical illnesses.
Jordan et al.
(2014)
To study the association between
multiple chronic illnesses and
receiving adequate depression
treatment.
receive adequate treatment with
antidepressants compared to patients
with depression alone. Patients with
alcohol/substance abuse were less likely
to receive either adequate
antidepressant care or continuation
phase treatment compared to those
patients with depression alone.
Administrative data
from 43,189 Veterans
Affairs patients with a
new episode of
depression.
Chronic conditions were examined
singularly. Those patients with
cardiovascular disease, peptic
ulcers/gastroesphageal reflux disease, or
arthritis were 8-13% more likely to
32


Author(s) Year Objective(s) Population Outcome(s)
Menear, Duhoux, To identify primary care practice Patient surveys from 61 Likelihood of having depression
Roberge, & characteristics associated with primary care clinics in recognized was higher in clinics with
Fournier quality of depression care in Quebec, Canada access to mental health professionals,
(2014) patients with comorbid chronic medical and/or psychiatric conditions. totaling 824 adults with depression and comorbid chronic conditions. and having at least one general practitioner at the clinic devoted to mental health was also associated with improved treatment for depression.
Stanners, Barton, A thematic analysis from interviews 12 semi-structured The interviews elicited descriptions of
Shakib, & with multimorbid patients to interviews were multimorbid patient contexts for
Winefield explore experiences of depression conducted with patients developing depression, and their
(2014) diagnosis and treatment. with two or more chronic conditions and a diagnosis of depression using a metropolitan multidisciplinary outpatient clinic. experiences of the detections and management of depression. Common themes identified included: the loss of identity, denial about the presence of depression, low self-efficacy with treatment regimens, and coping skills like exercise and pet ownership. Recommendations from the study include advising general practitioners raising the subject of mood and suggesting psychotherapy as part of their treatment.
Theoretical Approach
The patients sampled for this study were largely unemployed, poor, and marginalized,
and suffering from resource scarcity and economic and social demands that can supersede health
behavior change and treatment adherence. Consistent with the existing research on prevention
and health promotion interventions for low-income populations, it is important to recognize the
economic, social, and psychosocial contexts that potentially hinder health treatment and illness
improvement. The uses of theoretical models when developing interventions can help illuminate
the potential impacts of these tensions as well as guide the practices around recognizing and
validating patients experiences that could improve clinical communication and improve health
outcomes (e.g., depression symptoms).
33


Because I am interested in exploring the impacts of coexisting illnesses on depression
care and also the potential factors hindering or facilitating depression symptom improvement, I
have identified two theoretical models that help me frame and understand results from the
analyses, including my hypotheses.
Competing Demands
Competing demands often hinder care and lead to suboptimal communication between
patients and providers (Williams, 1998; Rost, Nutting, Smith, Coyne, & Cooper-Patrick, 2000).
Poor depression recognition and treatment in primary care has been attributed to the concept of
competing demands which influence how physicians and patients decide which problems to
address during a given visit, or over a sequence of visits (Henke, Zaslavsky, McGuire, Ayanian,
& Rubenstein, 2009; Jaen, et al., 1994; Klinkman, 1997; Nutting, et al., 2000; Rost, et al., 2000).
Given most depression care is handled in primary care settings, and most if not all patients with
depression will present at least one coexisting illness, it can be difficult to prioritize treatment, as
well as cause disagreement between the patient and provider on which illness is most important.
Within biomedicine and social medicine, one dominant conceptual model to explain suboptimal
clinical encounters and outcomes is the presence of competing demands consisting of clinical (in
the form of more pressing coexisting illnesses), and social (e.g., daily hassles, economic
adversity) (Jaen, Stange, & Nutting, 1994; Klinkman, 1997; Williams, 1998; Nutting et.al.2000).
Competing demands suggests that providers and patients each have their own (often
conflicting) priorities or agendas (Nutting, et al., 2000) they each want addressed during the
visit(s). The structure of the Competing Demands Model for the Delivery of Psychosocial Care
(Klinkman, 1997) is comprised of three domains directly influencing clinical encounters:
clinician (provider), patient, and practice ecosystem (FIGURE II.2). For the purposes of this
project, patient-level demands are the primary focus in identifying potential influences that inhibit
behavioral health care. The patient domain focuses on elements that directly influence the clinical
34


encounter. For example, at the provider-level, PCPs are required to have some level of knowledge
about the symptoms of depression and differentiating depression from other possible disorders
and along with attitudes, time constraints, and personal knowledge can influence decision-making
and quality of care (Jaen, et al., 1994; Jaen, Stange, Tumiel, & Nutting, 1997; Klinkman, 1997;
Nutting, et al., 2000; Rost, et al., 2000). Patients are expected to have similar levels of
competency about their illnesses, and be able to understand how their personal knowledge and
beliefs can be an asset or a hindrance for symptom improvement.
Though the presence of coexisting illnesses is found to affect both domains (patient and
provider), the specific dynamics are quite different. For providers, complex patients require
fundamentally different approaches to care. Treating complex patients may require more time and
a higher level of skills in order to recognize and treat physical ailments as well as mental health
problems; both identified as competing demands for adequate care. For patients, coexisting
illnesses may dictate their care agendas, conflicting with their providers list of priorities for the
visit. Having multiple illnesses can also influence patients attitudes towards their health, limiting
their level of self-efficacy creating a feeling of hopelessness, affecting treatment adherence and
the quality of the patient-provider interaction.
35


Patient Domain
Knowledge
Beliefs and attitudes
Characteristics
Coexisting illnesses
Expectations
Personal knowledge of providers

1 f
Provider Domain
Knowledge
Beliefs and attitudes
Skills
Time constraints
Alternative demands (e.g., coexisting illness)
Personal know ledge of patients
FIGURE II.2: The Competing Demands of Psychosocial Care (Klinkman 1997)
Using the model of competing demands (adapted for patient-level demands), I posit that
competing demands are one means through which patients with higher levels of coexisting illness
severity using the Cumulative Illness Rating Scale (CIRS), along with social and economic
tensions, will experience significantly less depression symptom improvement over time and be
less responsive to the intervention compared to patients with lower illness severity.
The Illness Narrative
The qualitative component of this project was guided by Kleinmans construction of
chronic illness as a disease and how patients define and experience illness (Kleinman, 1988).
What we know from qualitative and narrative research about chronic illness and its effects on the
human condition is the individuality of each experience. The domains that make up much of a
36


patients life including their work, family life, personal autonomy, self-efficacy, and psychosocial
health are influenced by the course of chronic illness, and the manifestations of that influence can
adversely impact treatment regimens and the healing process. Kleinman focuses on the illness
experience which he defines as the:
categorizing and explaining, in common-sense ways accessible to all lay
persons in the social group, the forms of distress caused by those
pathophysiological processes. And when we speak of illness, we must include
the patients judgments about how best to cope with the distress and with the
practical problems in daily living it creates. Illness behavior consists of
initiating treatment (for example, changing diet and activities, taking over-the-
counter medication or on-hand prescription drugs) and deciding when to seek
care from professionals or alternative practitioners (p. 4).
Additionally, Kleinman discusses the debilitating ways illness affects a patients life,
how a disabling illness confines, frustrates, and disappoints the patient often leading to significant
loss of interest and pleasure in participating in everyday activities a key indicator for major
depressive disorder. This experience, Kleinman argues, must be legitimized during the clinical
encounter through empathetic dialogue by the biomedical specialist (provider) in order to
effectively treat illness and establish the connection necessary for chronically ill patients. By
having dialogue with their patients, providers can gain a better perspective of how illness is
experienced both from a contextual and social perspective, assisting them in adopting the
appropriate attitudes, knowledge, and skills necessary to treat complex patients.
The collection of interviews, the tape-recorded visits, and data sources (e.g., demographic
variables, medical charts) used for this dissertation provided an understanding of patient needs
and barriers when it comes to their care. Findings from the qualitative interviews of patients
helped to inform a more comprehensive approach to patient behavioral health care, including a
better understanding of the bi-directional effects of depression and chronic illness. The
qualitative phase of this dissertation, using Kleinmans Illness Narratives as a framework, aimed
37


to collect sometimes sensitive data on the patients lived experience with their illnesses including
depression, and to better understand their needs and barriers when it comes to their care. This
supports Kleinmans assertion that treatment should begin with the systematic evaluation of the
psychosocial crises in (a patients) life experience. It should include therapeutic interventions
directed at each (emphasis added) of the major problems and integrated within a comprehensive
clinical approach (p. 73). By giving primary care patients the opportunity to voice their
experience, I hope to gain rich descriptions from the data, and information critical to the design of
future interventions for training primary care providers in mental health recognition and
treatment.
Summary of the Theoretical Frameworks
The frameworks and theories reviewed in this chapter serve as guides for much of this
dissertation, offering rationale for the aims, hypotheses, and methods purposed. For example,
though competing demands has been studied mostly at the provider-level, clinical interventions
need to account for patient-level demands contributing to health behaviors and outcomes.
Coexisting illnesses, a construct of the patient domain of competing demands, can contribute
adversely to the clinical encounter by creating a communicative disconnect between the patient
and provider as a result of conflicting agendas. Kleinmans Illness Narrative helps describe the
importance of understanding the experiences of each patient and how those experiences influence
decision-making around health and health compromising behaviors, treatment adherence and
preferences, as well as how health is prioritized.
38


CHAPTER III
METHODS
This chapter describes the general framework for this mixed-methods study, beginning
with an overview of the randomized control trial (RCT), which guided both the quantitative and
qualitative components of the study. Next, the research design, study samples, data collection
procedures, and analyses for both the quantitative and qualitative portions of this study are
discussed. Given the complexities surrounding mental health, a mixed-methods approach was
deemed most appropriate for explaining the bi-directional relationship between depression and
physical illness. Though a purely quantitative design is often appropriate for social science
inquiry, qualitative data offers rich and empathic descriptions potentially resulting in a better
understanding about social/health phenomena (Johnson & Onwuegbuzie, 2004).
Description of the Original Randomized Controlled Trial
Both quantitative and qualitative data for this study came from a RCT funded by the
National Institutes of Mental Health (Grant #s K23MH0829972 & 3K23082997-S1) conducted
at Denver Health and Hospital Authority (DHHA) from May 2010 to November 2012 (Total n =
168) (Keeley & Brody, 2007). The purpose of the RCT was to test a psychosocial strategy using
an adapted form of Motivational Interviewing (MI) delivered by trained primary care providers
(PCPs) to improve dialogue around depression with patients, and improve depressive symptoms.
The control is national standard of care treatment for depression in clinical settings which
includes assessment of suicide risk, assessment of substance abuse, and discussions with patients
around depression treatment options (Mitchell et al., 2003). MI is an empirically-based
counseling method designed to improve medication adherence for various chronic conditions.
The investigators of the RCT originally hypothesized that MI delivered by PCPs would improve
depressive symptoms for patients with a new treatment episode of depression by increasing the
39


adherence of antidepressant medication. The trial involved two phases: Phase 1 included brief
structured interviews delivered by trained providers using MI to improve treatment adherence for
depression; collection of data on the patients and providers feasibility ratings about using MI
during the clinical encounter; and assessment of treatment integrity. Phase 2 compared the MI to
standard care (Guideline Based Medical Management) to determine if MI improved adherence
and outcomes. Additional aims of the RCT investigated factors that potentially mediate change
in depression outcomes, and to determine what characterizes those who adhere to treatment and
recover from depression (Keeley & Brody, 2007). For this dissertation, secondary quantitative
data from the RCT study pertaining to depressive symptoms improvement was analyzed at 6-,
12-, and 36-weeks after baseline. Additionally, primary data from patients medical history (chart
reviews) for all 168 patients and patient interviews were collected and analyzed for this
dissertation study only. Table HIT summarizes the RCT (primary study) and the dissertation
study.
40


TABLE III. 1: Description of the Primary and Dissertation Studies
Primary Study
Title Motivational interviewing (MI) adapted
to improve depression treatment in
primary care
Description/Phases A randomized control trial using an
adapted form of MI to improve
antidepressant adherence and depressive
symptoms for patients with a new episode
of depression funded by the National
Institutes of Mental Health (NIMH).
The RCT had two phases:
Phase 1 included training primary care
providers in MI
Phase 2 compared MI to standard care to
determine if MI improved treatment
adherence and depressive symptoms
Patient populations All participants received their health care
from Denver Health and were screened
for depression using the Patient Health
Questionnaire-9 (PHQ-9).
Inclusion criteria included:
18 years or older at baseline
PHQ-9 score of >10
Confirmation of depression
Informed consent
Exclusion criteria included:
Receiving specialty mental healthcare
during the past 90 days at of recruitment
Females either pregnant or nursing
High risk for suicide
Evidence of perinatal depression, bipolar
disorder, psychosis, or active substance
abuse; and cognitive, language, or
hearing impairment severe enough to
preclude participation
Dissertation Study
Depression treatment for complex
patients in primary care: The role illness
severity in depression symptom
improvement
A secondary (quantitative aims) and
primary (qualitative aims) study the
potential effects coexisting illness
severity has on depression symptom
improvement in primary care settings.
Narrative data from complex patients
concerning their lived experience with
depression and chronic illnesses.
The dissertation had three phases:
Phase 1 scored each participants
coexisting illness severity using the
Cumulative Illness Rating Scale (CIRS)*
Phase 2 tested if CIRS had an impact on
depression symptom improvement over
time
Phase 3 conducted and analyzed semi-
structured interviews with purposefully
selected patients to assess their lived
experience with depression including
symptoms, causes, competing demands,
illness treatment priority and their overall
attitudes about living as a complex
patient and depression
All participants that completed all phases
of the RCT, and inclusion and exclusion
criteria followed the RCT protocol.
For the qualitative phases, participants
were selected based on their CIRS score
and availability to gain a broad sample of
perspectives. Informed consent was
collected for every participant.
41


Primary Study Dissertation Study
Aims Assessment of patient and provider Explored the potential effect coexisting
ratings of feasibility and acceptability of illness severity has depression over time;
an adapted form of MI Identified how patients perceive the bi-
Compare MI to enhanced usual care for directional relationship between
increasing treatment adherence and depression and chronic illness
improving depressive symptoms for Identify the competing demands that
patients experiencing a new episode of influence patients beliefs, attitudes, and
depression treatment preferences in receiving care
To explore moderators and mediators of the effect of MI on adherence and outcomes. Does not include a comprehensive measure of comorbidity severity for depression and physical illnesses
Study Setting
The study population for the RCT consisted of patients receiving their care from Denver
Health and Hospital Authority (DHHA), a comprehensive health care network that cares for about
one-quarter of Denver residents (Denver Health: About Us, 2012). DHHA is a network of eight
family health centers, two hospital-based urgent care centers, and 15 school-based health centers.
Participants were recruited from seven of the eight family health centers and from the main
hospital. The adult patient population that utilized services is approximately, 16% non-Hispanic
Black, 57% Hispanic, 19% non-Hispanic White, and 8% other, with 10% being 65 years or older.
Many of the patients at DHHA have incomes that fall below the federal poverty line and most are
enrolled in either Medicaid or the Colorado Indigent Care Program (CICP). Given that DHHA
provides health care regardless of ability to pay, services are restricted to priority care and mental
and/or behavioral health problems are often under-diagnosed and untreated. The RCT recruited
patients from seven DHHA clinics located around the Denver Metro area.
Study Participants
The RCT enrolled participants from DHHA. All participants were required to be at least
18 years of age and have had contact with a primary care provider within 12-months prior to the
study. Additionally, they had to have been diagnosed with moderate to severe depression at
42


baseline. Those patients with serious alcohol or drug addictions, no access to a telephone, or
without some proficiency in English were excluded from the RCT. All participants gave their
informed consent. Women and men 18 years of age or older who were receiving care at Denver
Health and had received a diagnosis of moderate to severe depression were included in the study.
Patients were excluded from the study for the following reasons:
Receipt of an antidepressant medication in the previous 90 days other than a low-dose
tricyclic antidepressant for pain or Trazodone for sleep
Receiving interpersonal or cognitive behavioral psychotherapy focusing on depression
Pregnant or nursing
Drug or alcohol dependency or abuse (excluding caffeine or nicotine)
High risk for suicide
Inability to communicate in English
Lifetime bipolar disease
Psychosis
History of autism, mental retardation, or pervasive developmental disorders
Cognitive, language, or hearing impairment severe enough to preclude participation.
Measures
Patient Health Questionnaire-2 and -9. The Patient Health Questionnaire-2 (PHQ-2)
(Appendix B) was used as one part of the screening process and to assess baseline depressive
symptoms. The PHQ-2 is a 2-item measure that inquires about the frequency of depressed mood
and anhedonia, defined as loss of interest in previously interesting or enjoyable activities
(Kroenke, Spitzer, & Williams, 2003). The stem question of the PHQ-2 is
Over the last 2 weeks, how often have you been bothered by any of the following
problems?
The 2 items are:
Little interest or pleasure in doing things (i.e., anhedonia) and
Feeling down, depressed, or hopeless.
43


The Patient Health Questionnaire -9 (PHQ-9) (Appendix C) is a 9-item scale that includes the two
items that make up the PHQ-2 along with the following additional items:
Trouble falling or staying asleep, or sleeping too much
Feeling tired or having little energy
Poor appetite or overeating
Feeling bad about yourself or feeling that you are a failure or have let yourself or your
family down
Trouble concentrating on things such as watching television or reading
Moving or speaking so slowly that other people have noticed
Thoughts that you would be better off dead (Suicidality)
As measures, the PHQ-2 score can range from 0-6, and the PHQ-9 can range from 0-27, each
item scored as follows:
0= not at all
1 = several days
2 = more than half the days
3 = nearly every day
The PHQ-9 can be a useful tool for diagnosing and helping patients to manage their
depression in primary care settings, mainly because it both offers a continuous measure of
depressive symptoms and a method to monitor change and treatment outcomes (Arroll, et al.,
2010). In a report of the largest validation study of the PHQ-2 and -9 in a primary care setting,
Arroll and colleagues concluded the PHQ-2 was very sensitive (0.86) and had a specificity of
0.78 for a diagnosis of depression when compared to the Composite International Diagnostic
Interview (CIDI) (2010). The PHQ-9 had similar sensitivity (0.74), but a higher specificity (0.91)
with a cutoff score of >10 (Arroll, et al., 2010), which was the cutoff score used for this study. A
2004 study of the reliability and validity of the PHQ-9 compared to the Hopkins Symptom
Checklist Depression Scale (SCL-20) reported a high rating for test-retest reliability (0.81 for
worst-case sample and 0.96 for best-case sample), and a significantly greater responsiveness at 3
months; -1.3 (95% confidence interval [Cl]) versus -0.9 (Lowe, Unutzer, Callahan, Perkins, &
44


Kroenke, 2004). Two other studies (Lowe, Grafe, et al., 2004; Lowe, Spitzer, et al., 2004) have
shown the PHQ-9 having superior criterion validity as a diagnostic measure (Lowe, Unutzer, et
al., 2004, p. 1195) compared to other self-report measures for depression.
Description of the Dissertation Study
As previously mentioned, the primary quantitative data from this study came from the
two-year prospective RCT (discussed above), as well as the patients selected for the qualitative
phase. The quantitative portion of the study reported here is a secondary analysis of the RCT data
(n=168). The primary goal of the quantitative component of this study was to determine the
potential influence coexisting illness(s) and their severity have on depression symptom
improvement over time.
Measures
Cumulative Illness Rating Scale. For this study, the Cumulative Illness Rating Scale
(CIRS) (Linn, Linn, & Gurel, 1968) (APPENDIX E) was added as a primary data collection
source to evaluate the severity of each participants coexisting illnesses, in addition to their
depression. Data derived from the CIRS are useful for quantifying and summarizing medical
illness burden.(Kemp, et al., 2010). Using medical record review and International Statistical
Classification of Diseases and Related Health Problems (ICD)-IO diagnostic and procedure code
data, physical and behavioral morbidities were identified at baseline. The comorbid diseases of
the patient was rated using the CIRS, which classifies comorbidities by 14 organ systems affected
and rates them according to their severity from 0 to 4 (Linn, et al., 1968). The scores are as
follows:
0 = no problem
1 = current mild or past significant problem
2 = moderate disability requiring first-line treatment
3 = uncontrollable chronic problems or significant disability
4= end-organ failure requiring immediate treatment (for this study, no 4s was scored
because presence of severe illnesses are including in the exclusion criteria)
45


Existing literature supports the use of the CIRS calculated from chart reviews along with high
recommendations as an available method of measuring comorbidity (de Groot, et al., 2003). CIRS
scores were only collected once, at baseline, because it was not the aim of this project to evaluate
improvement in any illnesses other than depression as measured by the PHQ-9.
In accordance with existing literature CIRS scores was stratified into three categories based
from each patients total score:
Low (0-7)
Moderate (8-14)
High (15+)
It was assumed that because patients were seeing their primary care provider for a separate
reason when recruited for the RCT, they will have at least one coexisting illness in addition to
their depression. Then, CIRS scores were categorized into three ratings for each patient:
Total Score.
Number of categories endorsed.
Severity index (total score/number of categories endorsed)
Levels of severity rating for each of the 14 systems of the CIRS were measured using a 0-
4 scoring system. Every single disease listed in the patients medical chart was classified into the
appropriate system and then given a score using the specific guidelines for each disease/system
(Hudon, Lortin, & Soubhi, 2007). Lor reference, the general rules for each potential score are
listed below (Hudon, et al., 2007):
0- No problem affecting that particular system (no disease listed)
1- Current mild problem.
2- Moderate disability or morbidity and/or requires first line therapy.
3- Severe problem and/or constant and significant disability and/or hard to control chronic
problems.
Extremely severe problem and/or immediate treatment required and/or organ failure and/or severe
functional impairment.
46


Sociodemographic information was used as explanatory variables and included age,
household income, race/ethnicity, gender, and employment status all which were collected at
baseline and self-reported. The demographic variables were included as covariates in the
statistical models. Additionally, they were used to assist with the stratified purposeful sampling
for the qualitative interviews.
Data Collection Procedure
The RCT study first randomized PCPs into MI-training (treatment) or enhanced usual
care (control) and the patients were contacted for participation in the study. The RCT and the
amendments necessary for the secondary analysis and qualitative interviews were approved by
the Colorado Multiple Institutional Review Board (Protocol# 08-1282). The research team
contacted primary care patients who met the inclusion criteria first by telephone to discuss the
study, consent to the study, and to make an appointment in accordance with a future visit to meet
in order to complete the screening questionnaires. At their clinic visit, all participants were again
asked to consent to the study. If consented, patients completed a demographic survey
(APPENDIX D), as well as the Patient Health Questionnaire (PHQ-9) (APPENDIX C), a more
thorough instrument for measuring major depression compared to the PHQ-2. The demographic
data collected included employment status, marital status, income and housing status (e.g., rent,
own), level of education completed, and race/ethnicity. The PHQ-9 was used to ascertain major
and minor depression at baseline and follow-up. PHQ-9 scores were collected at four time points,
baseline, 6-, 12-, and 36-weeks. In practice, the testing dates for each patient varied and not
every patient was screened at each time point.
The CIRS were scored using medical and pharmaceutical records by a licensed
Physicians Assistant (PA) employed at Denver Health (in accordance with the suggested
credentials necessary to score the CIRS appropriately). A physician also scored 25 patients
chosen randomly in order to compare scores with the PA to ensure fidelity along with allowing
for dialogue around scoring procedures and consistency.
47


The independent variables selected for the quantitative analyses were determined by the
literature reviewed in Chapter 2, as well as the existing RCT dataset that was collected. Since
identifying competing demands of primary care patients was an aim for this analysis, independent
variables were selected based on factors that potentially influence depression care, particularly for
complex patients.
Data Analysis
The purpose of this dissertation study was to test whether severity of coexisting medical
illnesses is related to depression symptom improvement over time for depressed complex primary
care patients. Both within-person changes in depression (i.e., intraindividual change) and
between- person changes in depression (interindividual differences) were analyzed using
hierarchical linear modeling (HLM). By using HLM, the potential differential impacts of illness
severity on depression trajectory over time were tested. Applying HLM to longitudinal data
offers the opportunity to explore theoretically different questions, which is not possible with
regression analyses or cross-sectional data (Taylor, Ntourmanis, Standage, & Spray, 2010).
To address the Aim 1, quantitative data were analyzed using Hierarchical Linear Modeling
(HLM). HLM was selected because of its fit for studying the predictors of individual change as
well as analyzing hierarchically structured data accounting for the nested structure of these
relationships including the individual, ecological-contextual, and individual-contextual
relationships often neglected in logistic and regression models (Raudenbush & Bryk, 2002;
Subrahmanian, Jones, & Duncan, 2003). Given the data collected for this study were collected at
various time points, HLM allowed nesting repeated scores within persons as an individual (Level
1) measurement in order to show individual differences in growth curves (Kreft & De Leeuw,
1998; Tabachnick & Fidell, 2007). By organizing the data into hierarchies (e.g., repeated
measures of depression nested within patients), HLM allowed for individual-level variables, as
well as group-level variables to be included in the analyses (Kreft & De Leeuw, 1998).
Additionally, because not every patient was screened at every time point, HLM was optimal
48


method of analysis of change (e.g., repeated measures ANOVA) because it did not require an
equal number of responses from each patient, meaning every patient could be included in the
analysis even with missing values (Raudenbush & Bryk, 2002; Taylor, et al., 2010).
Each time measurement of depression was nested within each patient. Equation 1
outlines the basic rudiments of the within-patient (i.e., Level 1) models used in the linear growth
models:
Yti] = noij + UPTIME P01NT)ti + eti (1)
where:
Ytij is the outcome at time t for patient
(TIME POINT)tij takes on a value of 0 at baseline, a value of 6 at 6-weeks, a value of 12
at 12-weeks, and a value of 36 at 36-weeks;
noij is the initial status of patient /. that is, the expected outcome for that patient at
baseline (when t =0);
7T, ij is the depression score change rate for patient j during the study;
and e is the error term.
The estimate parameter from the within-patient model (Level 1) will then be used as the outcome
variable in the between-patient equations (Level 2):
k0 ij = Poo + i?oi (comorbidity severity)! + p02X2 + p03X3 + eoi (2)
ij = Pio + rlt (2a)
where:
P is the regression parameter;
Xis a time-invariant predictor variable (e.g., patients level of comorbidity severity, and
gender);
and e is the error term.
Degree of variance in the study variables was explored using intercept-only models (i.e.,
49


no predictor variables were included) for all study variables, separated into two parts: variance
associated with Level-1 errors (within-patient), and Level-2 errors (between-patient). From these
models, interclass correlations coefficients (ICCs) were computed to describe the proportion of
variance associated with the between-patient level. Additionally, statistical assumptions
associated with multilevel modeling were assessed (e.g., error terms are normally distributed,
homoscedasticity, and independence of observations) (Raudenbush & Bryk, 2002). In addition to
the HLM, a Repeated Measures ANOVA was used to test the correlations between the main
variables (i.e. PHQ-9 and CIRS).
Qualitative Primary Data Research Design
In general, the qualitative aim of this dissertation was to triangulate the quantitative
findings particularly around complex primary care patients experience with depression as a
coexisting illness, as well as the competing demands expressed by patients including illnesses,
issues with employment and income, and other identified daily hassles. An additional intent of the
interviews was to inform future depression interventions about how patients regard their
depression in terms of treatment preferences and communication with their provider during the
clinical encounter, primarily among socially and economically disadvantaged adults with minimal
access to psychiatric services. Although the aims of the study are supported by the literature, the
paucity of contextual knowledge about bi-directional relationship between depression and
physical illness and their treatments from the patients perspective justified use of qualitative
research strategies, which can gamer information often not obtainable with quantitative research
methodologies (Addison, 1999; Pope & Mays, 1995).
Data Collection Procedure
Though the first draft of the interview guide (Appendix F) was developed prior to the
first interview, the guide was amended as necessary to best elicit the information to address the
aims, which is often advisable in qualitative research (King & Horrocks, 2010). The interview
50


guide was developed using an iterative process that included the author of this dissertation, three
of the dissertation committee members (KL, DM, & RK) as well as the data manager of the
original RCT. Four pilot interviews were completed followed by a modification of the question
guide that included changing particular words, order of the questions, and the number of
questions. Two more interviews were completed before the final question guide was completed.
A copy of the final interview guide is included as Appendix F. Interviews were only conducted
in English.
Using purposeful sampling techniques, patients were stratified by CIRS score (i.e., low,
moderate, high). Nineteen patients were interviewed to collect descriptive accounts about
patients attitudes about depression and depression treatment. Six of the interviews were
conducted by the author of this dissertation and the remaining 13 by the data manager of the
original RCT. Interviews were done by telephone and lasted approximately 10-20 minutes each.
The telephonic method was selected given that the patients could not be directly observed, and
the patient population at DHHA suffers disproportionately from challenges around transportation
(e.g., some patients have identified that they take several bus routes to get to their primary care
center) and childcare. The patients interviewed were compensated for their participation with
$25.00 gift cards for a local supermarket.
The interviews were audiotaped and fully transcribed for analysis. The semi-structured
format was selected to enable the opportunity for patients to elaborate about their experiences and
for the interviews to be less restrictive, while still maintaining a level of fidelity through the use
of a question guide to provide consistency in the question asked during each interview.
Patton (2001) identified six types of interview questions, each looking to elicit certain
information from the participant. The interview guides for this study will focus on three of the
categories:
51


Experience/behavior questions: These questions focused on specific actions and
reactions experienced by the patient that could be observed such as reactions to a
physical or psychological diagnosis; reactions to questions asked by the PCP about
depression and other psychological distress; and how (if at all) the patient initiates
dialogue with their provider about their concerns.
Opinion/values questions: These questions focused on what the patient thinks about
the topic of interest (e.g., depression and physical illness) and/or how their thoughts
about the topic relate to their values, goals, and intentions. For example, the
interview script will include questions about what the patient thinks is the best way to
treat depression and his/her physical illnesses; how the patient sees the relationship
between his/her psychological and physical health; and what the patients intentions
during a primary care visit given his/her current health status.
Feeling questions: These questions focused on the emotional experiences felt by the
patients. It is important to phrase these questions with the term emotion in the
question not only to elicit the patients feelings but also to differentiate them from
opinion/values questions, which Patton warns can happen without intention.
Examples include what feelings does your health status provoke in you?
Data Analysis
Data collected from the interviews (text transcribed from audio recordings) were
interpreted using a blending of suggested techniques for systematically examining qualitative
findings. Consistent with the immersion and crystallization techniques as identified by
Crabtree and Miller (1999), the initial techniques consisted of multiple, in-depth readings of the
transcripts (immersion), with a different focus for each reading depending on assumptions and
observations made during initial readings (crystallization). Other reflective techniques consisted
of dialogue with colleagues and advisors about patterns and emerging themes and to cross-check
interpretations and codes derived from the interviews (Addison, 1999; Creswell, 2009). Members
of the dissertation committee met on several occasions to discuss the themes identified from the
interviews as well as to check for compliance with the research questions posed for this
dissertation research. The analysis was concerned with both the language and the content of what
52


the patients shared during the interviews to understand their lived experience dealing with illness
from their perspective (King & Horrocks, 2010).
For this study, a variation of the three-stage thematic analysis system (FIGURE III. 1) was
adapted to code the transcribed interviews (King & Horrocks, 2010). Though the process has a
sequential order, changing and rethinking codes occurred and stages repeated to best interpret the
data. The first stage identified sections of the interviews that were relevant to the study aims.
Initial coding consisted of relatively broad a priori codes developed from the existing literature
and the qualitative aims of the study while highlighting any quotes that are relevant to aims with
comments that will help organize the text into themes. The second stage organized the broad
coding and the highlighted sections into descriptive codes, using single words or short phrases
relevant to the study aims. The descriptive codes were applied to all of the interviews and elicited
the strength of the concepts in the data and tied together any similarities across the interviews.
The third and final stage of coding entailed identifying key themes in the text by grouping
together descriptive codes and any emerging themes that shared common meanings, along with
creating a diagram that exhibited the relationships within the text (APPENDIX E). The initial
results from the qualitative analysis were reported by theme then organized by more specific
themes, and finally linked and corroborated using Deys Qualtitative Anaylsis as an Iterative
Process model (1993) (1993) (FIGURE III.l). The stages of this model emphasized a spiral
design for qualitative research analysis, less sequential and linear, with the data looping back
and forth through various phases within the broader progress of the analysis (Dey, 1993).
53


FIGURE III. 1: Iterative Process Model (Dey 1993)
Qualitative Validity
In order to improve the validity of the qualitative findings, a clear explanation of the data
collection methods and analysis was included as part of this study along with the demographic
and health data of each participant (patient) to show differences among the interviewees which
can be conducive to obtaining different perspectives (Mays & Pope, 2000). Additionally,
academic advisors and colleagues reviewed the findings to refine key themes and codes. In order
to ensure accuracy and credibility of the qualitative results, I employed procedures including
(Creswell, 2009):
Checking transcripts for mistakes made during transcription of the interviews.
Staying consistent with using codes derived from the literature and the interviews,
and documenting any changes to the code book.
Sharing the analysis with committee members familiar with the study.
Utilizing an external auditor with qualitative analysis experience but with little to no
knowledge of the study.
54


FIGURE III. 1: Description of the Qualitative Analysis
55


CHAPTER IV
RESULTS: QUANTITATIVE ANALYSIS
This chapter reports the quantitative results for the secondary analyses of the original
RCT dataset, including descriptive statistics of the variables, correlations, and hierarchical linear
regression analyses. Descriptive statistics and correlations were run using SPSS, Version 22.0
(IBM Corp, 2013), while all hierarchical analyses were run using the student version of HLM,
Version 7 (Scientific Software International, 2013).
Quantitative Patient Population
Table IV. 1 summarizes the sociodemographic characteristics of the sample. All 168
patients from the original RCT were eligible for this secondary analysis. Patient characteristics
including age, gender, income, and employment status are summarized in Table IV. 1. Of the 168
patients, the mean age of the sample was 48.9 years of age (SD=12.3) and 118 (70.2%) were
female. Non-white and Hispanics and African Americans represented the largest racial/ethnic
groups and over half of the patients (50.3%) reported an income of less than $10,000 in the past
year at baseline; over half of the sample also reported being unemployed (53.4%), and were either
looking for employment or unable to work due to physical or mental illness(es). As shown by the
demographics, a majority of the sample is low-income, unemployed (many due to mental or
physical disability), and racial/ethnic minorities, all groups disproportionately affected by
untreated mental illness and concomitant illnesses.
56


TABLE IV. 1: Patient Demographics
Total
Characteristic (n=168)
Mean fSDt
Age 48.1 (13.2)
n (%)
Female 118(70.2)
Education
Less than high school 86 (51.3)
High School 12(7.1)
Some College 14(8.3)
Vocation/Trade 40 (23.8)
Race
African American 57(33.9)
Hispanic 57(33.9)
Non-Hispanic white 39(23.2)
Other 15 (9.0)
Income
< $10,000 82 (54.7)
<$15,000 29(19.3)
< $25,000 25 (16.7)
< $35,000 8 (5.3)
> $35,000 6 (4.0)
Employment status
Working for wages 40 (23.8)
Unemployed (undisclosed reason) 26(15.5)
Unemployed (due to mental or physical illness) 62 (36.9)
Student/homemaker/other 40 (23.8)
57


Research Question 1
Research question 1 (RQ1) aimed to determine if the severity of coexisting medical
illnesses impacts depression symptom improvement over time for primary care patients
experiencing a new episode of depression over time. It was hypothesize that patients with high
competing illness severity would experience significantly less depression symptom improvement
over time and would be less responsive to the RCT than those with lower illness severity.
Depression and Illness Severity
The Patient Health Questionnaire (PHQ-9) for all 168 patients were collected and were
the main outcome of interest for the original RCT and used to measure depression change over
time for this study. The range for the PHQ-9 is 0-27, with data collected at four time points:
baseline, 6-, 12-, and 36-weeks. From baseline to 36-weeks, the mean PHQ-9 went from 15.99 to
11.15 for the entire sample (TABLE IV.2).
TABLE IV.2: PHQ-9 Mean Scores at all Four Time Points
Mean (SD) Standard Deviation 95% Confidence Interval of the Difference
Time point
Baseline 16.0 (4.2) 4.2 15.4-16.6
6 weeks 11.5 (5.3) 5.3 10.6-12.4
12 weeks 11.8(5.6) 5.6 10.9-12.8
36 weeks 11.2 (5.6) 5.6 10.2-12.2
Given that RQ1 is designed to test the potential impact of illness severity, some of the data
collected to measure illness severity are reported below: the mean scores for illness severity, as
well as data concerning the body systems affected by illness (e.g., psychiatric, vascular). For the
quantitative analysis, CIRS are reported using two separate categories:
58


Total systems affected (14 possible systems)
Total CIRS score (0-54)
Total CIRS scores were first measured by the total sample (n=168) and stratified into one
of three categories, low, moderate, or high. Among the patients in the RCT, CIRS scores were
fairly evenly distributed. All but one patient had at least a CIRS score of 1. Of the 168 patients,
35.7% reported a CIRS scored between 0 and 7 (low), 33.3% scored between 8 and 12
(moderate), and 31% scored between 13 and 56 (high) (FIGURE IV. 1). Since the distribution was
fairly evenly distributed, additional analyses are reported here to look for differences among the
sample including age, gender, and level of impairment.
FIGURE IV. 1: Total CIRS (0-56) Score Reported By Percent of die Sample
Table IV.3 reveals that there is a statistically significant relationship between age an total
CIRS score, such that as age increases so does the burden of coexisting illness severity.
59


TABLE IV.3: Total CIRS by Age (n=168)
Total Age 18-39 Age 40-59 Age > 60 x2
n % n % n % n % -
Low (0-7) 52 31.0 28 59.6 21 23.3 3 9.7 36.1***
Moderate (8-12) 56 33.3 15 31.9 30 33.3 11 35.5
High (13-56) 60 35.7 4 8.5 39 43.3 17 54.8
***p <0.001, Test: Chi-Square (Likelihood ratio statistic)
Figure IV.2 presents the means and standard deviations of the PHQ-9 scores (depression
scale) at all four time points stratified by the three levels of total CIRS scores (low, moderate,
high). As can be seen, depression scores for all three groups improved at 36 weeks. Those
patients in the moderate CIRS range (8-12) saw the most improvement over time.
60


Baseline
Total CIRS
Low
D Moderate
a High
FIGURE IV.2: Change in depression overtime stratified by total CIRS
Mean PHQ-9 at baseline was 16.0 and 11.2 at 36-weeks with a mean change of 3.47,
while the mean CIRS score was 10.8. The correlation between change in PHQ-9 at 36-weeks and
CIRS score was -0.4 (TABLE IV.4).
TABLE IV.4: Main Variable Correlations
Mean (SD) Pearson Correlation (PHQ-9 Change Over 36-weeks)
PHQ-9 Change Over 36-weeks 3.5 -
(5.5)
CIRS Total 10.8 -0.4
(5.3)
Table IV.5 reports the results from the linear regression looking at change in depression
score over 36-weeks and total CIRS. Less than 1% of the variance is accounted for change in
61


depression is accounted by total CIRS score at baseline (-0.04). This does not support the
hypothesis of the expected strong affect CIRS would have on change in depression.
TABLE IV.5: Results of Regressing Change in Depression on Illness Severity (n=125)
Step P Adj.iL F P
1 Total CIRS Score at Baseline -0.04 -0.04 0.29 0.59
Results from the Hierarchical Linear Modeling
The first HLM model that was tested was the baseline model with no predictors. The
interclass correlation (ICC), which describes the percent of variance in depression scores between
patients, was 0.69. ICC is the proportion of the between-individual variance to the sum of the
between- and within-individual variances of an outcome variable and generally ranges between 0
and 1. Hox (2002) interpreted ICC as the proportion of the variance explained by the grouping
structure in the population (p. 15). ICC can also be (roughly) viewed as the average relation
between any pair of observations (i.e., the PHQ-9 scores) within a cluster (i.e., a patient). With
this model, the intercept,ng and time,nt were examined for reliability (TABLE IV.6). The
reliability estimates represent the proportion of the variance in the Level-1 estimates that is
parameter variance. The reliability of the random effect of the level 1 intercept is the average
reliability of the level 2 units. It measures the overall reliability of the OLS estimates for each of
the intercepts.
The reliability estimates are .42 for intercept and .05 for slope. These indicate that the
slope (change in depression) is not a reliable estimate. In other words, PHQ-9 score is not a
reliable measure of a patients depression change time (TABLE IV.6).
62


