A Patient Opinion Survey to Identify Perceived Barriers to the Introduction of a Screening Program for
Depression in a Hemodialysis Population
by
Farhat Farrokhi
A thesis submitted in conformity with the requirements for the degree of Masters of Science
Clinical Epidemiology and Health Care Research Institute of Health Policy, Management and Evaluation
University of Toronto
©Copyright by Farhat Farrokhi 2013
ii
A Patient Opinion Survey to Identify Perceived Barriers to the Introduction of a
Screening Program for Depression in a Hemodialysis Population
Farhat Farrokhi
Masters of Science, 2013
Institute of Health Policy, Management and Evaluation
University of Toronto
ABSTRACT
Patient-related barriers may reduce the effectiveness of screening for depression. This study
aimed to explore perceived barriers to participation in a Screening Program for Depression
by hemodialysis patients. In a cross-sectional study of hemodialysis patients, the Perceived
Barriers to Psychological Treatment questionnaire was used to measure barriers to the
Screening Program. Of 160 participants, 73.1% perceived at least one barrier (95% CI,
66.2% to 80.0%). The most common barriers were concerns about the side effects of
antidepressant medications (40%), concerns about having more medications (32%), feeling
that the problem is not severe enough (23%), and perceiving no risk of depression (23%). A
high depression score was an independent predictor of barriers related to perceiving no
benefit of the Screening Program and psychological, social, and practical barriers. We
believe that patient-related barriers need to be addressed before implementing any case
identification and treatment program for depression.
iii
ACKNOWLEDGEMENTS
First and foremost, I would like to sincerely thank my supervisor, Dr Alexander Logan, for his expertise and guidance throughout my thesis process. I would also like to express my debt of gratitude to Dr Vanita Jassal, my committee member, for her patience, guidance, encouragement, and mentorship during the 2 years I spent completing my Master’s thesis. I am also indebted to Dr Paul Kurdyak (my committee member) and Dr Heather Beanlands for their invaluable advice and feedback throughout the entire process.
In addition, I would like to extend my thanks to George Tomlinson for his biostatistical advice and Dr Sara Davison and Dr Elizabeth Lin for reviewing my thesis. I am also grateful for the help and support provided by Dr Joan Bargman, Dr Sagar Parikh, and Diane Watson. I would also like to acknowledge the Clinical Epidemiology Program, the Institute of Health Policy, Management and Evaluation, and the physicians and staff of the dialysis units at Toronto General Hospital and Sunnybrook Health Sciences Centre. Finally, I would like to thank my wife, Ghazal, for her unwavering love and support. Without her, I could not do any of this.
iv
TABLE OF CONTENT
THESIS OVERVIEW ---------------------------------------------------------------------------------------------------------- xi 1. BACKGROUND --------------------------------------------------------------------------------------------------------- 1 1.1. DEPRESSION --------------------------------------------------------------------------------------------------------- 1 1.2. CHRONIC KIDNEY DISEASE AND RENAL REPLACEMENT THERAPY-------------------------------------- 2 1.3. DEPRESSION IN PATIENTS WITH END-STAGE RENAL DISEASE ------------------------------------------ 3
1.3.1. Overview --------------------------------------------------------------------------------------------------- 3 1.3.2. Epidemiology ---------------------------------------------------------------------------------------------- 5 1.3.3. Depression and patient outcomes ------------------------------------------------------------------- 5 1.3.4. Previous work on the association between depression and all-cause mortality --------- 6
1.4. MANAGEMENT OF DEPRESSION IN PATIENTS WITH END-STAGE RENAL DISEASE ----------------- 7 1.4.1. Screening and diagnosis -------------------------------------------------------------------------------- 7 1.4.2. Treatment -------------------------------------------------------------------------------------------------- 8 1.4.3. Current challenges in management of depression in dialysis patients ---------------------- 9 1.4.4. Enhancement of care: screening program --------------------------------------------------------- 9
1.5. BARRIERS TO SCREENING FOR DEPRESSION --------------------------------------------------------------- 11 1.5.1. Sources of barriers ------------------------------------------------------------------------------------- 11 1.5.2. Literature review of patient-related barriers to mental health care ---------------------- 11
2. RATIONALE, OBJECTIVES, AND HYPOTHESES ---------------------------------------------------------------- 21 2.1. RATIONALE --------------------------------------------------------------------------------------------------------- 21 2.2. OBJECTIVES -------------------------------------------------------------------------------------------------------- 22 2.3. HYPOTHESES ------------------------------------------------------------------------------------------------------- 22 3. IDENTIFICATION OF BARRIERS AND MEASUREMENT TOOLS ------------------------------------------- 23 3.1. OVERVIEW AND PURPOSE ------------------------------------------------------------------------------------- 23 3.2. IDENTIFICATION OF POSSIBLE BARRIERS ------------------------------------------------------------------- 23
3.2.1. Conceptualizing participation in a screening program for depression -------------------- 24 3.2.2. Literature review for identification of barriers -------------------------------------------------- 29 3.2.3. Expert opinion on possible barriers ---------------------------------------------------------------- 29 3.2.4. Summarizing and categorizing barriers ----------------------------------------------------------- 29
3.3. CRITICAL REVIEW OF AVAILABLE BARRIER SCALES ------------------------------------------------------- 30 4. METHODS ------------------------------------------------------------------------------------------------------------- 40 4.1. STUDY DESIGN, SETTING, AND PARTICIPANTS ------------------------------------------------------------ 40 4.2. MAIN OUTCOME MEASURE ----------------------------------------------------------------------------------- 40
4.2.1. Definitions ------------------------------------------------------------------------------------------------ 40 4.2.2. Measurement tool ------------------------------------------------------------------------------------- 41 4.2.3. Adaptation of perceived barriers to psychological treatment ------------------------------ 41 4.2.4. Content validity ----------------------------------------------------------------------------------------- 42
v
4.2.5. Subscales ------------------------------------------------------------------------------------------------- 42 4.3. EXPLANATORY FACTORS ---------------------------------------------------------------------------------------- 43
4.3.1. Depression scale ---------------------------------------------------------------------------------------- 43 4.3.2. Covariates ------------------------------------------------------------------------------------------------ 43
4.4. RESEARCH PROCEDURE ----------------------------------------------------------------------------------------- 44 4.4.1. Informed consent -------------------------------------------------------------------------------------- 44 4.4.2. Non-participation--------------------------------------------------------------------------------------- 44 4.4.3. Study visits ----------------------------------------------------------------------------------------------- 44 4.4.4. Data management ------------------------------------------------------------------------------------- 44
4.5. SAMPLE SIZE ------------------------------------------------------------------------------------------------------- 45 4.6. FEASIBILITY --------------------------------------------------------------------------------------------------------- 45 4.7. DATA ANALYSIS --------------------------------------------------------------------------------------------------- 45
4.7.1. Psychometrics of the adapted PBPT --------------------------------------------------------------- 45 4.7.2. Descriptive analysis of patients with barriers (Objective 1) --------------------------------- 46 4.7.3. Descriptive analysis of barriers (Objective 2) ---------------------------------------------------- 46 4.7.4. Sensitivity analysis: comparison with the original barrier questionnaire ----------------- 46 4.7.5. Non-participants’ data -------------------------------------------------------------------------------- 47 4.7.6. Patient characteristics associated with barriers (Objective 3) ------------------------------ 47 4.7.7. Correction of significance level for multiple testing ------------------------------------------- 48 4.7.8. Analysis software --------------------------------------------------------------------------------------- 49
4.8. ETHICS --------------------------------------------------------------------------------------------------------------- 49 5. RESULTS---------------------------------------------------------------------------------------------------------------- 51 5.1. BASELINE DATA --------------------------------------------------------------------------------------------------- 51
5.1.1. Participation --------------------------------------------------------------------------------------------- 51 5.1.2. Baseline characteristics ------------------------------------------------------------------------------- 51 5.1.3. Depressive symptoms --------------------------------------------------------------------------------- 52 5.1.4. History of depression ---------------------------------------------------------------------------------- 52
5.2. RELIABILITY AND VALIDITY OF THE ADAPTED PBPT ------------------------------------------------------ 52 5.3. BARRIERS ----------------------------------------------------------------------------------------------------------- 53
5.3.1. Barriers scores and dichotomized results -------------------------------------------------------- 53 5.3.2. Sensitivity analysis: barriers using the PBPT without additional items -------------------- 53 5.3.3. Relationship between barriers and PHQ-2 scores ---------------------------------------------- 54
5.4. UNIVARIABLE ANALYSES ---------------------------------------------------------------------------------------- 54 5.5. MULTIVARIABLE ANALYSES ------------------------------------------------------------------------------------ 54 6. DISCUSSION----------------------------------------------------------------------------------------------------------- 81 6.1. COMMENTARY ---------------------------------------------------------------------------------------------------- 81
6.1.1. Summary of study results ---------------------------------------------------------------------------- 81 6.1.2. Proportion of patients with barriers to screening for depression -------------------------- 81 6.1.3. Barriers to screening for depression --------------------------------------------------------------- 82
vi
6.1.4. Correlation between depressive symptoms and barriers ------------------------------------ 83 6.1.5. Correlation between time on renal replacement therapy and barriers ------------------ 83 6.1.6. Correlation between socio-demographic factors and barriers ------------------------------ 84 6.1.7. Correlation between comorbidities and barriers ----------------------------------------------- 84
6.2. LIMITATIONS ------------------------------------------------------------------------------------------------------ 85 6.2.1. Study population --------------------------------------------------------------------------------------- 85 6.2.2. Measurement of primary outcome ---------------------------------------------------------------- 85 6.2.3. Measurement of covariates ------------------------------------------------------------------------- 88 6.2.4. Analysis of data ----------------------------------------------------------------------------------------- 89 6.2.5. Interpretation of data --------------------------------------------------------------------------------- 89
6.3. IMPLICATIONS ----------------------------------------------------------------------------------------------------- 90 6.4. FUTURE DIRECTIONS -------------------------------------------------------------------------------------------- 91 6.5. CONCLUSIONS ----------------------------------------------------------------------------------------------------- 91 7. REFERENCES ---------------------------------------------------------------------------------------------------------- 92 8. APPENDICES -------------------------------------------------------------------------------------------------------- 103
vii
LIST OF TABLES
Table 1.1. Depressive symptom scales validated in the CKD and dialysis patients ------------------------ 15 Table 3.1. Constructs of a conceptual framework for health care utilization ------------------------------- 32 Table 3.2. Possible barriers arising from each construct of mental health care utilization -------------- 33 Table 3.3. Studies on barriers to mental health care utilization ------------------------------------------------ 34 Table 3.4. Barriers to mental health care identified through review of literature, theory, and expert opinion -------------------------------------------------------------------------------------------------------------- 36 Table 3.5. Scales for measurement of perceived barriers to mental health care utilization ------------ 38 Table 4.1. Added items to PBPT ---------------------------------------------------------------------------------------- 50 Table 5.1. Baseline characteristics of participants and non-participants------------------------------------- 57 Table 5.2. Internal consistency of the modified PBPT subscale subscores ----------------------------------- 58 Table 5.3. Corrected item-total correlations (Pearson correlation coefficient) and internal consistency of subscales with each item when that item is omitted. ----------------------------------------- 59 Table 5.4. Median subscores and percentage of participants with barriers for each barrier subscale60 Table 5.5. Ten most frequently perceived barriers ---------------------------------------------------------------- 61 Table 5.6. Percentage of participants with barriers for each subscale based on the original PBPT --- 62 Table 5.7. Median barrier scores in patients with and without depressive symptoms ------------------ 63 Table 5.8. Ten most frequent barriers perceived by the patients with depressive symptoms --------- 64 Table 5.9. Univariable analysis of the association between the subscore of perceiving no threat and covariates -------------------------------------------------------------------------------------------------------------- 65 Table 5.10. Univariable analysis of the association between the subscore of perceiving no benefit and covariates -------------------------------------------------------------------------------------------------------------- 66 Table 5.11. Univariable analysis of the association between the subscore of psychological barriers and covariates -------------------------------------------------------------------------------------------------------------- 67 Table 5.12. Univariable analysis of the association between the subscore of social barriers and covariates -------------------------------------------------------------------------------------------------------------- 68 Table 5.13. Univariable analysis of the association between the subscore of practical barriers and covariates -------------------------------------------------------------------------------------------------------------- 69 Table 5.14. Multivariable regression model for the subscore of perceiving no benefit (log transformed) -------------------------------------------------------------------------------------------------------------- 70 Table 5.15. Multivariable regression model for the subscore of psychological barriers (log transformed) -------------------------------------------------------------------------------------------------------------- 71 Table 5.16. Multivariable regression model for the subscore of social barriers (log transformed) --- 72 Table 5.17. Multivariable regression model for the subscore of practical barriers (log transformed) 73