TABLE IV.6: Results from the Random Coefficient Estimates
Random level-1 coefficient Reliability estimate
Intercept, nQ 0.420
TIME Slope, 7i i 0.054
For the Level-1 model (Table IV.7), each time measurement of depression was nested
within each patient. Equation 1 outlines the basic rudiments of the within-patient (i.e., Level 1)
models used in the linear growth models:
Ytij = noij + nUj(TIME POINT)tij + etij
[EQUATION 1]
Table V. 10 shows the results from a random coefficient model. The fixed results indicate
that the average PSQ-9 score at baseline was 14.1 and that the average change in slope across
patients over the four time points (BL, 6-weeks, 12-weeks, 36-weeks) was -0.1 (which indicates a
statistically significant decrease in depression overtime). It can be inferred from the random
effects results that the relationship between CIRS and change in depression over time does not
vary significantly across the patient population; that there are no differences between patients in
terms of change in depression over time slopes. This was expected given that the model did not
have reliable estimates (TABLE IV.7).
63


TABLE IV.7: Results from Depression Over Time Model
Coeff (SE) Fixed Effects t P Random Effects Variance x2 P
Mixed Model
Intercept, n0 14.13 43.34 <0.001 7.88 260.20 <0.001
Intercept, p0o (0.33)
TIME Slope, 7i, -0.10 -7.00 <0.001 0.001 159.94 >0.500
Intercept, p10 (0.01)
Deviance 3398.49
* Significant if p < 0.05
Given that the Level-1 showed no difference between patients in terms of depression
symptom improvement over time, theoretically, a Level-2 HLM model is unnecessary because it
is not expected to yield any difference from the Level-1 results. The Level-2 model was
completed for illustrative purposes (TABLE IV.8).
The estimate parameter from the within-patient model (Level 1) was used as the outcome
variable in the between-patient equation:
k0 ij = Poo j + Poij (comorbidity severity), y + /?02yX2y + /?03yX3y + eoij
[EQUATION 2]
64


TABLE IV.8: Results from the Depression Change by CIRS Total
Coeff (SE) Fixed Effects t P Random Effects Variance x2 P
Model for CIRS predicting depression at baseline
Intercept, n0 Intercept, p0o 14.1 (0.3) 43.3 <0.00 7.9 259.5 <0.001
TIME Slope, Tii Intercept, p10 -0.1 (0.0) -3.9 <0.00 0.00 159.5 >0.500
CIRS_TOT, p 0.00 (0.00) 0.7 0.47
Deviance 3409.9
* Significant if p < 0.05
As expected, the CIRS was not a statistically significant predictor of change in depression
over time, which was the first research question.
Response to Research Question 1
Given the reliability of the estimates offer little variability (TABLE IV.6), it is
unnecessary to expand the between-subjects model in order to examine the possible factors
associated with depression symptom improvement and chronic illness severity. If most of the
variability is due to error, there would likely be no systematic relations between the estimates and
the second model estimates. It could then be falsely concluded that there are no relationships
when in fact the data are incapable of detecting such relationships. For the individual growth
parameter reliability coefficients, averaging the estimates across the n individuals provides a
summary index of the instrument's reliability in measuring each of the growth parameters on this
population of subjects. The estimated variance components for the PHQ-9 reliabilities were low,
especially for the variation in the growth rate parameters for the depression assessment overtime,
meaning the depression levels for individual patients changed at a relatively constant rate across
individuals.
65


CHAPTER V
RESULTS: QUALITATIVE ANALYSIS
Cause my illnesses keep me down. Depression keeps me down. So it is just a
different form to me. It is just another added on disease, so to speak
Yeah. I mean you know, pain. Like ifI have aches and pains more, it (depression)
hurts more. If I'm you know, ifI'm sick with a cold or something, things feel like
harder. I don't know... it just... everything is harder. Let's just say everything is
bigger and harder.
This chapter reports the qualitative results from semi-structured interviews conducted
with a sample of participants from the original RCT. The chapter includes identified themes,
sample quotes, as well as simple counts of all the codes from every interview. All handling of the
data was done using ATLAS.ti Version 7.0, a software package designed to organize
qualitative data, as well as assist the process of assigning and managing codes assigned to the
text.
The primary aim of the qualitative results was not to only offer context to the quantitative
results, but to increase the overall strength of the study, and to derive data to compare with the
quantitative results as well as to maximize any similarities and the differences of information
(Creswell, 2009). The qualitative data was used to answer two of the three study research
questions:
How do complex patients describe the lived experience with depression and concurrent
illnesses?
How does a diagnosis of a new episode of depression change the management and
prioritization of concurrent illness for complex patients?
66


Qualitative Patient Population
The sample for the qualitative analysis included 19 participants using Denver Health for
their primary care with 4 being male; CIRS ranging from 6-21, ages ranging from 36-68; and
baseline PHQ-9 scores ranging from 10-25, which is representative for the larger quantitative
sample (TABLE V.l).
67


TABLE V. 1: Interview Participant Characteristics
Interview ID Gender Age BMI PHQ-9 Baseline (0-27) CIRS (0-56)
101 F 54 45 14 8
102 F 52 43 16 21
103 F 36 46 10 9
104 M 68 22.9 10 7
105 F 46 45.4 20 13
106 F 56 33.4 17 12
107 F 56 49.7 17 19
108 F 52 23.9 20 6
109 F 51 60 15 10
110 F 52 39 10 16
111 F 60 26.2 19 14
112 F 54 36.3 17 15
113 M 54 30 14 16
114 F 48 46 16 13
115 F 56 28.9 25 19
116 F 44 19.8 13 7
117 M 56 27.1 10 11
118 F 42 28.7 12 20
119 M 58 37.6 20 21
Means 52.4 36.3 15.5 13.5
The remainder of this chapter describes the major themes interpreted and analyzed from
the interview data. The qualitative research questions and specific aims are weaved within the
sections along with emergent themes that help to describe the narrative offered by participants
with depressive symptoms. The analysis resulted in eight themes and 23 categories (TABLE
VI.2) describing the participants experiences with a new episode of depression and as a complex
68


patient in primary care. The selected quotes include the respondents reported gender, age range,
and CIRS range (i.e., Female, age range 30-35, CIRS range 10-15).
Research Question 2
In order to assess participants lived experience with depression and concurrent illnesses,
participants were asked a range of questions to elicit their narrative. Questions were focused on
how participants understand, experience, and negotiate depression, along with their perceptions of
the relationship between their mental and physical symptoms and their health. For the purposes
of this study, Kleinmanns definition of the illness experience aided the development of this
section of the question guide, as well as the analyses:
Illness is the lived experience of monitoring bodily processes such as respiratory
wheezes, abdominal cramps, stuffed sinuses, or painful joints. Illness involves
the appraisal of those processes as expectable, serious, or requiring treatment.
The illness experience includes.. .the forms of distress caused by those
pathophysiological processes. And when we speak of illness, we must include the
patients judgments about how best to cope with distress and with the practical
problems in daily living it creates (emphasis added). (p.4)(Kleinman, 1988)
Themes emerged from across the data providing insight on how depression is viewed, felt
(symptoms), and if it has any impact on physical health and overall quality of life.
Perceived Causes of Depression
Given the characteristics of the study population, a priori themes were hypothesized
when analyzing the perceived causes of depressive symptoms. For example, all 168 patients
sampled for the original RCT had at least one other concomitant illness, therefore it was
hypothesized that coexisting illnesses could be contributing to the onset of depression symptoms.
Additionally, given over half of the 168 patients sampled for the initial RCT were considered
living in poverty, financial and employment burdens were also hypothesized as contributing to
depression. When asked about causes of their depressive symptoms, participants did not struggle
69


for answers. Many offered precise origins for their depression. Below, the most common causes
attributed to depression by participants are reported.
Physical health. The burden of physical illnesses was a contributing source of depression
reported by more than half of the participants interviewed. Eleven of the 19 participants
attributed their feelings of depression to the presence of physical comorbidities and these
impressions did not vary by age or CIRS. Some of the participants mentioned not only were
having multiple illnesses causing their depression, the increasing severity of those illnesses
causing feelings of distress and hopelessness.
I was depressed about my medical problems, I think its because, you know, I
have arthritis, and its genetic and I think that you know, my condition has gotten
worse. (F, 50-54, 20-24)
Well, I've been diagnosed with Addison's Disease. Quite a bit in the last year or
two years. So with that, and not being able to work... being denied a few times for
disability already, is very depressing. First of all I never thought I would be this
sick especially this early in life, somewhat. And it is really debilitating. It makes
life for me a lot more depressing... and I keep all my appointments, see all my
specialists. And work here and just maintaining with medication. It is just not
improving. And so having to deal with it every day, it is overwhelming. I am
thinking at that time, that's when the diagnoses started. I mean getting one thing
told after another that I had. And so that was the beginning of the snowball effect
for me... .1 don't get a remission. So... Yeah. Pretty depressing. (F, 50-54, 15-19)
Pain. One of the more consistent comorbidities mentioned as a hindrance to participating
and completing daily tasks was chronic pain. Chronic pain was also a significant contributor to
the participants overall feelings about their current physical and emotional states. Participants
mentioned feelings of hopelessness, failure, frustration, as well as the social and physical
limitations as a result of living in constant pain, and as a causal factor for depressive symptoms.
Just this arthritis and this aching and not feeling good all the time, you can't help
but be depressed at times. (F, 50-54, 15-19)
And you know, it (pain) just causes depression. Sometimes my kids... It's like my
gosh. I can't believe it and I start getting like real sick. And I can't even hardly
talk... you know, things that happened before and just... I don't know. Trying to
70


think the right thoughts and try to help them and stuff. It bothers me a lot. (F, 45-
49, 10-14)
But still again, those same things the aches and pains and all that stuff is
enough to bring anybody down. So... I don't know if it brings me down to a point
of being depressed, but I know it is certainly a downer to my feeling good. (F, 55-
59, 10-14)
I think I was one of those people who was always waking up and feeling in pain
and not happy with anything. Feeling hopeless. You know, I think that was me,
you know... thats depression ...it doesn t seem like anything is getting better (F,
40-45, 20-25)
Being in pain. It just puts your life on hold... its frustrating when you are in pain,
you get depressed. (F, 50-55, 20-25)
For one, failure. And you know now that we are talking about this, that's one of
the times that the pain and all that kind of stuff started affecting my body. When I
started going through some of the depression and stuff, and the loss and
everything. It just felt like my whole body not just my mind and my mood and
everything but my whole body physically was attacked. And um...Iwas so in
pain all through my body. (F, 50-55, 5-10)
Two participants acknowledged a causal pathway between their pain leading to
restrictions in their daily lives, then to feelings of sadness, frustration and depression.
Well, you know, I was depressed about my medical problems I was having and
then I was depressed about that, and that's how it came up. Being in pain. It just
puts your life on hold. And so, that's how it came up. It's frustrating. When you
are in pain, you get depressed. Yeah. Well, when I'm in pain, of course I can't do
certain things. Like my exercise is walking. It prevents me from walking if I'm in
a lot ofpain. So I get depressed. So that's one of my main problems. And it makes
me very depressed. If I can't get out(side). If I'm in pain and I can't do certain
things, I get depressed. And anyone who is feeling a lot of pain, I mean they get
depressed. It makes you not want to do anything. It just frustrates you. Do you
understand what I'm saying? (F, 50-54, 20-24)
Feeling pain all the time waking up. I dont want to do anything. Im not happy
anymore. Just pretty doubtful (F, 45-49, 10-14)
Having multiple illnesses. While some participants mentioned specific illnesses
contributing to their depression (e.g., pain), other patients explained their depression
being the result of the frustrations, stresses, and difficulties associated with suffering
71


from multiple chronic illnesses. The feelings mentioned by participants are supported by
the literature looking at complex patients. For example, many of the treatment
interventions aimed at high-utilizers of health care resources did not address multiple
conditions simultaneously and typically focused and defined by single diseases.
(Newcomer, et al., 2011)
I have a lot of medical issues and I didn't think I was going to pull through a lot
of them. And I think they had me just down and out, out of it... Like now. Every
Monday, Wednesday, and Friday ... Ugh. I don't want to go to dialysis hut I know
I need it to stay alive. (F, 55-59, 15-19)
I have Emphysema. And it has kind of worsened as of today's date. Besides I am
now a diabetic and I go to the doctor today. I don't know if they are going to start
me on pills and I also have Glaucoma. So I got a few little stresses going on here,
ok? (laughs) How I'm dealing with them, I'm just praying and just to be strong in
this... It is hurting me. (F, 60-64, 10-14)
I had a total knee replacement. Hip replacement. I had hernia surgery a month
ago Well, you know, I also have hypertension. Uncontrollable hypertension. I
had a stroke in my eye and you know, trying to keep my blood pressure under
control. The list goes on. Plus I have arthritis, osteoporosis... being in pain. It
just puts your life on hold. And so, that's how it (depression) came up. It's
frustrating. When you are in pain, you get depressed. (F, 50-54, 20-24)
One participant, when referring to how illnesses impact her overall health, stated
I am always in a state where my body hurts me which illustrates the potential
frustrations and stresses associated with being a complex patient.
Other participants reported causes of depression focused on the hassles
associated with coping with depression and concurrent illnesses on a daily basis. Kanner
et al. (1980) defined these hassles as the:
Irritating, frustrating, distressing demands that to some degree
characterize everyday transactions with the environment. They include
annoying practical problems such as losing things or traffic jams and
fortuitous occurrences such as inclement weather, as well as arguments,
disappointments, and financial and family concerns (emphasis added).
(1980, p.3)
It is these, sometimes ignored, events that when they accumulate, can have a significant
72


impact on health outcomes important to understanding a patients illness narrative. Also
referred to as life stress (Guthrie, 2014), this can both lead to depression and be an
outcome of chronic illness, causing a vicious circle for complex patients.
Financial troubles. Eight of the 19 participants reported personal finances as a
contributing to their depression. This was anticipated given the population studied is
considered living in poverty and the stresses associated with finances can be felt daily
and affect not only day to day to life, but also access to health care including physical and
psychiatric care.
Well, depression ... well the stress is not being able to afford the rent but you still
pay it. Ok? But you have to give up things. I have to choose my medications, you
know? I can't keep the best medications but for a couple months a year, because
of my deductible. And I have to settle for off-brand drugs and when I take them I
get worse through the year and then after 2 months I can straighten back up with
the good medications. And then I have to settle back down because I can't afford
it again. That's a stress. Going around looking for a place to live, cause you can't
afford where you live and the lists are like 2 years out every place you go, even
for Senior Citizens. So you get stuck in a place you can't afford, but you have to
deal with it. (F, 55-60, 10-15)
Cause they have things they want to do and I'm not able to do it money wise.
So... I just don't do it. I don't. I don't know ...it is kind of hard to put into words. I
feel inadequate a lot. Just, I don't know. Why me, you know? Can't I have a little
luck? A little stroke of luck? Can't I get one thing cured? Can't I just get enough
money, which is my money, social security, so I can just be self-sufficient and not
have to depend on others so much. (F, 50-54, 15-19)
Five participants mentioned not just financial struggles causing depressive symptoms, but the
consequences of depression as it pertains to employment and normalcy.
Because I got sick and couldn't work and I'm used to working. And uh...I was
really depressed when I got ...I knew it was a change in my life and I had to
adapt to my new circumstances. You know? (M, 55-60, 20-25)
With my oxygen no one will give me a job. That's really the thing... instead of
them looking at I really have the will to do it and I can pass all my tests, I just
have this oxygen and it's in your face. And so that is a little bit of depression
there. So it causes that no one is willing to give me a chance. And I guess that's
it. (M, 65-69, 5-9)
73


You feel better about yourself when you are making money and you can do what
you want to do and stuff like that. And not feel like people are helping me maybe
when they don't want to. I don't like when people are in my situation, but I have
no choice right now. Or putting them to the point where they have to help me or
be outdoors. They don't want that for me and neither do I. (F, 50-54, 15-19)
... when I can't work. My body also doesn't allow me to do a 40-hour work week.
And that in turn, has ...you know, I have very little income. I get (disability),
which is 175.00 a month. Which is nothing. And it just makes me feel like I'm
not... Well, I know I'm not able to take care of myself by myself. I'd like to be self-
sufficient. So that is not happening right now and I don't know when. And that is
more depressing. (F, 50-55, 15-19)
Just cause I couldn't do what I used to do before. And I don 't . .I don't make as
much money as I used to before. You know, I used to be pretty active, but now I'm
not active anymore. (M, 55-59, 20-24)
Familv/interpersonal stressors. Additional stressors associated with daily hassles
were those at the interpersonal level such as family or intimate partner relationships.
Two participants, when asked about their perception of the causes of their depressive
symptoms, reported difficulties their social environments.
(My partner) stopped helping me with child support ...I thought he would be there
to help with my children, and he didnt. I had to do it all on my own. So I think
little by little I was getting into depression about it because when the girls were
babies I used to cry a lot and them what am I going to do? (F, 50-54, 5-9)
The first time I was ever pregnant the first and only pregnancy I've ever had, I
actually had a miscarriage. And the doctors told me it was due to stress... and the
father of my child, during that week, did things above and beyond that stressed
me out no end. And because of the way the miscarriage happened, they came to
the conclusion that it was my PTSD that was the stress and the things that he was
doing purposefully that week. (F, 35-39, 5-9)
A male participant attributed his feelings of depression to loneliness due to changes in
family dynamics and a lack of communication with his children.
Yeah. So... but since I've gotten older and stuff and I'm by myself now. My kids
are grown and I dont really talk to them anymore. Or they don't call me. You
know ...I'm just pretty much by myself. (M, 55-59, 20-24)
74


Bereavement/personal loss. Of all the perceived causes of depression for those
participants interviewed, the most commonly mentioned cause was bereavement/complicated
grieving, with eight of the 19 participants mentioning death of a loved one as a major contributing
factor for their depression. Often noted as one of lifes most stressful events, bereavement,
defined as the grief that manifests from a traumatic event, in most cases death of a loved one, can
potentially affect psychosocial, social, and physical well-being and increase the risk of major
depressive disorders as well as suicidality (Latham & Prigerson, 2004).
Three of the participants shared specifically about the loss of a parent, both parents, or a
child as a major contributor to their depression. Some of the participants had been dealing with
their bereavement for some time and did not mention receiving any counseling or any other
treatment to deal with their loss.
I recently lost well, not recently but I lost my mother who I was very close
with. Who passed two days before my birthday... So it was really terrible. It got
the best of me. Cause I lost my mother. I didn Y have my mommy to go to. I didn't
have anybody. (F, 50-54, 5-9)
Because I been living like that for so long. That ain't nothing new to me. You
know what I mean? When I was first truly depressed was after my father
committed suicide and I was 19 years old. Ok? That's when I noticed it was like I
didn't want to live any longer either. And it's been since then that you know, it
just... it's just always been there, so it is not a big deal anymore. You know what
I'm saying? (F, 40-44, 5-9)
I first felt depression come on when I lost my daughter ...I became a full-time
parent to two small kids and my whole life turned around, so I was just stressed
and lost and all that stuff just got me in a down place. (F, 55-59, 10-14)
I lost a son and it started my depression really bad. I couldn't cope with it. I
didn't know how to cope with it. I had to get a lot of help to deal with it. It's been
10 years now and it still bothers me a lot so...yeah. It is a little bit better, but not
a whole lot. But... it's pretty hard. Pretty rough... when I lost my son, it was just
like I was hit by a train or something. It almost killed me like, I didn Y know. I
was like in shock Like I didn't know what to do or how to deal with it. (F, 45-49,
10-14)
75


Perceived Physical Symptoms of Depression
A significant component of understanding the patients illness narrative is how they
describe symptoms of depression. Kleinman describes symptoms as a kind of illness meaning
and these meanings to be standardized truths in a local cultural system (Kleinman, 1988, p.
10). Given the significance local meanings have on illness descriptions, it cannot be assumed that
patients descriptions will match clinical descriptions and this disconnect can impact the clinical
encounter. Though it cannot be concluded that the patients somatization around physical illness
is rooted in the psychological domain, many symptoms are the result of the integration between
physiological, psychological, and social meanings (Kleinman, 1988, p. 14). Consistent with
Kleinmans assertion, when participants were asked about their perceived symptoms of
depression, in themselves or others, described both emotional and physical responses, often
concomitantly. Though the symptoms described by participants do not differ from how
depression is clinically described including feelings of sadness or hopelessness, loss of pleasure
participating or completing activities, reduced ability to concentrate, fatigue, changes in weight,
and changes in mood, (American Psychiatric Association, 2000) participants shared an
understanding of how they classify their symptoms and an grasp of how those symptoms
perpetuate their adverse health status.
Nearly three quarters of the participants interviewed (14 of 19) focused on physical
symptoms as indicators of depression with most among the males (three of the four) mentioning
physical symptoms. A pattern also emerged by CIRS score with nine of the 10 participants with a
CIRS score of 10 or higher reporting physical symptoms. Three distinct physical symptoms
emerged from the interviews, though many had more than one type. Those symptoms are
described below.
76


Exhaustion/Fatiguc. Four of the 14 participants who mentioned physical symptoms
contributing to their depression reported exhaustion or fatigue that is persistent and hinders
participating in daily activities.
Being exhausted and not able to sleep... Just being exhausted all the time, Waking
up exhausted and not getting enough sleep. And waking up during the night. (F,
55-59, 10-14)
I knew something was wrong but I didn Y know it was depression. I would have
been saying I was just tired. For some reason, just tired all the time. You know.
Not very happy with life. Just tired. (F, 45-49, 10-14)
One participant described feeling sluggish as a symptom along with apathy and negative
feelings.
I don Y know. I don Y know how to describe it... not caring, sluggish... sometimes
bad thoughts, not much to say. I can Y explain it\ (F, 55-59, 15-19)
Pain. Another physical symptom of depression mentioned by participants was pain, not
just as a cause of depression (discussed above) but also as a manifestation of being depressed.
When comparing the causes and symptoms of depression, pain is a clear example of the bi-
directional relationship between physical and mental health, both cause and effect.
Just the way I feel and stuff, it (depression) causes me the pain and stuff. Like in
the mornings I wake up and sometimes I can Y get out of bed. And my legs are just
like paralyzed and stuff like that. I can Y even move. So I just start and like it's
coming on with pain. And try not to think of things that are going on and
happening in life. And you know that we are talking about this, thats one of the
times that the pain and all that kind of stuff started affecting my body. When I
started going through some of the depression... it just felt like my whole body- not
just my mind and my mood but my whole body physically was attacked... I was in
so much pain throughout my body. (F, 55-59, 10-14)
Perceived Emotional Symptoms of Depression
Three distinct responses emerged from the interviews in explaining how participants
perceive the emotional symptoms of depression. Along while describing the various emotions
connected with depression, participants also mentioned the social and physical consequences of
the emotional symptoms. Though the emotional responses reported by participants did not
77


deviate from the symptoms already reported in the literature, as well as the focus for instruments
aimed at diagnosing depression, the results are reported here to offer context specific to this
patient population.
Sadness. Five participants mentioned how the symptom of sadness equated to social
isolation, as well as to a lack of motivation to be active.
Being sad, not wanting to talk to anybody ...you just want to be by yourself. You
dont want nobody to bother you. Thats how I think it means, you know, just
wanting to be by yourself. (F, 40-44, 20-24)
Not wanting to do anything. Not motivated by anything. You know. Just
sad... gloomy... Its everyday life for me you know ...I mean I can tell cause I don Y
want to be bothered by nobody. I do want to do anything. I don Y want to talk to
people ...I don twant to see people. I don Y want to do anything. (F, 40-44, 5-9)
Mood swings. Insecurity. Some times negative thoughts. Not feeling your best.
Just lot of things. Cause it affects everything. (F, 50-54, 15-19)
The main thing is I will cry at the drop of a hat. Not wanting to get out of bed.
Not wanting to talk to people. Not wanting to be involved with anybody. There
are a lot of little things that go on. (F, 50-54, 5-9)
Actually, I'm pretty depressed all the time. Cause I mean... there is nothing to
look forward to. (M, 50-54, 15-19)
When you are depressed, you are more susceptible to colds and you just don Y
care and you don Y take care of yourself, so you just don't... you know... Oh for
sure. Not wanting to get out of bed, you know? So depressed and sad and
miserable. I would stay in bed for a whole day. I don Y eat. It's not good for you.
(F, 50-54, 5-9)
Hopelessness, participants also described their depression symptoms as feelings of
desperation, hopelessness, and helplessness. These symptoms are not uncommon with depression
and are including in many of the instruments to detect depression and other mental illnesses.
Referring to both depression in general and how it is experienced personally, one female
participant expressed how depression has many symptoms, including desperation, but also how
those feelings manifest into social isolation, negativity, and physical pain.
78


Well, what I understand it (depression) is feeling pretty low, that there is no help.
Feeling pretty desperate. Feeling tired all the time. Not wanting to he part of
anything. Just like living in your own little world... not wanting to he around too
many people and not wanting to do too many functions and just not feeling good
when you wake up every day- just not feeling good. Just mopey... if they are
mopey. If they have a negative attitude where everything is always negative and
stuff like that. They are not... you know... they are not happy. There is something
going on, you know ? I think I was one of those people who was from always
waking up and feeling in pain and not happy with anything. Feeling hopeless.
You know. I think that was me. You know. That's depression. (F, 45-49, 10-14)
For some of the participants, the feelings of hopelessness, and helplessness stemmed from having
to cope with their illnesses, and difficulty of coping or accepting their situations contributed to
defeated feelings.
So I might cry for a minute, you know ? And then I have to pray. And have God
just cover me so I can come hack out of it, you know? Because I'm telling you. I
do have times I sit down and there is nothing I can do. (F, 60-64, 10-14)
I just take my medications. Do my blood tests and do what I'm supposed to do,
you know? There is no cure for what I have. I mean there are things that
suppress it or ease it up. But it's there and it comes back so... it's just... I don't
know. Again, you know, I hate to keep saying Why me? But it is like just so much.
Why do I have so many things going on? I don't get a remission. So... Yeah.
Pretty depressing... my illnesses keep me down. Depression keeps you down. So it
is just a different form to me. It is just another added on disease, so to speak. (F,
50-54, 15-19)
Living as a Complex Patient
In addition to understanding how patients describe symptoms as part of their lived
experience with depression, questions were included to elucidate life experiences as a complex
patient, particularly how having multiple chronic illnesses affects a patients quality of life (QOL)
(SA3). The purpose for better understanding how patients see the relationship between their
illnesses and their QOL are two-fold: First, to recognize how having illnesses in both domains
affect the indicators associated with QOL (e.g., day-to-day life experiences). For example, if a
patient is ill and does not want to socialize though healthy social interactions are associated with
improved health outcomes; now there is a cycle contributing to adverse health. Second, to better
79


inform clinical interventions aimed at depression symptom improvement about the day-to-day
consequences of being a complex patient and how opening dialogue concerning those
consequences could lead to better outcomes. For example, if patients identify a lack of
motivation to engage in social interactions, and given the influence socializing has on health
outcomes (Kawachi & Berkman, 2001; Thoits, 2011) both clinicians and researchers working
with complex patients could include aims at increasing social interactions and community
engagement.
For the purposes of this analysis, the themes around quality of life were derived from the
literature that focused on functioning, well-being, and the patients subjective perception of their
day-to-day life experiences (Mendlowicz & Stein, 2000). For complex patients, dealing with
multiple health problems has social, physical, and behavioral implications. Having multiple
illnesses concurrently is associated with being more functionally impaired (Vogeli, et al., 2007),
and when certain chronic conditions are experienced with a mental health problem, there is a
significantly lower quality of life reported (Mujica-Mota, et al., 2015). When asked about their
own assessment of their quality of life, participants interviewed shared a similar experience,
expressing feelings of frustration, high levels of stress and anxiety, and feelings associated with
depression.
A common connection made by the participants when asked about the relationship
between their mental and physical health was the limitations and hindrances caused by dealing
with illnesses in both domains on daily activities. The responses around daily functioning were
focused on physical or social functioning or in some case a mix of both.
Physical functioning. Participants described an inability to participate in activities
viewed as a large part of a persons normalcy such as house duties, walking, daily exercise
routines, and hygiene practices. In describing how physical functioning is affected by illnesses,
including depression, shows the participants comprehension of the bi-directional relationship
80


between physical functioning and mental illness and how daily activities are important to their
quality of life.
One participant mentioned how they associate their physical activity/functioning with
normalcy, being a nobody in society and how the loss of being able to do basic, everyday
activities is a determinate of their self-value.
Of course it (physical health) affects your mental health, you know? When you go
from being a very active, productive person in society to a nobody, who can
barely take care of yourself. Of course that affects you mentally. What can you do
without your hands? Not much. Before long I won't be able to cook. I won't even
be able to wipe my own ass. What do you think it would be? I mean come
on!...you can handle anything when you are able to function like a normal human
being. I mean ... I just think about one day being on top of the world you know ?
And training for a marathon... walk and run around the lake a couple times for
my training. And then one day it stopped. It all stopped for me... if I could do all
those things again hell when I'd get upset I'd go for a run. If I was depressed
I'd go grab my bike and go for a couple mile bike ride you know? So yeah.
Getting rid of the pain would be an awesome thing. (F, 45-49, 10-14)
The following participants shared how they view the bi-directional relationship between physical
and mental illness, for example no sense of enjoyment, loss of autonomy, and mostly how
limiting their illnesses are and the frustrations they experience because of it.
Well, it slows me down, that is a prime example of mental/emotional over
physical... I still want to do things I want to do and need to do. I still try to do
those things. Whether it is chores or taking walks to the park, walking the dog,
really nice things to do, but yeah. It just slows me down. I get tired easier.
Um... thirsty more often. Sweating a lot more because everything slows down. I
get exhausted. (F, 35-39, 5-9)
... mental health affects my physical health is from the pain because I can't, you
know, do certain things and it just frustrates me. It's like getting older. So
shoot... it sucks! Well, when I'm in pain, of course I can't do certain things. Like
my exercise is walking. It prevents me from walking ifI'm in a lot ofpain. So I
get depressed. So that's one of my main problems. If I can't get out, if I'm in pain
and I can't do certain things, I get depressed. I do nothing. I just like... I don't
know what other people do, but I just like sleep... because the pain doesn't allow
me to do certain things And anyone who is feeling a lot ofpain, I mean they get
depressed. It makes you not want to do anything. It just frustrates you. Do you
understand what I'm saying? (F, 40-44, 20-24)
81


Well, I used to be able to walk around the block before I got sick I was walking
around my block you know ? Walking around and I walked around and I can't do
that now. And then now I have Osteoporosis and my back and my bones. My
wrist would shatter really easy. (M, 55-59, 20-24)
I mean it is just like a struggle to get out of bed. A struggle to do what you have
to do. Shower. Take your medications. You know... be active. Try to be active
with others and stuff like that. (F, 45-49, 10-14)
Dav-to-dav activities. When talking about having multiple conditions, three of the
participants connected their illnesses with a diminished interest or incapacity to do day-to-day
activities in a more general sense. Though similar to some of the findings associated with
physical functioning, these participants mentioned a diminished sense of motivation in a general
sense and not necessarily specific activities.
It (being depressed and physically ill) just makes me not want to do anything. (F,
45-49, 10-14)
Well, some days I just don't feel like doing nothing. So... if you don't feel like
doing anything that can affect you physically because we are designed to move.
(F, 50-54, 10-14)
It (having multiple illnesses) bothers me a lot. I have a lot of pain through the
day. And I don't know... it is hard to get stuff done. I take my medicine and it
helps me get stuff done. And I have diabetes so I have to kind of watch what I eat
and stuff. It's not easy, you know ? I don't know. I've had diabetes for over a year
and then I take medicine for that, twice a day. (F, 45-49, 10-14)
I just ache like that all the time... You know. I do what Ido and come back home.
If I don't, you know ...for instance I had a better offer today than the sale
yesterday. And now I wasn't able to do it because now I don't feel like it first of
all. And I know if I get out there, I'd be having to deal with my leg. Dragging my
leg all over town. I'm not going to do that. I don't do very much. I mean I do the
same thing. If I don't feel like doing it, I'm not going to do it. And it ends up being
90% of my daily living. It really does. Cause ifI don't want to do it, I'm not going
to do it. If I don't want to do it the next day, I'm still not going to do it. So it
just... I don't do it. I dont do anything. Oh, I'll take a shower every day. Well, two
or three times a day sometimes. But like I comb my hair. I do change clothes. I do
make my bed, but that's about it. (F, 40-44, 5-9)
Social functioning. Though participants tended to focus on how having multiple
illnesses impacted their physical health and diminished their quality of life, when asked about the
82


social implications posed by depression (and in some cases in conjunction with other illnesses),
participants often mentioned feelings of social isolation and having little to no motivation to be
around others.
Eight of the 19 participants interviewed talked about how having multiple conditions
along with depression can lead to social isolation. When referring to the effects of having to take
various medications to control her illnesses, one 50-54-year-old female expressed how having to
see her doctor frequently, she does not get out of her home very often and with my medicines,
some of them make me not able to get up...and do things... its just not a good thing at all. Two
patients talked about having to hide her depressive feelings and manifestations from friends and
family.
I mean, you know. There are times that you got friends that call and they want to
do something and you just don't want to do it. I just kind of keep a smile on and
don't say much of anything and just try to he cheerful you know ? And when it's
time to go into another room, I go into another room and cry and let it out and
stuff like that... It (depression) just makes me not want to do anything. It just
makes me not really want to do much of anything. Other than just stay by myself,
other than just be alone. (F, 45-49, 10-14)
You know... it is hard and around family, you have to put that happy face on and
try to act normal and it's hard. It's hard... (F, 45-49, 10-14)
One male participant described how his physical illness made him feel like a burden, therefore
limiting his social time with family.
Sometimes my brother might invite me to their house and I kind of get a little
irritated because I'm on that oxygen and which I could go. But I don't have a
ride. And I don't like bugging people to give me a ride. Give me a ride. Cause my
oxygen, to carry it, it only holds so much oxygen, a couple hours or so. And then,
you know, I start running out. Then I kind of panic because I'm running out of
oxygen see. And I got to get back on. And stuff like that. So I get really irritated
about that. (M, 65-69, 5-9)
Other participants expressed various ways depression (and other illnesses) pose social challenges
including how their mood impacts their interactions, staying in their homes from extensive
periods of time, and just a constant feeling of not wanting to be in social situations.
83


Um... Well, I don't know what to say about that. But I do notice when I am really
stressed and depressed a lot, I won't go out the house. And sometimes it is hard
to come out the bedroom. And that could last. That lasts for a little while, you
know? Until I can break myself out of it. It doesn't last for like a week, you
know? Because I make myself get up... and it gets hard cause it is like I can't
walk. Sometimes it gets that way, where I can't walk, but I don't know if that's
from the diabetes or what? I'm fine today. (F, 60-64, 10-14)
It does. It does. I try to say... I'm always happy when I'm out in public. I put my
best food forward. I don't let the mood swings get me too much. Sometimes when
I'm alone the mood swings come out. Or they come out at home around my dear
ones. I don't let everybody see it, but I have mood swings and I got to get me
some mental health. (F, 50-54, 15-19)
And I think they (illnesses) had me just down and out, out of it... .1 cant do a lot
ofplanning, By the time I get ready and get done with dialysis, most of the time I
just have to come home. And lay down. Go to bed early. (12)
Five of the participants felt that interactions with others were bothersome, both because
they were made aware of their limitations because of their conditions and seeing people
reminded them of that, or because their current health state negatively impacted any
desire to interact socially.
It means that I don't feel... let's see how to put it. I don't want to do anything. I
mean I do. But knowing I am not able. And since I can't, I just shut in. You know
what I mean? I mean I can't do it anyway. So... I just... plus I'm an only child. So I
was always by myself, so I'm used to being by myself. And that was what I would
prefer to do. I mean I'd like to be able to get out, but knowing that I can't just
makes me depressed more. So... it looks bad. I don't have suicidal thoughts. Iam
not on that road. I'm just on the self-mode. Closed in. Like pretty much dont
want to be bothered with anybody, because you know, I can't do anything with
them. And I'm not going to have people come and visit me all the time cause I
don't want to get out. You know what I mean? (F, 50-54, 15-19)
It is everyday life for me, you know ? I mean I can tell cause I don't want to be
bothered by nobody. I don't want to do anything. I don't want to talk to people. I
don't want to see people. I don't want to do anything. (F, 40-44, 5-9)
I'm just wanting to cave and not be around people. Or it would be fine with me if
I didn't even have to associate with people. (F, 55-59, 15-19)
If my legs worked properly or if I didn't have a constant ache in my back. You
know what I'm saying? It goes throughout my legs and it's been doing this for
quite a few years now. So... I mean I don't really want to do anything. I have to
84


take pain medication of course. And when you are on that in high doses, like I
am, what if I'm out in the public and something might happen? Cause I know I'm
not there. You see what I'm saying? (M, 55-59, 10-14)
Well, I don't know what to say about that. But I do notice when I am really
stressed and depressed a lot, I won't go out the house. And sometimes it is hard
to come out the bedroom. And that could last. That lasts for a little while, you
know? Until I can break myself out of it. It doesn't last for like a week, you
know ? Cause I make myself get up. And it gets hard cause it is like I can't walk.
Sometimes it gets that way, where I can't walk, but I don't know if that's from the
diabetes or what? I'm fine today. (F, 60-64, 10-14)
Response to Research Question 2
How do medically-indigent depressed patients describe their lived experience with
concurrent illnesses?
When describing their lived experience with depression and often multiple coexisting
illnesses, participants were able to articulate their experience with depression, giving insight and
voice to their experience, providing a useful narrative for clinicians to use for treatment and
communication (Barry, Stevenson, Britten, Barber, & Bradley, 2001). Participants described a
variety of causes for their depression including those hypothesized including coexisting illnesses
(living as a complex patient, further explored later in the chapter), and financial and employment
burdens. Other causes emerged from interviews including factors like bereavement, and
family/relationship stress. When describing symptoms of depression, participants by and large
reported those factors used as proxies in many of the instruments to measure depression,
including feelings of hopelessness, fatigue, and sadness. Though not all of these factors were
reported by a majority of the 19 participants interviewed, the disparities show a need for being
more responsive to personal narratives in the clinical encounter, aiding personalized treatment for
depressive symptoms.
Participants responses around life as a complex patient presented a rich representation of
how the participants day-to-day lives are impacted by their health. Their descriptions offer an
understanding of cyclical relationship between their physical and mental health, how it impacts
their social and psychosocial health, and in turn adds to their depression symptoms. The
85


participants mentioned a loss of self-value with losing their capacity to participate in daily
activities often viewed as important proxies for normalcy.
Research Question 3
This portion of the qualitative analysis aims to answer the third research question, how
does a diagnosis of a new episode of depression change the management and prioritization of
concurrent illness for complex patients? This set of questions looked at how patients manage
and prioritize their illnesses. Themes emerged providing some detail to how illnesses are
treated, prioritized, and the competing demands and barriers patients identify that hinder
physical and depression symptom improvement.
Illness Priority
In analyzing interviews, a salient theme emerged related to how patients prioritize
illness treatment after a new episode of depression. Participants mentioned priorities for both
themselves and their providers. Given that patients and providers were aware of the new
diagnosis of depression prior to the first visit while participating in the RCT, it was hypothesized
that depression would become a treatment priority, and though patients were interviewed after
the completion of the RCT, remain a priority even after the completion of the RCT. However,
when asked about which illnesses the patients and providers prioritized, the majority of the
participants interviewed were more likely to report that their physical ailments took precedent
over their mental health.
Patient treatment priority. When discussing their own illness priorities, participants
mainly focused on physical ailments. One female participant was explicit about not wanting her
depression symptoms to be a priority for her health care stating, I dont go to doctors for my
depression. Two participants mentioned that they prioritized their physical illnesses mainly
because they have learned to live with their mental health conditions.
86


No I was just thinking. I'd have to say my weight. Blood work is always good. I'm
healthy on the inside. So it is my weight. It hurts my joints and impedes on my life
more than anything. My mental health -1 make the choice to he happy and I
make the choice to look for the positive and the good. So I'm aware and in
control of my mental health, is the best I can do. But the PTSD -1 don't even
notice it sometimes. I'm sure other people notice it first. But yeah, I think it's just
the fat to be honest. (F, 35-39, 5-9)
Probably the other issues... the problems with my hands... I'm slowly losing the
use of my hands. What can you do without your hands? Not much. Because like I
said, my depression I've had so long that it only becomes a problem on occasion.
(F, 50-54, 5-9)
Some participants were adamant about their physical health taking priority when it came to
treatment. Two participants in particular viewed the relationship between their physical and
mental health as hierarchical.
I think physical. I would treat my physical before I treated my mental. Cause I
think my physical has a lot to do with my mental. I think if I were to go to . I have
my recreation card... I think if I were to go to the recreation center a little bit
more, when I get in those depressed states; I think if I would go and work out, it
would help a lot. So that's why I say my physical probably would overrule my
mental. (F, 50-54, 15-19)
I think probably physical health. That's the main thing. I been going through a
period of depression. And um... but I'm feeling a lot better, at this point. So I think
the muscular stuff is probably the main thing. Um... for I) I think if my physical
health were improved, then I would feel better on a daily basis and do a lot of
different things that sometimes I'm not doing because of the health issues. (F, 55-
59, 10-14)
One participant mentioned specific physical illnesses as treatment priorities.
My physical, I guess. Like now, every Monday, Wednesday and Friday ... Ugh. I
don't want to go to dialysis but I know I need it to stay alive. So... And the fact
that my diabetes has my eyesight... I can't see at night. (F, 55-59, 15-19)
Provider treatment priority, participants also described which illness(es) their providers
prioritized. The most common response given by patients was that mental health was not a
priority. Though providers were aware of their patients depressive symptoms, participants did
not always believe it was their providers priority. An obese participant mentioned that her
provider was aware of her depressive symptoms but still focused on her weight problem, shes
trying to control my weight because the weight is what is making me depressed. Two
87


participants described there being no discussion around mental health because of there being
other health problems to manage.
Pretty much when I call to make an appointment, we pretty much talk about the
things I think are important. The last time I talked to him it was for... what did I?
Oh, they thought I might have... what do you call that? They thought I might have
a hernia! Well, right now we haven't even discussed it (depression). We haven't
had no discussion in a long time. It is well over a year. I dont think we've even
discussed it (depression). We've been focused on so many other things. (F, 50-54,
15-19)
Thats to the forefront. I got other issues of course, but we ain't really worried
about those. It is just getting through this pain on a constant basis, all day, every
day. (F, 40-44, 5-9)
Another participant seemed open to changing her priorities and discuss her mental health but did
not want to trouble her provider so focused on the main issues as not to take up too much time.
Yeah, well, she (primary care provider) basically prioritizes whatever is
happening with my health at the time. You know ? And she sends me for tests and
all of that. I don't... I don't feel like I want to put that (depression) on her a whole
lot. Because she has so many patients that she needs to deal with as well. So I
don't want to take up a lot of her time, even though she would be willing to sit
there with me. But I just don't want to take up a lot of her time. You know ? Just
sitting back and discussing everything that is going on with me, I just go in and I
want her to take care of my main issues that are going on with me, like pain and
all that. (F, 55-59, 15-19)
A male participant mentioned that his providers main concern was preventing a heart attack and
was focused on managing his energy exertion in the workplace.
Not really any other illness. Well... I think my doctor, he doesn 't want me to
work. Cause he says... he used to tell me he was worried about me having a
heart attack. You know ? Because of in my body ...he was afraid I would push it
too hard. (M, 55-59, 20-24)
Depression Treatment Preferences
A particular salient finding from the qualitative interviews was that more than half of the
participants interviewed did not want to be treated for depression in a primary care setting.
Twelve of the 19 participants interviewed shared their preferences ranging from seeing a
therapist/counselor, to social support.
88


Spiritual/praver. Three of the 12 participants shared how spirituality, specifically prayer,
was an important modality for self-treatment for their depression.
I just don't want this to defeat me. You know? And I'm sure that my faith helps me
to get up in the morning, you know ? I don't want to give up regardless of my
stresses and my circumstances. I just... just have faith. I'm going to still keep my
faith and I'm still in faith for recovery, you know? Thats how I recovered from
this diabetic thing. I just... I'm just going to do what it takes to try to alleviate it.
You know ? And it helps with my stresses and my circumstances and my
depression. You know ? I don't take medication for it. I just have to pray. Pray
and praise. That's me... Yes. It helps a lot with the depression. It helps a lot with
the stresses. If you can get your mind off of yourself. I learnt that since I been on
that study. I learned how to try to just go about it another way. Read to enlighten
myself. But I'd say I'm doing much better in my depression and stresses than I
was. Cause I have a lot... I am in a place I can't afford but I can't find a place to
live. Ok? But I have to keep positive that sooner or later something is going to
happen. Well, I guess Jesus. Because I found the faith you know ? I've got
something else to release my thoughts to. And it is helping. My faith is helping
me. (F, 60-64, 10-14)
I start praying and I pray every day. I don't get to church every week like I
should because of the oxygen and stuff. But I pray every day. And my problems
and whatever, I leave it in the hands of God. Let him take care of it. And so far,
he's been taking care of me. (M, 65-69, 5-9)
Family support. Three participants mentioned support from a family member (one
mentioned it in conjunction with prayer) as a way to deal with his depression symptoms.
Well, being a Jehovah's Witness, of course, I pray about it. And I have many
friends that I can talk with and I know... you know... when I talk with my friends,
that's what happens when you are depressed. You got to have some way to vent.
And I have people that I can talk with. And then I have my oldest son, you know ?
My relationship with him is where I feel comfortable in talking with him. (F, 50-
54, 10-14)
I have friends and family that help me out. (F, 50-54, 15-19)
Well, actually, I did have a lot of depression, but now I'm a lot better. I learned
to just like walk and pretty much and talk with family and that makes me feel
better. It was bad. But then I had to snap out of it though. From feeling sorry for
myself. And make myself a better person. Getting out of the house. I go walking
to the park with my son. Just having a good day. (F, 40-44, 20-24)
89


Professional treatment (not primary care provider). Three of the twelve participants
mentioned wanting to get professional treatment for their depression in the form of counseling or
therapy.
Grief Counseling. It did. Yep. It helped. It helped. It helped a whole lot. It was
just that every year when that month comes around it seems like I go try to go
into it, depression, again. Ido counseling with my Pastor too. (F, 55-59, 15-19)
Oh, therapy. I don't believe in prescription drugs. (F, 35-39, 5-9)
I get counseling and stuff. And just talking to people and opening up instead of
closing down and stuff. Just trying to talk to somebody about it. And it helps
me... Breathe and stuff. I try to breathe. (F, 45-49, 10-14)
Opposition to medications. Two participants expressed a strong opposition to medications
for depression, both had first-hand experience with anti-depressants but either had their provider
take them off the medication or expressed concerns about their current medication regime.
I wish I didn't have to take all these pills. And like before. I was taking the wrong
kind of medications for my depression. I would sit there and count my pills I don't
know how many times. Cause I don't know, that particular pill seemed to be
affecting my way of thinking. Because it was interfering with concentration. (F, 55-
59, 15-19)
I just had him take me of my medication because I'm better without the medication.
(F, 50-54, 5-9)
These findings suggest that some participants experiencing a new episode of depression do not
want to be treated in primary care settings for their depression, which can be problematic for
interventions in these settings.
Primary care/medication. Not all participants interviewed were opposed to their primary
care provider providing depression treatment or to taking medications for their symptoms. When
asked about her treatment preference, one female participant expressed, I would love for him
(primary care physician) to treat it. That would not be a problem. Another could not see why a
medical doctor would not be responsible for mental health care.
I don't think you separate them. I've often thought that it is ridiculous that you
have to see a therapist aside from your medical doctor. I think medical doctors in
90


Full Text

PAGE 1

A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Health and Behavioral Sciences 2015 08 Fall : THE ROLE OF COMPETING ILLNESS IN DEPRESSION SYMPTOM IMPROVEMENT By ERNESTO ANDRES MORALEZ B.H.C.S., New Mexico State University 2007 M.P.H., New Mexico State University 2009

PAGE 2

ii This thesis for the Doctor of Philosophy degree by Ernesto Andres Moralez has been approved for the Health and Behavioral Science Program by Karen Lutfey, Chair Debbi Main, Advisor Elizabeth A. Bayliss Robert D. Keeley Jini Puma Date July 24, 2015

PAGE 3

iii Moralez, Ernesto, Andres ( Ph.D., Health and Behavioral Sciences ) Depression Treatment for Complex Patients in Primary Care: The Role of Illness Severity in Depression Symptom Improvement Thesis directed by Professor Debbi Main ABSTRACT Background: Growing evidence shows the co e xistence of physical chronic illnesses and mental health disorders like depression can confound and hinder physical symptom improvement and depression treatment Objectives: The purpose of this mixed methods study was examine the potential relationship between coexisting illness severity and change in depression over time ; and to identify contextual factors that influence beliefs, attitudes, and treatment preferences for depression and coexisting illness care for primary care patients experiencing a new episode of depression. Participants and Setting: 168 primary care patients experiencing a new episode of depression recruited from a primary care network in Denver, Colorado. Nineteen participants were interviewed for the qualitative portion of the stud y. Measures: Patient Health Questionnaire 9 (PHQ 9) to measure depression and the Cumulative Illness Rating Scale (CIRS) to measure illness severity Results: H ierarchical L inear M odeling analyses reveled CIRS was not a statistically significant pred ictor of change in depression over time give t he estimated variance components for the PHQ 9 reliabilities were low (Intercept = 0.42; Time S lope = 0.054) Participants described a variety of causes for their depression including coexisting illnesses as well as describing coexisting illness as a competing demand to seeking mental health care. Additionally, i n spite of a new diagnosis of depression, participants for the most part continued to focus on physical ailments and controlling existing chronic dis ease

PAGE 4

iv Conclusions: Although this study did not show a statistical relationship between chronic illness severity and change in depression, qualitative findings did suggest that illness severity has some in fluence on depression symptoms which should be fur ther tested The interviews offer some challenges to the potential assumptions of depression interventions in primary care settings that treatment. It can be th eorized that the patients interviewed for this study see their depression as the result of many of the difficulties associated with illness complexity. Though this study was unable to identify a relationship between chronic illness severity and depression symptom improvement, additional efforts are needed to improve the understanding of chronic illness burden on depression outcomes The form and content of this abstract are approved. I recommend its publication. Approved: Debbi Main

PAGE 5

v ACKNOWL EDGEMENTS First of all I want to thank my father, Alejandro Moralez, for being my hero and my biggest fan. I owe all my success to your success as a dad. I would like to thank my dissertation committee for their support and patience. Dr. Keeley for brin ging me onto his project and being a friend and supporter throughout this endeavor Dr. Bayliss for being so positive and supportive of my ideas. Dr. Lutfey for being the right kind of motivator. All the teachers, professors, and researchers that have c ontributed to my professional development ; Jeanine Voller, Mark Halling, Satya Rao, Yvonne Kellar Guenther (thank you for being not just a mentor but a wonderful friend) Jean Scandlyn, Erin Wright, Howard Waitzkin, David Tra cer and Arthur Klein man. My family. Thank you Alex for being proud of your little brother. Thank you Isabelle, Bobby, and Lilly for always having my back and supporting this long endeavor Countless friends (who tolerated my incessant self doubt and were always there to keep me on it without you doing it first); Joe Beling, Tres Morley, and Dave Troyer; Christine, Ryan, Whitn ey Jones, Debbie, Melanie, Sarah, and all the HBS doctoral student s I met along the way. Abby Fitch and Jessica Halliday (special thanks for everything); Carlos and Sonia Ray, Jerry, Pete Robert, Tomas and Angela, and all my friends in Albuquerque who w ill never let me forget what matters most ; t he Kaufmans, Jennifer Cantrell (always there to listen and get me out of trouble) ; Mike Anderson for help uploading 2.0. and seeing me through at the finish. A very special thank you to Natalie Slevin for knowi ng this day would come, even when I was A very special thank you to Jini Puma! for always supporting, helping, and smiling. Could not have done it without you A very special thank you to Ashley Foldes and the Foldes family for being unco nditionally supportive and giving me the belief in myself to keep going. A very special thank you to Gilbert Escarcida for always being there to offer a certain kind of empathy and understanding (and a place to live). A very special thank you to Indy for b eing my best friend. A very special thank you to Matthew Engel for doing so much for me. Special people and places: Commonground Golf Course for being my sanctuary, Sugar Bake Shop, Tattered Cover and Hooked on Colfax for the coffee and free Wi Fi. Ment ors along the way: Basil Walter, Max Contreras, Jon Stewart Robert F. Kennedy, and Ernesto Galarza.

PAGE 6

vi DEDICATION To my mother, Blanca Velasco, who I know is proud of her son and even more proud that I have spent so much time studying a disease that needs attention and care. I love you very much and I hope to continue to make you proud. I would not be who I am today without the love from my grandmother Lucy Caballero. me finish but call ed me doctor To Debbi Main, for rescuing me when I had no idea what I was doing, or where I was going. You took me on and made sure I found something I was passionate about and that I was developing the fortitude to finish. Thank you Debbi, I t ruly would not have done this without you.

PAGE 7

vii TABLE OF CONTENTS CHAPTER I INTRODUCTION ................................ ................................ ................................ ....................... 1 Study Pur pose and Rationale ................................ ................................ ................................ ... 3 Research Questions ................................ ................................ ................................ .............. 3 Specific Aims ................................ ................................ ................................ ...................... 4 Overview of the Research Methods ................................ ................................ ......................... 4 Study Setting and Participants ................................ ................................ ............................. 5 Significance of the Study ................................ ................................ ................................ ..... 6 Terminology ................................ ................................ ................................ ........................ 9 Structure of the Dissertation ................................ ................................ ................................ .. 11 II REVIEW OF THE LITERATU RE ................................ ................................ ......................... 12 Chronic Illness and Depression among Medically Indigent Populations in Primary Care ................................ ................................ ................................ 16 Medical Illness Comorbidity and Depression ................................ ................................ ........ 20 Prevalence of Depression Comorbid with Specific Medical Illnesses .............................. 21 Studies comparing depression treatment outcomes for patients with and without coexisting illnesses ................................ ................................ ................................ ............. 23 Theoretical Approach ................................ ................................ ................................ ............. 33 Competing Demands ................................ ................................ ................................ ......... 34 The Illness Narrative ................................ ................................ ................................ ......... 36 Summary of the Theoretical Frameworks ................................ ................................ ......... 38 III METHODS ................................ ................................ ................................ ............................ 39 Description of the Original Randomized Controlled Trial ................................ ................... 39 Study Setting ................................ ................................ ................................ ..................... 42 Study Participants ................................ ................................ ................................ .............. 42 Measures ................................ ................................ ................................ ............................ 43 Description of the Dissertation Study ................................ ................................ .................... 45 Measures ................................ ................................ ................................ ............................ 45 Data Collection Procedure ................................ ................................ ................................ 47 Data Analysis ................................ ................................ ................................ ..................... 48 Qualitative Primary Data Research Design ................................ ................................ ....... 50 Data C ollection Procedure ................................ ................................ ................................ 50 Data Analysis ................................ ................................ ................................ ..................... 52 Qualitative Validity ................................ ................................ ................................ ........... 54 IV RESULTS: QUANTITATIVE ANALYSIS ................................ ................................ .......... 56 Quantitative Patient Population ................................ ................................ ............................. 56 Research Question 1 ................................ ................................ ................................ .............. 58 Depression and Illness Severity ................................ ................................ ......................... 58 Results from the Hierarchical Linear Modeling ................................ ................................ 62 Response to Research Question 1 ................................ ................................ .......................... 65

PAGE 8

viii V RESULTS: QUALITATIVE ANALYSIS ................................ ................................ ............... 66 Qualitative Patient Population ................................ ................................ ............................... 66 Research Question 2 ................................ ................................ ................................ .............. 69 Perceived Causes of Depression ................................ ................................ ........................ 69 Perceived Physical Symptoms of Depression ................................ ................................ .... 76 Perceived Emotional Symptoms of Depression ................................ ................................ 77 Living as a Complex Patient ................................ ................................ .............................. 79 Response to Resear ch Question 2 ................................ ................................ ...................... 85 Research Question 3 ................................ ................................ ................................ .................. 86 Illness Priority ................................ ................................ ................................ ................... 86 Depression Treatment Preferences ................................ ................................ .................... 88 Competing Dem ands/ Barriers to Treatment and Improving Symptoms .......................... 91 Response to Research Question 3 ................................ ................................ ...................... 95 VI DISCUSSION ................................ ................................ ................................ .......................... 99 Change in Depression Over Time in Comple x Patients ................................ ...................... 100 Key Findings from the Qualitative Analyses ................................ ................................ ...... 101 Patient Description of Depression ................................ ................................ ................... 101 Perceived Relationship between Depression and Chronic Illness ................................ ... 103 Illness Priority and Competing Demands ................................ ................................ ........ 104 Assessing Methodology ................................ ................................ ................................ ....... 106 Conclusion ................................ ................................ ................................ ........................... 106 Limitations ................................ ................................ ................................ ........................... 108 Limitations of the Quantitative Analyses ................................ ................................ ........ 109 Limitations of the Qualitative Analyses ................................ ................................ .......... 109 Dissemination of Results ................................ ................................ ................................ ..... 110 REFERENCES ................................ ................................ ................................ ............................ 111 APPENDIX ................................ ................................ ................................ ................................ .. 124 A. Glossary of Terms ................................ ................................ ................................ .......... 124 B. Patient Health Questionnaire 2 ................................ ................................ ....................... 128 C. Patient Health Questionnaire 9 ................................ ................................ ....................... 129 D. Demographic Variables Questionnaire ................................ ................................ ........... 130 E. Cumulative Illness Rating Scale (CIRS) ................................ ................................ ......... 133 F. Q ualitative Interview Guide ................................ ................................ ............................ 134 E. Diagram of the relationships between levels of coding in the qualitative analysis ......... 137

PAGE 9

ix LIST OF TABLES TABLE II.1 Diagnostic Criteria for Depression/Depressive Disorder ................................ .... 13 II.2 Summaries of the Existing Literature Since 2002 on the Effect of Coexisting Illness on Depression ................................ ................................ .......................... 29 III.1 Description of the Primary and Dissertation Studies ................................ .......... 41 IV.1 Patient Demographics ................................ ................................ ......................... 57 IV.1 PHQ 9 Mean Scores at all Four Times Points ................................ .................... 59 IV.3 Total CIRS by Age ................................ ................................ ............................... 60 IV.4 Main Variable Correlations ................................ ................................ ................. 61 IV.5 Results of Regressing Change in Depression on Illness Severity ....................... 62 IV.6 Results From the Random Coefficient Estimates ................................ ............... 63 IV.7 Results From Depression Ov er Time Model ................................ ...................... 64 IV.8 Results From the Depresion Change by CIRS Total ................................ .......... 65 V.1 Interview Participant Characteristics ................................ ................................ .. 68 V.2 Total Counts of the Codes Identified ................................ ................................ .. 96

PAGE 10

x LIST OF FIGURES FIGURE I.1 Order of the Data Collection Procedure ................................ ................................ 5 I.2 Structure of the Disseration ................................ ................................ ................... 8 II.1 Factors Contributing to Depression Outcomes in Primary Care .......................... 19 II.2 The Competing Demands of Psychosocial Care ................................ ................. 36 III.1 Iterative Process Model ................................ ................................ ........................ 54 III.2 Description of the Qualitative Analysis ................................ .............................. 55 IV.1 Total CIRS Score Reported by Percent of Samp le ................................ ............. 59 IV.2 Change In Depression Over Time Stratified by CIRS ................................ ........ 61

PAGE 11

1 CHAPTER I INTRODUCTION The Centers for Disease Control a nd Prevention (CDC) estimate that by 2020, depression will be the second leading cause of disease burden in the U.S. behind cardiovascular disease ( Centers for Disease Control and P revention, 2013 ) Depression is a debilitating illness typically associated with a loss of pleasure in activities and perceptible changes in mood and behavior such as sleeping patterns, appetite, and concentration, as well as feelings of hopelessness ( Gilbody, 2011 ; National Institute for Health and Clinical Excellence (NIHCE), 2009 ) Bot h the CDC (2010) and the NIHCE (2009) confirm that depression is more prevalent among people with chronic illnesses and the coexistence of chronic physical health problems can cause and complicate depression. approximately two to three times more common in patients with a chronic physical health problem then in people who ( 2009 ) Growing evid ence shows the co existence of physical chronic illnesses and mental health disorders like depression can confound and hinder physical symptom improvement and depression treatment ( Gonzales, Esbitt, Schneider, Osborne, & Kupperman, 2011 ; Katon & Schulberg, 1992 ; Koike, Unutzer, & Wells, 2002 ; Nutting, Rost, Smith, Werner, & Elliot, 2000 ; Pagoto, Schneider, Appelhans, Curtin, & Hajduk, 2011 ; Rost, et al., 2000 ; Rutledge, Reis, Linke, Greenberg, & Mills, 2006 ) D epression has a bi directional relationship with physical illness in that it adversely affects the severity of physical illness, but similarly the presence of physical illness exacerbates the severity of depressive symptoms ( Clarke, 2009 ) complicating the management of both physical and mental health conditions ( Morris, et al., 2012 ) Report s from the World Health Organization indicate that of any chronic disease, depression produces the greatest diminution of overall personal health and as a coexisting condition, is more debilitating

PAGE 12

2 than any other disease combination contributor ( Moussavi, et al., 2007 ) The prevalence of depression is reported anywhere from three to five times higher among complex patients than the general population ( Bair, Robinson, Katon, & Kroenke, 2003 ; Teh, Reynolds, & Cleary, 2008 ) Additionally, physical illness can disguise the sy mptoms of depression since some symptoms are common to both physical and mental disorders, complicating the assessment of depression (NICHE, 2009). Furthermore, diagnostic decisions, depression treatment adherence and symptom improvement are further compl icated by the economic, socio cultural, and environmental factors often e xperienced by individuals with lower socioeconomic status (SES) ( Alegria, et al., 2008 ; Everson, Maty, Lynch, & Kaplan, 2002 ; Lorant, et al., 2003 ) For example, studies show higher odds of persisting depression ( Lorant, et al., 2003 ) and lower rates of service use for d epression ( Alonso, et al., 2004 ) a mong low SES populations Given most patients (particularly low SES populations) with poor mental health and coexisting illnesses are cared for in primary care settings ( Arnow, et al., 2006 ) developing and targeting more effective interventions for primary care settings is critical for reducing disparities in the diagnos is and treatment of depression, especially for low income primary care patients with limited ac cess to mental healthcare Providing adequate behavioral health treatment in primary care is challenging, especially for medically indigent populations suffering from poorer mental and physical health. Even when depressi on recognition is adequate for low SES population s gaps remain in the quality of the with depression car e and improved symptomatology. Research related to improving behavioral health treatments and interventions for medically indigent complex patients are not only a research priority, but also a clinical imperative. Fortunately, there is evidence that treating depression in patients with chronic physical health problems can potentially increase life expectancy and their overall quality of life (NICHE, 2009); however, we lack understanding about how the severity of chronic illnesses moderates the effect of depression treatment on

PAGE 13

3 symptom improvement ( Teh, et al., 2008 ) particularly for patients who are medically indigent and economically disadvantaged. The purpose of this two phase sequ ential mixed methods study was two fold: examine the potential relati onship between coexisting illness severity and change in depression over time ; and to identify contextual factors that influence beliefs, attitudes, and treatment preferences for depression and coexisting illness care for primary care patients experiencing a new episode of depression. This study will build upon the findings of a randomized control trial (RCT) that trained primary care providers in a communication strategy to improve clin ical dialogue around depression ( Keeley & Brody, 2012 ) The setting for the study was an integrated safety net health care network called Denver Healt h and Hospital Authority (DHHA) in Denver, Colorado that primarily tre ats economically indigent patients. Though existing research has identified the adverse effect of chronic illness on depression symptom improvement gaps remain in understanding how the severity of chro nic illnesses as well i nterplay with depression care and improved symptomatology. Study Purpose and Rationale Research Questions Given the lack of evidence concerning the potential relationship between physical illness severity and depression symptom improvement as well as the patient narrative around living with depression and coexisting illnesses, this dissertation addressed t he following research questions: 1. Does the severity of coexisting medical illnesse s impact depression symptom improvement for primary care patients experi encing a new e pisode of depression? 2. How do complex depression patients d escribe their lived experience with concurrent illnesses? 3. How does a diagnosis of a new episode of depression impact the management and prioritization of concurrent illnesses for c omplex patients?

PAGE 14

4 Influenced by the model of competi ng demands particularly patient level demands ( Jaen, Stange, & Nutting, 1994 ) I hypothesize d that patients with high competing illness severity would experience significantly less depression symptom improvement over time Specific Aims Specifically, this dissertation research addressed the following aims: Aim 1 Determine whether severity of coexisting illnesses were associated with improvement of depressive symptom outcomes for moderately to severely depressed primary care patients with coexisting illnesses (complex patients) Aim 2 between depression and physical illness Aim 3 Describe the ef fects of illness co life including, but not limited to, physical and social functioning. Aim 4 Describe the contextual factors (competing demands) that preferences in receiving car e for their depression and coexisting illnesses in a primary care setting. Overview of the Research Methods This mixed methods study used a combination of quantitative analysis of secondary data, as well as in depth qualitative interview data to address the study aims (FIGURE I.1 ) For the quantitative analysis, I used Hierarchical Linear Modeling (HLM) to examine the association between the number and severity of coexisting medical illnesses and depression symptom improvement, over time, for complex pa tients with a major depressive disorder (MDD) in a clinical trial study. For the qualitative analysis, I conducted a thematic analysis of semi structured interviews I selected a mixed method design because it provided a way to place statistical findings in context allowing for a rich explanation of related phenomena ( Johnson & Onwuegbuzie, 2004 )

PAGE 15

5 Equal Priority (weighting) Study Setting and Participants This secondary analysis used data from 168 patients screened for depression from eight primary care health clinics in the Denver Health system that provide health care services to low income and underinsured people (medicall y indigent) in Denver, Colorado. Patients were enrolled in Medicare, Medicaid or the Colorado Indige nt Care Program (CICP) or both, and of various race/ethnicities, and at least 18 years of age For the qualitative interviews, and g iven the intent to under stand the potential effect of comorbid illnesses on depression symptom improvement purposeful sampling techniques were used to identify patients with low, medium and high illness severity, and with increased, neutral and decreased depression symptom impro vement at 6 12 and 36 weeks after baseline Data Collectio n Procedures Sequential Implementation Quantitative (First Section) Qualitative (Second Section) F IGURE I .1 : Order of the Data Collection Procedure

PAGE 16

6 Significance of the Study A review of the literature on treating depres sion in primary care highlights that al though the impact of coexisting illness on depression symptom improvement has been examined, mo st studies have focused on either the impacts of antidepressant medications ( Koike, et al., 2002 ; Kurdyak & Gn am, 2004 ; Lin, et al., 2012 ) isolating only one co existing chronic illness for comp arison (e.g., cancer, diabetes), or included a sample consisting of only older adults, or all or some of the above ( Iosifescu, 2007 ) (all relevant studies are reviewed in Chapter II ) Additionally, a disproportionate number of the studies used a si mple count of morbidities, which may not reflect severity of symptoms or the chronicity of specific illnesses, which are limitations when assigning equal weight to all illnesses. A more nuanced approach was to assess the severity of each illness ( Fortin, Bravo, Hudon, Vanass e, & Lapointe, 2005 ; Huntley, Johnson, Purdy, Valderas, & Salisbury, 2012 ) Thus, a primary aim of this dissertation was to test whether severity of coexisting illnesses are associated with improvemen t of depressive symptom outcomes among moderately to severely depressed medically indigent primary care patients ages pharmaceutical treatment focused on improving clinical dialogue around depression delivered by primary care pro viders (PCPs). Improving our understanding the pathways through which coexisting illnesses influence depression symptom improvement in primary care is imperative. From a clinical perspective, complex patients (patients with multiple chronic medical condi tions occurring simultaneously) represent the major users of health care services in the U.S. accounting for more than two thirds of health care spending ( Tinetti, Fried, & Boyd, 2012 ) Furthermore, d epression as a coexisting illness with the more prevalent chronic disorders (e.g., coronary heart disease, diabetes) is associated with poor self care, and increased complications (Katon 2010), which can lead to more severe illness severity and earlier death. Thus, the identification of effective interventions to treat patients with multiple conditions, particularly the coexistence of physical and psychological disorders, is time ly and important

PAGE 17

7 A second aim of this dissertation was to collect in depth qualitative data from semi structured interviews to identify specific contextual factors (e.g., competing demands) that rences for their depression and coexisting illness care. Though competing demands at the provider level have been previously explored (Jaen et. al., 1994; Nutting et.al., 2008; Rost et. al., 2000; Stange et.al., 1994), less understood are the competing de mands at the patient level (e.g., financial strain, poor living conditions, co occurring illnesses), and what those demands mean for chronic illness care and depression symptom improvement. A third aim was to explore how patients perceive the relationship between depression and chronic physi cal illness, specifically which, between depression and chronic illness, patients prioritize as crucial for im mediate care Though coexisting depression and chronic illness is common, it is not satisfactory to conclude that they have a linear relationship but in fact a more complex bi directional one, where depression influences chronic illness severity and treatment, or vice versa. A conceptual model including descriptions of the three phases, along with sampling st rategies, brief descriptions of the quantitative and qualitative methods, and how they associate with each of the aims is included below (FIGURE I. 2).

PAGE 18

8 Description of Original Study (Keeley, 2012) urrent list of patients. Recruited by telephone. All patients completed the Patient Health Questionnaire 2 during initial telephone call. Informed consent and baseline data collected prior to visit in clinic waiting areas. Quantitative (Specific Aim 1) Us depression symptom improvement over time. PHASE I (Secondary Data) n=168 illness severity using the Cumulative Illness Rating Scale (CIRS) Qual itative (Specific Aims 2, 3, & 4) Criterion sampling based on baseline CIRS score; stratified purposeful sampling varying on four parameters: CIRS score, PHQ 9 score, gender, and intervention/control. Iterative process including review of the question guid e and script Data analyzed using ATLAS.ti. PHASE III (Primary Data) n=19 Telephonic, semi structured interviews with purposefully selected patients to better understand their lived experience with a new episode of depression as well as their beliefs abo ut the relationship between depression and coexisting illnesses PHASE II Test for the potential effect of coexisting illness severity on depression symptom improvement FIGURE I .2 : Structure of the Dissertation Figure 0 1

PAGE 19

9 Terminology Given the number of terms used in this study and the complexity of some of those terms, definitions of the various terms along with citations are included to improve understanding. These citations include existing studies published in peer reviewed academic journals, the Diagnostic and Statistical Manual I V (DSM IV) published by the American Psychiatric Association, the Centers for Disease Control and Prevention (CDC), and the National Institutes of Mental Health (NIMH). A glossary of terms (Appendix A ) is included with the definitions of terms central to t his study. The DSM IV offers a series of symptoms associated with depressive disorder including interest in most daily activities including those activities nec essary to maintain manageable living conditions (e.g., employment, family and other social interactions), change in sleep patterns, lack thoughts of hurting oneself or suicidal ideation with or without a specific plan ( American Psychiatric Association, 2000 ) For the purposes of this study, the term depression is used to refer to a mental health condition that meets diagnostic criteria such as those publishe d in the DSM IV. Though major depressive disorder (MDD) is typically reported as a dichotomous variable (YES or NO), this study aims to measure improvement in depr essive symptoms necessitating a continuous measure instrument. While acknowledging that MDD can be difficult to quantify or interpret uniformly across populations of different socioeconomic status, cultural traditions, and communities, there are several instruments for detecting MDD the aimed at accurately assessing status. The study assumed that distinctions exist in the severity of depression (Aim 1) critical to fully understanding not only the current depressive state of each patient, but to better assess the effectiveness of depression interventions at multiple levels of depressive states. Specifically, depression was given diagnost ic scores listed as moderate (10

PAGE 20

10 14) moderately severe (15 19), and severe ( > 20) depression in accordance with a 9 item depression module called the Patient Health Questionnaire 9 ( PHQ 9) ( Kroenke, Spitzer, & Williams, 2001 ) The PHQ 9 is a questionnaire that assesses depressive symptom criteria in accordance with the symptoms described in the DSM little interest in doing thin , and better off dead or hurting yourself in some way ( Kroenke, et al., 2001 ) As a measure of depression severity, the PHQ 9 scores range from 0 27 with each of the nine items having possible individual scores of 0 (not at all) to 3 (nearly every day). A copy of the PHQ 9 is included in the appendix. I used the term comorbidity in this research as an encompassing term to refer to both physical a ( Valderas, Starfield, Sibbald, Salisbu ry, & Roland, 2009 ) Comorbidity can often be either the cause or the consequence of an index disease and can affect disease detection, therapy, and desired outcomes or changes to behavior ( de Groot, Beckerman, Lankhorst, & Bouter, 2003 ) Additionally, evidence suggests that comorbid illnesses affect the effectiveness of treatments and interventions. For the purposes of my research comorbidity was counted weighted and reported by severi ty of the illnesses using the Cumulative Illness Rating Scale (CIRS) which rates comorbidity on a five point severity scale based on 13 anatomical domains (e.g., cardio vascular respiratory system, neurological etc. ), making it a methodological superior ap proach to measuring comorbidity to using strictly a simple count This study used complex patient as a term to refer to those patients with multiple chronic conditions that often require unique services and care management strategies ( Bayliss, Ellis, & Steiner, 2005 ; Fortin, Soubhi, Hudon, Bayliss, & van den Akker, 2007 ) Complex patients can experience advers e health outcomes, higher healthcare costs, as well as a decreased quality of life, and psychological distress ( Bayliss, Edwards, Steiner, & Main, 2008 ; Bayliss, et al., 2005 ; Fortin, Soubhi, et al., 2007 ; Newcomer, Steiner, & Bayliss, 2011 )

PAGE 21

11 This study use d several descriptors to refer to social issues and stressors experienced by the patient populations. The term socially and economically disadvantaged is used as an encompassing term to describe individuals and groups experience conditions of poverty, ina dequate resources, and deprivation as we ll as limited political capital ( Centers for Disease Control and Prevention, 2011c ; U.S. Department of Health and Human Services, 2001 ) This study includ ed patients that are socially and economically disadvantaged with limited access to health services along with transportation issues, language barriers, or other issues hindering adequate medical care and are referred throughout this research as medically indigent Lastly, this study referred to providers as Primary Care Providers (PCPs) which nternists. Specific titles are delineated when necessary. Structure of the Dissertation This dissertation follows a six chapter format and includes a Table of Contents, List of Tables and Figures, Glossary of Terms, Append ices, and a List of References. Following this introduction (Chapter 1), Chapter 2 discusses related literature suppor interpretation of the results. Chapter 3 outlines the research methods and design used for data collection (both quantitative and qualitative respectively), the analysis of the findings, and methods to verify the findings. The results are presented in Chapters 4 (Quantitative) and 5 (Qualitative). Chapter 6 summarizes and conclude s the study by discussing the findings relative to existing liter ature, the identified theoretical concepts (both introduced in Chapter 2), implications of this work, study limitations, and how the findings contribute to the current body of knowledge, as well as suggestions for future research.

PAGE 22

12 CHAPTER I I REVIEW OF T HE LITERATURE In the United States, mental disorders a re common, and often serious, significantly impact ing the population. In 2010, the Substance Abuse and Mental Health Services Administration (SAMHSA) estimated 45.9 million adults (ages 18 and older) had a mental illness 1 representing 20% of the total U.S. adult population ( Substance Abuse and Mental Health Services Administration (SAMHSA), 2012 ) Globally, mental illnesses account for more than 11% of disease burden ( Ustun, Ayuso Mateos, Chatterji, Mathe rs, & Murray, 2004 ) with that number increasing to 15% for developed nations, including the United States ( Murray & Lopez, 1996 ) Depression disorders 2 in particular can result in seri ous impairment and societal costs ( Katon, 2009 ; Substance Abuse and Mental Health Services Administration (SAMHSA), 2012 ) The economic toll depression has in the United States is estimated at over $83 billion annually, of which $52 billion are attributed to workplace costs inclu ding absenteeism or productivity impairment ( Greenberg & Birnbaum, 2005 ) The intra personal toll of depression includes continuous feelings of diminished interest in previously enjoyed activities, recurring feelings of guilt, chronic fatigue, diffic ulties concentrating, sleep disorders, weight gain or loss, and recurrent thoughts of inflicting pain on oneself and/or suicide ( Centers for Disease Control an d Prevention, 2011a ) Both the World Health Organization International Classification of Diseases (ICD 10) and the American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders (DSM IV) list criteria for depression with some gen eral agreement s on the etiology of the disorder (TABLE II .1). 1 Mental Illness is defined as currently or at any time in the past year having a diagnosable mental, behavioral, or emotional disorder (excl uding substance abuse) of sufficient duration to meet diagnostic criteria specified within the Diagnostic and Statistical Manual of Mental Disorders (DSM IV; American Psychiatric Association). 2

PAGE 23

13 T ABLE II.1: Diagnostic Criteria for Depression/Depressive D isorder from the World Health Organization ICD 10 (WHO, 1990) and the American Psychiatric Association DSM IV (APA, 2000) Adapted from Gilbody (2011) In the US, c hronic conditions which require ongoing treatment and self management, are the most common ly seen in health care settings ( Grumbach, 2003 ; Ridgeway, et al., 2014 ) and are responsible for nearly 70% of all mortality affecting approx imately 133 million adults ( National Center for Chronic Disease Prevention and Health Promotion, 2012 ) The rate of peopl e living ICD 10 Depressive disorder DSM IV Major depressive disorder Clinical significance Some difficulty in continuing with ordinary work and social activities, but will probably not cease to functi on completely in mild depressive episode; considerable distress or agitation, and unlikely to continue with social, work, or domestic activities, except to a very limited extent in severe depressive episode Symptoms cause clinically significant stress or i mpairment in social, occupational, or other important areas of functioning Duration of symptoms A duration of at least 2 weeks is usually required for diagnosis for depressive episodes of all three grades of severity A major duration of the day, nearly ev ery day, for at least 2 weeks Severity Depressed mood, loss of interest and enjoyment, and reduced energy leading to increased fatigue and diminished activity in typical depressive episodes; other common symptoms are: (1) Reduced self esteem/self confidence (2) R educed concentration and attention (3) Recurring feelings of guilt and unworthiness (4) Bleak and pessimistic views of the future (5) Suicidality/self harm (6) Not enough or too much sleep (7) Change in appetite (8) Recognizable changes in speech For mild depressive episode 2 of most typical symptoms of depression and of the other symptoms are required For moderate depressive episode 2 of 3 of most typical symptoms of depression and at least 3 of the other symptoms are required. For severe depression episode all 3 of the typ ical symptoms noted for mild and moderate depressive episodes are present and at least 4 other symptoms of severe intensity are required Five or more of the following symptoms; at least one symptom is either depressed mood or loss of interest or pleasure: (1) Depressed mood (2) Loss of interest (3) Significant weight loss or gain or decrease or increase in appetite (4) Insomnia or hypersomnia (5) Psychomotor agitation (6) Fatigue or loss of energy (7) Feelings of worthlessness or excessive or inappropriate guilt (8) Diminished ability to think or concentrate (9) Recurring Suicidality/self harm

PAGE 24

14 with chronic conditions is increasing faster than origina lly predicted, supporting the critical need for more research around improving chronic care. In 2010, approximately 75 million people in the U.S. fit the definition of complex patients defi ned as having multiple ( two or more) coexisting chronic illnesses characterized by ongoing ( Eton, et al., 2015 ; Parekh & Barton, 2010 ; Warshaw, 2006 ; World Health Organization, 2013 ) Chronic illnesses are co stly accounting for nearly 65 80 % of total national health care expenditures in the U.S., ( Anderso n, 2010 ; Guthrie, 2014 ; Parekh & Barton, 2010 ; Robert Wood Johnson Founda tion, 2010 ; Ward, Schiller, & Goodman, 2014 ) and contribute an undue individual medical burden by decreasing quality of life, increasing levels of psychological distress and social deprivation ( Creed, et al., 2002 ; Fortin, Bravo, Hudon, Lapointe, Almirall, et al., 2006 ; Fortin, Bravo, Hudon, Lapointe, Dubois, et al., 2006 ; Fortin, Soubhi, et al., 2007 ) The bi directional effects of depression and chro nic illness are well documented with depression contribut ing to the development and progression of various physical illnesses, and physical illnesses increasing the risk of depression ( Steptoe, 2007 ) Depression is prevalent among patients with competing chronic illnesses and adversely impact s self care (e.g., patient initiated behaviors like exercise, diet, medication adherence) of c hronic disease resulting in increase d resource utilization (e.g., expensive secondary care referrals) ( Gilbody, 2011 ; Guthrie, 2014 ) De pression in people with coexisting illnesses impairs functioning by (1) amplifying reactions to somatic symptoms; (2) reducing motivation to care for physical ailments; and (3) limit ing the energy, and cognitive capacity to cope with physical illness while increasing subjective senses of sh ame or social stigma ( Creed & Dickens, 2006 ; Creed, et al., 2002 ; Martucci, et al., 1999 ) The prevalence of depression coexisting with other chronic conditions is staggering, a problem compounded by the fact that ( Creed & Dickens, 2006, p. 3 ) Though evidence based treatment s for depression are available and have been shown to improve

PAGE 25

15 depressive symptoms for the c hronically ill, many, if not most, complex patients continue to suffer from untreated depression. Further complicating the deleterious effects of coexisting depression and chronic physical illness are economic disadvantages, which are associated with the initial on set of depression symptoms and worse prognosis ( Gilman, Kawachi, Fitzmaurice, & Buka, 2002 ; Gilman, T rinh, et al., 2013 ) Although m ental illness is a ubiqui tous problem affecting all races and ethnicities ( U.S. Department of Health and Human Services, 2001 ) e conomic and social disadvantages magnify the consequences o f menta l illnesses, and limit a ccess to a dequate care, leading to mistrust of the medical system ( Alegria, et al., 2002 ; Alegria, et al., 2008 ; Smedley, Stith, & Nelson, 2003 ; U.S. Department of Health and Human Services, 2001 ) One way that mental illness comp that accrue including estimates showing between a 33 to 170 percent increase in monthly medical costs for complex patients with a psychological disorder (e.g., depression and/or anxiety) ( Guthrie, 2014 ; Melek & Norris, 2008 ) Being economically disadv antage d is also linked to reduced access to mental health treatment and poorer clinical depression outcomes. ( Gilman, Bruce, et al., 2013 ; Gilman, Fitzmaurice, et al., 2013 ) Potentially, the relationship between being economically disadvantaged and poorer mental health is mediated by chronic exposure to adverse life situations (e.g., unemployment, low wages, ho using conditions), and little or no social support ( Clarke, 2009 ; Clarke & Currie, 2009 ; Wilhelm, Mitchell, Slade, Brownhill, & Andrews, 2003 ) The World Federation for Mental Health (2012) identified additional risk factors for depression such as low education level and exposure to violence, which is more pre valent in economically disadvantaged populations. The problems faced by patients with dep ression and chronic illness have highlighted the need for research to address issues around (1) access to adequate mental health care, (2) equity of access t o distribu te mental health care, and (3) the effectiveness of mental health services including better collective management of mental illnesses and eliminating inappropriate

PAGE 26

16 allocation and use of psychotropic medication s ( Gilbody, 2011 ) Concurrently, the U.S. Department of Health and Human Services identified areas to address including cultural, social, and economic factors contributing to mental health care disparities incl uding stigma, poverty, and the overall financing of services ( U.S. Department of Health and Human Services, 2001 ) along with integrating mental health services into primary care ( Galson, 2009 ) Consistent with these recommendations, this dissertation research explore d the potential effect coexisting illness severity has on depression symptom improvement in economically disadvantaged primary care patients using secondary data from a completed RCT on depression treatment ( Keeley & Brody, 2007 ) The following background sections include a review what is known about the connection between chronic hea lth conditions and depression, delivery of behavioral h ealth services in primary care, challenges in caring for me dically and economically indigent c omplex patients with depression, the complications faced by primary care provid ers caring for medically and economically indigent complex patients, as well as the theoretical assumptions guiding the Chronic Illness and Depression among Medically Indigent Populations in Primary Care Given the increased susceptibility to mental disorders, lack of access to m ental health resources, and numerous barriers to care, developing and implementing strategie s to better manage depressive disorders for economically and medically indigent 2 individuals is a public health imperative One setting that can have a significant impact on mental health disparities seen in populations of low socioeconomic status (SES) is primary care. However, t he barriers to adequate detection and treatment of depression in primary care can be challenging for PCPs caring for socially and economically disadvantaged; disadvantages typically associated with 2 Some of the literature on medically indigent populations only includes those individuals with no health insurance or en rolled in safety net programs for indigent populations including but not limited to MEDICAID.