viii
LIST OF FIGURES
Figure 1.1. A conceptual model for interaction between depression and ESRD --------------------------- 16 Figure 1.2. Presence of depressive symptoms as a risk factor of mortality among dialysis patients (adjusted risk estimates using hazard ratios). ---------------------------------------------------------------------- 17 Figure 1.3. Depression scale score as a risk factor of mortality among dialysis patients (adjusted risk estimates using hazard ratios per score) ----------------------------------------------------------------------------- 18 Figure 1.4. Sources of barriers to a screening program for depression that may affect different stages of a screening program --------------------------------------------------------------------------------------------------- 19 Figure 1.5. Diagrams outline 2 studies on hemodialysis and peritoneal dialysis patients that illustrate some patient-perceived barriers to diagnosis and treatment of depression. ------------------------------- 20 Figure 3.1. Conceptual framework for participation in a screening program for depression based on the Health Belief Model, complemented by other social and behavioural models. ----------------------- 39 Figure 5.1. Recruitment of hemodialysis patients ----------------------------------------------------------------- 74 Figure 5.2. PHQ-2 score of the participants ------------------------------------------------------------------------- 75 Figure 5.3. History of diagnosis and treatment of depression -------------------------------------------------- 76 Figure 5.4. A: Diagrams on the left side demonstrate histograms of barrier subscores. B: Diagrams on the right side show the distribution of participants by the number of perceived barriers regarding each barrier construct. --------------------------------------------------------------------------------------------------- 77 Figure 5.5. Percentage of participants who perceived one barrier or more by PHQ-2 results ---------- 79 Figure 5.6. Most frequently perceived barriers to screening for depression among participants without depressive symptoms, as compared to those in participants with depressive symptoms --- 80
ix
LIST OF APPENDICES
Appendix A. The systematic review manuscript on the association of depression with mortality among dialysis patients ------------------------------------------------------------------------------------------------- 103 Appendix B. Models describing human behaviour for conceptualizing barriers to mental health care utilization ------------------------------------------------------------------------------------------------------------ 126 Appendix C. Modified Perceived Barriers to Psychological Treatment ------------------------------------- 134 Appendix D. Patient Health Questionnaire-2 --------------------------------------------------------------------- 142 Appendix E. Consent forms -------------------------------------------------------------------------------------------- 143 Appendix F. Sample Sizes and Confidence Intervals ------------------------------------------------------------- 152 Appendix G. Barriers perceived by the participants, sorted by prevalence ------------------------------- 153
x
LIST OF ABBREVIATIONS
BDI Beck Depression Inventory
BHMSS Barriers to Mental Health Services Scale
CESD Center for Epidemiological Studies Depression Scale
CI Confidence interval
CKD Chronic kidney disease
DSM-IV Diagnostic and Statistical Manual of Mental Disorders-version 4
ESRD End-stage renal disease
HBM Health belief model
MDD Major depressive disorder
PBPT Perceived Barriers to Psychological Treatment
PHQ-2 Patients Health Questionnaire-2
RRT Renal replacement therapy
SHSC Sunnybrook Health Sciences Centre
SPD Screening program for depression
UHN University Health Network
xi
THESIS OVERVIEW
Chronic physical illnesses are highly associated with depressive disorders. Patients with
chronic kidney disease are no exception. Non-specific complaints, such as fatigue, poor
sleep, lack of concentration, and anorexia are common amongst patients undergoing regular
dialysis treatments, yet few undergo evaluation for, or are given a formal diagnosis of
depression. Even fewer patients on dialysis are offered treatments such as counselling,
antidepressant medications or other psychological support.
Under ideal circumstances, patients needing mental health support would be identified and
treated through implementation of a dialysis unit-based program. Such a program (termed
Screening Program in this thesis) would offer screening, further assessment, and treatment
for depression where needed. In this thesis, I explore the potential patient-related barriers that
limit acceptability of such a Screening Program. Using a questionnaire-based cross-sectional
study, I have identified the most common barriers that may limit the effectiveness of a
dialysis unit-based Screening Program and determined the characteristics of those patients
most likely to refuse participation.
• Chapter 1 is a review of the literature on depression, the interactions between
depression and kidney disease, epidemiology of depression among end-stage renal
disease patients, and the current approaches and gaps in the diagnosis and treatment
of depression.
• Chapter 2 illustrates the rationale of this thesis project and the objectives and
hypotheses.
• Chapter 3 supports the thesis project by a summary of the possible barriers and the
available tools for their measurement. This chapter includes preliminary work to
identify possible barriers to the Screening Program. Patient-related barriers were
identified through a review of the literature and theories around the health care
utilization. Currently available scales for measurement of barriers to mental health
care were also critically appraised in this chapter.
• Chapter 4 describes the research design and analytic methods used.
• Chapters 5 and 6 present the results and the discussion.
1
1. BACKGROUND
The purpose of this chapter is:
• To describe, briefly, the clinical spectrum of disease associated with depressive symptoms and to define clinical depression in the general population
• To summarize the relationship between depressive symptoms and end-stage renal disease • To provide evidence for an association between depression and outcomes in the population
undergoing routine dialysis • To provide the rationale for why there may be reduced effectiveness if a Screening Program
was to be implemented across dialysis-units
1.1. DEPRESSION
Depression is a clinical spectrum of disease that is represented, amongst others, by feelings
of lack of interest (anhedonia), loss of energy, and low mood in association that affect an
individual’s quality of life. At one end of the spectrum, individuals experience mild
symptoms of low mood or disinterest, while at the other extreme, patients are diagnosed with
a major depressive disorder with or without suicidal intent.