PAGE 27

17 poorer access to care and limi ted health care resources including little or no health insurance (i.e., medically indigent) ( Bovbjerg & Kopit, 1986 ) Health outcomes follow SES gradients related to socioeconomic status (SES) from those living in poverty to those individuals with relatively high SES with the most affluent having better health outcomes compared to the less advantaged ( Glazier, Agha, Moineddin, & Sibley, 2009 ; Macintyre, 1994 ; Pickett & Pearl, 2001 ) The relationship between chr onic disease prevalence and SES is clear, as is the correlation between SES and the risk factors for chronic d isease ( Adler & Ostrove, 1999 ) Low SES is associated with increased risk for multiple health problems leading to an increase risk in pr emature morbidity and mortality ( Inglis, Ball, & Crawford, 2005 ; Lantz, et al., 1998 ; Marmot, Kogevinas, & Elston, 1987 ; Pampel, Krueger, & Denney, 2010 ; Stringhini, et al., 2010 ) Economically disadvantaged populations not only suffer disproportionally from adverse health outcomes and behaviors ( Stringhini, et al., 2010 ; Walsh, Seen, & Carey, 2013 ) but concurrently face personal and contextual barriers to successful treatment of those diseases ( Mansyur, Pavlik, Hyman, Taylor, & Goodrick, 2013 ) Existing resea rc h suggests psychosocial factors have an impact on mental disorders (e.g., depression ) even after controlling for genetic factors ( Piccinelli & Wilkinson, 200 0 ; Stansfeld & Rasul, 2006 ) Individuals in the lower socioeconomic strata (i.e., poverty) experience significantly higher rates o f psychological stress and mental disorder s including depression ( Akhtar Danesh & Landeen, 2007 ; Hudson, Neighbors, Geronimus, & Jackson, 2012 ; Kohn, Dohrenwend, & Mirotznik, 1998 ; Piccinelli & Wilkinson, 2000 ; Saraceno, Levav, & Kohn, 2005 ) In 2010, The Office o f Minority Health reported (as determined using the U.S. Census Bureau computations) were three times more likely to have ( The Office of Minority Health, 2013 ) The predominance of mental disorders, especially depressive disorders, among people of lower S ES has been hypothesize d to be explained by financial disadvantages, high rates of

PAGE 28

18 unemployment ( Mossakowski, 2 008 ) limited educational opportunities and low standard of living ( Stansfeld & Rasul, 2006 ) all demonstrated to have proximal eff ect s on ri sk for depression. Exacerbating these disparities for less affluent populations are lower access and availability of mental health services ( U.S. Department of Health and Human Services, 2001 ) for certain racial/ethnic groups. Recent reports demonstrate the red uced utilization rates and adherence to mental health treatment for black and Latino groups in rates, noting that the needs of minority racial/ethnic groups remain unmet ( Chow, Jaffee, & Snowden, 2003 ; U.S. Department of Health and Human Services, 2001 ; US Department of Health and Human Services, 1999 ) Alt hough the relationship between racial /ethnic populations, low SES, and mental health utilization is complex, several factors are clear: Poorer communities inhabited by historically margi nalized racial/ethnic groups do not have the community resources to recognize and treat mental health ( Chow, et al., 2003 ; U.S. Department of Health and Human Services, 2001 ) ; Patients may expect mistreatment due to perceived discrimination and prejudice ( Wang, et al., 2005 ) ; Patients may not trust the medical system ( Katon, 2003 ) ; and Provider level factors including competing demands, insufficient reimbursement protocols for mental health care, and inadequate training and experience treating mental health disorders are exacerbated in poorer communities ( Klinkman, 1997 ; Nutting, et al., 2000 ; Wang, et al., 2005 ) Figure II .1 illustrates the relationships between those levels and how they potentially hinder depression symptom improvement including the potential relationship posed by patients not receiving adequate care on the overall health care system (e.g., increased cost, increased emergency room visits)

PAGE 29

19 F IGURE II. 1: Facto rs Contributing to Depression Outcomes in Primary C are Addressing system provider and patient level factors contributing to health outcomes of medically indigent populations, including the Medicaid populations, becomes even more important with recen t legislation projects expand ing Medicaid coverage to 15 million more people by 2019 ( National Association of Community Health Centers, 2010 ) System level Lack of mental health services for medically indigent populations Overburdened social service staff Insufficient reimbursement protocols for mental health care Provider l evel Feeling unprepared to adequately discuss and/or treat mental illness (clinical inertia) Inadequate detection and treatment of mental illness Competing demands caring for patients' other illness(es) Burnout Patient level Competing demands from oth er chronic illness(es) Health service utilization Psychological distress Sociodemographics Perceived discrimination Allostatic load Poorer Outcomes Direct Effect Feedback Effect

PAGE 30

20 Medical Illness Comorbidity and Depression A substantial amount of evidence demonstrates that individuals wi th psychological disorders (e.g., depression) are disproportionately affected by chronic disease, often contributing to poorer outcomes for both chronic disease management and mental health improvement ( Benton, Staab, & Evans, 2007 ) Concomitant physical and mental illnesses are associated with poor treatment (both physical and psychological) response, lost work productivity, occupational disability, lower reporte d quality of life, increase d participation in health compromising behaviors, and increased healthcare costs ( Benton, et al., 2007 ; Gonzales, et al., 2011 ; Iosifescu, 2007 ; Katon, 2003 ; Pagoto, 2011 ; Steptoe, 2007 ) Though research suggests the existence of a bi directional relationship between depression and chronic illness, u nderstanding the contextual and latent factors that make up the relationship is imperative to improve both physical and psychological health outcomes Recently, research on the relationship between depression and physical illness has increased along with interventions designed to improve depression recognition and treatment for PC Ps. However, these studies did not aim to understand the potential bi directional relationship between chronic illness and depression focusing rather on depression as a cause and/or consequence of a chronic illness as well as depression being associated with poor outco mes of medical illness and increased mortality ( Iosifescu, 2007 ) Examples of this unidirectional relationship include depression caused by obesity ( Clark, Cargill, Medeiros, & Pera, 1996 ; Linde, et al., 2004 ; Pagoto, et al., 2007 ) type 2 diabetes ( Anderson, Freedland, Clouse, & Lustman, 2001 ; Black, Markides, & Ray, 2003 ; de Groot, And erson, Freedland, Clouse, & Lustman, 2001 ; Katon, et al., 2005 ; Zhang, et al., 2005 ) cardiovascular disease a disease that depression independent ly predicts ( Barth, Schumacher, & Herrmann Lingen, 2004 ; Kronish, Rieckmann, Schwartz, Schwartz, & Davidson, 2009 ; van Melle, et al., 2004 ) and cancer ( Spiegel & Giese Davis, 2003 ) Since understanding the association between depression and the overall burden of comorbid medical illness is critical, an equally important area of research pertains to the course of

PAGE 31

21 d epression in complex patients. Though Benton and colleagues (2007) as well as Pagoto (2011) have extensively studie d the relationships between psychiatric illness and physical illness this dissertation includes a review of some of those findings with concentration on relevant illnesses seen in the population recruited for this study. Prevalence of Depression Comorbi d with Specific Medical Illnesses A large body of literature describes the prevalence of depression among complex patients in primary care. Most of the studies look at the relationship between depression and one particular chronic disease, primarily the most common illnesses seen in primary care (i.e., obesity, type II diabetes, and chronic pain). The studies identified as relevant to this dissertation research are described below to demonstrat e the coexistence of depressive disorders and chronic medica l illness The prevalence of obesity in patients with diagnosed psychiatric illnesses has rapidly increased compared to the general population ( Allison, et al., 2 009 ) Obese individuals not only suffer disproportionately from psychiatric illnesses, but obesity appears to be a risk factor for various mental problems like depression ( Pagoto, et al., 2011 ) The association between depressive disorders and obesity is sh own in both epidemiological and clinical studies. Simon and colleagues using a U.S. representative sample showed a significantly higher percentage of obese adults met the criteria for depression compared to their non obese counterparts ( Simon, et al., 2006 ) One important finding from studies on obesity and depression is the relationship between increasing degrees of obesity with increasing severity of depression Psychological factors could potentially explain the association between obesity and depression. Given the association of depression and obesity, assessments of depression for obese individuals in clinical settings (e.g., primary care) should be part of s tandard care, especi ally given depression could contribute to further weight gain, as well as patient decision making around weight loss treatments, and patients confidence for weight loss success and management. Though

PAGE 32

22 antidepressant medication treatmen t options are available for obese patients with depressive symptoms, psychotropic medication has been linked as a residual contributor to the growing obesity epidemic in t he U.S., given weight gain is a side effect of antidepressants which can lead to med ication noncompliance ( McAllister, et al., 2009 ) and further psychological frus tration felt by the patient. Type 2 diabetes (T2DM) continues to be a major pub l ic health problem estimated to affect more than 435 million adults worldwide, and a 20% increase is projected in developed nations over the next 20 years ( Gonzales, et al., 2011 ) Optimal treatment for T2DM is significant ly dependent on patient behavior (lifestyle factors like diet and exercise) and treatment adherence to blood glucose monitoring, and prescribed medications related to their counterparts without depression Individuals with T2DM commonly show symptoms of d epression and distress, which has been associated with poorer treatment adherence and increased mortality ( Black, et al., 2003 ; Gonzalez, et al., 2008 ) Furthermore, depression and diabetes mellitus seem to have a bi directional relationship, with depression symptoms preceding the development of diabetes ( Gonzales, et al., 2011 ; Knol, et al., 2006 ) and with development of T2DM followed by an increase in depressive symptoms ( Bogner, Morales, de Vries, & Cappola, 2012 ) Though available literature does not offer definitive evidence of biological pathways mediating the relationship between diabetes and depression, it does show th at the health compromising behaviors (HCBs) often associated with the development of T2DM are prevalent in those individuals either clinically diagnosed with depression or exhibiting depressive symptoms. These HCBs include tobacco use, poor diet, lack of physical activity, and medication non adherence ( Gonzales, et al., 2011 ; McClave, et al., 2009 ) Additio nally, while pharmacological interventions for depression have been recommended for patients with diabetes (including T2DM) ( Goodnick, 2001 ; Lustman, et al., 2000 ) antidepressant medications frequently lead to undesired side effects including weight gain and hyperglycemia both serious complications for patients with diabetes ( Gonzal es, et al., 2011 )

PAGE 33

23 Chronic pain continues to be one of the most commonly seen medic al conditions in primary care and ranks as the 3 rd most prevalent complaint for all primary care visits ( Upshur, Luckmann, & Savageau, 2006 ; Van Dorsten & Weisberg, 2011 ) Adding to the burden of pain are the feelings of depression commonly associated with chronic pain ( Fishbain, Cutler, Rosomoff, & Rosomoff, 1997 ) The symptoms of depression (e.g., loss of pleasure in everyday activities, sadness, hopele ssness, and fatigue) are often the residual effect of persistent pain, and of debilitating episodes experienced regularly by patients with chronic pain ( Van Puymbroeck, Zautra, & Harakas, 2006 ) The strong association between pain and depression also contributes to adverse immune fu nctioning critical to physical and mental illness improvement. L iterature on illnesses coexisting with depression identifies pain as the most frequently seen concomitant complaint ( Van Dorsten & Weisberg, 2011 ) with some rates as high as 30 56% of patients with depression. Concurrentl y patient s with depression experience symptoms of pain more than patients with cardiac disease, cancer, diabetes, and neurological disorders ( Campbell, Clauw, & Keefe, 200 3 ; Van Dorsten & Weisberg, 2011 ; Van Puymbroeck, et al., 2006 ; Von Korff & Simon, 1996 ) Studies comparing depression treatment outcomes for patients with and without coexisting illnesses Some of the studies on the relationship of coexisting illness with depression symptom improvement focused on the effects of comorbidity on antide pressant medication tolerance and adherence ( Iosifescu, et al., 2003 ; Koike, et al., 2002 ; Oslin, et al., 2002 ; Papakostas, et al., 2003 ; Perlis, et al., 2004 ) Of those, two studies had an open label design using one single antidepressant ( Iosifescu, et al., 2003 ; Papakostas, et al., 2003 ) one looke d at a single antidepressant ( Perlis, et al., 2004 ) and three looked at different antidepressant medications to test for rates of prescription use and adherence ( Koike, et al., 2002 ; Oslin, et al., 2002 ; Simon, Von Korff & Lin, 2005 )

PAGE 34

24 Six additional studies examined the differences in response to depression care interventions between individuals with depression and coexisting chronic illnesses ( Bogner, et al., 2005 ; Duhoux & et. al., 2009 ; Harpole, et al., 2005 ; Koike, et a l., 2002 ; Morris, et al., 2012 ; Vera, et al., 2010 ) Teh and colleagues (2008) evaluated the effect of chronic medical conditions on depression diagno sis and care including outcomes related to depression diagnosis, patient satisfaction, patient/provider communication and continuity of care. The i nstruments used to measure depre ssion in these studies include d the Composite International Diagnostic Inter view (Short Form, and standard versions) ( Kurdyak & Gnam, 2004 ; Teh, et al., 2008 ) Hamilton Rating Scale for Depression ( Bogner, et al., 2005 ; Iosifescu, et al., 2003 ; Morris, et al., 2012 ; Papakostas, et al., 2003 ; Perlis, et al., 2004 ) Centers for Epidemiological Studies Depression scale (CES D) ( Koike, et al., 2002 ) mental component sco re of the Short Form 12 ( Harpole, et al., 2005 ) Structured Clinical Interview for DSM IV/ Hopkins Symptom Checklist ( Simon, et al., 2005 ; Vera, et al., 2010 ) Geriatric Depression Scale ( Oslin, et al., 2002 ) and the Patient Health Questionnaire 2 and 9 ( Katon et al., 2010 ; Vera, et al., 2010 ) The strategies/instruments to measure medical illness varied by study and included noting the presence of medical illness ( Koike, et al., 2002 ) survey data completed by patients (count of medical illness) ( Harpole, et al., 2005 ; Kurdyak & Gnam, 2004 ; Morris, et al., 2012 ; Oslin, et al., 2002 ; Teh, et al., 2008 ; Vera, et al., 2010 ) the Cumulative Illness Rating Scale (CIRS) which classifies comorbidities by organ system along with the severity of each illness ( Iosifescu, 2007 ; Papakostas, et al., 2003 ; Perlis, et al., 2004 ) the Charlson Comorbidity Index ( Bogner, et al., 2005 ) and computerized chart review s of existing medical illness ( Kat on, et al., 2010 ; Simon, et al., 2005 ) Differences among the study designs and interventions make comparison s of their findings difficult. For example, four of the studies found that coexisting il lness had little or no effect on depression remission and symptom improvement ( Bogner, et al., 2005 ; Harpole, e t al., 2005 ; Papakostas, et al., 2003 ; Perlis, et al., 2004 ) Of those studies, two ( Papakostas, et al., 2003 ; Perlis, et al., 2004 ) included only patients who had treatment resistant depression and had small

PAGE 35

25 sample sizes (n=101 and n=92) limiting their ability to detect a difference. Two of the studies ( Iosifescu, et al., 2003 ; Koike, et al., 2002 ) sh owed that comorbid medical disorders had an impact on depression symptom improvement for both antidepressant and behavioral interventions. Iosifescu et al. (2003) investigated the role of comorbid medical illness on severity of depression and antidepressa nt outcome s in depressed patients, hypothesizing that patients with depression would experience more severe symptoms of depression which would lead to lower depres sion score at the end of the trial) compared to patients with depression but no medical comorbidity. The sample consisted of 384 outpatients meeting DSM III R criteria for depression, enrolled in a n 8 week open treatment with fluoxetine. Using the CIRS t o measure comorbidity, the authors concluded that the total burden of medical illness (CIRS overall score) and the number of organ systems affected response to fluoxetine treatment and clinical remission. Additionally, patients with higher burden of medic al comorbidity had significantly higher depression scores at the end of the 8 Koike, Unutzer, and Wells (200 2) examined two quality improvement programs for depression, specifically a comparative analysis of treatment rates and outcomes for depressed participants with and without coexisting medical conditions. The authors hypothesized that 1) depressed patients with coexisting medical illness(es) would have worse outcomes compared to depressed patien ts without a coexisting illness; 2) treatment rates would not vary si gnificantly between groups; and 3) a quality improvement program for depression would benefit bo th groups improving treatment rates and health related outcomes. The sample consisted of 1,336 patients with depression randomized into one of three groups; usual care, quality improvement program with medication, or quality improvement program with thera py. The findings showed that though depressed patients with coexisting medical illness had similar rates of treatment they had worse depression outcomes compared to patients without coexisting medical illness. H owever,

PAGE 36

26 improving the quality of treatment d id show some improvements for depressed patients with coexisting medical illness, offering evidence of the importance of making greater efforts to include the medically ill when implementing quality improvement programs for depression. Two studies ( Kurdyak & Gnam, 2004 ; Teh, et al., 2008 ) examined the effects of chronic medical conditions on the quality of de pression care for persons with coexisting illness and are described below. Teh and colleagues (2008) analyzed data from the National Survey of Alcohol, D rug, and Mental Health Problems, a survey of community use and qua lity of depression care for people with one or more chronic medical condition provider relationship. The sample consisted of 1,309 patients with de pression. The findings indicated that patients with a CMC and depression were more likely to have their depression recognized by a health care provider compared to patients without a CMC, bu t the existence of a CMC did not impact depression treatment. A second finding from the study showed the patient provider relati onship potentially mediates depression recognition among patients with chronic conditions suggesting improving the patient provider relationship could increase the probability of having depression recognize d by a health care provider. Kurdyak and Gnam (2004) compared utilization of mental health services as well as the quality of medication management delivered in health care settings between depressed patients with and without CMCs. The sample consisted o f 278 individuals with the diagnosis of major depression. The data showed that depressed persons with CMCs are more likely to receive guideline level antidepressant treatment compared to depressed persons without a competing CMC. Katon et al. (2010) con ducted a randomized controlled trial consisting of primary care patients diagnosed with diabetes, coronary heart disease, or both and concomitant depression. Patients all reported at least one measure of poor disease control including high blood pressure, high levels of low density lipoprotein (LDL) cholesterol, and glycated hemoglobin level of

PAGE 37

27 symptoms would improve using a care management intervention delivered c ollaboratively by nurses and PCP s in a primary care setting. After baseline biomarker measures were collected, nurse practitioners initiate d depression treatment, which included self care strategies, pharmacotherapy to control depression, and motivational coaching to help patients set treatment goals and handle barriers impeding treatment. At 12 months, the intervention group had glycated hemoglobin levels, LDL cholesterol, systolic blood pressure, and dep ression outcomes, as well as significant between group differences on three of the four disease control measures ( Katon, et al., 2010, p. p. 2616 ) Additionally, i ntervention patients were more satisfied with their care, and reported higher scores of quality of life, contributing to the evidence that training health care providers in communication strategies (i.e., motivational coaching) can potentially affect depre ssion and disease outcomes, especially around creating dialogue around goal setting, and competing demands. Common suggestions from the existing research include the need for further research on how best to recognize and treat depressed patients with coe xisting illness. Some of the studies suggest a collaborative approach to treatment including antidepressants, behavioral therapy, and quality improvement programs being effective in this population. Though the previous studies emphasized the importance of treating complex patients with depression, it is important to reiterate that those reasons are amplified for medically indigent populations given issues around access to care, lower social support around mental health, and daily hassles competing with sel f care of medical conditions. Important to this dissertation research is to examine the hypothesis that competing dem ands experienced by the patient, including their competing illnesses, may not affect depression treatment ( Ani, et al., 2009 ; Vyas & Sambamoorthi, 2011 ) A summary of the studies describ ed is included below (TABLE II.2 ). In conclusion, studies show the potential impact of coexisting medical illness on depression and depression treatments (both behavioral and psychotropic). Alt hough the described

PAGE 38

28 studies support the rationale and study d esign of the p roposed analysis, they also help identify important areas this research will address. For example, the proposed study included the following suggestions made by previous research including methods and gaps to be addressed : C omplex patients ( Bogner, et al., 2005 ; Harpole, et al., 20 05 ; Oslin, et al., 2002 ) M edical chart review s to compile a list of coexisting illness compared to self report ( Koike, et al., 2002 ; Morris, et al., 2012 ; Teh, et al., 2008 ) With the use of the CIRS, most medical illnesses included compared to specific illne sses identified pre study. Additionally, the severity of those illnesses compared to a simple count ( Bogner, et al., 2005 ; Harpole, et al., 2005 ; Katon, et al., 2010 ; Koike, et al., 2002 ; Kurdyak & Gnam, 2004 ; Morris, et al., 2012 ; Simon, et al., 2005 ; Teh, et al., 2008 ; Vera, et al., 2010 ) Whereas some of the mentioned studies only examined the potential effect of comorbid ity on medication adherence, ( Iosifescu, e t al., 2003 ; Koike, et al., 2002 ; Morris, et al., 2012 ; Papakostas, et al., 2003 ; Perlis, et al., 2004 ) and given the potentially harmful effects of antidepressant medications these may not be viable option s for all complex patients the proposed study is a secondary analysis evaluating a com munication intervention versus a pharmacologic regimen alone Attention to the patient provider relationship and to the continuity of care as it pertains to depression improvement for complex patients. Primarily, this study aimed to address the gap s of ( 1) including the medically ill (especially giving notice to the severity of their health conditions) in an intervention to improve the quality of depression care ( Koike, et al., 2002 ) and (2) to identify ill ness severity (opposed to a count of illness) as a potential causal mechanism hindering depression symptom improvement ( Benton, et al., 2007 ) Additionally, this study explore d influence on clinical encounters and communication, and on depression care in order to gain perspective on the competing demands and treatment preferences that matter to patients ( Bayliss, 2012 )

PAGE 39

29 TABLE II.2 : Summaries of the Existing Literature since 2002 on th e Effect of Coexisting Illness on Depression Author( s) (Year) Objective(s) Population Outcomes Oslin, et al. (2002) To examine the relationship between specific medical illnesses and the outcomes of treatment for late life depression. 671 older adult patients receiving inpatient treatment for depression fr om one of 71 psychiatric treatment facilities in the U.S. Physical disability and the total number of medical illnesses were significantly related to change in depressive symptoms. Certain somatic disorders play a role in the treatment response of late lif e depression suggesting that the effect of specific illnesses on depression may be mediated by the presence of functional disability. Koike, Unutzer, & Wells (2002) Compare treatment and outcomes for depressed primary care patients with and without como rbid medical conditions and assess the impact of quality improvement programs with medication and with therapy for these patients. Depression was measured using the CES D Scale and chronic medical conditions using self report. 1,356 depressed primary care patients from six managed care organizations (46 primary care clinics) from five U.S. states. At 6 and 12 month follow up, he likelihood of having a probable depressive disorder was higher but the rates of use of antidepressant medication and specialty counseling were similar, for depressed patients with comorbid medical disorders than for depressed patients without a coexisting illness. Quality improvement programs (some were an additional cost to the patient) resulted in greater use of antidepressant medications and psychotherapy and lower rates of probable depressive disorder compared to usual care. Papakostas, Petersen, Iosifescu, Roffi, Alpert, et al. (2003) To test whether the presence of comorbid medical conditions can predict clinical response in patients with treatment resistant major depressive disorder treated with open label nortriptyline (NT). Tested whether comorbidity (using the CIRS G) predicted clinical response or depression severity at endpoint. 92 patients with treatment resistant major depressive disorder starting a 6 week trial of NT. The results failed to confirm the relationship between comorbid medical conditions and poor outcome in the treatment of major depressive disorder for patients with treatment resistant depression. Iosifescu, et al. (2003) Investigate the impact of medical comorbidity on the acute phase of antidepressant treatment in patients with major depressive disorder. 384 adult depressed (determined using the Hamilton Rating Scale) outpatients enrolled in an 8 week open treatment with fluoxetine, 20 mg/day. The total burden of medical illness (using the CIRS G instrument) and the number of organ systems affected by medical illness had a significantly negative predictive value for clinical outcome in the acute phase of treatment in major depressive disorder. Harpole, et al. (2004) RCT to determine if the presence of multiple coexisting medical illnesses (average was 3.8 conditions) affects patient response to a multidisciplinary depression treatment program. 1 ,801 depressed older age) from 18 primary care clinics from eight health care organizations in five U.S. states. The presence of multiple coexisting medical conditions did not affect patient response to a multidisciplinary depression treatment program (a trained nurse or psychologist working directly with the patient to determine a treatment option for depression that included either antidepressants or psychotherapy).

PAGE 40

30 Author (s) (Year) Objective (s) Population Outcomes Kurdyak, & Gnam (2004) To examine the difference in quality of care for depression between depressed persons with and without a chronic m edical condition. 287 adults ages 18 to 64 with the diagnosis of major depression evaluated using the Composite International Diagnostic Interview. Depressed persons with comorbid medical conditions are more likely to receive guideline level care for depr ession than are depressed persons without comorbid medical illnesses. However, the association did not persist once high utilizing patients were excluded. Perlis, Iosifescu, Alpert, Nierenberg, Rosenbaum, & Fava (2004) To examine the moderating effect of general medical illnesses on treatment outcome in a controlled trial with patients whose major depressive disorder failed to respond to an 8 week trial of fluoxetine (treats depression, obsessive compulsive disorder, and other mood disorders; commonly known as Prozac). 386 outpatients (mean age 39.9 years) who met the criteria for major depressive disorder using the Hamilton Depression Rating Scale and had coexisting illnesses as determined by the Cumulative Illness Rating Scale. Using logistic regres sion analysis, CIRS score was not associated with likelihood of depression remission. Coexisting medical illness(es) does not appear to be associated with significantly poorer outcomes among patients whose major depressive disorder failed initially to res pond to an initial trial of fluoxetine. Simon, VonKorff, & Lin (2005) A longitudinal study of depressed primary care patients with and without specific co morbid chronic medical conditions (ischemic heart disease, diabetes, chronic obstructive lung diseas e) to assess differences in baseline characteristics, course of depressive symptoms following initiation of antidepressant treatment, and course of functional impairment and disability. 204 primary care patients identified using health plan administrative data to identify those patients initiating antidepressant treatment. Depression severity in patients with diabetes at baseline was not affected by comorbidities but was in patients with ischemic heart disease. All groups were not significantly different in terms of social and emotional functioning, but those patients with coexisting illnesses reported greater physical impairment. Improvement in depression during treatment was strongly associated with change in disability. Bogner, et al. (2005) To descri be the influence of specific medical conditions on clinical remission of major depression in a clinical trial evaluating a care management intervention among care patients. 324 older adults and were randomly assigned to ei ther usual care or to an intervention consisting of depression care managers offering algorithm based depression care. Usual care showed mixed results of remission depending on the illness (e.g., patients with myocardial infarction reported faster remissio n compared to patients without; but slower for patients with chronic pulmonary disease). Intervention patients showed no significant associations between treatment and remission. Results suggest that the association of medical comorbidity and treatment ou tcomes for major depression for older adults may be determined by the intensity of treatment for depression.

PAGE 41

31 Author(s) (Year) Objective(s) Population Outcomes Teh, Reynolds, & Cleary (2008) Determine of the effect of chronic medical conditions (CMCs) on the use of and quality of depression care and to understand if the patient/provider relationship mediates the relationship between CMCs and depression care quality. Severity of depression was measured by the presence of suicidal ideation and CMCs were identified by self report. 1309 adults in the U.S. with Major Depressive Disorder (MDD) identified from the National Survey of Alcohol, Drug, and Mental Health Problems. 6.1% Medicaid 16.1% uninsured 60.5% had private insurance (employer or indivi dual). Depressed people with at least one comorbid CMC were more likely to have their depression recognized than those without a CMC, though were no more likely to receive adequate depression care or patient satisfaction. Additionally, aspects of the pa tient/provider relationship including trust, and continuity of care may help explain the increased rate of depression recognition among patients with CMCs. Ani et al. (2009) Cross sectional study using survey data to compare guideline concordant treatmen t, and follow up care between primary care patients with chronic medical conditions and depression, and depression alone. 315 primary care patients recruited from 3 separate public primary care clinics with depression at baseline using the PHQ 9. Comorbid ities were measured using the Charlson Comorbidity Index. No significant difference in the likelihood of depression diagnosis, guideline concordant treatment, or follow up care in individuals with depression alone compared to those with concomitant chroni c illnesses was found. Severity of depression did contribute to being diagnosed with depression. The authors concluded that physician depression care in primary care settings is not influenced by competing demands for care for other coexisting chronic ill nesses. Katon et al. (2010) An RCT designed to determine whether a primary care based, care management intervention (self care and pharmacotherapy) f or complex patients delivered in collaboration by nurses and primary care providers, could improve medica l outcomes (i.e., diabetes coronary heart disease, or both) and depressive symptoms. A portion of the intervention consisted of training nurses in motivational coaching to help patients set goals and solve problems around medica tion adherence and self con trol around diabetes management. 214 primary care patients identified to have a diagnosis of diabetes, coronary heart disease or both according to the ICD 9 and a PHQ 9 score of were White, and ~12% were unemployed or disabled. Those p atients in the intervention group had greater overall one year improvement across measurements of diabetes control and depressive symptoms. Other improvements for intervention patients included higher ratings of diabetes, coronary heart disease, and depre ssion care compared to control patients as well as overall rating of quality of life.

PAGE 42

32 Author(s) Year Objective(s) Population Outcome(s) Vera et al. (2010) To examine if using a collaborative care model (including cognitive behavioral therapy and anti depressant medication) would improve clinical and functional outcomes compared to usual care (care managers encouraging depressed patients to talk to their provider about their mental health) for complex primary care patients with depression in Puerto Rico (specifically to replicate other findings outside the U.S.) age), Spanish speaking primary care patients with depression (PHQ 9 and the Hopkins Symptom Checklist 20) and at least one other chronic illness (e.g., diabetes, heart disease, hypertension). Depressed patients with coexisting me dical illnesses offered an evidence based antidepressant treatment or cognitive behavioral therapy reported significant reduction in depressed symptoms and improved functioning within a collaborative care model that included psychiatric consultation, case management, and patient provider education compared to their usual care counterparts. Furthermore, those patients with depression and chronic illness use of depression care significantly increased when given the opportunity to receive care in primary care settings. Vyas & Sambamoorthi (2011) Adults with depression and at least one chronic physical condition were clustered into body systems (e.g., cardiometabolic, respiratory) to compare treatment for depression among individuals with multiple chronic i llnesses to a single chronic physical condition. 1,376 adults with depression and at least one chronic condition; 45% had at least two conditions, 80% identified as white, and less 10% were uninsured. It was concluded that presence of multimorbidity in t he fully adjusted model including total number of outpatient visits was not associated with depression treatment. The authors concluded that competing demands from other illnesses did not affect depression treatment. Morris et al. (2012) To determine if differences exist in overall antidepressant treatment outcomes based on the number of general medical illnesses (using self report) in terms of depression symptom severity, medication tolerability, and psychosocial functioning. Adult (18 75 years of age) primary care and psychiatric patients enrolled in a RCT comparing single medication and multiple medications to treat depression. Number of coexisting illnesses had little or no effect on antidepressant treatment response Those patients with 3 or more coexisting illnesses reported higher rates of social and occupational functioning ( using the Work and Social Adjustment Scale). The study showed complex patients can be effectively treated with antidepressants regardless of the existence and total number of coexisting medical illnesses. Jordan et al. (2014) To study the association between multiple chronic illnesses and receiving adequate depression treatment. Administrative data from 43,189 Veterans Affairs patients with a new episode of depression. Ch ronic conditions were examined singularly. Those patients with cardiovascular disease, peptic ulcers/gastroesphageal reflux disease, or arthritis were 8 13% more likely to receive adequate treatment with antidepressants compared to patients with depressio n alone. Patients with alcohol/substance abuse were less likely to receive either adequate antidepressant care or continuation phase treatment compared to those patients with depression alone.

PAGE 43

33 Theoretical Approach The patients sampled for this study were largely unemployed, poor, and marginalized, and suffering from resource scarcity and economic and social demands that can supersede health behavior change and treatment adherence. Consistent with the existing research on prevention and health promotion interventio ns for low income populations, it is important to recognize the economic, social, and psychosocial conte xts that potentially hinder health treatment and illness improvement. The uses of theoretical models when developing interventions can help illuminate the potential impacts of these tensions as well as guide the practices around recognizing and experiences that could improve clinical communication and improve health outcomes (e.g., depression symptoms). Author(s) Year Objective(s) Population Outcome(s) Men ear, Duhoux, Roberge, & Fournier (2014) To identify primary care practice characteristics associated with quality of depression care in patients with comorbid chronic medical and/or psychiatric conditions. Patient surveys from 61 primary care clinics in Q uebec, Canada totaling 824 adults with depression and comorbid chronic conditions. Likelihood of having depression recognized was higher in clinics with access to mental health professionals, and having at least one general practitioner at the clinic devo ted to mental health was also associated with improved treatment for depression. Stanners, Barton, Shakib, & Winefield (2014) A thematic analysis from interviews with multimorbid patients to explore experiences of depression diagnosis and treatment. 12 se mi structured interviews were conducted with patients with two or more chronic conditions and a diagnosis of depression using a metropolitan multidisciplinary outpatient clinic. The interviews elicited descriptions of multimorbid patient contexts for devel oping depression, and their experiences of the detections and management of depression. Common themes identified included: the loss of identity, denial about the presence of depression, low self efficacy with treatment regimens, and coping skills like exer cise and pet ownership. Recommendations from the study include advising general practitioners raising the subject of mood and suggesting psychotherapy as part of their treatment.

PAGE 44

34 Because I am interested in exploring the impacts of coexisting illnesses on depression care and also the potential factors hindering or facilitating depression symptom improvement, I have identified two theoretical models that help me frame and understand results from the analyses, including my hypotheses. Competing Demands Competing demands often hinder care and lead to suboptimal communication between patients and providers (Williams, 1998; Rost, Nutting, Smith, Coyne, & Cooper Patrick, 2000). Poor depression recognition and t reatment in primary care has been attributed to the concept of competing demands which influence how physicians and patients decide which problems to address during a given visit, or over a sequence of visits ( Henke, Zaslavsky, McGuire, Ayanian, & Rubenstein, 2009 ; Jaen, et al., 1994 ; Klinkman, 1997 ; Nutting, et al., 2000 ; Rost, et al., 2000 ) Given most depression care is handled in primary care settings, and most if not all patients with depression will pres ent at least one coexisting illness, it can be difficult to prioritize treatment, as well as cause disagreement between the patient and provider on which illness is most important Within biomedicine and social medicine, one dominant conceptual model to e xplain suboptimal clinical encounters and outcomes is the presence of competing demands consisting of clinical (in the form of more pressing coexisting illnesses), and social (e.g., daily hassles, economic adversity) (Jaen, Stange, & Nutting, 1994; Klinkma n, 1997; Williams, 1998; Nutting et.al. 2000). Competing demands suggests that providers and patients each have their own (often conflicting) priorities or agendas ( Nutting, et al., 2000 ) they each want addressed during the visit(s). The structure of the Competing Demands Model for the Delivery of Psychosocial C are ( Klinkman, 1997 ) is comprised of three domains directly influencing clinical encounters: clinician (provider), patient, and practice ecosystem (FIGURE II.2 ). For the purposes of this project, patient level demands are the primary focus in identifying potential influences that inhibit behavioral health care. The patient domain focuses on elements that directly influence the clinical

PAGE 45

35 encounter. For example, at the provider level, PCPs are required to have som e level of knowledge about the symptoms of depression and differentiating depression from other possible disorders and along with attitudes, time constraints, and personal knowledge can influence decision making and quality of care ( Jaen, et al., 1994 ; Jaen, Stange, Tumiel, & Nutting, 1997 ; Klinkman, 1997 ; Nutting, et al., 2000 ; Rost, et al., 2000 ) Patients are expected to have similar levels of competency about their illnesses, and be able to understand how t heir personal knowledge and beliefs can be an asset or a hindrance for symptom improvement. Though the presence of coexisting illnesses is found to affect both domains (patient and provider) the specific dynamics are quite different. For providers, comp lex patients require fundamentally different approaches to care. Treating complex patients may require more time and a higher level of skills in order to recognize and treat physical ailments as well as mental health problems; both identified as competing demands for adequate care. For patients, coexisting g their level of self efficacy creating a feeling of hopelessness, affecting treatment adherence and the quality of the patient provider interaction.

PAGE 46

36 Using the model of competing demands (adapted for patient level demands), I posit that competing demands are one means through which patients with higher levels of coexisting illness severity using the Cumulative Illnes s Rating Scale (CIRS) along with social and economic tensions, will experience significantly less depression sym ptom improvement over time and be less responsive to the intervention compared to patients with lower illness severity. The Illness Narrative The qualitative component of this project was chronic illness as a disease and how patients define and experience illness (Kleinman, 1988). What we know from qualitative and narrative research about chronic illness an d it s e ffects on the human condition is the individuality of each experience. The domains that make up much of a Provider Domain Knowledge Beliefs and attitudes Ski lls Time constraints Alternative demands (e.g., coexisting illness) Personal knowledge of patients Patient Domain Knowledge Beliefs and attitudes Characteristics Coexisting illnesses Expectations Personal knowledge of providers Provider delivery of psych osocial care FIGURE II.2: The Competing Demands of Psychosocial Care (Klinkman 1997)

PAGE 47

37 efficacy, and psychosocial health are influenced by the course of chronic illness, and the manifestations of that influence can adversely impact treatment regimens and the healing p rocess. Kleinman focuses on the illness experience which he defines as the: categorizing and explaining, in common sense ways accessible to all lay persons in the social group, the forms of distress caused by those pathophysiological processes. And when we speak of illness, we must include practical problems in daily living it creates. Illness behavior consists of initiating treatment (for example, changing diet and activities, ta king over the counter medication or on hand prescription drugs) and deciding when to seek care from professionals or alternative practitioners (p. 4). Additionally, Kleinman discusses the debilitating ways illness affects a p h ow a disabling illness confines, frustrates, and disappoints the patient often leading to significant loss of interest and pleasure in partic ipating in everyday activities a key indicator for major depressive disorder. This experience Kleinman argues m ust be legitimized during the clinical effectively treat illness and establish the connection necessary for chronically ill patients. By having dialogue with their patients, providers can gain a better perspective of how illness is experienced both from a contextual and social perspective, assisting them in adopt ing the appropriate attitudes, knowledge, and skills necessary to treat complex patients. The collection of interviews, the tape recorded visits, and data sources (e.g., demographic variables, medical charts) used for this dissertation provided an understanding of patient needs and barriers when it comes to their care. Findings from the qualitative intervie ws of patients helped to inform a more comprehensive approach to patient behavioral health care, including a better understanding of the bi directional effects of depression and chronic illness. The Illness Narratives as a framework, aimed

PAGE 48

38 to collect sometimes sensitive data on depression, and to better understand their needs and barriers when it comes to their care. This supports Kleinma directed at each (emphasis added) of the major problems and integrated withi n a comprehensive experience, I hope to gain rich descriptions from the data, and information critical to the design of future interventions for training primary ca re providers in mental health recognition and treatment. Summary of the Theor etical Frameworks The frameworks and the ories reviewed in this chapter serve as guides fo r much of this dissertation offering rationale for the aims, hypotheses and methods p urposed. For example, though competing demands has been studied mostly at the provider level, clinical interventions need to account for patient level demands contributing to health behaviors and outcomes. Coexisting illnesses, a construct of the patient domain of competing demands, can contribute adversely to the clinical encounter by creating a communicative disconnect between the patient and provider as a result of conflicting agendas. describe the importance of und erstanding the experiences of each patient and how those experiences influence decision making around health and health compromising behaviors, treatment adherence and preferenc es as wel l as how health is prioritized.

PAGE 49

39 CHAPTER III METHODS This chap ter describes the general framework for this mixed methods study, beginning with an overview of the randomized control tria l (RCT), which guided both the quantitative and qualitative components of the stu dy. Next, the research design, study samples, data c ollection procedures, and analyses for both the quantitative and qualitative portions of this study are discussed. Given the complexities surrounding mental health a mixed methods app roach was deemed most appropriate for explain ing the bi directional rel ationship between depression and physical illness. Though a purely quantitative design is often appropriate for social science inquiry, qualitative data offers rich and empathic descriptions potentially resulting in a better understanding about social /hea lth phenomena ( Johnson & Onwuegbuzie, 2004 ) Description of the Original Randomized Controlled Trial Both quantitative and qualitative data for thi s study came from a RCT funded by the National Institutes of Mental Health (Grant # s K23MH0829972 & 3K23082997 S1 ) conducted at Denver Health and Hospital Authority (DHHA) from May 2010 to November 2012 (Total n = 168) ( Keeley & Brody, 2007 ) The purpose of the RCT was to test a psychosocial strategy using an adapted form of Motivational Interviewing (MI) delivered by tra ined primary care providers (PCP s) to improve dialogue around depression with patients, and improve depressive symptoms. The control is national standard of care treatment for depression in clinical settin gs which includes assessment of suicide risk, assessment of substance abuse, and discussions with patients around depressio n treatment options (Mitchell et al., 2003) MI is an empirically based counseling method designed to improve medication adherence for various chronic conditions. The investigators of the RCT originally hypothesized that MI delivered by PCPs would improv e depressive symptoms for patients with a new treatment episode of depression by increasing the

PAGE 50

40 adherence of antidepressant medication The trial involved two phases: Phase 1 included brief structured interviews delivered by trained providers using MI to improve treatment adherence for during the clinical encounter; and assessment of treatment integrity. Phase 2 compared the MI to standard care (Guideline Base d Medical Managem ent) to determine if MI improved adherence and outcomes Additional aims of the RCT investigated factors that potentially mediate change recove ( Keeley & Brody, 2007 ) For this dissertation, secondary quantitative data from the RCT study pertaining to depressive symptoms improvement was analyzed at 6 12 and 36 medical history (chart reviews) for all 168 patients and patient interviews were collected and analyzed for this dissertation study only. Table III .1 summarizes the RCT (primary study) and the dissertation study

PAGE 51

41 TABLE III 1 : Description of the Primary and Dissertation Studies Primary Study Dissertation Study Title Motivational interviewing (MI) adapted to improve depression treatment in primary care Depression treatment fo r complex patients in primary care: The role illness severity in depression symptom improvement Description/Phases A randomized control trial using an adapted form of MI to improve antidepressant adherence and depressive symptoms for patients with a new e pisode of depression funded by the National Institutes of Mental Health (NIMH). The RCT had two phases: Phase 1 included training primary care providers in MI Phase 2 compared MI to standard care to determine if MI improved treatment adherence and depressi ve symptoms A secondary (quantitative aims) and primary (qualitative aims) study the potential effects coexisting illness severity has on depression symptom improvement in primary care settings N arrative data from complex patients concerning their lived experience with depression and chronic illnesses. The dissertation had three phases: Phase 1 scored each participants coexisting illness severity using the Cumulative Illness Rating Scale (CIRS)* Phase 2 tested if CIRS had an impact on depression sympto m improvement over time Phase 3 conducted and analyzed semi structured interviews with purposefully selected patients to assess their lived experience with depression including symptoms, causes, competing demands, illness treatment priority and their overa ll attitudes about living as a complex patient and depression Patient populations All participants received their health care from Denver Health and were screened for depression using the Patient Health Questionnaire 9 (PHQ 9). Inclusion criteria includ ed: 18 years or older at baseline PHQ 9 score of > 10 Confirmation of depression Informed consent Exclusion criteria included: Receiving specialty mental healthcare during the past 90 days at of recruitment Females either pregnant or nursing High risk fo r suicide Evidence of perinatal depression, bipolar disorder, psychosis, or active substance abuse; and cognitive, language, or hearing impairment severe enough to preclude participation All participants that completed all phases of the RCT, and inclusion and exclusion criteria followed the RCT protocol. For the qualitative phases, participants were selected based on their CIRS score and availability to gain a broad sample of perspectives. Informed consent was collected for every participant.

PAGE 52

42 Primar y Study Dissertation Study Aims Assessment of patient and provider ratings of feasibility and acceptability of an adapted form of MI Compare MI to enhanced usual care for increasing treatment adherence and improving depressive symptoms for patients exper iencing a new episode of depression To explore moderators and mediators of the effect of MI on adherence and outcomes. Does not include a comprehensive measure of comorbidity severity Explored the potential effect coexisting illness severity has depression over time ; bi directional depression and chronic illness Identify the competing demands that treatment preferences in receiving care for depression and physical illnesses Study Setting T he study population for the RCT consisted of patients receiving their care from Denver Health and Hospital Authority (DHHA), a comprehensive health care network that cares for about one quarter of Denver residents ( Denver Health: About Us, 2012 ) DHHA is a network of eight fa mily health centers, two hospital based urgent care centers, and 15 school based health centers. Participants were recruited from seven of the eight family health centers and from the main hospital. The adult patient population that utilize d services is approximately 16% non Hispanic Black, 57% Hispanic, 19% non Hispanic White, and 8% other, with 10% being 65 years or older. Many of the patients at DHHA have incomes that fall below the federal poverty line and most are enrolled in either Medicaid or the Colorado Indigent Care Program (CICP). Given that DHHA provides health care regardless of ability to pay, services are restricted to priority care and mental and/or behavioral health problems are often under diagnosed and untreated. The RCT recruited pati ents from seven DHHA clinics located around the Denver Metro area. Study Participants The RCT enrolled participants from DHHA. All participants were required to be at least 18 years of age and have had contact with a primary care provider within 12 mon ths prior to the study. Additionally, they had to have been diagnosed with moderate to severe depression at

PAGE 53

43 baseline Those patients with serious alcohol or drug addictions, no access to a telephone, or without some proficiency in English were excluded fr om the RCT. All participants gave their informed consent. Women and men 18 years of age or older who were receiving care at Denver Health and had received a diagnosis of moderate to severe depression were included in the study. Patients were excluded fro m the study for the following reasons: Receipt of an antidepressant medication in the previous 90 days other than a low dose tricyclic antidepressant for pain or Trazodone for sleep Receiving interpersonal or cognitive behavioral psychotherapy focusing on depression Pregnant or nursing Drug or alcohol dependency or abuse (excluding caffeine or nicotine) High risk for suicide Inability to communicate in English Lifetime bipolar disease Psychosis History of autism, mental retardation, or pervasive developmen tal disorders Cognitive, language, or hearing impairment severe enough to preclude participation. Measure s Patient Health Questionnaire 2 and 9 The Patient Health Questionnaire 2 ( PHQ 2) (A ppendix B) was used as one part of the screening process and to assess baseline depressive symptoms. The PHQ 2 is a 2 item measure that inquires about the frequency of depressed mood and anhedonia, defined as loss of interest in previously interesting or enjoyable activities ( Kroenke, Spitzer, & Williams, 2003 ) The stem question of the PHQ 2 is Over the last 2 weeks, how often have you been bothered by any of the following problems? The 2 items ar e: Little interest or pleasure in doing things (i.e., anhedonia) and Feeling down, depressed, or hopeless.

PAGE 54

44 The Patient Health Questionnaire 9 (PHQ 9) (A ppendix C) is a 9 item scale that includes the two items that make up the PHQ 2 along with the followi ng additional items: Trouble falling or staying asleep, or sleeping too much Feeling tired or having little energy Poor appetite or overeating Feeling bad about yourself or feeling that you are a failure or have let yourself or your family down Trouble con centrating on things such as watching television or reading Moving or speaking so slowly that other people have noticed Thoughts that you would be better off dead (Suicidality) As measures, the PHQ 2 score can range from 0 6, and the PHQ 9 can range from 0 27, each item scored as follows: The PHQ 9 can be a useful tool for diagnosing and helping patients to manage their depression in primary care settings mainly bec ause it both offers a continuous measure of depressive symptoms and a method to monitor change and treatment outcomes ( Arroll, et al., 2010 ) In a report of the largest validation study of the PHQ 2 and 9 in a primary care setting, Arroll and colleagues concluded the PHQ 2 was very sensitive (0.86) and had a specificity of 0.78 for a diagnosis of depression when compared to the Composite International Diagnostic Interview (CIDI) (2010). The PHQ 9 had similar sensitivity (0.74) but a higher specificity (0.91) ( Arroll, et al., 2010 ) which was t he cutoff score used for this study. A 2004 study of the reliability and validity of the PHQ 9 compared to the Hopkins Symptom Checklist Depression Scale (SCL 20) reported a high rating for test retest reliability (0.81 for worst case sample and 0.96 for best case sample), and a significantly greater responsiveness at 3 months; 1.3 (95% confidence interval [CI]) versus 0.9 ( Lowe, Unutzer, Callahan, Perkins, &

PAGE 55

45 Kroenke, 2004 ) Two other s tudies ( Lowe, Grafe, et al., 2004 ; Lowe, Spitzer, et al., 2004 ) have shown the PHQ ( Lowe, Unutzer, et al., 2004, p. 1195 ) compared to other self report measures for depression. Description of the Dissertation St udy As previously mentioned, the primary quantitative data from this study came from the two year prospective RCT (discussed above), as well as the patients selected for the qualitative phase. The quantitative portion of the study reported here is a seco ndary analysis of the RCT data (n=168). The primary goal of the quantitative component of this study was to determine the potential influence coexisting illness(s) and their severity have on depression symptom improvement over time. Measures Cumulativ e Illness Rating Scale For this study, the Cumulative Illness Rating Scale (CIRS) ( Linn, Linn, & Gurel, 1968 ) ( A PPENDIX E) was added as a primary data collection sses, in addition to their depression. Data derived from the CIRS are useful for quantifying and summarizing medical illness burden. ( Kemp, et al., 2010 ) Using medical record review and International Statistical Classification of Diseases and Related Health Problems (ICD) 10 diagnostic and procedure code data, physical and behavioral morbiditi es were identified at baseline. The comorbid diseases of the patient was rated using the CIRS, wh ich classifies comorbidities by 14 organ systems affected and rates them according to their severity from 0 to 4 ( Linn, et al., 1968 ) The scores are as follows: ability requiring first was scored because presence of severe illnesses are including in the exclusi on criteria)

PAGE 56

46 Existing literature supports the use of the CIRS calculated from chart reviews along with high recommendations as an available method of measuring comorbidity ( de Groot, et al., 2003 ) CIRS scores were only collected once, at baseline, because it was not the aim of this project to evaluate improvement in any illnesses other than depression as measured by the PHQ 9. In accordance with exis ting literature CIRS scores wa s stratified into three categories based total score: Low (0 7) Moderate (8 14) High (15 +) It was assumed that because patients were seeing their primary care provider for a separate reason when recruited for the RCT they will have at least one coexisting illness in add ition to their depression. Then, CIRS scores were categorized into three ratings for each patient: Total Score. Number of categories endorsed. Severity index (total scor e/number of categories endorsed ) Levels of severity rating for each of the 14 systems of the CIRS were measured using a 0 appropriate system and then given a score using the specific guidelines for each disease/system ( Hudon, Fortin, & Soubhi, 2007 ) For reference, the general rules for each potential score are listed below ( Hudon, et al., 2007 ) : 0 No problem affecting that particular system (no disease listed) 1 Current mild problem. 2 Moderate disability or morbi dity and/or requires first line therapy. 3 Severe problem and/or constant and significant disability and/or hard to control chronic problems. Extremely severe problem and/or immediate treatment required and/or organ failure and/or severe functional impairme nt.

PAGE 57

47 Sociodemographic information was used as explanatory variables and included age, household income, race/ethnicity, gender, and employment status all which were collected at baseline and self reported The demographic variables were included as covar iates in the statistical models. Additionally, they were used to assist with the stratified purposeful sampling for the qualitative interviews. Data Collection Procedure The RCT study first randomized PCPs into MI training (treatment) or enhanced usual care (control) and the patients were contacted for participation in the study. The RCT and the amendments necessary for the secondary analysis and qualitative interviews were approved by the Colorado Multiple Institutional Review Board (Protocol# 08 1282) The research team contacted primary care patients who met the inclusion criteria first by telephone to discuss the study, consent to the study, and to make an appointment in accordance with a future visit to meet in order to complete the screening quest ionnaires. At their clinic visit, all participants were again asked to consent to the study. If consented, patients completed a demographic survey ( A PPENDIX D), as well as the Patient Health Questionnaire (PHQ 9) ( APPENDIX C), a more thorough instrument f or measuring major depression compared to the PHQ 2. The demographic data collected included employment status, marital status, income and housing status (e.g., rent, own), level of education completed, and race/ethnicity The PHQ 9 was used to ascertain major and minor depression at baseline and follow up. PHQ 9 scores were collected at four time points, baseline, 6 12 and 36 weeks. In practice, the testing dates for each patient varied and not every patient was screened at each time point. The CIRS were s cored using medical and pharmaceutical records by a licensed credentials necessary to score the CIRS appropriately). A physician also score d 25 patients chos en randomly in order to compare scores with the PA to ensure fidelity along with allowing for dialogue around scoring procedures and consistency.

PAGE 58

48 The independent variables selected for the quantitative analyses were determined by the literature reviewed in Chapter 2, as well as the existing RCT dataset that was collected. Since identifying competing demands of primary care patients was an aim for this analysis, independent variables were selected based on factors that potentially influence depression car e, particularly for complex patients. Data Analysis The purpose of this dissertation study was to test whether severity of coexisting medical illnesses is related to depression symptom improvement over time for depressed complex primary care patients Bo th within person changes in depression (i.e., intraindividual change) and between person changes in depression (interindividual differences) were analyzed using hierarchical linear modeling (HLM). By using HLM, the potential differential impacts of illne ss severity on d epression trajectory over time were tested Applying HLM to longitudinal data offers the opportunity to explore theoretically different questions, which is not possible with regression analyses or cross sectional data ( Taylor, Ntourmanis, Standage, & Spray, 2010 ) To address the A im 1, quantitative data were analyzed using Hierarchical Linear Modeling (HLM). HLM was selected because of its fit for studying the predictors of i ndividual change as well as analyzing hierarchically structured data accounting for the nested structure of these relationships including the individual, ecological contextual, and individual contextual relationships often neglected in logistic and regress ion models ( Raudenbush & Bryk, 2002 ; Subra hmanian, Jones, & Duncan, 2003 ) Given the data collected for this study were collected at various time points, HLM allow ed nesting repeated scores within persons as an individual (Level 1) measurement in order to show individual differences in growth c urves ( Kreft & De Leeuw, 1998 ; Tabachnick & Fidell, 2007 ) By organizing the data into hierarchies (e.g., repeated measures of de pression nested within patients ), HLM allowed for individual level variables as well as group level variables to be included in the analyses ( Kreft & De Leeuw, 1998 ) Additionally, because not every patient was screened at every time point, HLM was optimal

PAGE 59

49 method of analysis of change (e.g., repeated measures ANOVA) because it did not require an equal number of responses from each patient, meaning every patient could be included in the analysis even with missing values ( Raudenbush & Bryk, 2002 ; Taylor, et al., 2010 ) Each time measurement of depression was nested within each patient Equation 1 outlines the basic rudiments of the within patient (i.e., Level 1) models used in the linear growth models: (1) where: is the outcome at time t for patient i ; takes on a valu e of 0 at baseline, a value of 6 at 6 weeks, a value of 1 2 at 12 weeks, and a value of 3 6 at 36 weeks; is the initial status of patient j that is the expected outcome for that patient at baseline (when t =0); is the depression score change rate for patient j during the study; and e is the error term. The estimate parameter from the with in patient model (Level 1) will then be used as the outcome variable in the between patient equations (Level 2): (2) (2a) where : is the regression parameter; X is a time of comorbidity severity, and gender ); and e is the error term. Degree of variance in the study variables was explored using intercept only models (i.e.,

PAGE 60

50 no pr edictor variables were included) for all study variables, separated into two parts: variance associated with Level 1 errors (within patient), and Level 2 errors (between patient ). From these models, interclass correlations coefficients (ICCs) were compute d to describe the proportion of variance associated with the between patient level. Additionally, statistical assumptions associated with multilevel modeling were assessed (e.g., error terms are normally distributed, homoscedasticity, and independence of observations) ( Raudenbush & Bryk, 2002 ) In addition to the HLM, a Repeated Measures ANOVA was used to test the correlations between the main variables (i.e. PHQ 9 and CIRS). Qualitative Primary Data Research Design In general, the qualitative aim of this dissertation was to triangulate the quantitative coexisti ng illness, as well as the competing demands expressed by patients including illnesses, issues with employment and income, and other identified daily hassles. An additional intent of the interviews was to inform future depression interventions about how pa tients regard their depression in terms of treatment preferences and communication with their provider during the clinical encounter primarily among socially and economically disadvantaged adults with minimal access to psychiatric services. Although the a ims of the study are supported by the literature, the paucity of contextual knowledge about bi directional relationship between depression and research strate gies, which can garner information often not obtainable with quantitative research methodologies ( Addison, 1999 ; Pope & Mays, 1995 ) Data Collection Procedure Though the first draft of the interview guide (A ppendix F ) was developed pri or to the first interview, the guide was amended as necessary to best elicit the information to address the aims, which is often advisable in qualitative research ( King & Horrocks, 2010 ) The interview

PAGE 61

51 guide was developed using an iterative process that included the author of this dissertation three of the dissertation committee members (KL, DM, & RK) as well as the data manager of the guide that included changing particular words, order of the questions, and the number of questions. Two more interviews were completed before the final question guide was completed. A copy of the final interview guide is included as Appendix F. Interviews were only conducted in English. Using purposeful sampling techniques, patients were stratified by CIRS score (i.e., low, moderate, high). Nineteen patients were interviewed to collect descriptive accounts about Six of the interviews were conducted by the author of this dissertation and the remaining 13 by the data manager of the original RCT. Interviews were done by telephone and lasted approximately 10 20 minutes each. The telephonic method was selected given that the patients could not be directly observed, an d the patient population at DHHA suffers disproportionately from challenges around transportation (e.g., some patients have identified that they take several bus routes to get to their primary care center) and childcare. The patients interviewed were comp ensated for their participation with $25.00 gift cards for a local supermarket The interviews were audiotaped and fully transcribed for analysis. The semi structured format was selected to enable the opportunity for patients to elaborate about their expe riences and for the interviews to be less restrictive, while still maintaining a level of fidelity through the use of a question guide to provide consistency in the question asked d uring each interview. Patton ( 2001 ) identified six types of interview questions, each looking to elicit certain information from the participa nt. The interview guides for this study will focus on three of the categories:

PAGE 62

52 Experience/behavior questions: These questions focused on specific actions and reactions experienced by the patient that could be observed such as reactions to a physical or ps ychological diagnosis; reactions to questions asked by the PCP about depression and other psychological distress; and how (if at all) the patient initiates dialogue with their provider about their concerns. Opinion/values questions: These questions focus ed on what the patient thinks about the topic of interest (e.g., depression and physical illness) and/or how their thoughts about the topic relate to their values, goals, and intentions. For example, the interview script will include questions about what th e patient thinks is the best way to treat depression and his/her physical illnesses; how the patient sees the relationship during a primary care visit given his/her curren t health status. Feeling questions: These questions focus ed on the emotional experiences felt by the patients. It is important to phrase these questions with the term emotion in the but also to different iate them from opinion/values questions, which Patton warns can happen without intention. Data Analysis Data collected from the interviews (text transcribed from audio recordings) we re interpreted using a blending of suggested techniques for systematically examining qualitative Crabtree and Miller (1999), the initial techniques consisted of mu ltiple, in depth readings of the transcripts (immersion), with a different focus for each reading depending on assumptions and observations made during initial readings (crystallization). Other reflective techniques consist ed of dialogue with colleagues a nd advisors about patterns and emerging themes and to cross check interpretations and codes derived from the interviews ( Addison, 1999 ; Creswell, 2009 ) Members of the dissertation committee met on several occasions to discu ss the themes identified from the interviews as well as to check for compliance with the research questions posed for this dissertation research. The analysis was concerned with both the language a nd the content of what

PAGE 63

53 the patients share d during the inter views to understand their lived experience dealing with illness from their perspective ( King & Horrocks, 2010 ) For this study, a variation of the three stage th ematic analysis system (FIGURE III .1) was adapted to code the transcribed interviews ( King & Horrocks, 2010 ) Though the process has a sequential order, changing and rethinking codes occur red and stages repeated to best interpret the data. The first stage identified sections of the interviews that we re relevant to the study aims. Initial coding consisted of relatively broad a priori codes developed from the existing literature and the qualitat ive aims of the study while highlighting any quotes that are relevant to aims with comments that will help organize the text into themes. The second stage organized the broad coding and the highlighted sections into descriptive codes, using single words o r short phrases relevant to the study aims. The descriptive codes were applied to all of the interviews and elicit ed the strength of the concepts in the data and tie d together any similarities across the interviews. The third and final stage of coding en tailed identifying key themes in the text by grouping together descriptive codes and any emerging themes that share d common meanings along with creating a diagram that exhibit ed the relationships within the text (APPENDIX E). The initial results from the qualitative analysis we re reported by theme then organized by more specific themes Qualtitative Anaylsis as an Iterative Process model (1993) (1993) (FIGURE III spiral ( Dey, 1993 )

PAGE 64

54 Qualitative Validity In order to improve the validity of the qualitative findings, a clear explanation of the data collection methods and analysis was i ncluded as part of this study along with the demographic and health data of each participant (patient) to show differences among the interviewees which can be conducive to obtaining different perspectives ( Mays & Pope, 2000 ) Additionally, academic advisors and colleagu es reviewed the findings to refine key themes and codes. In order to ensure accuracy and credibility of the qualitative results, I employed procedures including (Creswell, 2009): Checking transcr ipts for mistakes made during transcription of the interviews. Staying consistent with using codes derived from the literature and the interviews, and documenting any changes to the code book. Sharing the analysis with committee members familiar with the s tudy. Utilizing an external auditor with qualitative analysis experience but with little to no knowledge of the study FIGURE III.1: Iterative Process Model (Dey 1993)

PAGE 65

55 Quality Checks at all Stages of the Analysis Stage one: General Coding Read through entire transcript Remove any text that could reveal the identity of the patient Define and identify a priori codes Highlight relevant material and attach notes and comments Repeat for each transcript Stage two: Descriptive/Interpretive Coding Define descriptive codes Cluster descriptive codes Interpret the meaning of the clusters in relation to the research questions Apply interpretive codes to full data set Stage three: Overarching Themes Identify key themes for the data set as a whole (comparison analysis) Construct a diagram to r epresent relationships between levels of coding in the analysis Include quantitative findings for the final analysis FIGURE III. 1 : Description of the Qualitative Analysis

PAGE 66

56 CHAPTER IV RESULTS: QUANTITATIVE ANALYSIS This chapter re ports the quantitative results for the secon dary analyses of the original RCT dataset, including descriptive statistics of the variables, correlations, and hierarchical linear regression analyses. Descriptive statistics and correlations were run using SPSS Version 22.0 (IBM Corp, 2013) while all hierarchical analyses were run using the student version of HLM Version 7 (Scientific Software International, 2013) Quantitative Patient P opulation Table IV.1 summarizes the sociodemographic characteristics of the sample. All 168 patients from the ori ginal RCT were eligible for this secondary analysis. Patient characteristics including age, gender, income and employment status are summarized in Table IV.1. Of the 168 patients, the mean age of the sample was 48.9 years of age (SD=12.3) and 118 (70.2 % ) were female. Non white and Hispanics and African Americans represented the largest racial/ethnic groups and over half of the patients (50.3 % ) reported an income of less than $10,000 in the past year at baseline; over half of the sample also reported bein g unemployed (53.4 % ), and were either looking for employment or unable to work due to physical or mental illness(es). As shown by the demographics, a majority of the sample is low income, unemployed (many due to mental or physical disability), and racial/e thnic minorities, all groups disproportionately affected by untreated mental illness and concomitant illnesses.

PAGE 67

57 TABLE IV.1: Patient Demographics Characteristic Total (n=168) Age Mean (SD) 48.1 (13.2) n (%) Female 118 (70.2) Education Less than high school 86 (51.3) High School 12 (7.1) Some College 14 (8.3) Vocation/Trade 40 (23.8) Race African American 57 (33.9) Hispanic 57 (33.9) Non Hispanic white 39 (23.2) Other 15 (9.0) Income < $10,000 82 (54.7) < $15,000 29 (19.3 ) < $25,000 25 (16.7) < $35,000 8 (5.3) > $35,000 6 (4.0) Employment status Working for wages 40 (23.8) Unemployed (undisclosed reason) 26 (15.5) Unemployed (due to mental or physical illness) 62 (36.9) Student/homemaker/other 40 (23.8)

PAGE 68

58 Resea rch Question 1 Research question 1 (RQ1) aimed to determine if the severity of coexisting medical illnesses impacts depression symptom improvement over time for primary care patients experiencing a new episode of depression over time. It was hypothesize that patients with high competing illness severity would experience significantly less depression symptom improvement over time and would be less responsive to the RCT than those with lower illness severity. Depression and Illness Severity The Patient Health Questionnaire (PHQ 9) for all 168 patients were collected and were the main outcome of interest for the original RCT and used to measure depression change over time for this study The range for the PHQ 9 is 0 27 with data collected at four time points: baseline, 6 12 and 36 weeks. From baseline to 36 weeks, the mean PHQ 9 went from 15.99 to 11.15 f or the entire sample (TABLE IV.2 ). TABLE IV.2: PHQ 9 Mean Scores at all Four Time Points Mean (SD) Standard Deviation 95% Confidence Interval of the Difference Time point Baseline 16.0 (4.2) 4.2 15.4 16.6 6 weeks 11.5 (5.3) 5.3 10.6 12.4 12 weeks 11.8 (5.6) 5.6 10.9 12.8 36 weeks 11.2 (5.6) 5.6 10.2 12.2 Given that RQ1 is designed to test the potential impact of illness severi ty, some of the data collected to measure ill ness severity are reported below: the mean scores for illness severity, as well as data concerning the body systems affected by illness (e.g., psychiatric, vascular). For the quantitative analysis, CIRS are rep orted using two separate categories :

PAGE 69

59 Total systems affected (14 possible systems) Total CIRS score (0 54) Total CIRS scores were first measur ed by the total sample (n=168) and stratified into one of three categories, low, moderate, or high. Among the pa tients in the RCT CIRS scores were fairly evenly distributed. All but one patient had at least a CIRS score of 1. Of the 168 patients, 35.7% reported a CIRS scored between 0 and 7 (low) 33.3% scored between 8 and 12 (moderate) and 31% scored between 1 3 and 56 (high) (FIGURE IV.1). Since the distribution was fairly evenly distributed, additional analyses are reported here to look for differences among the sample including age, gender, and level of impairment. FIGURE IV .1: Total CIRS (0 56) Score Repor ted By Percent of t he Sample Table IV.3 reveals that there is a statistically significant relationship between age an total CIRS score, such that as age increases so does the burden of coexisting illness severity. 35.7 33.3 31 0 10 20 30 40 50 Low (0-7) Moderate (8-12) High (13-56) Pereent Reporting (n=168) Total CIRS Score (0 56 possible)

PAGE 70

60 TABL E IV.3 : Total CIRS by Age (n=16 8) Total Age 18 39 Age 40 59 2 n % n % n % n % Low (0 7 ) 52 31.0 28 59.6 21 23.3 3 9.7 36.1 *** Moderate (8 12) 56 33.3 15 31.9 30 33.3 11 35.5 High (13 56) 60 35. 7 4 8.5 39 43.3 17 54.8 ***p <0.001, Test: Chi Square ( Likelihood ra tio statistic) Figure IV.2 presents the means and standard deviations of the PHQ 9 scores (depression scale) at all four time points stratified by the three levels of total CIRS scores (low, moderate, high). As can be seen, depression scores for all thr ee groups improved at 36 weeks. Those patients in the moderate CIRS range ( 8 12) saw the most improvement over time

PAGE 71

61 FIGURE IV.2 : Change in depression over time stratified by total CIRS Mean PHQ 9 at baseline was 16.0 and 11.2 at 36 weeks with a me an change of 3.47, while the mean CIRS score was 10.8 The correlation between change in PHQ 9 at 3 6 weeks and CIRS score was 0.4 (TABLE IV.4) TABLE IV.4 : Main Variable Correlations Mean (SD) Pearson Correlation (PHQ 9 Change Over 36 weeks) PHQ 9 Chan ge Over 36 weeks 3.5 (5. 5 ) CIRS Total 10.8 (5.3 ) 0.4 Table IV.5 reports the results from the line ar regression looking at change in depression score over 36 weeks and total CIRS. L ess than 1% of the variance is ac counted for change in 12.0 12.7 12.1 16.0 10.9 11.8 12.0 16.5 10.4 10.5 10.4 15.3 0 5 10 15 20 36-weeks 12-weeks 6-weeks Baseline PHQ 9 Mean Score Time Point Low Moderate High Total CIRS

PAGE 72

62 depression i s accounted by total CIRS score at baseline ( 0.04) This does not support the hypothesis of the expected strong affect CIRS would have on change in depression. TABLE IV.5 : Results of Regressing Change in Depression on Illness Severity (n=125) Step Adj. R 2 F p 1 Total CIRS Score at Baseline 0.0 4 0.04 0.29 0.59 Results from the Hierarchical Linear Modeling The first HLM model that was tested was the baseline model with no predictors. The interclass correlation (ICC) which describes the percen t of variance in depression scores between patients, was 0.6 9. ICC is the proportion of the between individual variance to the sum of the between and within individual variances of an outcome variable and g enerally ranges between 0 and 1. Hox ( 2002) int explained by the grouping as the average relation between any pair of observations (i.e., the PHQ 9 scores) within a cluster (i.e., a pati ent). With this model, the intercept 0 and time 1 were exa mined for reliability (TABLE IV.6) The reliability estimates represent the proportion of the variance in the Level 1 estima tes that is parameter variance. The reliability of the random effect of the level 1 intercept is the average reliability of the level 2 units. It measures the overall reliability of the OLS estimates for each of the intercepts. The reliability estimates are .42 for intercept and .05 for slope. These indicate that the slope (change in depression) is not a relia ble estimate. In other words, P HQ 9 score is not a reliable measure depression change time (TABLE IV.6)

PAGE 73

63 TABLE IV. 6 : Results from the Random Coefficient Estimates Random level 1 coefficient Reliability estimate Intercept, 0 0.420 TIME Slope, 1 0.054 For the Level 1 model (Table IV.7), e ach time measurement of depression was nested within each patient Equation 1 outlines the basic rudiments of the within patient (i.e., Level 1) models used in the linear growth models : [EQUATION 1] Table V.10 shows the results from a r andom c oefficient model The fixed results indicate that the average P SQ 9 score at baseline was 14.1 and that the average change in slope across patients over the four time points (BL, 6 weeks, 12 weeks, 36 weeks) was 0.1 (which indicates a statistically significant decrease in depression over time). It can be inferred from the random effects results that the relationship b etween CIRS and change in depression over time does not vary significantly across the patient population; that there are no differences between patients in terms of change in depression over time slopes. This was expected given that the model did not have reliable estimates (TABLE IV.7).

PAGE 74

64 TABLE IV.7 : Results from Depression Over Time Model Fixed Effects Coeff t p (SE) Random Effects Variance 2 p Mixed Model 0 Intercept, 00 14.13 (0.33) 43.34 <0.001 7.88 260.20 <0.001 1 Intercept, 10 0.10 (0.01) 7.00 < 0.001 0.001 159.94 >0.500 Deviance 3398.49 Given that the Level 1 showed no difference between p atients in terms of depression symptom improvement over time, theoretically, a Level 2 HLM model is unnecessary because it is not expected to yield any difference from the Level 1 results. The Level 2 model was completed for illustrative purposes (TABLE I V.8). The estimate parameter from the with in patient model (Level 1) was used as the outcome variable in the between patient equation : [EQUATION 2] Significant if p < 0.05

PAGE 75

65 TABLE IV.8 : Results from the Depression Change by CIRS Total Fixed Effects Coeff t p (SE) Random Effects Variance 2 p Model for CIRS predicting depression at baseline 0 Intercept, 00 14.1 (0.3 ) 43.3 <0.00 7.9 259.5 <0.001 1 Intercept, 10 0.1 (0.0 ) 3.9 < 0.00 0.00 159.5 >0.500 CIRS_TOT, 11 0.00 (0.00 ) 0.7 0.47 Deviance 3409.9 As expected, the CIRS was not a statistically significant predictor of change in depressio n over time, which was the first research question. Response to Research Question 1 Given the reliability of the estimates offer little variability (TABLE IV.6), it is unnecessary to expand the between subjects model in order to examine the possible fac tors associated with depression symptom improvement and chronic illness severity. If most of the variability is due to error, there would likely be no systematic relations between the estimates and the second model estimates. It could then be falsely con cluded that there are no relationships when in fact the data are incapable of detecting such relationships. For the individual growth parameter reliability coefficients, averaging the estimates across the n individuals provides a summary index of the instr ument's reliability in measuring each of the growth parameters on this population of subjects. The estimated variance components for the PHQ 9 reliabilities were low, especially for the variation in the growth rate parameters for the depression assessment over time, meaning the depression levels for individual patients changed at a relatively constant rate across individuals. Significant if p < 0.05

PAGE 76

66 CHAPTER V RESULTS: QUALITATIVE ANALYSIS Cause my illnesses keep me down. Depression keeps me down. So it is just a different form to me. It is just another added on disease, so to speak. Yeah. I mean you know, pain. Like if I have aches and pains more, it (depression) hurts more. If I'm you know, if I'm sick with a cold or something, things feel like rything is harder. Let's just say everything is bigger and harder. This chapter reports the qualitative results from semi structured interviews conducted with a sample of participants from the original RCT. The chapter includes identified themes, sampl e quotes, as well as simple counts of all the codes from every interview. All handling of the data was qualitative data, as well as assist the process of assigning and managing code s assigned to the text. The primary aim of the qualitative results was not to only offer context to the quantitative results but to increase the overall strength of the study, and to derive data to compare with the quantitative results as well as to ma ximize any similarities and the differences of information (Creswell, 2009). The qualitative data was used t o answer two of the three study research questions: H ow do complex patients describe the lived experience with depre ssion and concurrent illnesses? How does a diagnosis of a new episode of depression change the management and prioritization of concurrent illness for complex patients?