The diagnostic nomenclature of major depressive disorder (MDD), minor depressive
disorder, and dysthymia helps guide the treatment.1 The Diagnostic and Statistical Manual of
Mental Disorders-version 4 (DSM-IV)2 defines an episode of MDD as the presence of at
least 5 of 9 depressive symptoms (depressed mood, anhedonia, sleep disturbance, appetite or
weight change, decreased energy, altered psychomotor activity, decreased concentration,
guilt or feelings of worthlessness, and suicidal ideation). If 2 to 4 depressive symptoms are
present, the depressive episode is called minor depressive disorder. To fulfil diagnostic
criteria for MDD and minor depressive disorder, patients must have at least one of the two
core symptoms (depressed mood and anhedonia) for at least 2 weeks that are severe enough
to cause an impairment in the individual’s social, occupational, or personal function.1
Dysthymia is a chronic course of milder depression (< 5 symptoms lasting for ≥ 2 years).2
2
Structured clinical interviews based on the DSM or the International Classification of
Diseases diagnostic criteria are used to confirm clinical diagnosis. Most experts, however,
agree that the disease spectrum is wide and the diagnosis is best made through clinical
assessment by trained individuals such as psychiatrists or clinical psychologists. The initial
step can be the use of self-report depression scales. Patients with scores higher than a
specified threshold should be further assessed through clinical interview to establish disease
severity, and the way symptoms have impacted the daily functioning of the patient.
Assessment should also rule out medical conditions as the cause of the symptoms.1
Based upon the World Health Organization’s Composite International Diagnostic Interview,
it has been estimated that approximately 8% of the general population in Canada suffered
from MDD in 2003.3 Another national report shows that the 2-year incidence of MDD has
been increasing from 2.9% in 2002-2003 to 7.2% in 2006-2007.4 The literature, in the
general population, suggests that depression, documented as either a clinical diagnosis of
depression or self-report depressive symptoms, is associated with a poorer quality of life,
increased risk of hospitalization, and increased mortality risk. Systematic review of study
data suggests an independent association between depression and mortality with an overall
1.8-fold increased risk of death in those with clinical depression.5 Canadian data suggest
similar associations, although a community-based study of Canadians demonstrating a 2-fold
higher risk of mortality among those with MDD, was unable to show an association after
adjustment for other mortality risk factors.6
Most studies suggest that individuals with chronic diseases are at a higher risk of
depression.4,7 In Canada, the incidence of depression is 1.5-fold higher in adults with any
chronic condition.4 The risk varies between chronic diseases, among which end-stage renal
disease (ESRD) is associated with the highest risk of depression.7 The association between
depression and poor outcomes has also been documented in patients with chronic conditions,
including end-stage renal disease.8-11
1.2. CHRONIC KIDNEY DISEASE AND RENAL REPLACEMENT THERAPY
Chronic kidney disease (CKD) is defined as either kidney damage or decreased kidney
function for at least 3 months.12 Numerous diseases may affect the kidney. Over time these
3
diseases follow a common final pathway with the end result being renal scarring and CKD.
In advanced stages, patients present with symptoms associated with fluid or toxin
accumulation and start onto renal replacement therapy (RRT). The term RRT refers to any
therapy that may be used to help clear toxins or fluid accumulation in individuals with
ESRD. Therapies include kidney transplantation, peritoneal dialysis, or hemodialysis.
According to the 2011 annual report of the Canadian Organ Replacement Registry, close to
37 744 Canadians were on RRT in 2009, and more than half of these patients were receiving
dialysis, predominantly hemodialysis.13
Hemodialysis may be performed in the home, in the hospital setting, or in community-based
specialized units. Within Canada, most patients undergo hemodialysis in hospital-based units
in the form of 3 treatment sessions lasting on average 4 hours per week. These patients are
closely followed by a team of doctors and nurses, and on average are seen by a health care
professional at least three times a week.
1.3. DEPRESSION IN PATIENTS WITH END-STAGE RENAL DISEASE
1.3.1. Overview
End-stage renal disease causes a significant change in the daily life of patients. The
knowledge that the disease has progressed to a stage where treatment is needed and the
treatment itself can make patients prone to depression.14 Depression in this population
appears to be secondary to several factors such as the loss of a primary role in family,
inability to continue working, decreased physical function, cognitive impairment, sexual
dysfunction, medication load, dietary restrictions, and tethering to the ‘lifesaving’ dialysis
machine.15-17 Somatic symptoms commonly seen in depression are also commonly attributed
to uremia and dialysis treatment.18,19
Katon proposed a conceptual model of the interaction between chronic medical illnesses and
depression as an independent clinical entity.20
4
Figure 1.1 illustrates the theoretical model adapted for ESRD. This model shows that in
addition to underlying risk factors seen in the general population (genetic characteristics,
personality, life events, etc), depression can be the result of the burden and consequences of a
disease. The model describes three levels of interactions between depression and chronic
illnesses (such as ESRD):
Level 1 - Role of depression in progression to ESRD: Depression can influence the course of
CKD. Patients with depression are more likely to progress to ESRD and are more likely to
initiate dialysis with higher glomerular filtration rates than non-depressed patients.21,22 One
of the explanations is that patients with depression are more likely to have a poor diet and to
be non-adherent to the treatments used to delay chronic progression.23-31 Factors associated
with depression such as inflammation,32,33 compromised immunity,34-36 lack of social
support,37 low threshold to physical symptoms,38 and altered perception of physical
symptoms39 may accelerate the progression of CKD.