PAGE 77

67 Qualitative Patient Population The sample for t he qualitative analysis included 19 participants using Denver Health for their primary care with 4 being male; CIRS ranging from 6 21, ages ranging from 36 68; and baseline PHQ 9 scores ranging from 10 25, which is representative for the larger quantitative sample (TABLE V .1)

PAGE 78

68 TABLE V.1: Interview Participant Charact eristics Interview ID Gender Age BMI PHQ 9 Baseline (0 27) CIRS (0 56) 101 F 54 45 14 8 102 F 52 43 16 21 103 F 36 46 10 9 104 M 68 22.9 10 7 105 F 46 45.4 20 13 106 F 56 33.4 17 12 107 F 56 49.7 17 19 108 F 52 23.9 20 6 109 F 51 60 15 10 110 F 5 2 39 10 16 111 F 60 26.2 19 14 112 F 54 36.3 17 15 113 M 54 30 14 16 114 F 48 46 16 13 115 F 56 28.9 25 19 116 F 44 19.8 13 7 117 M 56 27.1 10 11 118 F 42 28.7 12 20 119 M 58 37.6 20 21 Means 52.4 36.3 15.5 13.5 The remainder of this chapter describes the major themes interpreted and analyzed from the interview data. The qualitative research questions and specific aims are weaved within the sections along with emer gent themes that help to describe the narrative offered by participants with d epressive symptoms. The analysis resulted in eight themes and 23 categories ( TABLE VI.2 ) describing the experiences with a new episode of depression and as a complex

PAGE 79

69 patient in primary care. The selected repo rted gender, age range, and CIRS range (i.e., Female, age range 30 35, CIRS range 10 15). Research Question 2 In order to assess lived experience with depression and concurrent illnesses, participants were asked a range of questions to eli cit their narrative. Questions were focused on how participants understand, experience, and negotiate depression, along with their perceptions of the relationship between their mental and physical symptoms and their health For the purposes of this study illness experience aided the development of this section of the question guide, as well as the analyses: Illness is the lived experience of monitoring bodily processes such as respiratory wheezes, abdominal cramps, stuffed sinuses, or painful joints. Illness involves the appraisal of those processes as expectable, serious, or requiring treatment. pathophysiological processes. And when we speak of illness, we must include the problems in daily living it creates (emphasis added). (p.4) ( Kleinman, 1988 ) Themes emerged from across the data providing insight on how depression is viewed, felt (symptoms), and if it has any impact on physical healt h and overall quality of life. Perceived Causes of Depression Given the characteristics of the study population, a priori themes were hypothesized when analyzing the perceived causes of depressive symptoms. For example, all 168 patients sampled for the original RCT had at least one other concom itant illness, therefore it was hypothesized that coexisting illnesses could be contributing to the onset of depression symptoms. Additionally, given over half of the 168 patients sampled for the initial RCT were considered living in poverty, financial and employment burdens were also hypothesized as contributing to depression. When asked about causes of their depressive symptoms, participants did not struggle

PAGE 80

70 for answers. Many offered precise origins for their depression. Below, the most common causes att ributed to depression by participants are reported. Physical health The burden of physical illnesses was a contributing source of depression reported by more than half of the participants interviewed. Eleven of the 19 participants attributed their feeli ngs of depression to the presence of physical comorbidities and these impressions did not vary by age or CIRS. Some of the participants mentioned not only were having multiple illnesses causing their depression the increasing severity of those illnesses c ausing feelings of distress and hopelessness. have arthritis, and its genetic and I think that you know, my condition has gotten worse. (F, 50 54, 20 24) Well, I've been diagnose d with Addison's Disease. Quite a bit in the last year or disability already, is very depressing. First of all I never thought I would be this sick especially this early in li fe, somewhat. And it is really debilitating. It makes specialists. And work here and just maintaining with medication. It is just not improving. And so having to deal with it eve ry day, it is overwhelming. I am thinking at that time, that's when the diagnoses started. I mean getting one thing told after another that I had. And so that was the beginning of the snowball effect sing. (F, 50 54, 15 19) Pain One of the more consistent comorbidities mentioned as a hindrance to participating and completing daily tasks was chronic pain. Chronic pain was also a significant contributor to the overall feelings about the ir current physical and emotional states Participants mentioned feelings of hopelessness, failure, frustration, as well as the social and physical limitations as a result of living in constant pain, and as a causal factor for depressive symptoms. Just thi s arthritis and this aching and not feeling good all the time, you can't help but be depressed at times. (F, 50 54, 15 19) eal sick. And I can't even hardly

PAGE 81

71 think the right thoughts and try to help them and stuff. It bothers me a lot. (F, 45 49, 10 14) But still again, those same things the aches and pains and all that stuff is of being depressed, but I know it is certainly a downer to my feeling good. (F, 55 59, 10 14) I think I was one of those people who was always waki ng up and feeling in pain and not happy with anything. Feeling hopeless. You know, I think that was me, (F, 40 45, 20 25) ing when you are in pain, you get depressed. (F, 50 55, 20 25) For one, failure. And you know now that we are talking about this, that's one of the times that the pain and all that kind of stuff started affecting my body. When I started going through some of the depression and stuff, and the loss and everything. It just felt like my whole body not just my mind and my mood and everything pain all through my body. (F, 50 55, 5 10) Two particip ants acknowledged a causal pathway between their pain leading to rest rictions in their daily lives, then to feelings of sadness, frustration and depression. Well, you know, I was depressed about my medical problems I was having and then I was depressed ab out that, and that's how it came up. Being in pain. It just puts your life on hold. And so, that's how it came up. It's frustrating. When you are in pain, you get depressed. Yeah. Well, when I'm in pain, of course I can't do certain things. Like my exercis e is walking. It prevents me from walking if I'm in a lot of pain. So I get depressed. So that's one of my main problems. And it makes me very depressed. If I can't get out(side). If I'm in pain and I can't do certain things, I get depressed. And anyone w ho is feeling a lot of pain, I mean they get depressed. It makes you not want to do anything. It just frustrates you. Do you understand what I'm saying? (F, 50 54, 20 24) anymo re. Just pretty doubtful (F, 45 49, 10 14) Having multiple illnesses While some participants mentioned specific illnesses contributing to their depression (e.g., pain), other patients explained their depression being the result of the frustrations, stres ses, and difficulties associated with suffering

PAGE 82

72 from mu ltiple chronic illnesses. The feelings mentioned by participants are supported by the literature looking at complex patients For example, many of the treatment interventions aimed at high utilizers o f health care resources did not address multiple conditions simultaneously and typically focused and ( Newcomer, et al., 2011 ) I have a lot of medical issues and I didn't think I was going to pull through a lot Mon I need it to stay alive. (F, 55 59, 15 19) I have Emphysema. And it has kind of worsened as of today's date. Besides I am now a diabetic and I go to the doctor today. I don't know i f they are going to start me on pills and I also have Glaucoma. So I got a few little stresses going on here, ok? (laughs) How I'm dealing with them, I'm just praying and just to be strong in (F, 60 64, 10 14) I had a total knee repl acement. Hip replacement. I had hernia surgery a month ago Well, you know, I also have hypertension. Uncontrollable hypertension. I had a stroke in my eye and you know, trying to keep my blood pressure under control. The list goes on. Plus I have arthrit just puts your life on hold. And so, that's how it (depression) came up. It's frustrating. When you are in pain, you get depressed. (F, 50 54, 20 24) One participant when referring to how illnesses impact her overall he alth stated am frustrations and stresses associated with being a complex patient. Other participants reported causes of depression focused on the hassles associated with coping with depression and concurrent illnesses on a daily basis. Kanner et al. (1980) defined these hassles as the: Irritating, frustrating, distressing demands that to some degree characterize everyday transactions with the environment. They include annoying p ractical problems such as losing things or traffic jams and fortuitous occurrences such as inclement weather, as well as arguments, disappointments, and financial and family concerns (emphasis added). (1980, p.3) It is these, sometimes ignored, events tha t when they accumulate, can have a significant

PAGE 83

73 impact on health outcomes important to understanding Also referred to as life stress ( Guthrie, 2014 ) this can both lead to depression and be an outcome of chronic illness, causing a vicious circle for complex patients. Financial troubles Eight of the 19 participants reported personal finances as a contributing to their depression. This was anticipated given the populati on studied is considered living in poverty and the stresses associated with finances can be felt daily and affect not only day to day to life, but also access to health care including physical and psychiatr ic care. being able to afford the rent but you still pay it. Ok? But you have to give up things. I have to choose my medications, you know? I can't keep the best medications but for a couple months a year, because of my deductible. And I have to settle for off br and drugs and when I take them I get worse through the year and then after 2 months I can straighten back up with the good medications. And then I have to settle back down because I can't afford it again. That's a stress. Going around looking for a place t o live, cause you can't afford where you live and the lists are like 2 years out every place you go, even for Senior Citizens. So you get stuck in a place you can't afford, but you have to deal with it (F, 55 60, 10 15) Cause they have things they want t o do and I'm not able to do it money wise. feel inadequate a lot. Just, I don't know. Why me, you know? Can't I have a little luck? A little stroke of luck? Can't I get on e thing cured? Can't I just get enough money, which is my money, social security, so I can just be self sufficient and not have to depend on others so much. (F, 50 54, 15 19) Five participants mentioned not just financial struggles causing depressive sympt oms, but the consequences of depression as it pertains to employment and normalcy. adapt to my new circu mstances. You know? (M, 55 60, 20 25) them looking at I really have the will to do it and I can pass all my tests, I just have this oxygen and it's in your face. And so that is a little bit of depression there. So it causes that no one is willing to give me a chance. And I guess that's it. (M, 65 69, 5 9)

PAGE 84

74 You feel better about yourself when you are making money and you can do what you want to do and stuff like that. And not feel l ike people are helping me maybe when they don't want to. I don't like when people are in my situation, but I have no choice right now. Or putting them to the point where they have to help me or be outdoors. They don't want that for me and neither do I. (F 50 54, 15 19) hour work week. which is 175.00 a month. Which is nothing. And it just makes me feel like I'm Well, I know I'm not able to take care of myself by myself. I'd like to be self sufficient. So that is not happening right now and I don't know when. And that is more depressing. (F, 50 55, 15 19) Just cause I couldn't do what I used to do before. And I do much money as I used to before. You know, I used to be pretty active, but now I'm not active anymore. (M, 55 59, 20 24) Family/interpersonal stressors Additional stressors associated with daily hassles were those at the interpersonal level such as family or intimate partner relationships. Two participants when asked about their perception of the causes of their depressive symptoms reported difficulties their social environments. thought he would be there little by little I was getting into depression about it because when the girls were (F 50 54, 5 9) The first time I was ever pregnant the first and only pregnancy I've ever had, I father of my child, during that week, did things above and beyond that stresse d me out no end. And because of the way the miscarriage happened, they came to the conclusion that it was my PTSD that was the stress and the things that he was doing purposefully that week. (F, 35 39, 5 9) A male participant attributed his feelings of dep ression to loneliness due to changes in family dynamics and a lack of communication with his children. You (M, 55 59, 20 24)

PAGE 85

75 Bereavement/personal loss Of all the perceived causes of depression for those participants interviewed, the most common ly mentioned cause was bereavement/complicated grieving with eight of th e 19 participants mentioning death of a loved one as a major contributing defined as the grief that manifests from a traumatic event in most cases death of a lo ved one, can potentially affect psychosocial, social, and physical well being and increase the risk of major depressive disorders as well as suicidality ( Latham & Prigerson, 2004 ) Thre e of the participants shared specifically about the loss of a parent, both parents, or a child as a major contributor to their depression. Some of the participants had been dealing with their bereavement for some time and did not mention receiving any coun seling or any other treatment to deal with their loss. I recently lost well, not recently but I lost my mother who I was very close the best of me. Cause I lost my mother have anybody (F, 50 54, 5 9) Because I been living like that for so long. That ain't nothing new to me. You know what I mean? When I was first truly depressed was after my father committed suicide and I was 19 years old. Ok? That's when I noticed it was like I didn't want to live any longer either. And it's been since then that you know, it I'm saying? (F, 40 44, 5 9) I first felt time parent to two small kids and my whole life turned around, so I was just stressed and lost and all that stuff just got me in a down place. (F, 55 59, 10 14) I lost a son and it started my depre ssion really bad. I couldn't cope with it. I didn't know how to cope with it. I had to get a lot of help to deal with it. It's been like I was hit by a train or something. It almost killed me like, I didn't know. I was like in shock. Like I didn't know what to do or how to deal with it (F, 45 49, 10 14)

PAGE 86

76 Perceived Physical Symptoms of Depress ion ( Kleinman, 1988, p. 10 ) Given the significance local meanings have on illnes s descriptions, it cannot be assumed that patients descriptions will match clinical descriptions and this disconnect can impact the clinical is rooted in the physiological, psyc ( Kleinman, 1988, p. 14 ) Consistent with when participants were asked about their perceived symptoms of depression, in themselves or others, described both emotional and physical respo nses, often concomitantly. Though the symptoms described by participants do not differ from how depression is clinically described including feelings of sadness or hopelessness, loss of pleasure participating or completing activities, reduced ability to c oncentrate, fatigue, changes in weight, and changes in mood, ( American Psychiatric Association, 2000 ) participants shared an understanding of how they classify their symptoms and an grasp of how thos e symptoms perpetuate their adverse health status. Nearly three quarters of the participants interviewed (14 of 19) focused on physical symptoms as indicators of depression with most among the males (three of the four) mentioning physical symptoms A patt ern also emerged by CIRS score with nine of the 10 participants with a CIRS score of 10 or higher reporting physical symptoms. Three distinct physical symptoms emerged from the interviews, though many had more than one type. Those symptoms are described below.

PAGE 87

77 Exhaustion/Fatigue. Four of the 14 participants who mentioned physical symptoms contributing to their depression reported exhaustion or fatigue that is persistent and hinders participating in daily activities. Being exhausted and not able to sleep up exhausted and not getting enough sleep. And waking up during the night. (F, 55 59, 10 14) been saying I was just tired. For some reas on, just tired all the time. You know. Not very happy with life. Just tired (F, 45 49, 10 14) One participant feelings. (F, 55 59, 15 19) Pain Another physical symptom of depression mentioned by participants was pain, not just as a cause of depression (discussed above) but also as a manifestation of being depressed. When comparing the causes and symptoms of depression, pain is a clear example of the bi directional relationship between physical and mental health, both cause and effect. Just the way I feel and stuff, it (depression) causes me the pain and stuff. Like in the mornings I wake up and sometimes I can't get out of bed. And my legs are just like paralyzed and stuff like that. I can't even move. So I just start and like it's coming on with pain. And try not to think of things that are going on and happening in life. times that the pain and all that kind of stuff started affecting my body. When I not just my mi so much pain throughout my body. (F, 55 59, 10 14) Perceived Emotional Symptoms of Depression Three distinct responses emerged from the interviews in explaining how participants perceive t he emotional symptoms of depression. Along while describing the various emotions connected with depression, participants also mentioned the social and physical consequ ences of the emotional symptoms. Though the emotional responses reported by participant s did not

PAGE 88

78 deviate from the symptoms already reported in the literature, as well as the focus for instruments aimed at diagnosing depression, the results are reported here to offer context specific to this patient population. Sadness. Five participants me ntioned how the symptom of sadness equated to social isolation, as well as to a lack of motivation to be active. you know, just wanting to be by yourself. (F, 40 44, 20 24) Not wanting to do anything. Not motivated by anything. You know. Just want to be bothered by nobody. I do want to do a (F, 40 44, 5 9) Mood swings. Insecurity. Some times negative thoughts. Not feeling your best. Just lot of things. Cause it affects everything. (F, 50 54, 15 1 9) The main thing is I will cry at the drop of a hat. Not wanting to get out of bed. Not wanting to talk to people. Not wanting to be involved with anybody. There are a lot of little things that go on. (F, 50 54, 5 9) Actually, I'm pretty depressed all the look forward to. (M, 50 54, 15 19) When you are depressed, you are more susceptible to colds and you just don't sure. Not wanting to get ou t of bed, you know? So depressed and sad and miserable. I would stay in bed for a whole day. I don't eat. It's not good for you. (F, 50 54, 5 9) Hopelessness participants also described their depression symptoms as feelings of desperation, hopelessness, and helplessness. These symptoms are not uncommon with depression and are including in many of the instruments to detect depression and other mental illnesses. Referring to both depression in general and how it is experienced personally, one female part icipant expressed how depression has many symptoms, including desperation, but also how those feelings manifest into social isolation, negativity, and physical pain.

PAGE 89

79 Well, what I understand it (depression) is feeling pretty low, that there is no help. Fee ling pretty desperate. Feeling tired all the time. Not wanting to be part of many people and not wanting to do too many functions and just not feeling good when you wake up ev ery day mopey. If they have a negative attitude where everything is always negative and going on, you know? I think I was one of t hose people who was from always waking up and feeling in pain and not happy with anything. Feeling hopeless. You know. I think that was me. You know. That's depression. (F, 45 49, 10 14) For some of the participants the feelings of hopelessness, and helpl essness stemmed from having to cope with their illnesses, and difficulty of coping or accepting their situations contributed to defeated feelings. So I might cry for a minute, you know? And then I have to pray. And have God just cover me so I can come bac k out of it, you know? Because I'm telling you. I do have times I sit down and there is nothing I can do. (F, 60 64, 10 14) I just take my medications. Do my blood tests and do what I'm supposed to do, you know? There is no cure for what I have. I mean th ere are things that know. Again, you know, I hate to keep saying Why me? But it is like just so much. is just a different form to me. It is just another added on disease, so to speak. (F, 50 54, 15 19) Living as a Complex Patient In addition to understanding how patients desc ribe symptoms as part of their lived experience with depression, questions were included to elucidate life experiences as a complex patient, particularly how having multiple chronic illnesses affects quality of life (QO L) (SA3 ). The purpose for better understanding how patients see the relationship between their i llnesses and their QO L are two fold: First, to recognize how having ill nesses in both domains affect the indicators associated with QOL (e.g., day to day life experiences). For example, if a patient is ill and does not want to socialize though healthy social interactions are associated with improved health outcomes; now there is a cycle contributing to adverse health Second, to better

PAGE 90

80 inform clinical interventions aimed at depression sy mptom improvement about the day to day consequences of being a complex patient and how opening dialogue concerning those consequences could lead to better outcomes. For example, if patients identify a lack of motivation to engage in social interactions, a nd given the influence socializing has on health outcomes ( Kawachi & Berkman, 2001 ; Thoits, 2011 ) both cl inicians and researchers working with complex patients could include aims at increasing social interactions and community engagement. For the purposes of this analysis, the themes around quality of life were derived from the literature that focused on func tioning, well eir day to day life experiences ( Mendlowicz & Stein, 2000 ) For complex patients, dealing with multiple health problems has s ocial, physical, and behavioral implications. Having multiple illnesses concurrently is associated with being more functionally impaired ( Vogeli, et al., 2007 ) and when certain chronic conditions are experienced with a mental health problem, there is a significant ly lower quality of life reported ( Mujica Mota, et al., 2015 ) When asked about their own assessment of their quality of life, participants interviewed shared a similar experience, expressing feelings of frustration, high levels of stress and anxiety, and feelings associated with depression. A common connection made by the particip ants when asked about the relationship between their mental and physical health was the limitations and hindrances caused by dealing with illnesses in both domains on daily activities. The responses around daily functioning were focused on physical or soci al functioning o r in some case a mix of both. Physical functioning Participants described an inability to participate in activities routines, and hygiene practice s In describing how physical functioning is affected by illnesses including depression shows the participants comprehension of the bi directional relationship

PAGE 91

81 between physical functioning and mental illness and how daily activities are important to the ir quality of life. One participant mentioned how they associate their physical activity/functioning with activities is a determinate of their self value. Of cour se it (physical health) affects your mental health, you know? When you go from being a very active, productive person in society to a nobody, who can barely take care of yourself. Of course that affects you mentally. What can you do without your hands? Not much. Before long I won't be able to cook. I won't even be able to wipe my own ass. What do you think it would be? I mean come on!...you can handle anything when you are able to function like a normal human on top of the world you know? those things again hell when I'd get upset I'd go for a run. If I was depressed I'd go grab my bike and go for a couple mile bike ride you know? So yeah. Getting rid of the pain would be an awesome thing. (F, 45 49, 10 14) The following participants shared how they view the bi directional relationship between physical and mental illness, for example no sense of enjoyment, loss of autonomy, and mostly how limiting their illnesses are and the frustrations they experience because of it. Well, it slows me down, that is a prime example of mental/emotional over till want to do things I want to do and need to do. I still try to do those things. Whether it is chores or taking walks to the park, walking the dog, really nice things to do, but yeah. It just slows me down. I get tired easier. ating a lot more because everything slows down. I get exhausted. (F, 35 39, 5 9) know, do certain things and it just frustrates me. It's like getting older. So Well, when I'm in pain, of course I can't do certain things. Like my exercise is walking. It prevents me from walking if I'm in a lot of pain. So I get depressed. So that's one of my main problems. If I can't get out, if I'm in pain and I can't do certain me to do certain things And anyone who is feeling a lot of pain, I mean they get depressed. It makes you not want to do anything. It just frustrates you. Do you understand what I'm saying? (F, 40 44, 20 24)

PAGE 92

82 Well, I used to be able to walk around the block before I got sick. I was walking around my block you know? Walking around and I walked around and I can't do that no w. And then now I have Osteoporosis and my back and my bones. My wrist would shatter really easy. (M, 55 59, 20 24) I mean it is just like a struggle to get out of bed. A struggle to do what you have ve. Try to be active with others and stuff like that. (F, 45 49, 10 14) Day to day activities When talking about having multiple conditions, t hree of the participants connected their illnesses with a diminished interest or incapacity to do day to day a ctivities in a more general sense. Though similar to some of the findings associated with physical functioning, these participants mentioned a diminished sense of motivation in a general sense and not necessarily specific activities. It (being depressed and physically ill) just makes me not want to do anything. (F, 45 49, 10 14) doing anything that can affect you physically because we are designed to move. (F, 50 54, 10 14) It (having multiple illnesses) bothers me a lot. I have a lot of pain through the helps me get stuff done. And I have diabetes so I have to kind of watch what I eat and stuff. It's not easy, you know? I don't know. I've had diabetes for over a year and then I take medicine for that, twice a day. (F, 45 49, 10 14) a better offer today than the sale yesterday. And now I wasn't able to do it because now I don't feel like it first of all. And I know if I get out there, I'd be having to deal with my leg. Dragging my leg all over town. I'm not going to do that. I don't d o very much. I mean I do the same thing. If I don't feel like doing it, I'm not going to do it. And it ends up being 90% of my daily living. It really does. Cause if I don't want to do it, I'm not going to do it. If I don't want to do it the next day, I'm still not going to do it. So it or three times a day sometimes. But like I comb my hair. I do change clothes. I do make my bed, but that's about it. (F, 40 44, 5 9) Soci al functioning Though participants tended to focus on how having multiple illness es impacted their physical health and diminished their quality of life, when asked about the

PAGE 93

83 social implications posed by depression (and in some cases in con junction with o ther illnesses), participants often mentioned feelings of social isolation and having little to no motivation to be around others. Eight of the 19 participants interviewed talked about how having multiple conditions along with depression can lead to soc ial isolation. When referring to the effects of having to take various medications to control her illnesses, one 50 54 year old female expressed how having to so patients talked about having to hide her depressive feelings and manifestations from friends and family. I mean, you know. There are times that you got friends that call and they want to do something and you just don't want to do it. I just kind of keep a smile on and don't say much of anything and just try to be cheerful you know? And when it's time to go into another room, I go into another room and cry and le t it out and makes me not really want to do much of anything. Other than just stay by myself, other than just be alone. (F, 45 49, 10 14) you have to put that happy face on and try to act normal and it's hard. It's hard 49, 10 14) One male participant described how his physical illness made him feel like a burden, therefore limiting his social time with family. Sometimes my brother might invite me to their house and I kind of get a little irritated because I'm on that oxygen and which I could go. But I don't have a ride. And I don't like bugging people to give me a ride. Give me a ride. Cause my oxygen, to carry it, it only holds so much oxygen, a couple hours or so. And then, you know, I start running out. Then I kind of panic because I'm running out of oxygen see. And I got to get back on. And stuff like that. So I get really irritated about that. (M, 65 69, 5 9) Other participants expressed various ways depression (and other illnesses) pose social challenges including how their mood impacts their interactions, staying in their homes from extensive periods of time, and just a constant feeling of not wanting to be in social situation s.

PAGE 94

84 stressed and depressed a lot, I won't go out the house. And sometimes it is hard to come out the bedroom. And that could last. That lasts for a little while, you know? Until I can break myself out of it. It doesn't last for like a week, you walk. Sometimes it gets that way, where I can't walk, but I don't know if that's from the diabetes or what? I' m fine today. (F, 60 64, 10 14) best food forward. I don't let the mood swings get me too much. Sometimes when I'm alone the mood swings come out. Or they come out at home aro und my dear ones. I don't let everybody see it, but I have mood swings and I got to get me some mental health. (F, 50 54, 15 19) of planning, By the time I get ready and get done with dialysis, most of the time I just have to come home. And lay down. Go to bed early. (12) Five of the participants felt that interactions with others were bothersome, both because they were made aware of their limitations because of their condit ions and seei ng people reminded them of that or because their current health state negatively impacted any desire to interact socially. mean I do. But knowing I am not abl e. And since I can't, I just shut in. You know was always by myself, so I'm used to being by myself. And that was what I would prefer to do. I mean I'd like to be able to get out, but knowing that I can't just not on that road. I'm just on the self want to be bothered with anybody, because you know, I can't do anything with them. And I'm not going to have people come and visit me all the time cause I don't want to get out. You know what I mean? (F, 50 54, 15 19) It is everyday life for me, you know? I mean I can tell cause I don't want to be bothered by nobody. I don't want to do anything. I don't want to talk to people. I don't want to see people. I don't want to do anything. (F, 40 44, 5 9) I'm just wanting to cave and not be around people. Or it would be fine with me if I didn't even have to associate with peo ple. (F, 55 59, 15 19) If my legs worked properly or if I didn't have a constant ache in my back. You know what I'm saying? It goes throughout my legs and it's been doing this for to

PAGE 95

85 take pain medication of course. And when you are on that in high doses, like I am, what if I'm out in the public and something might happen? Cause I know I'm not there. You see what I'm saying? (M, 55 59, 10 14) Well, I don't know what to say about that But I do notice when I am really stressed and depressed a lot, I won't go out the house. And sometimes it is hard to come out the bedroom. And that could last. That lasts for a little while, you know? Until I can break myself out of it. It doesn't last for like a week, you know? Cause I make myself get up. And it gets hard cause it is like I can't walk. Sometimes it gets that way, where I can't walk, but I don't know if that's from the diabetes or what? I'm fine today (F, 60 64, 10 14) Response to Resea rch Question 2 How do medically indigent depressed patients describe their lived experience with concurrent illnesses? When describing their lived experience with depression and often multiple coexisting illnesses, participants were able to articulate th eir experience with depression, giving insight and voice to their experience providing a useful narrative for clinicians to use for treatment and communication ( Barry, Stevenson, Britten, Barber, & Bradley, 2001 ) Participants described a variety of causes for their depression including t hose hypothesized including coexisting illnesses (living as a complex patient, further explored later in the chapter), and financial and employment burdens. Other causes emerged from interviews including factors like bereavement, and family/relationship s tress. When describing symptoms of depression, participants by and large reported those factors used as proxies in many of the instruments to measure depression including feelings of hopelessness, fatigue, and sadness. Though not all of these factors we re reported by a majority of the 19 participants interviewed, the disparities show a need for being more responsive to personal narratives in the clinical encounter aiding personalized treatment for depressive symptoms. responses around l ife as a complex patient presented a rich representation of how the day to day lives are impacted by their health. Their descriptions offer a n understanding of cyclical relationship between their physical and mental health, how it impacts th eir social and psychosocial health, and in turn adds to their depression symptoms. The

PAGE 96

86 participants mentioned a loss of self value with losing their capacity to participate in daily activities often viewed as important proxies for normalcy. Research Qu estion 3 This portion of the qualitative analysis aims to answer the third research question, how does a diagnosis of a new episode of depression change the management and prioritization of concurrent illness for complex patients? This set of questions l ooked at how patients manage and prioritize their illnesses. Themes emerged providing some detail to how illnesses are treated, prioritized, and the competing demands and barriers patients identify that hinder physical and depression symptom improvement. Illness Priority In analyzing interviews, a salient theme emerged related to how patients prioritize illness treatment after a new episode of depression Participants mentioned priorities for both themselves and their providers. Given that patients and providers were aware of the new diagnosis of depression prior to the first visit while participating in the RCT, it was hypothesized that depression would become a treatment priority, and though patients were interviewed after the completion of the RCT remain a priority even after the completion o f the RCT. However, w hen asked about which illnesses the patients and providers prioritized, the majority of the participants interviewed were more likely to report that their physical ailments took precedent over their mental health. Patient treatment priority When discussing their own illness priorities, participants mainly focused on physical ailments. One female participant was explicit about not wanting her depression symptoms to be a priority for he Two participants mentioned that they prioritized their physical illnesses mainly because they have learned to live with their mental health conditions.

PAGE 97

87 No I was just thinking. I'd have to s ay my weight. Blood work is always good. I'm healthy on the inside. So it is my weight. It hurts my joints and impedes on my life more than anything. My mental health I make the choice to be happy and I make the choice to look for the positive and the go od. So I'm aware and in control of my mental health, is the best I can do. But the PTSD I don't even notice it sometimes. I'm sure other people notice it first. But yeah, I think it's just the fat to be honest (F, 35 39, 5 9) use of my hands. What can you do without your hands? Not much. Because like I said, my depression I've had so long that it only becomes a problem on occasion. (F, 50 54, 5 9) Some participants were adamant about their physical health taking priority when it came to treatment. Two participants in particular viewed the relationship between their physical an d mental health as hierarchical. I think physical. I would treat my physical before I treated my mental Cause I more, when I get in those depressed states; I think if I would go and work out, it would help a lot. So that's why I say my physical probably would overrule my mental. (F, 50 54, 15 19) I think probably physical health. That's the main thing. I been going through a oint. So I think health were improved, then I would feel better on a daily basis and do a lot of different things that sometimes I'm not doing because of the health issues. (F 55 59, 10 14) One participant mentioned specific physical illnesses as treatment priorities. t (F, 55 59, 15 19) Provider treatment priority participants also described which illness(es) their providers prioritized. The most common response given by patients was that mental health was not a priority. participants did An obese participant mentioned that her provider was aware of her depressive symptoms but still focused on trying to control my weight because the weight is w Two

PAGE 98

88 participants described there being no discussion around mental health because of there being other health problems to manage. Pretty much when I call to make an appointment, we pretty much talk about the t? They thought I might have a h ernia! Well, right now we haven 't even discussed it ( depression ). We haven't discussed it (depression) We've been focused on so many other things. (F, 50 54, 15 19) ssues of course, but we ain't really worried about those. It is just getting through this pain on a constant basis, all day, every day. (F, 40 44, 5 9) Another participant seemed open to changing her priorities and discuss her mental health but did not wan t to trouble her provider so focused Yeah, well, she (primary care provider) basically prioritizes whatever is happening with my health at the time. You know? And she sends me for tests and all of tha lot. Because she has so many patients that she needs to deal with as well. So I don't want to take up a lot of her time, even though she would be willing to sit there with me. Bu t I just don't want to take up a lot of her time. You know? Just sitting back and discussing everything that is going on with me, I just go in and I want her to take care of my main issues that are going on with me, like pain and all that. (F, 55 59, 15 19 ) A male participant mentioned that his provider main concern was preventing a heart attack and was focused on managing his energy exertion in the workplace. Not really any o octor, he doesn't want me to work. Cause he says h e used to tell me he was worried about me having a too hard. (M, 55 59, 20 24) Depression Treatment Preferences A particular salient finding from the qualitative interviews was that more than half of the participants interviewed did not want to be treated for depression in a primary care setting Twelve of the 19 participants interviewed shared their preferences ranging from seeing a therapist/counselor, to social support

PAGE 99

89 Spiritual/ prayer Three of the 12 participants shared how spirituality, specifically prayer, was a n important modality for self treatment for their depression. I just don't want this to defeat me. You know? And I'm sure that my faith helps me to get up i n the morning, you know? I don't want to give up regardless of my this diabetic thing. I just You know? And it helps with my stresses and my circumstances and my depression. You know? I don't take medication for it. I just have to pray. Pray the depression. It helps a lot with the stresses. If you can get your mind off of yourself. I learnt that since I been on that study. I learned how to try to just go about it another way. Read to enlighten myself. But I'd say I'm doing much better in my d epression and stresses than I live. Ok? But I have to keep positive that sooner or later something is going to happen. Well, I guess Jesus. Because I found the faith you kno w? I've got something else to release my thoughts to. And it is helping. My faith is helping me. (F, 60 64, 10 14) I start praying and I pray every day. I don't get to church every week like I should because of the oxygen and stuff. But I pray every day. And my problems and whatever, I leave it in the hands of God. Let him take care of it. And so far, he's been taking care of me. (M, 65 69, 5 9) Family support Three participants mentioned support from a family member (one mentioned it in conjunction with prayer) as a way to deal with his depression symptoms. Well, being a Jehovah's Witness, of course, I pray about it. And I have many that's what happens when you are depressed. You got to have some way to vent. And I have people that I can talk with. And then I have my oldest son, you know? My relationship with him is where I feel comfortable in talking with him. (F, 50 54, 10 14) I have friends and family that help me out. (F, 5 0 54, 15 19) Well, actually, I did have a lot of depression, but now I'm a lot better. I learned to just like walk and pretty much and talk with family and that makes me feel better. It was bad. But then I had to snap out of it though. From feeling sorry f or myself. And make myself a better person. Getting out of the house. I go walking to the park with my son. Just having a good day. (F, 40 44, 20 24)

PAGE 100

90 Professional treatment (not primary care provider) Three of the twelve participants mentioned wanting to get professional treatment for their depression in the form of counseling or therapy. Grief Counseling. It did. Yep. It helped. It helped. It helped a whole lot. It was just that every year when that month comes around it seems like I go --try to go into it, depression, again. I do counseling with my Pastor too. (F, 55 59, 15 19) Oh, therapy. I don't believe in prescription drugs. (F, 35 39, 5 9) I get counseling and stuff. And just talking to people and opening up instead of closing down and stuff. Just trying to talk to somebody about it. And it helps (F, 45 49, 10 14) Opposition to medications Two participants expressed a strong opposition to medications for depression, both had first hand experience with anti depressants but either had their provider take them off the medication or expressed concerns about their current medication regime. I wish I didn't have to take all these pills. And like before. I was taking the wrong kind of medications for my depr ession. I would sit there and count my pills I don't know how many times. Cause I don't know, that particular pill seemed to be affecting my way of thinking. Because it was interfering with concentration. (F, 55 59, 15 19) I just had him take me of my med ication because I'm better without the medication. (F, 50 54, 5 9) These findings suggest that some participants experiencing a new episode of depression do not want to be treated in primary care settings for their depression, which can be problematic fo r interventions in these settings Primary care/medication Not all participants interviewed were opposed to their primary care provider providing depression treatment or to taking m edications f or their symptoms. When asked about her treatment preferenc e, one female participant I would love for him (primary care physician) to treat it. That would not be a problem. medical doctor would not be responsible for mental health care. I don't think you separate them. I'v e often thought that it is ridiculous that you have to see a therapist aside from your medical doctor. I think medical doctors in

PAGE 101

91 general should be trained in the psychology part of their patient's lives. I just don't see how you can separate them. You rea lly can't separate any of it. (F, 35 39, 5 9) Two participants reported either being open to medication, or were currently taking medication for the depression symptoms. I figured I'm depressed and know I know what is going on. We can do medications. We c an try to get this under hand and maybe everything, you know, will look a lot brighter. You know, everything won't be so gloomy. (F, 45 49, 10 14) Well, the drugs don't hurt. They help a little. But I try to keep my mind, these days, where I just don't eve n think about the depression thing. You know, like not think. (F, 55 59, 15 19) Competing Dema nds/ Barriers to Treatment and Improving Symptoms Competing demands is defined as the factors that pull patients and providers in many directions and include but are not limited to, personal demands from family and friends, discrepancies in medical information, financial burdens, conflicting agendas of care, concurrent illne sses, employment demands, transportation, and time constraints ( Jaen, et al., 1994 ; Klinkman, 1997 ; Nutting, et al., 2000 ) Though research has looked at these demands and how they impact the clinical encounter, most research has focused on the demands in the clinician domain ( Klinkman, 1997 ; Nutting, et al., 2000 ; Rost, et al., 2000 ; Stange, Fedirko, Zyzanski, & Jaen, 1994 ; Williams, 1998 ) For the purposes of this study, participants were asked about if and what factors impeded them receiving care, openi ng dialogue about, and starting and maintaining a treatment regimen for their depression symptoms. It was hypothesized that because the participants interviewed were dealing with concurrent illnesses, that competing illness would be a significant factor c ompeting with depression care. Three main demands emerged from the int erviews and are described below.

PAGE 102

92 Concurrent illnesses As predicted, concurrent illnesses was a common competing demand described by the participants interviewed. Seven of the 19 part icipants reported trying to manage their various illnesses, other than their depression, as their primary priority. earlier in the interview), but we ain't really worried about those. It is just getting through this pain on a constant basis, all day, every day. (F, 40 44, 5 9) (F, 55 59, 15 19) One participant her depression symptoms and causing other health problems exacerbating the demand put on by the weight. What gets in the way? Being overweight. That stops a lot of things. You know, I t on arthritis in my knees, fluid on the knees. I'm just falling apart over here. (F, 55 59, 15 19) When asked about her competing demands, one participant answered by listing her illnesses as her answer. Degenerative joint disease, asthma, vertigo. What else? Sciatic, weakness in the knees, carpal tunnel, rotator cuff injury, clavicle injury, arthritis throughout the out, but I don't know. (F, 50 54, 15 19) Family/social relationships Seven of the 19 participants interviewed mentioned their family or other social relationships as taking priority over seeking treatment not only for their depressive symptoms but also for their overall health care (ofte n along with other d emands). One participant described that her overall health was not prioritized stating, Another shared an experience that occurred the same day she was i nterviewed where other commitments tend ed to take precedence over seeking care, in this case for her depression.

PAGE 103

93 I was going to call today (mental health counseling), but I didn't get a chance. My daughter came and I ran around with her and my granddaughte health. (F, 50 54, 15 19) Along with family commitments, two participants also shared additional competing demands impeding their own care. One participants not her relationship with him, but also grief, and her own competing illnesses as not only competing with seeking care for her depression, but also the reasons to not accept her depression symptoms. Well, my p arents passed away. My son being sick. He has Spina bifi da the closure of the spine and that was bothering me too. And I'm epileptic so that bothered me a lot, you know? Now I have to deal with it. So it's not going away. I been epil eptic since I was 13 years old. Just you know, try to deal with life, you know? Cause I figure like my son, you know, he doesn't need to see me being depressed and being sad and all that. He needs his mom. He needs me to be there for him. Being depressed i and he gets sad because my son tells me, I wish I could walk like my friends. And it makes me sad. But then you know, I have to think, my son can do s ome things in his wheelchair. Just you know, try to deal with life, you know? (F, 40 44, 20 24) Similarly, another participant mentioned the health of a family member as a priority, but concurrently she is dealing with competing illnesses, financial probl ems, and desire for independence. She (mother) has early stages of Alzheimer's. So I have to make sure she gets to her doctor's appointments and get what she needs as far as hygiene and all that kind of stuff. So, I think between that and doctor's appoint ments and stuff, I keep I need to have my other knee done. I had one knee replaced already. They are d kind of hard to put into words. I feel inadequate a lot. Just, I don't know. Why me, you know? Can't I have a little luck? A little stroke of luck? Can't I get one thing cured? Can't I just get enough money, which is my money, social security, so I

PAGE 104

94 can just be self sufficient and not have to depend on others so much. (F, 50 54, 15 19) Finances Another type of competing demand mentioned by the participants was concerns a bout finances including costs of care (e.g., medications, and insurance), and daily living, in some cases not only hindering seeking depression seeking behavior but also adding to the feelings of stress and hopelessness. Well, first of all cause I ain't go t no money to plan with. So basically, it is not just (F, 55 59, 15 19) It is money that comes up. For me it's money. Because for me to treat all 3, the nce it because the money is just not there. And you have to work for it and you have to take the time away from what is really important and put it into your job. I'd have to say money and time. (F, 35 39, 5 9) When asked about the demands affecting their health outcomes, two participants talked about how the cost of health care and insurance along with other environmental stresses make it difficult to focus on treating their depression symptoms. I have to choose my medications, you know? I can't keep the best medications but for a couple months a year, because of my deductible. And I have to settle for off brand drugs and when I take them I get worse through the year and then after 2 months I can straighten back up with the good medications. And then I ha ve to settle back down because I can't afford it again. That's a stress. Going around looking for a place to live, cause you can't afford where you live and the lists are like 2 years out every place you go, even for Senior Citizens. So you get stuck in a place you can't afford, but you have to deal with it. Cause I take so much they put me on more medication with my Glaucoma and stuff. (F, 60 64, 10 14) One participant priorit ized finding ways to pay for her medication for her pain, though she was suffering from multiple illnesses including depression. get my deductible paid down. When my deductible is paid I can get a 3 month packet of that for like 90.00 ok?

PAGE 105

95 Other than that there, I can't afford it. (F, 60 64, 10 14) Response to Research Question 3 How does a diagnosis of a new episode of depression change the management and prioritization of concurrent illnesses for complex patients? align with the hypothesis originally purposed stating that patients dealing with a new episode of depression would want to make treating those symptoms a priority. Both patients and providers continued to focus on physical ailments and controlling existing disease. When discussing their treatment for depression symptoms, partici pants offered several modalities preferences including medication, prayer, social support, and counseling/therapy. One emergent theme from the treatment preferences was participants did not want to initiate or continue a prescribed medication regimen for t heir symptoms either because the patient was already on enough medications for other illnesses, or because they did not believe it offered any relief. Responses to which competing demands impacted depression treatment seeking behavior focused on three typ es, competing illnesses, social relationships, and financial burdens, some a mix of different competing demands occurring simultaneously. The code counts for each category are reported in Table V.2.

PAGE 106

96 TABLE V .2 : Total Counts of the Codes Identifie d Totals: T otal interviews (number of patients ) 1. Patient Demographics (code) 27 19 2. Patient Overall Health 10 5 3. FAM Clinical Encounter 3. Communication/treatment preferences 13 5 3. Discussion around mental health 35 13 3. Initi ating Priorities 2 1 3. Medications 7 4 3. Negative Response to RCT 11 5 3. Positive Response to RCT 25 12 3. Provider Inquiry (Clinical) 6 4 3. Provider Inquiry (non clinical) 0 0 3. RCT effects 2 1 3. Time with provider 8 5 4. FAM MH >PH Relationship 4. Daily Activities 23 9 4. Medications/Side effects/combination 4 3 4. Physical Health Hindrance 10 6 4. Physical consequences 30 10 4. Social consequences 22 10 5. FAM Physical Illness > MH 5. Alternating Moo d 7 3 5. Daily Activities 25 10 5. Disability 2 1 5. PH causing depression 9 5

PAGE 107

97 Totals: T otal interviews (number of patients ) 5. Physical Consequences 22 9 5. Social consequences 17 9 6. FAM Illness description 6. Behavior Change 9 2 6. Causes/Symptoms 91 17 6. Co occurring illness 65 16 6. Complex patient descript 20 7 6. Depression 130 17 6. Depression diagnosis 20 7 6. Hospitalizations 1 1 6. Illness consequences 15 4 6. Medications 16 9 6. More Import ant MH/PH 26 9 6. Pain 40 11 6. Physical causes of depression 2 1 6. Severity 4 4 6. Social Causes of depression 17 6 6. Stress (other descriptions of depression) 8 5 6. Treatment preference 27 11 6. Worry 7 3 7. FAM Barriers 7 Communication/Knowledge 4 1 7. Depression 5 4 7. Financial 16 6 7. Improving Health 8 4 7. Lack of support 5 3

PAGE 108

98 Totals: T otal interviews (number of patients ) 7. Locus of Control 1 1 7. Medications/side effects 11 3 7. Physical Hea lth 7 6 7. Time 5 2 7. Transportation Issues 3 1 8. FAM Competing Demands 1 1 8. Illness importance (patient perspective) 23 10 8. Illness importance (provider perspective) 2 1 8. Patient Level 55 16 8. Provider Level 2 2 8. Treatme nt 6 2 9. FAM Treatment 9. Behavior Change 1 1 9. Mental health 44 13 9. Physical health 24 10 9. Spiritual/Other 6 4

PAGE 109

99 CHAPTER VI DISCUSSION Depression is a difficult disease to diagnose, treat and often left out of clinical dialogue between patient and provider ( Weihs & Wert, 2011 ) ; and it is responsible for signific ant functional impairment and disease burden globally (CDC National Center for Health Statistics 2013), high utilization of health resources (Hasselman, 2013; Guthrie, 2014), and overall medical costs ( Katon, 2011 ) Because of deleterious effects caused by depression, better interventions are necessary to improve depression diagnosis and treatment, includ ing those that address its co occurrence with physical health. Increasingly, it is recognized that depression symptoms occurring concurrently with chronic illness conditions can increase the severity of medical illnesses (National Institute for Health and Clinical Excellence ( National Institute for Health and Clinical Excellence (NIHCE) ) 2009), as well as contributing to more somatic complaints during primary care visits ( Thielke, Vannoy, & Unutzer, 2007 ) Additionally, given the ubiquity of antidepressant p rescriptions, specifically SSRIs, interventions aimed at developing treatment strategies either in conjunction with, or as an alternative to, prescribed medications are imperative given that the presence of physical health disorders can lead to complicatio ns with interactions with other medications, and the potential of antidepressants exacerbating chronic illnesses (National Institute for Health and Clinical Excellence (NIHCE), 2009). This phenomenon polypharmacy reaction from health care to medicalize Shakib, & Winefield, 2012). When setting out to examine the potential effect of coexisting illness severity on depression symptom improvement, I hypothesized patients with less coexisting illness severity (CIRS score)

PAGE 110

100 would see greater depression symptom improvement compared to those patients with high illness severity. I also sought to illuminate the bi directional relation ship between depression and chronic illness particularly the relationship between depression and the management and prioritization of coexisting illnesses (e.g., competing demands), particularly because the research regarding competing demands and depressi on is mixed ( Ani, et al., 2009 ; Klinkman, 1997 ; Nutting, e t al., 2000 ; Rost, et al., 2000 ; Vyas & Sambamoorthi, 2011 ) The goal of the qualitative phase of this dissertation research was to add to the understa nding of the lived experience of patients with depression, including how depression and other illnesses impact their daily living, and their treatment priorities, addressing an important gap in the literature about depression in complex patients (Stanners, et al., 2012). The hope is that these findings can inform both the assessment and treatment of depression in the context of co morbid chronic conditions and complexity of issues that emerge in the lived experience of medically indigent complex patients. This dissertation study sought to help fill additional gaps mentioned in the literature concerning depression among c omplex patients. Change in Depression Over Time in Complex Patients Though the original RCT was not designed to look at illness severity as a potential predictor for depression change over time, the patients approached for the study were not excluded because of their current or past physical health conditions (only if they were currently on a depression medication regimen) affording the op portunity to look at the medically ill and their response to a depression intervention, historically a population not always included in depression interventions (Koike, Unutzer, & Wells, 2002). Though most of the research on multimorbidity and depression does mention higher prevalence of depression among those with chronic conditions, (Bair, Robinson, Katon, & Kroenke, 2003; Benton, Staab, & Evans, 2007; Clarke, 2009; Koike, et al., 2002), for this study those patients with moderate to high CIRS had highe r PHQ 9 scores at each time point suggesting a relationship between physical illness

PAGE 111

101 severity and depression severity. This also potentially supports the claims of the bi directional ity of depression and physical illness (Evans & Charney, 2003; Evans, et al., 2005). Though I hypothesized that complex patients with high levels of chronic illness severity would experience significantly less depression symptom improvement over time and would be less responsive to the RCT than those with lower illness severit y, there was not enough observed variation in the scores over time to test this hypothesis adequately (Raudenbush & Bryk, 2002). The HLM analyses of the relationship between illness severity and change in depression over time did not vary significantly ac ross the patients in the study. Though the findings do not refute the relationship between illness severity (rather than simple counts) and depression symptom improvement it cannot be inferred that illness severity is a causal mechanism hindering depress ion improvement (Benton, et al., 2007). What did predict depression change over time was baseline PHQ 9 score; though intuitively this would be due to patients experiencing more severe depression at baseline have a larger range of potential improvement. These findings suggest an important need for future research to test for potential factors contributing to depression symptom improv ement in primary care settings. Key Findings from the Qualitative Analyses In contrast to quantitative findings, the qualita tive portion of this work offered a rich description of how patients with multiple chronic conditions describe and live with depression. Patients openly described how depression impacts their quality of life, their coexisting illnesses and current health status, descriptions of what they consider to be the symptoms and causes of depression, and how they prioritize their treatments and the competing demands that hinder receiving or maintaining treatment fo r their depressive symptoms Patient Description of Depression One of the factors contributing to the difficulties behind diagnosing, discussing, and treating depression for patients with multiple chronic illnesses is the myriad causes for depression

PAGE 112

102 as well as how depression is described (symptoms). Unlik e most chronic illnesses, the causes of depression are many and experienced differently depending on the individual, as are the symptoms, which do not always sync with clinical descriptions of depression. As expected, the patients interviewed described th eir physical health as one cause of their depression including their overall physical health (e.g., having to live with multiple illnesses), and living with chronic pain. Other causes mentioned by the patients included financial problems because of unempl oyment, dependency on financial assistance programs, and health care costs (especially the costs associated with having multiple illnesses); and dealing with loss including the death of their children, parents, and close friends and the manifestations of t he bereavement including their own mortality, compounded by their poor health. The causes of depression reported by patients could help explain part of the difficulty treating depression in primary care some of which have not been reported before in the literature For example, causes like physical health including diseases with no remission, financial difficulties, unemployment, and bereavement are not necessarily ideal for conventional pharmaceutical treatments, necessitating different approaches depen In contrast to patients identifying many different causes of their depression, self reported descriptions of symptoms were actually quite consistent. Most of the symptoms centered on emotional responses to trauma resembling the clinical symptoms for de pression (e.g., hopelessness, sadness), along with physical symptoms of fatigue, and physical pain. Since depression is measured using validated, survey type instruments (e.g., PHQ 9, Hamilton Depression Scale), patients are not typically afforded the opp ortunity to share any additional symptoms outside of the clinical symptoms such as chronic pain, problems with diet and digestion, and defeated feelings around having chronic illnesses.

PAGE 113

103 Perceived Relationship between Depression and Chronic Illness One of t he aims of the qualitative interviews was to describe how patients view the relationship between depression and chronic illness, an unexplored aspect of the patient narrative in the existing literature (Bayliss, 2012). When asked about the causes of depre ssion, patients described their overall health status as well as specific illnesses as substantial contributors, including feelings of distress particularly around diagnosis, feelings of hopelessness particularly around prognosis, high levels of chronic st ress when dealing with treatment regimens and the various appointments with healthcare providers and healthcare costs common for complex patients. For example, some chronic conditions can be felt as a burden with no hope for remission, and though patients are diligent about appointments and treatment plans, general The relationship between depression and chronic illness was also noted by patients while dis cussing their symptoms of depression. One of the most substantial ways patients experienced the bi directional relationship between their depression and illness is the effects of a symptom like sadness has on their social functioning. Social functioning and engagement are crucial for both the prevention and management of depression symptoms and chronic illness. For example, literature around social functioning shows the importance of social relationships and interactions as crucial to improving depressiv e symptoms (Cruwys, et al., 2014; Gleibs, et al., 2011), and conversely the relationship between social isolation (i.e., loneliness) and depression (Cacioppo, Hawkley, & Thisted, 2010; Trivedi, Morris, Pan, Grannemann, & John Rush, 2005). Patients shared their desires to be alone as not to be a burden on loved ones, socializing being reserved for healthcare visits, not wanting to be around others due to their chronic mood swings. Patients also experienced the bi directional relationship in how their chr onic illnesses (e.g., pain, physical complications from diabetes) physical functioning that included daily tasks, exercise and activity, and household chores with those limitations causing frustration, and depression. These feelings are further exacerba ted by the loss of autonomy caused by having

PAGE 114

104 physical illnesses, forcing dependence on others, and inability to work for wages, all adding to the a diminish ed quality of life and self value. Living with multiple illnesses impacts the daily functions for patients associated with normalcy, leading to a low level of interest in tasks, or sometimes a fear felt by the risks associated with daily functions that cou ld potentially compound their chronic illnesses ( Fortin, Dubois, Hudon, Soubhi, & Almirall, 2007 ; Fortin, et al. 2004 ) Illness Priority and Competing Demands An important theme from the interviews concerned how patients, given their illness complexity, prioritize their illnesses in terms of clinical importance (agenda setting) and treatment. Though I hypothes ized patients would want to prioritize t heir new episode of depression, the patients interviewed for this study put importance on their physical ailments. This could be directed by the expectation of treating physical conditions only in primary care, and m ental health detection and treatment being less emphasized ( Kravitz & Ford, 2008 ; Nutting, et al., 2002 ; Nutting, et al., 2000 ) as well as patients not reporting depression symptoms to their primary care provider ( Bel l, et al., 2011 ) Problems with weight, chronic pain, hypertension, and diabetes complications were among those illnesses mentioned as priorities for treatment in spite of being aware of depression. When it comes to treatment priorities, patients overw helming preferred focusing on their chronic conditions explaining that having more control over their chronic illnesses would improve their mental health by increasing their autonomy, activity levels, social interactions, having to live with less pain, and having less worry about illness prognosis. An important aspect of this research was to elucidate the role of competing demands for patients who are seeking or receiving care for their depression. The three main themes identified concerning competing de mands for the patients interviewed were the treatment and control of other competing illnesses, family responsibilities including care of sick individuals and parental responsibilities, and economic demands including having to prioritize medications becaus e of costs, feeling the costs of depression care being too much to take on, and being on a fixed

PAGE 115

105 income. There is some evidence that because complex patients have several conditions needing management simultaneously, those illnesses compete with one anothe r and affect the quality of care ( Ani, et al., 2009 ; Jaen, et al., 1994 ; Stange, et al., 1994 ; Vogeli, et al., 2007 ; Vyas & Sambamoorthi, 2011 ) including mental health care ( Klinkman, 1997 ; Rost, et al., 2000 ) Three studies since 2005 have concluded that depression care in primary care settings is not affected by the competing demands due to multiple chronic conditions ( Ani, et al., 2009 ; Harman, Edlund, Fortney, & Kallas, 2005 ; Vyas & Sambamoorthi, 2011 ) Though the quantitative analyses support this assertion, the findings from the interviews show a strong connection between patients coexisting illnesses and either seeking tre atment for, or managing their depression. Patients mentioned trying to live with chronic pain, problems with weight, along with the burdens felt from being a complex patient. One methodological reason for the conflicting findings could be that this disse rtation used semi structured interviews to gauge the influence of competing demands on depression as opposed to a more rigid survey approach without much latitude for context found in the studies mentioned earlier. Additionally, the interviews offer some challenges to the potential assumptions of depression interventions in primary care settings that may see depression as a separate condition interviewed fo r this study see their depression as the result of many of the difficulties associated with illness complexity. If this is true, then interventions aimed at improving symptoms of depression should focus on treatment adherence and management of chronic ill ness and dialogue around the social implications of being a complex patient, not necessarily recommending therapy or medication for depression specifically. By controlling illnesses enough to increase social activity and patient autonomy, decreasing pain and worry about prognosis, as well as partnerships with professionals such as social workers could have a potential impact on symptoms of depression for primary care patients.

PAGE 116

1 06 Assessing Methodology The strength of this dissertation was the use of mixed m ethodology, particularly given that the quantitative portion of this research did not yield significant findings. Interviews provided a compelling forum for patients to offer a narrative of their lived experience with depression. The prevalence of depress ion in primary care and the difficulties with adequately treating depression suggest the need for a broader, more contextual approach than the limited understanding and objectivity of quantitative research alone. Understanding how depression manifests dif ferently in populations, how it is experienced and described, and how it impacts the everyday lives of individuals necessitates additional explanation, and combination of approaches to increase the depth of understanding of a very complex health condition (Wisdom, Cavaleri, Onwuegbuzie, & Green, 2012). Interestingly, though the HLM showed no relationship between chronic illness severity and change in depression, patients described a strong relationship between their illnesses and their symptoms of depre ssion, including that their illnesses hindered their seeking mental health treatment. The addition of semi structured interviews offered important perspectives and a rich context of the lived experience of depression and compelling discrepancies between th e quantitative and qualitative results that also raised important, additiona l questions for future study. Conclusion Providing quality depression treatment in primary care settings for complex patients is challenging and though the severity of chronic i llnesses competing with depression could impede depression symptom improvement. Though this study was unable to identify a relationship between chronic illness severity and depression symptom improvement, additional efforts are needed to improve the under standing of chronic illness burden on depression outcomes. Given this, interventions aimed at improving mental health care in these settings is imperative and must

PAGE 117

107 include support and training for primary care providers to deal with any ambivalence about discussing and treating mental health illnesses. Additionally, we lack studies of potential confounders of depression symptom improvement; qualitative findings suggest that factors such as health status, economic, socio cultural and environmental influenc es could also impact depression but they are not usually included in our measures. Another important aim for this dissertation was to improve understanding of how patients, in their own words, experience, describe and impact their daily living, and how patients prioritize their treatment. Findings suggest that, although competing illnesses did not impact depression change in quantitative models, patients reported dealing with a multitude of problems that complicate and sometimes adversely contribute to t heir health. Though the importance of effective communication between primary care providers and patients has shown to be crucial for patients to have an accurate understanding about depression as well as treatment expectations ( Malpass, et al., 2009 ; Stanners, Barton, Shakib, & Winefield, 2014 ) I propose that there is information beneficial for providers f rom patients sharing their narratives. The interviews offered patients the chance to share their narrative to better understand their lived experience with depression and coexisting illness. This affords the opportunity to record the descriptions correct ly and understand the key antecedents and consequences of the course of illness ( Kleinman, 1988 ) A pati competing demands influencing decision making, and illness behavior, and knowing this can only cli from being encouraged to share biographical information. during the clinical encounter has been studied previously (Barry, Stevenson, Britten, Barber, & Bradley, 2001; Coventry, Dickens, & Todd, 2014), further research is needed to better measure the quality of dialogue during the clinical encounter, testing patient satis faction with visits where

PAGE 118

108 their narrative is shared, and the potential impact sharing their narrative with providers has on their depressive symptoms and other chronic illnesses. In addition, although this study did not show a statistical relationship betw een chronic illness severity and change in depression, qualitative findings did suggest that illness severity has some in fluence on depression symptoms which should be further tested using methods that measure illness severity differently tha n the CIRS, in cluding a self reported illness severity instrument. Future research should use different methods to measure coexisting illness including self reported measurements as opposed to medical chart reviews used for this study. For example, this dissertation study would have potentially benefitted from using self reported illness severity ratings which could give a more current and accurate understanding of just how a patient feels and measures their illnesses compared to a medical chart, which though is more thorough, it still relies on socially constructed definitions of illness. Also, findings from this research revealed a need for additional research on the impact of the differences in how patients with depression describe the causes of their symptoms. The interview data revealed a variety of causes of depressive symptoms and it could be that how patients view these causes impacts their treatment priority and response. For example, patients who see their depression as almost normal reaction to having mult iple chronic illnesses are prioritizing, and could benefit from, treatment on better controlling and improving their chronic conditions than to their depression symptoms, including but not limited to, anti depressants. What this suggests is a possible ove r medicalizing of the emotional responses to illness complexity and life struggles, especially among financially and medically indigent populations. Limitations Several limitations to this study should be recognized. First, the quantitative analyses were done using secondary data from a nearly completed RCT, the sample size for the project was small and only included patients using Denver Health and Hospitals for their health care,

PAGE 119

109 therefore compromising generalizability, not all of the 168 patients c ompleted depression scores for all four time points. Also, no data was collected concerning the patients approached to participate in the original RCT. Additional limitations for both the quantitative and qualitativ e analyses are described below. Limitat ions of the Quantitative Analyses There were several limitations to the quantitative phase that should be noted. First, there was a lack of statistical power due to small numbers of patients in the original RCT. Patients included in this study were Engli sh speakers and had to have access to a telephone and a physical address. This study only examined linear changes in depression over a relatively short time period, there is a relatively small sample size, and depression was measured using self report. A dditionally, CIRS was only measured once at baseline, which did not afford the opportunity to view trajectories of illness severity at various time points. The depression measures were self reported and therefore subject to recall bias. The reliability of the medical record abstractions may have also limited the findings, though steps were taken to ensure fidelity of the CIRS scoring. Also, the analyses did not control for demographic and social characteristics other than gender. Limitations of the Qualit ative Analyses Findings from the qualitative data are limited to the clinical settings from which patients were recruited, and less applicable to health care settings with different funding levels and patient demographics. The recruitment strategy or indi vidual primary care patients was limited to those with moderate depression and competing physical illness, which does not represent all primary care patients receiving their care in community clinics. Also, because of difficulty contacting patients and get ting consent to complete an interview, selection of the patients to be interviewed could not be as rigorous, therefore not having the kind of variability across patients that was proposed. Interviews were only done with a limited sample and those patients who spoke their illnesses.

PAGE 120

110 Dissemination of Results In addition to serving as the basis for my dissertation and doctoral degree, I will submit sections of t his study to peer reviewed academic journals and present my findings at local and national meetings and conferences. Some of the journals I anticipate submitting to include the Journal of Behavioral Medicine, Annals of Family Medicine, and the Journal of Primary Care & Community Health as all three of these journals publish articles associated with my topic. I also anticipate presenting my findings at the Society of Behavioral Medicine national conference in 2014 as well as the North American Primary Care Research Group national conference. Additionally, the results will be presented to Denver Health given the participants were recruited from their community health centers, and to research colleagues and medical students in the Colorado area.

PAGE 121

111 REFERE NCES Addison, R. B. (1999). A grounded hermeneutic editing approach. In B. F. Crabtree & W. J. Miller (Eds.), Doing Qualitative Research (2nd ed., pp. 145 161). Thousand Oaks, CA: Sage Publications, Inc. Adler, N. E., & Ostrove, J. M. (1999). Socioeconomic status and health: what we know and what we don't. Ann N Y Acad Sci, 896 3 15. Akhtar Danesh, N., & Landeen, J. (2007). Relation between depression and sociodemographic factors. Int J Ment Health Syst, 1 (1), 4. doi: 10.1186/1752 445 8 1 4 Alegria, M., Canino, G., Rios, R., Vera, M., Calderon, J., Rusch, D., et al. (2002). Inequalities in use of specialty mental health services among Latinos, African Americans, and non Latino whites. [Comparative Study]. Psychiatr Serv, 53 (12), 1547 15 55. Alegria, M., Chatterji, P., Wells, K., Cao, Z., Chen, C. N., Takeuchi, D., et al. (2008). Disparity in depression treatment among racial and ethnic minority populations in the United States. [Research Support, N.I.H., Extramura Research Support, Non U .S. Gov't Research Support, U.S. Gov't, P.H.S.]. Psychiatr Serv, 59 (11), 1264 1272. doi: 10.1176/appi.ps.59.11.1264 Allen, R. (1969). Legal Rights of the Disabled and Disadvantaged Washington, D.C.: Superintendent of Documents, U.S. Government Printing O ffice. Allison, D. B., Newcomer, J. W., Dunn, A. L., Blumenthal, J. A., Fabricatore, A. N., Daumit, G. L., et al. (2009). Obesity among those with mental disorders: a National Institute of Mental Health meeting report. Am J Prev Med, 36 (4), 341 350. doi: 1 0.1016/j.amepre.2008.11.020 Alonso, J., Angermeyer, M. C., Bernert, S., Bruffaerts, R., Brugha, T. S., Bryson, H., et al. (2004). Use of mental health services in Europe: results from the European Study of the Epidemiology of Mental Disorders (ESEMeD) proj ect. Acta Psychiatr Scand Suppl (420), 47 54. doi: 10.1111/j.1600 0047.2004.00330.x American Academy of Family Physicians. (2013). Primary Care Retrieved June 5, 2013, 2013, from http://www.aafp.org/online/en/home/policy/policies/p/primarycare.html American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders (IV TR ed.). Washington (DC): American Psychiatric Press. American Psychiatric As sociation (APA). (2012). DSM, from http://www.psychiatry.org/practice/dsm American Psychological Association (APA). (2013). Socioeconomic status, from http://www.apa.org/topics/socioeconomic status/ Anderson, G. (2010). Chronic Care: Making the Case for Ongoing Care: Robert Wood Foundation/Johns Hopkins Bloomber School of Public Health. Anderson, R. J., Freedland, K. E., Clouse, R. E., & Lu stman, P. J. (2001). The prevalence of comorbid depression in adults with diabetes: a meta analysis. Diabetes Care, 24 (6), 1069 1078. Ani, C., Bazargan, M., Hindman, D., Bell, D., Rodriguez, M., & Baker, R. S. (2009). Comorbid Chronic Illness and the Diag nosis and Treatment of Depression in Safety Net Primary Care Settings. J Am Board Fam Med (22), 123 135. Aragones, E., Pinol, J. L., & Labad, A. (2006). The overdiagnosis of depression in non depressed patients in primary care. Fam Pract, 23 (3), 363 368. d oi: 10.1093/fampra/cmi120

PAGE 122

112 Arnow, B. A., Hunkeler, E. M., Blasey, C. M., Lee, J., Constantino, M. J., Fireman, B., et al. (2006). Comorbid depression, chronic pain, and disability in primary care. [Research Support, Non U.S. Gov't]. Psychosom Med, 68 (2), 26 2 268. doi: 10.1097/01.psy.0000204851.15499.fc Arroll, B., Goodyear Smith, F., Crengle, S., Gunn, J., Kerse, N., Fishman, T., et al. (2010). Validation of PHQ 2 and PHQ 9 to screen for major depression in the primary care population. Ann Fam Med, 8 (4), 348 353. doi: 10.1370/afm.1139 Bair, M. J., Robinson, R. L., Katon, W., & Kroenke, K. (2003). Depression and Pain Comorbidity: A Literature Review. Arch Intern Med, 163 2433 2445. Bandura, A. (1977). Self efficacy: toward a unifying theory of behavioral cha nge. Psychol Rev, 84 (2), 191 215. Barry, C. A., Stevenson, F. A., Britten, N., Barber, N., & Bradley, C. P. (2001). Giving voice to the lifeworld. More humane, more effective medical care? A qualitative study of doctor patient communication in general pra ctice. [Research Support, Non U.S. Gov't]. Soc Sci Med, 53 (4), 487 505. Barth, J., Schumacher, M., & Herrmann Lingen, C. (2004). Depression as a risk factor for mortality in patients with coronary heart disease: a meta analysis. Psychosom Med, 66 (6), 802 813. doi: 10.1097/01.psy.0000146332.53619.b2 Bayliss, E. A. (2012). Simplifying care for complex patients. Ann Fam Med, 10 (1), 3 5. doi: 10.1370/afm.1352 Bayliss, E. A., Edwards, A. E., Steiner, J. F., & Main, D. S. (2008). Processes of care desired by eld erly patients with multimorbidities. Family Practice (25), 287 293. doi: 10.1093/fampra/cmn040 Bayliss, E. A., Ellis, J. L., & Steiner, J. F. (2005). Subjective assessments of comorbidity correlate with quality of life health outcomes: initial validation of a comorbidity assessment instrument. Health Qual Life Outcomes, 3 51. doi: 1477 7525 3 51 [pii]10.1186/1477 7525 3 51 Bell, R. A., Franks, P., Duberstein, P. R., Epstein, R. M., Feldman, M. D., Fernandez y Garcia, E., et al. (2011). Suffering in silence: reasons for not disclosing depression in primary care. Ann Fam Med, 9 (5), 439 446. doi: 10.1370/afm.1277 Benton, T., Staab, J., & Evans, D. L. (2007). Medical co morbidity in depressive disorders. [Review]. Ann Clin Psychiatry, 19 (4), 289 303. doi: 10.108 0/10401230701653542 Black, S. A., Markides, K. S., & Ray, L. A. (2003). Depression predicts increased incidence of adverse health outcomes in older Mexican Americans with type 2 diabetes. Diabetes Care, 26 (10), 2822 2828. Bogner, H., Dobransky, L. N., & W ittink, M. N. (2008). Patient ethnicity and perceptions of families and friends regarding depression treatment. Ethn Health, 13 (5), 465 478. doi: 10.1080/13557850802023125 Bogner, H. R., Cary, M. S., Bruce, M. L., Reynolds, C. F., 3rd, Mulsant, B., Ten Hav e, T., et al. (2005). The role of medical comorbidity in outcome of major depression in primary care: the PROSPECT study. [Randomized Controlled Trial]. Am J Geriatr Psychiatry, 13 (10), 861 868. doi: 10.1176/appi.ajgp.13.10.861 Bogner, H. R., & de Vries, H F. (2008). Integration of depression and hypertension treatment: a pilot, randomized controlled trial. Ann Fam Med, 6 (4), 295 301. doi: 10.1370/afm.843 Bogner, H. R., Morales, K. H., de Vries, H. F., & Cappola, A. R. (2012). Integrated management of type 2 diabetes mellitus and depression treatment to improve medication adherence: a randomized controlled trial. Ann Fam Med, 10 (1), 15 22. doi: 10.1370/afm.1344 Bovbjerg, R. R., & Kopit, W. G. (1986). Coverage and care for the medically indigent: public and private options. Indiana Law Rev, 19 (4), 857 917. Campbell, L. C., Clauw, D. J., & Keefe, F. J. (2003). Persistent pain and depression: a biopsychosocial perspective. Biol Psychiatry, 54 (3), 399 409.

PAGE 123

113 Centers for Disease Control and Prevention. (2011a). D epression Retrieved January 29, 2013, from http://www.cdc.gov/mentalhealth/basics/mental illness/depression.htm Centers for Disease Control and Prevention. (2011c). Socia l Determinants Definitions, from http://www.cdc.gov/socialdeterminants/Definitions.html Centers for Disease Control and Prevention. (2013). FastStats: Ambulatory Care Use and Physician Visits. Retrieved from http://www.cdc.gov/nchs/fastats/docvisit.htm Chow, J. C., Jaffee, K., & Snowden, L. (2003). Racial/ethnic disparities in the use of mental health services in poverty areas. Am J Public Health, 93 (5), 792 797. Clark, M. M., Cargill, B. R., Medeiros, M. L., & Pera, V. (1996). Changes in self efficacy following obesity treatment. Obes Res, 4 (2), 179 181. Clarke, D. M. (2009). Depression and physical illness: more complex than simple comorbidity. Med J Aust, 190 (7 Suppl), S52 53. doi: Cla10161_fm [pii] Clarke, D. M., & Currie, K. C. (2009). Depression, anxiety and their relationship with chronic diseases: a review of the epidemiology, risk and treatment evidence. Med J Aust, 19 0 (7 Suppl), S54 60. doi: cla10974_fm [pii] Craven, M. A., Cohen, M., Campbell, D., Williams, J., & Kates, N. (1997). Mental health practices of Ontario family physicians: a study using qualitative methodology. Can J Psychiatry, 42 (9), 943 949. Creed, F., & Dickens, C. (2006). Depression in the Medically Ill. In A. Steptoe (Ed.), Depression and Physical Illness Cambridge, UK: Cambridge University Press. Creed, F., Morgan, R., Fiddler, M., Marshall, S., Guthrie, E., & House, A. (2002). Depression and anxiet y impair health related quality of life and are associated with increased costs in general medical inpatients. Psychosomatics, 43 (4), 302 309. Creswell, J. W. (2009). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches Thousand Oaks, CA: Sage Publications, Inc. de Groot, M., Anderson, R., Freedland, K. E., Clouse, R. E., & Lustman, P. J. (2001). Association of depression and diabetes complications: a meta analysis. Psychosom Med, 63 (4), 619 630. de Groot, V., Beckerman, H., Lankhorst G. J., & Bouter, L. M. (2003). How to measure comorbidity. a critical review of available methods. J Clin Epidemiol, 56 (3), 221 229. doi: S0895435602005851 [pii] deGruy, F. V. (1996). Mental Health Care in the Primary Care Setting. In M. S. Donaldson, K. D. Yordy, K. N. Lohr & N. Vanselow (Eds.), Primary Care: America's Health in a New Era Washington, D.C.: National Academy Press. Delany, G. (2007). A farewell to heart sink? Br J Gen Pract, 57 (537), 330 331. Denver Health: About Us. (2012) Retrieved M ay 3, 2012, from http://denverhealth.org/AboutUs.aspx Dey, I. (1993). Qualitative Data Anaylsis London: Routledge. Donaldson, M., Yordy, M., & Vanselow, N. (Eds.). (1994). Defining Primary Care: An Inte rim Report Washington, D.C.: Institute of Medicine (IOM). Donaldson, M. S., Yordy, K. D., Lohr, K. N., & Vanselow, N. (Eds.). (1996). Primary Care: America's Health in a New Era Washington, D.C.: National Academy Press. Duhoux, A., & et. al. (2009). Guid eline concorance of treatment for depressive disorders Atlantic Canada. Soc Psychiatry Psychiatr Epidemiol, 44 385 392. Emmons, K. M., & Rollnick, S. (2001). Motivational interviewing in health care settings. Opportunities and limitations. [Research Supp ort, Non U.S. Gov't Research Support, U.S. Gov't, P.H.S. Review]. Am J Prev Med, 20 (1), 68 74. Everson, S. A., Maty, S. C., Lynch, J. W., & Kaplan, G. A. (2002). Epidemiologic evidence for the relation between socioeconomic status and depression, obesity, and diabetes. J Psychosom Res, 53 (4), 891 895.

PAGE 124

114 Fishbain, D. A., Cutler, R., Rosomoff, H. L., & Rosomoff, R. S. (1997). Chronic pain associated depression: antecedent or consequence of chronic pain? A review. [Research Support, U.S. Gov't, Non P.H.S. Revi ew]. Clin J Pain, 13 (2), 116 137. Fortin, M., Bravo, G., Hudon, C., Lapointe, L., Almirall, J., Dubois, M., et al. (2006). Relationship between multimorbidity and health related quality of life of patients in primary care. Qual Life Res (15), 83 91. Forti n, M., Bravo, G., Hudon, C., Lapointe, L., Dubois, M., & Almirall, J. (2006). Relationship between psychological distress and multimorbidity of patients in family practice. Annals of Family Medicine (4), 417 422. Fortin, M., Bravo, G., Hudon, C., Vanasse, A., & Lapointe, L. (2005). Prevalence of multimorbidity among adults seen in family practice. [Research Support, Non U.S. Gov't]. Ann Fam Med, 3 (3), 223 228. doi: 10.1370/afm.272 Fortin, M., Dubois, M. F., Hudon, C., Soubhi, H., & Almirall, J. (2007). Mult imorbidity and quality of life: a closer look. Health Qual Life Outcomes, 5 52. doi: 10.1186/1477 7525 5 52 Fortin, M., Lapointe, L., Hudon, C., Vanasse, A., Ntetu, A. L., & Maltais, D. (2004). Multimorbidity and quality of life in primary care: a systema tic review. Health Qual Life Outcomes, 2 51. doi: 10.1186/1477 7525 2 51 Fortin, M., Soubhi, H., Hudon, C., Bayliss, E. A., & van den Akker, M. (2007). Multimorbidity's many challenges. [Editorial]. BMJ, 334 (7602), 1016 1017. doi: 10.1136/bmj.39201.463819 .2C Galson, S. K. (2009). Mental health matters. Public Health Rep, 124 (2), 189 191. Gensichen, J., Beyer, M., Muth, C., Gerlach, F. M., Von Korff, M., & Ormel, J. (2006). Case management to improve major depression in primary health care: a systematic re view. Psychol Med, 36 (1), 7 14. doi: 10.1017/S0033291705005568 Gilbody, S. (2011). Depression in primary care. In S. Gilbody & P. Bower (Eds.), Depression in Primary Care: Evidence and Practice Cambridge UK: Cambridge University Press. Gilman, S. E., Bruc e, M. L., Ten Have, T., Alexopoulos, G. S., Mulsant, B. H., Reynolds, C. F., 3rd, et al. (2013). Social inequalities in depression and suicidal ideation among older primary care patients. Soc Psychiatry Psychiatr Epidemiol, 48 (1), 59 69. doi: 10.1007/s0012 7 012 0575 9 Gilman, S. E., Fitzmaurice, G. M., Bruce, M. L., Ten Have, T., Glymour, M. M., Carliner, H., et al. (2013). Economic inequalities in the effectiveness of a primary care intervention for depression and suicidal ideation. Epidemiology, 24 (1), 14 22. doi: 10.1097/EDE.0b013e3182762403 Gilman, S. E., Kawachi, I., Fitzmaurice, G. M., & Buka, S. L. (2002). Socioeconomic status in childhood and the lifetime risk of major depression. Int J Epidemiol, 31 (2), 359 367. Gilman, S. E., Trinh, N. H., Smoller J. W., Fava, M., Murphy, J. M., & Breslau, J. (2013). Psychosocial stressors and the prognosis of major depression: a test of Axis IV. Psychol Med, 43 (2), 303 316. doi: 10.1017/S0033291712001080 Glazier, R. H., Agha, M. M., Moineddin, R., & Sibley, L. M. (2009). Universal health insurance and equity in primary care and specialist office visits: a population based study. Ann Fam Med, 7 (5), 396 405. doi: 10.1370/afm.994 Gonzales, J. S., Esbitt, S. A., Schneider, H. E., Osborne, P. J., & Kupperman, E. G. (20 11). Psychological Issues in Adults with Type 2 Diabetes. In S. Pagoto (Ed.), Psychological Co Morbidities of Physical Illness: A Behavioral Medicine Perspective (pp. 73 122). New York: Springer + Business Media, LLC. Gonzalez, J. S., Peyrot, M., McCarl, L A., Collins, E. M., Serpa, L., Mimiaga, M. J., et al. (2008). Depression and diabetes treatment nonadherence: a meta analysis. Diabetes Care, 31 (12), 2398 2403. doi: 10.2337/dc08 1341

PAGE 125

115 Goodnick, P. J. (2001). Use of antidepressants in treatment of comorbi d diabetes mellitus and depression as well as in diabetic neuropathy. Ann Clin Psychiatry, 13 (1), 31 41. Greenberg, P. E., & Birnbaum, H. G. (2005). The economic burden of depression in the US: societal and patient perspectives. Expert Opin Pharmacother, 6 (3), 369 376. doi: EOP060302 [pii] 10.1517/14656566.6.3.369 Grumbach, K. (2003). Chronic illness, comorbidities, and the need for medical generalism. Ann Fam Med, 1 (1), 4 7. Guide to Community Preventive Services. (2011). Improving mental health and addr essing mental illness: collaborative care for the management of depressive disorders. Guthrie, E., Dickens, C., Blakemore, A., Watson, J., Chew Graham, C., Lovell, K., Afzal, C., Kapur, N., & Tomenson, B. (2014). Depression predicts future emergency hospit al admissions in primary care patients with chronic phyisical illness. Journal of Psychomatic Research doi: Center for Health Care Strategies, INC Harman, J. S., Edlund, M. J., Fortney, J. C., & Kallas, H. (2005). The influence of comorbid chronic medical conditions on the adequacy of depression care for older Americans. J Am Geriatr Soc, 53 (12), 2178 2183. doi: 10.1111/j.1532 5415.2005.00511.x Harpole, L. H., Williams, J. W., Jr., Olsen, M. K., Stechuchak, K. M., Oddone, E., Callahan, C. M., et al. (2005) Improving depression outcomes in older adults with comorbid medical illness. [Clinical Trial Randomized Controlled Trial Research Support, Non U.S. Gov't]. Gen Hosp Psychiatry, 27 (1), 4 12. doi: 10.1016/j.genhosppsych.2004.09.004 Haynes, B. R. (1979). In troduction. In B. R. Haynes, D. W. Taylor & D. L. Sackett (Eds.), Compliance in health care (pp. 1 10). Baltimore, MD.: The Johns Hopkins University Press. Henke, R. M., Zaslavsky, A. M., McGuire, T. G., Ayanian, J. Z., & Rubenstein, L. V. (2009). Clinical inertia in depression treatment. Med Care, 47 (9), 959 967. doi: 10.1097/MLR.0b013e31819a5da0 Hirschfeld, R. M., Keller, M. B., Panico, S., Arons, B. S., Barlow, D., Davidoff, F., et al. (1997). The National Depressive and Manic Depressive Association cons ensus statement on the undertreatment of depression. JAMA, 277 (4), 333 340. Hox, J. (2002). Multilevel Analysis: Techniques and Applications. London, UK.: Lawrence Erlbaum and Associates. Hudon, C., Fortin, M., & Soubhi, H. (2007). Abbreviated guidelines for scoring the Cumulative Illness Rating Scale (CIRS) in family practice. J Clin Epidemiol, 60 (2), 212. doi: S0895 4356(06)00287 3 [pii] 10.1016/j.jclinepi.2005.12.021 Hudon, C., Fortin, M., & Vanasse, A. (2005). Cumulative Illness Rating Scale was a reli able and valid index in a family practice context. J Clin Epidemiol, 58 (6), 603 608. doi: S0895 4356(04)00352 X [pii] 10.1016/j.jclinepi.2004.10.017 Hudson, D. L., Neighbors, H. W., Geronimus, A. T., & Jackson, J. S. (2012). The relationship between socioe conomic position and depression among a US nationally representative sample of African Americans. Soc Psychiatry Psychiatr Epidemiol, 47 (3), 373 381. doi: 10.1007/s00127 011 0348 x Huntley, A. L., Johnson, R., Purdy, S., Valderas, J., & Salisbury, C. (2012 ). Measures of Multimorbidity and Morbidity Burden for Use in Primary Care and Community Settings: A Systematic Review and Guide. Annals of Family Medicine, 10 134 141. doi: 10.1370/afm.1363. Inglis, V., Ball, K., & Crawford, D. (2005). Why do women of lo w socioeconomic status have poorer dietary behaviours than women of higher socioeconomic status? A qualitative exploration. Appetite, 45 (3), 334 343. doi: 10.1016/j.appet.2005.05.003 Iosifescu, D. V. (2007). Treating Depression in the Medically Ill. Psychi atr Clin N Am (30), 77 90.

PAGE 126

116 Iosifescu, D. V., Nierenberg, A. A., Alpert, J. E., Smith, M., Bitran, S., Dording, C., et al. (2003). The impact of medical comorbidity on acute treatment in major depressive disorder. Am J Psychiatry, 160 (12), 2122 2127. Jaen, C. R., Stange, K. C., & Nutting, P. A. (1994). Competing demands of primary care: a model for the delivery of clinical preventive services. [Review]. J Fam Pract, 38 (2), 166 171. Jaen, C. R., Stange, K. C., Tumiel, L. M., & Nutting, P. (1997). Missed opp ortunities for prevention: smoking cessation counseling and the competing demands of practice. [Comparative Study Research Support, Non U.S. Gov't]. J Fam Pract, 45 (4), 348 354. Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed Methods Research: A Reseac h Paradigm Whose Time Has Come. Educational Researcher, 33 (7), 14 26. Katon, W. (2003). Clinical and health services relationships between major depression, depressive symptoms. and general medical illness. Society of Biological Psychiatry (54), 216 226. Katon, W. (2009). The impact of depression on workplace functioning and disability. American Journal of Managed Care., 15 (11 Suppl), 5322 5327. Katon, W., & Schulberg, H. (1992). Epidemiology of depression in primary care. [Guideline Practice Guideline Re view]. Gen Hosp Psychiatry, 14 (4), 237 247. Katon, W. J. (2011). Epidemiology and treatment of depression in patients with chronic medical illness. Dialogues Clin Neurosci, 13 (1), 7 23. Katon, W. J., Lin, E. H., Von Korff, M., Ciechanowski, P., Ludman, E J., Young, B., et al. (2010). Collaborative care for patients with depression and chronic illnesses. N Engl J Med, 363 (27), 2611 2620. doi: 10.1056/NEJMoa1003955 Katon, W. J., Rutter, C., Simon, G., Lin, E. H., Ludman, E., Ciechanowski, P., et al. (2005) The association of comorbid depression with mortality in patients with type 2 diabetes. Diabetes Care, 28 (11), 2668 2672. Kawachi, I., & Berkman, L. F. (2001). Social ties and mental health. J Urban Health, 78 (3), 458 467. doi: 10.1093/jurban/78.3.458 K eeley, R. D., & Brody, D. (2007). Motivational Interviews to Improve Depressive Treatment in Primary Care. Denver Health: National Institutes of Mental Health. Keeley, R. D., & Brody, D. (2012). Motivational Interviewing to Improve Symptoms of Depression i n Primary Care. Denver, CO: NIMH. Kemp, D. E., Gao, K., Chan, P. K., Ganocy, S. J., Findling, R. L., & Calabrese, J. R. (2010). Medical comorbidity in bipolar disorder: relationship between illnesses of the endocrine/metabolic system and treatment outcome. Bipolar Disord, 12 (4), 404 413. doi: 10.1111/j.1399 5618.2010.00823.x King, N., & Horrocks, C. (2010). Interviews in Qualitative Research Thousand Oaks, CA.: SAGE Publications Inc. Kleinman, A. (1988). The Illness Naratives: Suffering, Healing, and the H uman Condition : Basic Books. Klinkman, M. S. (1997). Competing demands in psychosocial care. A model for the identification and treatment of depressive disorders in primary care. [Review]. Gen Hosp Psychiatry, 19 (2), 98 111. Knol, M. J., Twisk, J. W., Bee kman, A. T., Heine, R. J., Snoek, F. J., & Pouwer, F. (2006). Depression as a risk factor for the onset of type 2 diabetes mellitus. A meta analysis. Diabetologia, 49 (5), 837 845. doi: 10.1007/s00125 006 0159 x Kohn, M. L., Dohrenwend, B. P., & Mirotznik, J. (1998). Epidemiological findings on selected psychiatric disorders in the general population. In B. P. Dohrenwen (Ed.), Adversity, Stress and Psychopathology New York: Oxford University Press. Koike, A. K., Unutzer, J., & Wells, K. B. (2002). Improving the care for depression in patients with comorbid medical illness. Am J Psychiatry, 159 (10), 1738 1745.