Level 2 – Direct interaction of depression and ESRD consequences: Once ESRD has
developed, the physical and psychosocial burden associated with the disease can itself lead to
development of depression. Depression can also exacerbate the consequences of ESRD
through its impact on physical function, quality of life, and perceived burden of physical
symptoms.40,41 Depression is associated with biological complications that are also seen in
ESRD patients.33-36 Impaired immune system and elevated inflammatory markers are
recognized biological complications of both conditions. The link between depression and
poor immune system has been linked to the cytokine-induced dysregulation of the
hypothalamic-pituitary-adrenal axis.42 This neurohormonal imbalance results in high levels
of glucocorticosteriods.43 Dysregulation of cytokines in depressed patients also reduces
levels of 5-hydroxytryptamine, an essential component of neurocellular function.44 Increased
depressive symptoms have also been associated with the production of inflammatory
cytokines including interleukin-6 and C-reactive protein.45
Level 3 – Indirect effect of depression on ESRD consequences: Depression has a direct
impact on the ESRD patient’s self-care. Poor adherence to dialysis prescription, medications,
5
and diet limitations are expected to be common among depressed patients, and these
behavioural problems can exacerbate the consequences of ESRD.24,31
1.3.2. Epidemiology
Depression is more than twice as common in adults with chronic physical health problems as
in otherwise healthy individuals.7 Compared to other chronic conditions, ESRD is associated
with the highest rate of depression. The odds of depression is 3.56 times higher in ESRD
patients as compared to healthy adults, while this rate ranges from 1.96 to 3.21 in patients
with diabetes mellitus, chronic heart failure, hypertension, coronary artery disease,
cerebrovascular accident, and chronic obstructive pulmonary disease, respectively.7
According to self-administered questionnaires, up to half of the patients on dialysis suffer
from mild to severe depressive symptoms.46 In a recent report from Canada, Wilson et al
reported that 39% of 124 patients on hemodialysis in London, Ontario, screened positive for
depressive symptoms.47 In our systematic review (submitted for publication, December 2012)
of 14 studies across the world reporting the association of depression and mortality,11 we
found that self-reported depressive symptoms were present in 29.7% of 21 146 dialysis
patients (8.1% to 65.4%).18,48-60 A large variation in the estimates was noted, and attributed
to the use of a number of different measurement tools and cutoff criteria; but nevertheless
still speak to the significance of the problem.46
Data estimating the number of patients undergoing chronic dialysis with a current or previous
clinical diagnosis of depression are limited. Studies using structured clinical interviews for
diagnosis of depression reported an overall prevalence of 12% to 26% for depressive
disorders,61-63 and a prevalence of 17% to 21% for MDD.62-64 A documented clinical
diagnosis of depression, however, is usually less frequent (4.4% to 27.7%).18,55,65-70
1.3.3. Depression and patient outcomes
Depression is an independent predictor of poor outcomes in dialysis patients.40,41,49,71
Depression has been shown to be linked with poor health-related quality of life in ESRD
patients, affecting both physical and mental health function.40,41,72 Some recent longitudinal
studies have confirmed the association of depressive symptoms scores with increased
6
hospital admissions and mortality.10,13-15 Self-reported depression was associated with an
increased risk of mortality and hospitalization in hemodialysis patients followed in the multi-
national Dialysis Outcomes and Practice Patterns Study.16 In another prospective study,
physician-diagnosed depression was shown to increase the risk of death and hospitalization
by two fold.17 Furthermore, the presence of depression may be associated with an increased
risk of death in dialysis patients through the higher chance of patients attempting suicide or
requesting withdrawal from dialysis.54,73,74 Lacson et al demonstrated that each unit increase
in the depressive symptoms score of patients on hemodialysis was associated with 19%
higher risk of dialysis withdrawal (hazard ratio, 1.19; 95% confidence interval [CI], 1.08 to
1.31).54
1.3.4. Previous work on the association between depression and all-cause mortality
Data on the relationship between depression and mortality among dialysis patients are
conflicting. While earlier studies failed to document the link,75-77 some more recent
longitudinal studies have shown significant associations.49,78 We systematically reviewed
studies on the relationship between depressive symptoms or depression and mortality among
dialysis patients (full manuscript in Appendix A).11 Reviewing 2528 records retrieved
through systematic search of the MEDLINE, EMBASE, PsychINFO, and Proquest, we
identified 63 relevant publications. Based on well-defined inclusion criteria, 31 eligible
publications were selected for systematic review, 25 of which provided enough data for
meta-analysis. The studies were categorized according to their methods of depression
measurement: self-reported depressive symptoms dichotomized as presence or absence of the
symptoms, self-reported depressive symptoms scores reported as a continuum, and
physician-diagnosed depression.
Meta-analysis of the data showed a significant association between the presence of
depressive symptoms and mortality (12 studies; n = 21055; mean age, 57.6 years; males,
53%; hemodialysis, 99%; Figure 1.2). Adjusting for potential publication bias, it was shown
that the presence of depressive symptoms increased the risk of death by 45% (hazard ratio,
1.45; 95% CI, 1.27 to 1.65). Combining across 6 studies reporting adjusted hazard ratios for
depressive scores as a continuous variable (n = 7857; mean age, 61.3 years; males, 53%;
7
hemodialysis, 98%), each unit change in depressive scores was significantly associated with
mortality (adjusted hazard ratio, 1.04; 95% CI, 1.01 to 1.06; P = .002; Figure 1.3). This
effect size, however, was based on heterogeneous results (I2 = 74%).
Subgroup and sensitivity analyses were done to further explore the impact of the
heterogeneity between the included studies using depressive symptoms scales. The effect
sizes were significant and suggested a positive association between depressive symptoms and
mortality in all subgroup of studies. These included studies divided by follow-up duration,
incident versus prevalent dialysis patients, US-based versus non-US–based data, and single
baseline versus repeated measurements of depression. The exclusion of studies with small
sample size (n < 100) or those with a high risk of bias did not significantly change the results.
The meta-analysis results suggested the presence of an independent association between
depression and mortality risk among patients on chronic maintenance dialysis and perhaps
emphasize a need for further study looking at the effectiveness of strategies to diagnose and
treat depression in the dialysis population.
1.4. MANAGEMENT OF DEPRESSION IN PATIENTS WITH END-STAGE RENAL
DISEASE
1.4.1. Screening and diagnosis
As an initial step, the identification of high-risk patients using screening tools is
recommended across dialysis units. Several screening tools have been used in the CKD and
ESRD populations. These include the Beck Depression Inventory (BDI), Centre for
Epidemiological Studies Depression Scale (CESD), Mental Health Index (5 items of the
Short Form-36), Hospital Anxiety and Depression Scale, Quick Inventory of Depressive
Symptomatology Self-Report, Patient Health Questionnaire-9 , Geriatric Depression Scale,
and 2 items of the Short Form-36.18,48,49,55,57,62-64 A few have been validated in CKD and
dialysis patients although in relatively small cohorts. Currently validated tools are the BDI,
CESD, Geriatric Depression Scale, Quick Inventory of Depressive Symptomatology Self-
Report, and Patient Health Questionnaire-9 (Table 1.1). Many validation studies suggest an
8
adjustment is required to set a more appropriate cutoff score for ESRD patients. This
adjustment is justified by the high frequency of somatic symptoms among ESRD patients.