PAGE 127

117 Kraemer, H. C., Frank, E., & Kupfer, D. J. (2006). Moderators of treatment outcomes: clinical, research, and policy importance. JAMA, 296 (10), 1286 12 89. doi: 10.1001/jama.296.10.1286 Kraemer, H. C., Kiernan, M., Essex, M., & Kupfer, D. J. (2008). How and why criteria defining moderators and mediators differ between the Baron & Kenny and MacArthur approaches. Health Psychol, 27 (2 Suppl), S101 108. doi: 10.1037/0278 6133.27.2(Suppl.).S101 Kravitz, R. L., & Ford, D. E. (2008). Introduction: chronic medical conditions and depression -the view from primary care. Am J Med, 121 (11 Suppl 2), S1 7. doi: 10.1016/j.amjmed.2008.09.007 Kreft, I., & De Leeuw, J. (199 8). Introducing Multilevel Modeling London: Sage Publications, Inc. Kroenke, K., Spitzer, R. L., & Williams, J. B. (2001). The PHQ 9: validity of a brief depression severity measure. J Gen Intern Med, 16 (9), 606 613. doi: jgi01114 [pii] Kroenke, K., Spitz er, R. L., & Williams, J. B. (2003). The Patient Health Questionnaire 2: validity of a two item depression screener. Med Care, 41 (11), 1284 1292. doi: 10.1097/01.MLR.0000093487.78664.3C Kronish, I. M., Rieckmann, N., Schwartz, J. E., Schwartz, D. R., & Dav idson, K. W. (2009). Is depression after an acute coronary syndrome simply a marker of known prognostic factors for mortality? Psychosom Med, 71 (7), 697 703. doi: 10.1097/PSY.0b013e3181ad2abd Kurdyak, P. A., & Gnam, W. H. (2004). Medication management of d epression: the impact of comorbid chronic medical conditions. [Research Support, Non U.S. Gov't]. J Psychosom Res, 57 (6), 565 571. doi: 10.1016/j.jpsychores.2004.04.367 Lantz, P. M., House, J. S., Lepkowski, J. M., Williams, D. R., Mero, R. P., & Chen, J. (1998). Socioeconomic factors, health behaviors, and mortality: results from a nationally representative prospective study of US adults. JAMA, 279 (21), 1703 1708. Latham, A. E., & Prigerson, H. G. (2004). Suicidality and bereavement: complicated grief as psychiatric disorder presenting greatest risk for suicidality. Suicide Life Threat Behav, 34 (4), 350 362. doi: 10.1521/suli.34.4.350.53737 Lin, E. H., Katon, W., Von Korff, M., Tang, L., Williams, J. W., Jr., Kroenke, K., et al. (2003). Effect of improving depression care on pain and functional outcomes among older adults with arthritis: a randomized controlled trial. JAMA, 290 (18), 2428 2429. doi: 10.1001/jama.290.18.2428 Lin, E. H., Von Korff, M., Ciechanowski, P., Peterson, D., Ludman, E. J., Rutter, C. M., et al. (2012). Treatment adjustment and medication adherence for complex patients with diabetes, heart disease, and depression: a randomized controlled trial. [Multicenter Study Randomized Controlled Trial Research Support, N.I.H., Extramural]. Ann Fam Med, 10 (1), 6 14. doi: 10.1370/afm.1343 Linde, J. A., Jeffery, R. W., Levy, R. L., Sherwood, N. E., Utter, J., Pronk, N. P., et al. (2004). Binge eating disorder, weight control self efficacy, and depression in overweight men and women. Int J Obes Relat M etab Disord, 28 (3), 418 425. doi: 10.1038/sj.ijo.0802570 Linn, B. S., Linn, M. W., & Gurel, L. (1968). Cumulative illness rating scale. J Am Geriatr Soc, 16 (5), 622 626. Lorant, V., Deliege, D., Eaton, W., Robert, A., Philippot, P., & Ansseau, M. (2003). Socioeconomic inequalities in depression: a meta analysis. [Meta Analysis Research Support, Non U.S. Gov't Research Support, U.S. Gov't, P.H.S.]. Am J Epidemiol, 157 (2), 98 112. Lowe, B., Grafe, K., Zipfel, S., Witte, S., Loerch, B., & Herzog, W. (2004). Diagnosing ICD 10 depressive episodes: superior criterion validity of the Patient Health Questionnaire. Psychother Psychosom, 73 (6), 386 390. doi: 10.1159/000080393

PAGE 128

118 Lowe, B., Spitzer, R. L., Grafe, K., Kroenke, K., Quenter, A., Zipfel, S., et al. (2004). C omparative validity of three screening questionnaires for DSM IV depressive disorders and physicians' diagnoses. J Affect Disord, 78 (2), 131 140. Lowe, B., Unutzer, J., Callahan, C. M., Perkins, A. J., & Kroenke, K. (2004). Monitoring depression treatment outcomes with the patient health questionnaire 9. Med Care, 42 (12), 1194 1201. Lustman, P. J., Anderson, R. J., Freedland, K. E., de Groot, M., Carney, R. M., & Clouse, R. E. (2000). Depression and poor glycemic control: a meta analytic review of the lit erature. Diabetes Care, 23 (7), 934 942. Lynch, J. M., Askew, D. A., Mitchell, G. K., & Hegarty, K. L. (2012). Beyond symptoms: defining primary care mental health clinical assessment priorities, content and process. [Research Support, Non U.S. Gov't]. Soc Sci Med, 74 (2), 143 149. doi: 10.1016/j.socscimed.2011.08.043 Macintyre, S. (1994). Understanding the social patterning of health: the role of the social sciences. J Public Health Med, 16 (1), 53 59. Malpass, A., Shaw, A., Sharp, D., Walter, F., Feder, G. Ridd, M., et al. (2009). "Medication career" or "moral career"? The two sides of managing antidepressants: a meta ethnography of patients' experience of antidepressants. Soc Sci Med, 68 (1), 154 168. doi: 10.1016/j.socscimed.2008.09.068 Mansyur, C. L., Pa vlik, V. N., Hyman, D. J., Taylor, W. C., & Goodrick, G. K. (2013). Self efficacy and barriers to multiple behavior change in low income African Americans with hypertension. J Behav Med, 36 (1), 75 85. doi: 10.1007/s10865 012 9403 7 Marmot, M. G., Kogevinas M., & Elston, M. A. (1987). Social/economic status and disease. Annu Rev Public Health, 8 111 135. doi: 10.1146/annurev.pu.08.050187.000551 Martucci, M., Balestrieri, M., Bisoffi, G., Bonizzato, P., Covre, M. G., Cunico, L., et al. (1999). Evaluating ps ychiatric morbidity in a general hospital: a two phase epidemiological survey. Psychol Med, 29 (4), 823 832. Mays, N., & Pope, C. (2000). Assessing quality in qualitative research. BMJ, 30 (7226), 50 52. doi: PMC1117321 McAllister, E. J., Dhurandhar, N. V., Keith, S. W., Aronne, L. J., Barger, J., Baskin, M., et al. (2009). Ten putative contributors to the obesity epidemic. Crit Rev Food Sci Nutr, 49 (10), 868 913. doi: 10.1080/10408390903372599 McClave, A. K., Dube, S. R., Strine, T. W., Kroenke, K., Carabal lo, R. S., & Mokdad, A. H. (2009). Associations between smoking cessation and anxiety and depression among U.S. adults. Addict Behav, 34 (6 7), 491 497. doi: 10.1016/j.addbeh.2009.01.005 McEwen, B. S. (1998). Stress, adaptation, and disease. Allostasis and allostatic load. Ann N Y Acad Sci, 840 33 44. Melek, S., & Norris, D. (2008). Chronic Conditions and Comorbid Psychological Distress. Seattle: Millman. Mendlowicz, M. V., & Stein, M. B. (2000). Quality of life in individuals with anxiety disorders. Am J Psychiatry, 157 (5), 669 682. Mitchell, A. J., Vaze, A., & Rao, S. (2009). Clinical diagnosis of depression in primary care: a meta analysis. [Meta Analysis]. Lancet, 374 (9690), 609 619. doi: 10.1016/S0140 6736(09)60879 5 Mitchell J, Trangle M, Degnan B, G abert T, Haight B, Kessler D, Mack N, Mallen E, Novak H, Rossmiller D, Setterlund L, Somers K, Valentino N, Vincent S. Adult depression in primary care. Bloomington (MN): Institute for Clinical Systems Improvement (ICSI); 2013 Sep. 129 p. [334 references] Morris, D. W., Budhwar, N., Husain, M., Wisniewski, S. R., Kurian, B. T., Luther, J. F., et al. (2012). Depression treatment in patients with general medical conditions: results from the CO MED trial. Ann Fam Med, 10 (1), 23 33. doi: 10.1370/afm.1316

PAGE 129

119 Mossak owski, K. N. (2008). Is the duration of poverty and unemployment a risk factor for heavy drinking? Soc Sci Med, 67 (6), 947 955. doi: 10.1016/j.socscimed.2008.05.019 Moussavi, S., Chatterji, S., Verdes, E., Tandon, A., Patel, V., & Ustun, B. (2007). Depress ion, chronic diseases, and decrements in health: results from the World Health Surveys. Lancet, 370 (9590), 851 858. doi: S0140 6736(07)61415 9 [pii]10.1016/S0140 6736(07)61415 9 Mujica Mota, R. E., Roberts, M., Abel, G., Elliott, M., Lyratzopoulos, G., Rol and, M., et al. (2015). Common patterns of morbidity and multi morbidity and their impact on health related quality of life: evidence from a national survey. Qual Life Res, 24 (4), 909 918. doi: 10.1007/s11136 014 0820 7 Murray, C. J. L., & Lopez, A. D. (19 96). The global burden of disease: A comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 projected to 2020 Boston, MA: Havard University Press. National Association of Community Health Centers. (2010). He althcare reform impact at glance: What's in it for persons with mental and addiction disorders. In N. C. Magazine (Ed.). Washington D.C.: National Council for Community Behavioral Healthcare. National Center for Chronic Disease Prevention and Health Pro motion. (2012). Chronic Diseases and Health Promotion. Atlanta, GA: CDC. National Institute for Health and Clinical Excellence (NIHCE). (2009). Depression in Adults with a Chronic Physical Health Problem: Treatment and Management. London. National Institut es of Health (NIH). (2000). NIH strategic research plan to reduce and ultimately eliminate health disparities Bethesda, MD.: Retrieved from http://www.nimhd.nih.go v/strategicmock/our_programs/strategic/pubs/NCI.pdf Newcomer, S. R., Steiner, J. F., & Bayliss, E. A. (2011). Identifying subgroups of complex patients with cluster analysis. [Research Support, N.I.H., Extramural]. Am J Manag Care, 17 (8), e324 332. Nutt ing, P. A., Gallagher, K., Riley, K., White, S., Dickinson, W. P., Korsen, N., et al. (2008). Care management for depression in primary care practice: findings from the RESPECT Depression trial. Ann Fam Med, 6 (1), 30 37. doi: 10.1370/afm.742 Nutting, P. A. Rost, K., Dickinson, M., Werner, J. J., Dickinson, P., Smith, J. L., et al. (2002). Barriers to Initiating Depression Treatment in Primary Care Practice. J Gen Intern Med, 2002 (17), 103 111. Nutting, P. A., Rost, K., Smith, J., Werner, J. J., & Elliot, C. (2000). Competing Demands from Physical Problems: Effect on Initiating and Completing Depression Care Over 6 Months. Arch Fam Med, 9 1059 1064. Oslin, D. W., Datto, C. J., Kallan, M. J., Katz, I. R., Edell, W. S., & TenHave, T. (2002). Association bet ween medical comorbidity and treatment outcomes in late life depression. [Research Support, U.S. Gov't, P.H.S.]. J Am Geriatr Soc, 50 (5), 823 828. Pagoto, S. (Ed.). (2011). Psychological Co morbidities of Physical Illness: A Behavioral Medicine Perspectiv e New York: Springer + Business Media, LLC. Pagoto, S., Bodenlos, J. S., Kantor, L., Gitkind, M., Curtin, C., & Ma, Y. (2007). Association of major depression and binge eating disorder with weight loss in a clinical setting. Obesity (Silver Spring), 15 (11 ), 2557 2559. doi: 10.1038/oby.2007.307 Pagoto, S., Schneider, K., Appelhans, B. M., Curtin, C., & Hajduk, A. (2011). Psychological Co Morbidities of Obesity. In S. Pagoto (Ed.), Psychological Co Morbidities of Physical Illness: A Behavioral Medicine Persp ective (pp. 1 72). New York: Springer + Business Media, LLC. Palmer, V., Gunn, J., Kokanovic, R., Griffiths, F., Shrimpton, B., Hurworth, R., et al. (2010). Diverse voices, simple desires: a conceptual design for primary care to respond to depression and r elated disorders. Fam Pract, 27 (4), 447 458. doi: 10.1093/fampra/cmq016

PAGE 130

120 Pampel, F. C., Krueger, P. M., & Denney, J. T. (2010). Socioeconomic Disparities in Health Behaviors. Annu Rev Sociol, 36 349 370. doi: 10.1146/annurev.soc.012809.102529 Papakostas, G I., Petersen, T., Iosifescu, D. V., Roffi, P. A., Alpert, J. E., Rosenbaum, J. F., et al. (2003). Axis III disorders in treatment resistant major depressive disorder. Psychiatry Res, 118 (2), 183 188. Parekh, A. K., & Barton, M. B. (2010). The Challenge of Multiple Comorbidity for the US Health Care System. [Commentary]. JAMA, 303 (13), 1303 1304. Patton, M. Q. (2001). Qualitative Research & Evaluation Methods (3rd ed.). Thousand Oaks, CA: Sage Publications Inc. Perlis, R. H., Iosifescu, D. V., Alpert, J. Nierenberg, A. A., Rosenbaum, J. F., & Fava, M. (2004). Effect of medical comorbidity on response to fluoxetine augmentation or dose increase in outpatients with treatment resistant depression. Psychosomatics, 45 (3), 224 229. doi: 10.1176/appi.psy.45.3.2 24 Piccinelli, M., & Wilkinson, G. (2000). Gender differences in depression. Critical review. Br J Psychiatry, 177 486 492. Pickett, K. E., & Pearl, M. (2001). Multilevel analyses of neighbourhood socioeconomic context and health outcomes: a critical rev iew. J Epidemiol Community Health, 55 (2), 111 122. Pollard, K. M., & O'Hare, W. P. (1999). America's Racial and Ethnic Minorities Population Bulletin (Vol. 54). Washington, D.C: Population Reference Bureau. Pope, C., & Mays, N. (1995). Reaching the parts other methods cannot reach: an introduction to qualitative methods in health and human services research. British Medical Journal, 311 (6996), 42 45. doi: PMC2550091 Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical Linear Models: Applications and Data Analysis Methods (2nd ed.). Thousand Oaks: Sage Publications. Robert Wood Johnson Foundation. (2010). Chronic Care: Making the Case for Ongoing Care. Rost, K., Nutting, P. A., Smith, J., Coyne, J., Cooper Patrick, L., & Rubenstein, L. (2000). The Role of C ompeting Demands in the Treatment Provided Primary Care Patients with Major Depression. Arch Fam Med, 9 150 154. Rothman, A. A., & Wagner, E. H. (2003). Chronic illness management: what is the role of primary care? Ann Intern Med, 138 (3), 256 261. Rutle dge, T., Reis, V. A., Linke, S. E., Greenberg, B. H., & Mills, P. J. (2006). Depression in heart failure a meta analytic review of prevalence, intervention effects, and associations with clinical outcomes. J Am Coll Cardiol, 48 (8), 1527 1537. doi: 10.1016/ j.jacc.2006.06.055 Saraceno, B., Levav, I., & Kohn, R. (2005). The public mental health significance of research on socio economic factors in schizophrenia and major depression. World Psychiatry, 4 (3), 181 185. Scientific Software International (SSI). (20 13). Hierarchical Linear and Nonlinear Modeling (HLM). Skokie, IL. Shedler, J., Beck, A., & Bensen, S. (2000). Practical mental health assessment in primary care. Validity and utility of the Quick PsychoDiagnostics Panel. J Fam Pract, 49 (7), 614 621. Sim on, G. E., Von Korff, M., & Lin, E. (2005). Clinical and functional outcomes of depression treatment in patients with and without chronic medical illness. Psychol Med, 35 (2), 271 279. Simon, G. E., Von Korff, M., Saunders, K., Miglioretti, D. L., Crane, P K., van Belle, G., et al. (2006). Association between obesity and psychiatric disorders in the US adult population. Arch Gen Psychiatry, 63 (7), 824 830. doi: 10.1001/archpsyc.63.7.824 Simon, G. E., & VonKorff, M. (1991). Somatization and psychiatric diso rder in the NIMH Epidemiologic Catchment Area study. Am J Psychiatry, 148 (11), 1494 1500.

PAGE 131

121 Smedley, B. D., Stith, A. Y., & Nelson (Eds.). (2003). Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care Washington, D.C.: The National Ac ademies Press. Spiegel, D., & Giese Davis, J. (2003). Depression and cancer: mechanisms and disease progression. Biol Psychiatry, 54 (3), 269 282. Stange, K. C., Fedirko, T., Zyzanski, S. J., & Jaen, C. R. (1994). How do family physicians prioritize delive ry of multiple preventive services? J Fam Pract, 38 (3), 231 237. Stanners, M. N., Barton, C. A., Shakib, S., & Winefield, H. R. (2014). Depression diagnosis and treatment amongst multimorbid patients: a thematic analysis. BMC Fam Pract, 15 124. doi: 10.1 186/1471 2296 15 124 Stansfeld, S., & Rasul, F. (2006). Psychosocial factors, depression and illness. In A. Steptoe (Ed.), Depression and Physical Illness Cambridge, UK: Cambridge University Press. Starfield, B., Shi, L., & Macinko, J. (2005). Contributio n of primary care to health systems and health. [Research Support, U.S. Gov't, P.H.S. Review]. Milbank Q, 83 (3), 457 502. doi: 10.1111/j.1468 0009.2005.00409.x State of Colorado. Colorado Indigent Care Program (CICP) Retrieved from http://www.colorado.gov/cs/Satellite/HCPF/HCPF/1214299805914 Steptoe, A. (Ed.). (2007). Depression and Physical Illness Cambridge, UK: Cambridge University Press. Stringhini, S., Sabia, S., Shipley, M., Brunner, E., Nabi, H., Kivimaki, M., et al. (2010). Association of socioeconomic position with health behaviors and mortality. JAMA, 303 (12), 1159 1166. doi: 10.1001/jama.2010.297 Subrahmanian, K., Jones, K., & Duncan, C. (2003). Multilevel Methods for Public Health Research. In I. Kawachi & L. Berkman (Eds.), Neighborhoods and Health Oxford: Oxford Univeristy Press. Substance Abuse and Mental Health Services Administration (SAMHSA). (2012). Results for the 2010 National Survey on Drug Use and Health: Mental Health Findings. Rockville, MD. Swanson, J. G. (1994). Family physicians' approach to psychotherapy and counseling. Perceptions and practices. Can Fam Physician, 40 53 58. Tabachnick, B. G., & Fidell, L. S. (2007). Using Multivariate Statistics (5 th ed.). Boston, MA.: Pearson. Taylor, I. M., Ntourmanis, N., Standage, M., & Spray, C. M. (2010). Motivational Predictors of Physical Education Students' Effort, Exercise Intentions, and Leisure Time Physical Activity: A Multilevel Linear Growth Analysis Journal of Sport & Exercises Psychology, 2010 (32), 99 120. Teh, C. F., Reynolds, C. F., 3rd, & Cleary, P. D. (2008). Quality of depression care for people with coincident chronic medical conditions. [Research Support, N.I.H., Extramural]. Gen Hosp Psych iatry, 30 (6), 528 535. doi: 10.1016/j.genhosppsych.2008.07.002 The Johns Hopkins Primary Care Policy Center. (2013). Definitions: What is primary care? from http://www.jhsph.edu/research/centers and institutes/johns hopkins primary care policy center/definitions.html The Office of Minority Health. (2013). Mental Health Data/Statistics: U.S. Department of Health and Human Services Thielke, S., Vannoy, S., & Unutzer, J. (2007). Integrating mental health and primary care. [Research Support, N.I.H., Extramural Research Support, Non U.S. Gov't Review]. Prim Care, 34 (3), 571 592, vii. doi: 10.1016/j.pop.2007.05.007 Thoits, P. A. (2011) Mechanisms linking social ties and support to physical and mental health. J Health Soc Behav, 52 (2), 145 161. doi: 10.1177/0022146510395592 Tinetti, M. E., Fried, T. R., & Boyd, C. M. (2012). Designing health care for the most common chronic condition -m ultimorbidity. JAMA, 307 (23), 2493 2494. doi: 10.1001/jama.2012.5265

PAGE 132

122 U.S. Department of Health and Human Services. (2001). Mental health: Culture, race, and ethnicity a supplement to mental health: A report of the Surgeon General. Rockville, MD: U.S Depar tment of Health and Human Services, Substance Abuse and Mental Health Services Administration, Center for Mental Health Services. U.S. Department of Health and Human Services. (2013). Medicaid.gov: Keeping America Healthy, from http://www.medicaid.gov/ Upshur, C. C., Luckmann, R. S., & Savageau, J. A. (2006). Primary care provider concerns about management of chronic pain in community clinic populations. J Gen Intern Med, 21 (6), 652 655. doi: 10.1111/j.1525 1497.2006 .00412.x US Department of Health and Human Services. (1999). Mental Health: A Report of the Surgeon General. Rockville, MD: Substance Abuse and Mental Health Services Administration, Center for Mental Health Services. Ustun, T. B., Ayuso Mateos, J. L., Cha tterji, S., Mathers, C., & Murray, C. J. L. (2004). Global burden of depressive disorders in the year 2000. British Journal of Psychiatry (184), 386 392. doi: 10.1192/bjp.184.5.386 Valderas, J. M., Starfield, B., Sibbald, B., Salisbury, C., & Roland, M. (20 09). Defining comorbidity: implications for understanding health and health services. [Research Support, Non U.S. Gov't Review]. Ann Fam Med, 7 (4), 357 363. doi: 10.1370/afm.983 Van Dorsten, B., & Weisberg, J. N. (2011). Psychological Co morbidities in Pat ients with Pain. In S. Pagoto (Ed.), Psychological Co morbidities of Physical Illness: A Behavioral Medicine Perspective (pp. 275 310). New York: Springer Science + Business Media, LLC. van Melle, J. P., de Jonge, P., Spijkerman, T. A., Tijssen, J. G., Orm el, J., van Veldhuisen, D. J., et al. (2004). Prognostic association of depression following myocardial infarction with mortality and cardiovascular events: a meta analysis. Psychosom Med, 66 (6), 814 822. doi: 10.1097/01.psy.0000146294.82810.9c Van Puymbro eck, C. M., Zautra, A. J., & Harakas, P. P. (2006). Chronic pain and depression: twin burden of adaption. In A. Steptoe (Ed.), Depression and Physical Illness (pp. 145 164). Cambridge, UK: Cambridge University Press. Vera, M., Perez Pedrogo, C., Huertas, S E., Reyes Rabanillo, M. L., Juarbe, D., Huertas, A., et al. (2010). Collaborative care for depressed patients with chronic medical conditions: a randomized trial in Puerto Rico. Psychiatr Serv, 61 (2), 144 150. doi: 10.1176/appi.ps.61.2.144 Vogeli, C., Sh ields, A. E., Lee, T. A., Gibson, T. B., Marder, W. D., Weiss, K. B., et al. (2007). Multiple chronic conditions: prevalence, health consequences, and implications for quality, care management, and costs. J Gen Intern Med, 22 Suppl 3 391 395. doi: 10.1007 /s11606 007 0322 1 Von Korff, M., & Simon, G. (1996). The relationship between pain and depression. Br J Psychiatry, 168 101 108. Vyas, A., & Sambamoorthi, U. (2011). Multimorbidity and depression treatment General Hospital Psychiatry (33), 238 245. Wals h, J. L., Seen, T. E., & Carey, M. P. (2013). Longitudinal associations between health behaviors and mental health in low income adults. Translational Behavioral Medicine, 2013 (3), 104 113. doi: 10.1007/s13142 012 0189 5 Wang, P. S., Lane, M., Olfson, M., Pincus, H. A., Wells, K. B., & Kessler, R. C. (2005). Twelve month use of mental health services in the United States: results from the National Comorbidity Survey Replication. Arch Gen Psychiatry, 62 (6), 629 640. doi: 62/6/629 [pii] 10.1001/archpsyc.62.6. 629 Ward, B. W., Schiller, J. S., & Goodman, R. A. (2014). Multiple chronic conditions among US adults: a 2012 update. Prev Chronic Dis, 11 E62. doi: 10.5888/pcd11.130389

PAGE 133

123 Warshaw, G. (2006). Introduction: advances and challenges in care of older people wi th chronic illness. Generations, 30 (3), 5 10. Weihs, K., & Wert, J. M. (2011). A primary care focus on the treatment of patients with major depressive disorder. [Review]. Am J Med Sci, 342 (4), 324 330. doi: 10.1097/MAJ.0b013e318210ff56 Wilhelm, K., Mitche ll, P., Slade, T., Brownhill, S., & Andrews, G. (2003). Prevalence and correlates of DSM IV major depression in an Australian national survey. J Affect Disord, 75 (2), 155 162. doi: S016503270200040X [pii] Williams, J. W., Jr. (1998). Competing demands: Doe s care for depression fit in primary care? J Gen Intern Med, 13 (2), 137 139. Witko, K. D., Bernes, K. B., & Nixon, G. (2005). Care for psychological problems. Collaborative approach in primary care. Can Fam Physician, 51 799 801, 805 797. Woltman, H., F eldstain, A., MacKay, C., & Rocchi, M. (2012). An introduction to hierarchical linear modeling. Tutorials in Quantitative Methods of Psychology, 8 (1), 52 69. World Health Organization. (2001). World Health Report 2001: Mental Health: New Understanding, Ne w Hope. Geneva: World Health Organization, 2001. World Health Organization. (2013). Health Topics: Chronic Diseases. World Health Organization, & WONCA. (2008). Integrating Mental Health into Primary Care: A Global Perspective. Singapore: World Health Orga nization & World Organization of Family Doctors. World Health Organization (WHO). (2013). International Classification of Diseases (ICD), from http://www.who.int/classifications/icd/en/ Zhang, X., Norris, S. L., Gregg, E. W., Cheng, Y. J., Beckles, G., & Kahn, H. S. (2005). Depressive symptoms and mortality among persons with and without diabetes. Am J Epidemiol, 161 (7), 652 660. doi: 10.1093/aje/kwi089

PAGE 134

124 APPENDIX A. Glossary of Terms Use of terms in this report follow definitions from various peer reviewed articles, government and organization reports, and textbooks and were adopted to communicate to a broad audience about mental illness, primary care, and specific populations. Cita tions refer to the Adherence (medication/treatment) ( Haynes, 1979 ) Allostatsis/Allostatic Load Refers to an imbalance in systems that promote adaption, such as the autonomic nervous system and hypothalamo pituitary adrenal axis, to stress usually as a result of overexposure t o chronic or repeated stress which can cause physical damage (often manifesting as physical illness) or promote pathology. ( McEwen, 1998 ) Chronic Illness Chronic diseases are diseases of long duration and generall y slow progression. Chronic diseases, such as heart disease, stroke, cancer, chronic respiratory diseases and diabetes, are by far the leading cause of mortality in the world, representing 63% of all deaths. ( World Health Organization, 2013 ) Colorado Indigent Care Plan Provides funding to clinics and hospitals so that medical services can be provided at a discount to Colorado residents that meet the eligibility requirements. ( State of Col orado ) Competing Demands Factors considered by either the patient or provider that compete with each other for time on the agenda of the medical office visit, or potentially impede treatment adherence or prioritizing. These demands can include but are n ot limited to competing illness, values, cost, and reason for visit. ( Stange, et al., 1994 ) Complex Patient Patient s with multiple chronic conditions occurring simultaneously often resulting in decreased quality of life, longer hospital stays, psychological distress, and higher mortality. ( Fortin, Soubhi, et al., 2007 ) Cumulative Illness Rating Scale (CIRS) A tool used to measure morbidity tha t considers all medical problems encountered in primary

PAGE 135

125 care and allows for severity to be assessed based on a diagnostic manual. ( Hudon, Fortin, & Vanasse, 2005 ) D epression A common mental disorder, characterized by sadness, loss of interest or pleasure, feelings of guilt or low self worth, disturbed sleep or appetite, feelings of tiredness and poor concentration. ( World Health Organization, 2001 ) Diagnostic and Statistical Manual of Mental Disorders (DSM) The standard classification of mental disorders used by mental health professionals in the United States. ( American Psychiatric Association, 2000 ; American Psychiatric Association (APA), 2012 ) Health Disp arity Differences in the incidence, prevalence, mortality, and burden of diseases and other adverse health conditions that exist among specific population groups in the United States. ( National Institutes of Health (NIH), 2000 ) Hierarchical Linear Modeling St atistical modeling for social research that allows for individuals in a study to be classified or arranged in groups which themselves have qualities that influence the study, developed to allow for the study of relationships at any level in a single analys is, while not ignoring the variability associated with each level of the hierarchy. ( Scientific Software International (SSI), 2013 ) International Classification of Diseases (ICD) The standard diagnostic tool for epidemiology, health manageme nt and clinical purposes and is used to classify diseases and other health problems. ( World Health Organization (WHO), 2013 ) Medicaid A health and medical services program for individuals and families with low incomes and few resources. ( 2013 ) Medically Indigent The class of people who cannot afford necessary medical care from their own resources or from heal th insurance coverage, if any, including those individuals or families with medical expenses that exceed that of their income and assets. ( Bovbjerg & Kopit, 1986 ) Minority (race/ethnicity) Includes African Americans, Hispanics, Asians, Pacific Islanders, American India ns and any other race/ethnicity underrepresented usually resulting in economic and social implications. ( Pollard & O'Hare, 1999 ) Moderator (analysis) Factors that specify for whom and under what conditions a treatment works, helping to clarify the best choice for inclusion and exclusion criteria or the best choice of stratification, providi ng valuable information to inform future studies and interventions. These include sex,

PAGE 136

126 age, ethnicity, socioeconomic status, initial severity, genotype, biomarkers, or the subtype of the disorder under investigation. ( Kraemer, Frank, & Kupfer, 2006 ) Motivational Interviewing An evidence based practice that focuses on exploring and resolving ambivalence and centers on motivational processes within the individual that facilitate change. ( Emmons & Rollnick, 2001 ) Patient Health Questionnaire 2 and 9 Instruments for making criteria based diagnoses of depressive and other mental disorders commonly encountered in pri mary care. ( Kroenke, et al., 2001 ) Primary Care Primary care is that care provided by physicians specifically trained for and skilled in comprehensive first contact and continuing care for persons with any undiagnosed sign, sympto m, or health concern (the "undifferentiated" patient) not limited by problem origin (biological, behavioral, or social), organ system, or diagnosis. It includes health promotion, disease prevention, health maintenance, counseling, patient education, diagn osis and treatment of acute and chronic illnesses in a variety of health care settings (e.g., office, inpatient, critical care, long term care, home care, day care, etc.) and is performed and managed by a personal physician often collaborating with other h ealth professionals, and utilizing consultation or referral as appropriate. ( American Academy of Family Physicians, 2013 ) Primary Care Provider A primary care physician is a g eneralist physician who provides definitive care to the undifferentiated patient at the point of first contact and takes continuing responsibility for providing the patient's care. Such a physician must be specifically trained to provide primary care servi ces. Primary care physicians devote the majority of their practice to providing primary care services to a defined population of patients. The style of primary care practice is such that the personal primary care physician serves as the entry point for sub stantially all of the patient's medical and health care needs not limited by problem origin, organ system, or diagnosis. Primary care physicians are advocates for the patient in coordinating the use of the entire health care system to benefit the patient ( American Academy of Family Physicians, 2013 ) Self efficacy Self factors that predict m otivation. ( Bandura, 1977 ) Socially/Economically Disadvantaged Populations Individuals or groups including but not limited to those living in poverty, minority ethnic groups, mentally and physically disabled, older adults, and individuals dealing with addictions that are marginal ized and disenfranchised socially and politically. ( Allen, 1969 )

PAGE 137

127 Socioeconomic status (SES) Commonly conceptualized as the social standing or class of an individual or group; often measured as a c ombination of education, income, and occupation. ( American Psychological Association (APA), 2013 )

PAGE 138

128 B. Patient Health Questionnaire 2

PAGE 139

129 C. Patient Health Questionnaire 9

PAGE 140

130 D. Demographic Variables Questionnaire A. Do you consider yourself to be Hispanic or Latino? A. YES B. NO C. UNSURE D. PREFER NOT TO ANSWER If respondent is uncertain, read the following prompt: Where do y our ancestors come from? If respondent answers yes to any of the following, code as Hispanic. A. PUERTO RICO B. CUBA C. DOMINICAN REPUBLIC ( AMERICAN DOMINICAN) D. MEXICO (MEXICAN AMER ICAN) E. CENTRAL OR SOUTH AME RICAN COUNTRY F. OTHER LATIN AMERICAN COUNTRY G. OTHER HISPANIC OR LATINO CULTURE (S UCH AS SPAIN) H. I. PREFER NOT TO ANSWER B. What race do you consider yourself to be? Please select one or more, and choose all that apply. A. AMERICAN INDIAN OR ALASKAN NATIVE B. ASIAN C. BLACK OR AFRICAN AMERICAN D. NATIVE HAWAIIAN OR PACIFIC ISLANDER E. WHITE F. OTHER G. H. PREFER NOT TO ANSWER C. Are you: A. MARRIED B. DIVORCED C. WIDOWED D. SEPARATED E. NEVER MARRIED F. MEMBER OF AN UNMARRIED COUPLE G. PREFER NOT TO ANSWER D. What is the highest grade or year of school you completed? If respondent is uncertain, read the following prompt: A. NEVER ATTENDED SCHOOL or ONLY KINDERGARTEN B. GRADES 1 8 (Elementary) C. GRADES 9 11 (Some high school) D. GRADE 12 or GED (High school graduate) E. COLLEGE 1 YEAR TO 3 YEARS ( Some college or technical school)

PAGE 141

131 F. COLLEGE 4 YEARS or MORE (Colleg e Graduate) G. PREFER NOT TO ANSWER E. Are you currently: A. EMPLOYED FOR WAGES AT A BUSINESS THAT YOU DO NOT OWN B. SELF EMPLOYED/ BUSSINESS OWNER C. OUT OF WORK FOR MORE THAN A YEAR D. OUT OF WORK FOR LESS THAN A YEAR E. A HOMEMAKER F. A STUDENT G. RETIRED H. UNABLE TO WORK DUE T O ME NTAL OR PHYSICAL HEA LTH PROBLEMS I. UNABLE TO WORK DUE T O FAMILY OBLIGATIONS SUCH AS CHILDREARING OR CARE OF AGING PARENTS J. PREFER NOT TO ANSWER F. Is the annual income from all sources for this household: A. LESS THAN $10,000 B. LESS THAN $15,000 C. LESS THAN $20,000 D. LE SS THAN $25,000 E. LESS THAN $35,000 F. LESS THAN $50,000 G. LESS THAN $75,000 H. $75,000 OR MORE I. J. PREFER NOT TO ANSWER G. How many people, including yourself, depend on this income, even if they do not live with you? A. 1 B. 2 C. 3 D. 4 E. 5 F. 6 or more H. Are you Homeless, resi ding in a shelter or transitional housing program, or staying with friends or extended family? A. HOMELESS B. RESIDING IN A SHELTE R C. RESIDING IN A TRANSI TIONAL HOUSING PROGR AM D. STAYING WITH FRIENDS OR EXTENDED FAMILY ( IF YES, PROCEED TO QUESTION 9. IF NO, S KIP TO QUESTION 10). I. How many people currently live in your home? A. 1 B. 2 C. 3

PAGE 142

132 D. 4 E. 5 F. 6 or more J. If you have health insurance, does your health plan help cover the cost of prescription drugs? A. YES B. NO C. D. PREFER NOT TO ANSWER K. Was there a time in the past 12 months when you needed to see a doctor but could not? A. YES (IF YES, ASK FOR THE SPECIFIC REASON) B. NO C. D. PREFER NOT TO ANSWER L. Was there a time in the past 12 months when you needed to fill a prescription but could not fill it when you needed to? A. YES (IF YES, ASK FOR THE SPE CIFIC REASON) B. NO C. D. PREFER NOT TO ANSWER

PAGE 143

133 E. Cumulative Illness Rating Scale (CIRS)

PAGE 144

134 F. Qualitative Interview Guide Introduction script to be read after consent from the patient is given: Thank you very much for taki ng the time to talk with me. The purpose of this interview is to better understand how you communicate with your doctor about your illnesses, particularly feelings of stress and depression. When I use the term doctor, I mean whoever handles most of your h ealth care. Also I would like to ask you some questions about what other problems might get in the way of your treating your illnesses. I will tape record our interview and to ensure confidentiality, your name will not be on the tape. Is this okay with you ? I know I mentioned this already, but I want to remind you that you can stop this interview at any time and you do not need to answer every question. To get started, I am going to ask you a couple of background questions: Introductory What is your age? How long you been going to Denver Heath for care? How long have you been seeing your current doctor? I would like to know more about your experience as a patient here at Denver Health. 4. Can you tell me about your relationship with your doctor? PROBE: W hat do you like best? PROBE: What would you like to improve? 6. What illnesses you are currently dealing with? 7. Can you share with me which of those illnesses you mentioned worries you the most? 6A. Why that illness? 6B. What is your biggest worry? 8. How does having more than one illness impact your everyday life? PROBE: Does it make it more difficult? 8A. Can you describe some ways having more than illness affects your mental health?

PAGE 145

135 Patient Experience with Depression The purpose of the study you participated in was to help patients dealing with depressive symptoms. One of the purposes of this interview is to better understand how patients understand depression and how depression impacts a patient. 11. Earlier when I asked about you about you r current illnesses EITHER: b. you mentioned depression, Given your experience how does your depression affect your physical illnesses? 12. Describe to me what depression means to you? How do you recognize it in yourself? 12A. How would you recognize it in people you know? 13. Originally to be in the study, you had to show signs of depression. Looking back, why do you think you were identified as having depression? ED] 13A. How did you know? 13B. Where you surprised that you identified as having depression? 14. Had you been diagnosed with depression before? [IF YES] Was this time different? [IF YES OR NO] What did this new diagnosis mean to you? 15. You liste d (several) illnesses previously, how do you and your doctor determine which is most important to treat? 16. Have you ever talked to your doctor about depression? [IF YES] A. Who brings it up you or your doctor? B.Could you describe with as much detail as you remember a time when you talked with your doctor about your depression. 17. Earlier I asked you about which illnesses worried you. How did knowing you had signs of depression cha nge what you worry about? 18. Again going back to those illnesses you mentioned that most worry you, do you think those illnesses affect your overall health more than depression? 19. How do you feel about your doctor here at Denver Health treating your d epression?

PAGE 146

136 Depression impacts and lived experience descriptions 19. Tell me in what ways depression has affected your daily activities. 20. How about some ways being depressed has affected your health directly? PROBE: How often does it happen? PROBE: Can you remember the last time it happened? 21. Please describe an example of a time when your mood might have affected your physical health? 22. What barriers get in the way of you feeling better? PROBE: It could be during the visit or something experie nce personally or socially. 22A. Can you describe a specific time when one of those barriers kept you from doing something about your health? 23. Finally, between your physical illnesses and your mood (depression), which if treated would improve your o verall health most?

PAGE 147

137 E. Diagram of the relationships between levels of coding in the qualitative analysis

PAGE 148

138