Although the diagnostic accuracy of these screening tools seem good, the positive predictive
values are relatively low, and despite the high specificity, this raises the concern that there
will be a large number of individuals who screen positive who do not have depression (false-
positive results).
There is concern that somatic symptoms arising from uremia may cause a bias in the
depression scale results.42,79 One solution is to avoid using scales which include questions
relating to somatic symptoms. The use of scales with non-somatic symptom items, however,
is not supported by the evidence. Hedayati et al studied the validity of the Cognitive
Depression Index (a subscale of the BDI that excludes somatic items) by comparing the
results with a structured clinical interview in a group of patients on hemodialysis.62 The
diagnostic accuracy of the Cognitive Depression Index was significantly lower than that of
the BDI. Chilcot et al also confirmed the superiority of the whole BDI over the cognitive
subscale.80
Although the methods used to diagnose depression amongst ESRD patients are similar to
those used in the general population,15 patients with chronic illnesses are more likely to
somatise mental symptoms.81 To overcome this, some authors suggest placing more
emphasis on the presence of depressed mood and loss of interest (the core symptoms of
depression).79 However, it is common among patients with chronic illnesses to report only
somatic complaints as their presenting symptoms of depression.81 An alternative solution is
to consider depression when somatic symptoms appear to be out of proportion to the physical
health status.39,82 Assessment of the severity of the symptoms, rather than simply counting
the number of symptoms,81 and paying attention to the changes in physical symptoms
reported by family and caregivers is recommended.15
1.4.2. Treatment
Both pharmacological and non-pharmacological treatments are beneficial in CKD and
dialysis patients. Data on pharmacological treatment of depression in ESRD patients are
limited to small studies, some of which have methodological limitations.83-94 Safety and
9
effectiveness data are sparse. Fluoxetine has been studied in two clinical trials of 6 and 7
patients and found to be effective in reducing symptoms.84,90 Newer selective serotonin
reuptake inhibitors, including citalopram, sertraline, fluvoxamine, and paroxetine have also
been studied in observational studies and small clinical trials,83,85,87,89,91-94 and are reported to
be more favourable than fluoxetine because of fewer side effects and drug-drug interactions.
All of these studies have reported considerable success; however, lack of enough evidence
about drug-drug interactions and dose adjustments for dialysis patients are ongoing concerns
for many physicians.
Non-pharmacological therapies have also been shown to be successful in small series’. In one
recent randomized controlled trial, cognitive behavioral therapy given over 3 months
significantly improved depressive symptom scores.95 Another randomized controlled trial on
intradialytic exercise training reported promising results.96 The effectiveness of these
therapeutic options continues to be further investigated.
1.4.3.Current challenges in management of depression in dialysis patients
Published data suggest that there is a wide gap between the prevalence of depression and
those who receive treatment. In 2004, Lopes et al found that 43% of 9382 dialysis patients
from 12 countries had depressive symptoms suggestive of clinical depression. In contrast, the
prevalence of physician-diagnosed depression was 14%.55 Hedayati et al found depression in
27% of their hemodialysis patients using structured clinical interview of the entire cohort. Of
these individuals only 54% of them were known to have depression.71 In addition, data show
that many depressed patients on dialysis do not receive appropriate treatment. According to
the reports from the United States, only 16% to 42% of depressed patients on dialysis receive
treatment,62,97 and 45% of patients on antidepressant medications are on minimum doses.62
1.4.4. Enhancement of care: screening program
In order to improve the diagnosis and treatment of dialysis patients with depression, several
authors have proposed an active approach where patients undergo routine assessment.15,16,79
Hedayati et al have suggested screening at the initiation of dialysis, after 6 months, and
yearly thereafter.79 Screening is generally defined as “the systematic application of a test or
10
enquiry, to identify individuals at sufficient risk of a specific disorder to benefit from further
investigation or direct preventive action, amongst persons who have not sought medical
attention on account of symptoms of that disorder.98” Screening is most likely to be
successful if: (1) there is a high prevalence of the problem, (2) the problem is an important
health issue, (3) there is a high undetected rate of individuals with the problem if not
screened, (4) accurate screening tool is available, and (5) effective treatment is available.99
Screening for depression has been recommended in primary care setting. The Canadian Task
Force on Preventive Health Care recommends screening for depression only where there is a
system to follow up and provide care should referral and treatment be necessary (grade B
recommendation).100 Although endorsed also by the US Preventive Services Task Force,101
the recommendation has received criticism.99 Concerns are mainly about the accuracy of
screening tools, especially the likelihood of false positive results, identification of mild
depressive disorders that do not need intervention, lack of evidence showing the role of
screening in long-term improvement of outcomes, and diversion of scarce resources from
other more endeavours such as ensuring better care of those already diagnosed with
depression.99,102 Because of these concerns, the UK National Institute for Health and Clinical
Excellence recommended case identification only in those with a history of depression and
those with chronic health problems or functional impairment.102
Patients undergoing dialysis would likely benefit from screening for depression. Depression
is prevalent among these patients and is independently associated with outcomes, but it
largely remains undetected. On the other hand, patients undergoing dialysis are frequently in
contact with health care facilities and seen by their primary care physicians. Nonetheless,
despite its seemingly practical benefits, there is no evidence to support its effectiveness in
short-term and long-term. The effectiveness of screening for depression lies not just in the
number of patients identified, but in the number of depressed patients who would not
otherwise be identified and in the number of those who eventually receive treatment.103 Baas
et al have shown that screening for depression leads to treatment in a very small proportion
(1%) of some high-risk groups at the community level.104 There are similar concerns that
even in the medical settings in which screening is justifiable, a screening program might be
unsuccessful.103,104
11
1.5. BARRIERS TO SCREENING FOR DEPRESSION
1.5.1. Sources of barriers
The National Institute for Health and Clinical Excellence guidelines summarize barriers to
diagnosis and treatment of depression into 4 categories: patient-related, physician-related,
organizational, and societal factors.102 The 4 categories of barriers may be seen across all
steps of the screening procedure: screening assessment, diagnosis and referral processes, and
treatment (Figure 1.4). To be successful, a screening program needs to be acceptable to the
health care involved in patients care. Patients’ perception of depression and its treatment,
together with social and psychological issues such as stigma, can act as barriers to accepting
care for depression. Physicians may not recognize symptoms of the patients as depression
symptoms or may be reluctant to start treatment, especially in the setting of chronic medical
illnesses.105 External factors include organisational and societal barriers such as limitations of
case-identification tools and the effectiveness of the diagnosis and treatment options (as
discussed before), lack of access to mental health care, and deficiencies in the referral
system, as well as societal and cultural barriers such as poor economic status and
stigma.102,106
1.5.2. Literature review of patient-related barriers to mental health care
In the primary care setting, patient-related barriers may make the management of clinical
depression more challenging104,107,108,109 Although patient-related barriers have been studied
in the general population and a small number of clinical settings, there is limited information
on how best to systematically measure these barriers. In the ESRD population, the data is
even more limited, with only a small number of studies on this topic.110,111
Many patients with depression or other mental health disorders are hesitant to seek help or
accept treatment.106,108-128 Sareen et al122 surveyed a sample of Ontarians in 2007 and
reported that 6.5% of 6261 respondents perceived the need for treatment of a mental health
12
problem within the past year (mainly mood and anxiety disorders) without seeking help.
Similarly, data from the Canadian Community Health Survey–Mental Health and Well-being
demonstrated that 21% of individuals with mental or substance use disorders perceived the
need for help but did not seek help.123 Mohr et al108 used a questionnaire to assess barriers to
psychotherapy in the primary care setting and found that 55% of the respondents perceived
at least one substantial barrier to psychotherapy. To the best of our knowledge, a report by
Baas et al is the only study in the primary care setting that has quantified the effectiveness of
a screening program.104 They invited 1687 individuals at a high risk of depression, of whom
less than half (n = 780) participated. Screening documented presence of depressive
symptoms in 226 participants suggestive of depression. Twenty patients refused further
assessments, and of the 173 individual who eventually participated, 71 were diagnosed with
depression. Thirty-six patients were already receiving treatment, 14 refused treatment and 4
did not show up for an appointment. Overall, only 1% of the patients at risk started treatment
of depression, suggesting that the screening was not successful in identifying all individuals
who most likely needed follow-up and treatment.104
Canadian data shows that the most common reasons of patients who do not seek help for
mental health problems are around the theme of not feeling to have a problem or not feeling
that their problem is serious. Specific barriers to seeking help included: “I wanted to solve
the problem on my own” (47%), “I thought the problem would get better by itself” (41%),
and “The problem went away by itself, and I did not really need help” (22%). Other less
common reasons were time constraints or inconvenience of seeking help, not knowing where
to go for help, and feeling that help would not be effective.122 Perceived barriers were similar
in the Canadian Community Health Survey–Mental Health and Well-being; of a list of 13
possible barriers, preference to manage the problem by oneself, time constraint, and not
taking the problem seriously were the most common perceived barriers. In addition, stigma
was a concern, albeit not as common as other barriers; 4% of the respondents reported being
afraid to ask for help or “what others would think.123”
Lack of perception of the seriousness of a mental health disorder and the benefit of mental
health treatment has also been shown in some other large-scale studies. Mojtabai et al109
examined barriers to initiation of treatment in the National Comorbidity Survey Replication.
13
They demonstrated that among 1350 US patients with mental health disorders, 72.6%
perceived the desire to handle the problem on their own.109 In the survey by Mohr et al,108
other than some practical barriers to psychotherapy such as costs and availability of
counseling services, barriers related to misfit of therapy to needs were the most common
responses (eg, feeling that the problem is not bad enough).108
Studies focusing on the treatment of depression demonstrate similar themes of
barriers.106,117,118,121 Nutting et al106 surveyed physicians involved in diagnosis and treatment
for depression. Doctors reported that the patients underestimated the seriousness of the
problem or disagreed with the diagnosis, placed more importance on treating other medical
conditions, and were unwilling to consider counseling or not keen to start medication.106 A
qualitative study of men with a history of depression identified a series of themes as potential
barriers to diagnosis and treatment of depression, some of which were perceiving depression
as ‘normative to male role,’ not feeling to be depressed, preferring to solve the problem on
one’s own, perceiving incompetence of the physician, and becoming frustrated towards drug
treatment.121 In addition, another qualitative study of a group of men and women highlighted
concerns of the participants about competence, openness, and trustfulness of primary care
physicians.118
Barriers may be slightly different in older age groups.115,120 Older adults are less likely to
voluntarily report depressive symptoms, may perceive depression as a character flaw, and
may be more likely to attribute their symptoms to physical conditions. In addition, they
perceive their physical impairment as a practical barrier to access to therapy.115 In a
qualitative study of African-American older adults, the stigma of being diagnosed with
depression, lack of faith in treatment, lack of access to treatment, mistrust, ageism, and
disagreement with diagnosis of depression were noted.114
Studies looking populations with specific health conditions identify barriers related to the
medical condition as well as those mentioned above. For example, in a survey of pregnant
women asked about perinatal depression, more than 90% indicated that they would
participate in therapy if needed, but only 35% were willing to take medications during their
pregnancy.117 In physically ill patients, we anticipate concerns about the side effects of
14
antidepressant medications to be of the major barriers. In addition, both physician and
patients are likely to prioritize physical problems and feel depressive symptoms are “normal”
reaction to physical distress.116,125 Seventy-one percent of Japanese patients with lung cancer
believed that “emotional burden cannot be relieved by medication,” 56% had concerns about
the use of “medicines that act on the mind.” Patients were also found to be very concerned
about the effectiveness and side effects of psychiatric medications.116
Data on barriers to mental health care in dialysis patients are limited to two studies (Figure
1.5). Wuerth et al reported their experience with establishing a screening program in a
peritoneal dialysis centre.111 They showed that 49% of patients who screened positive for
depressive symptoms were unwilling to accept further assessment. Many patients did not
agree with the screening results (88%), while of those who accepted further assessment and
were diagnosed with depression, 74% were willing to take medications.111 Johnson and
Dwyer110 surveyed screen-positive dialysis patients for 14 possible barriers to treatment of
depression and reported that 71% had at least 1 barrier. Interestingly, the most common
barriers were “I do not feel anxious or depressed” (perceived by 41%) and “I am anxious or
depressed but I do not need help” (perceived by 16%). Of note, concerns about benefits and
harms of medications were not addressed in this study.
In summary, patient-perceived barriers include a poor understanding of the risk and
seriousness of the disease, disagreement with the diagnosis of mental disorders, concerns
about benefits and harms of the available treatment options, and several psychological,
social, and practical barriers. In this thesis, I propose to study these barriers using systematic
methods for identification of the barriers.
15
Table 1.1. Depressive symptom scales validated in the CKD and dialysis patients*
Scale/Study Year Sample Size Population Cutoff Sen, Spe PPV, NPV
BDI (21 items)
Balogun et al48 2010 62 HD patients ≥ 65 years ≥ 10 68%, 77% 57%, 85%
Hedayati et al64 2009 272 CKD patients (stages 2-4) ≥ 11 89%, 88% 67%, 97%
Hedayati et al62 2006 98 HD patients ≥ 14 62%, 81% 53%, 85%
Watnick et al63 2005 62 HD and PD patients ≥ 16 91%, 86% 59%, 98%
Craven et al61 1988 99 HD and PD patients ≥ 15 92%, 80% 39%, 99%
CESD (20 items)
Hedayati et al62 2006 98 HD patients ≥ 18 69%, 83% 60%, 88%
PHQ (9 items)
Watnick et al63 2005 62 HD and PD patients ≥ 10 92%, 92% 71%, 98%
GDS (15 items)
Balogun et al48 2010 62 HD patients ≥ 65 years ≥ 5 63%, 82% 60%, 83%
QIDS-SR (16 items)
Hedayati et al64 2009 272 CKD patients (stages 2-4) ≥ 10 91%, 88% 67%, 97%
*Sen indicates sensitivity; Spe, specificity; PPV, positive predictive value; NPV, negative predictive value; BDI beck Depression Inventory; CESD, Centre for Epidemiological Studies Depression scale; PHQ, Patient Health Questionnaire; GDS, Geriatric Depression Scale; QIDS-SR, 16-item Quick Inventory of Depressive Symptomatology Self-Report Scale; HD, hemodialysis; and PD, peritoneal dialysis.
16
Figure 1.1. A conceptual model for interaction between depression and ESRD. Three levels are recognised for the interaction between depression and ESRD. Level 1 shows the role of depression in progression to ESRD; level 2 illustrates the direct interaction of depression with the physical and mental health of and ESRD patient; and level 3 describes the indirect effect of depression on the symptoms through poor self-care of ESRD (adapted with permission from the model proposed by Katon20).
17
Figure 1.2. Presence of depressive symptoms as a risk factor of mortality among dialysis patients (adjusted risk estimates using hazard ratios).11
18
Figure 1.3. Depression scale score as a risk factor of mortality among dialysis patients (adjusted risk estimates using hazard ratios per score).11
19
Figure 1.4. Sources of barriers to a screening program for depression that may affect different stages of a screening program.
20
Figure 1.5. Diagrams outline 2 studies on hemodialysis and peritoneal dialysis patients that illustrate some patient-perceived barriers to diagnosis and treatment of depression (shown in dotted boxes).110,111
21
2. RATIONALE, OBJECTIVES, AND HYPOTHESES
The purpose of this chapter is:
• To summarise the rationale behind developing this project • To describe the objectives and hypotheses of the research project
2.1. RATIONALE
The high prevalence of depression in the dialysis population and the observed association
with poor quality of life, hospitalization, and mortality warrant active intervention. The first
step is to identify patients at risk through a process of screening. Since dialysis patients are in
frequent contact with health care services and they are at a high risk of depression, screening
for depression seems to be a practical and effective approach.15,46,72 Guidelines on screening
for depression support the idea by recommending screening high-risk groups when
appropriate follow-up and treatment is possible.100,102
However, the effectiveness of a screening program in these patients is yet to be proven,15 as
several patient-related, physician-related, and organizational factors may influence each stage
of a screening program.102 Patient-related barriers to management of depression are deemed
to be the primary challenge.107 Limited data in the dialysis literature shows that a
considerable proportion of dialysis patients with depressive symptoms are not willing to
undergo diagnostic assessments, seemingly because of not perceiving themselves as
susceptible to becoming depressed and having serious concerns about antidepressant
medications.110,111
I believe that a better understanding of dialysis patients’ perceived barriers to screening,
diagnosis, and treatment of depression is the crucial primary step. This may inform how a
screening program is implemented for depression in the dialysis units.
22
Note: In the following sections, the term “Screening Program for Depression (SPD),” is used
to mean a program incorporating routine questionnaire assessments of depressive symptoms,
referral, and treatment.
2.2. OBJECTIVES
a) To determine the proportion of patients on chronic hemodialysis who perceive one or
more barriers to participation in an SPD
b) To determine the most common perceived barriers to participation in an SPD
c) To determine the clinical characteristics of hemodialysis patients who perceive
barriers to participation in an SPD
2.3. HYPOTHESES
a) More than 50% of dialysis patients perceive barriers to participation in an SPD.
b) The most common perceived barriers by patients are
• failing to recognize the risk of being or becoming depressed, and
• concerns about being prescribed antidepressant medications.
c) Patients with depressive symptoms are more likely to perceive barriers to
participation an SPD as compared with those without depressive symptoms.
23
3. IDENTIFICATION OF BARRIERS AND MEASUREMENT TOOLS
The purpose of this chapter is:
• To describe sources of possible barriers to screening for depression • To conceptualize participation in an SPD and hypothesize barriers constructs and possible
barriers related to each • To summarise barriers introduced and studied in the literature • To identify and appraise tools for assessment of barriers to screening for depression
3.1. OVERVIEW AND PURPOSE
There are several studies measuring patient-related barriers to mental health care
utilization.106,108-128 However, very few have used systematically developed and validated
tools for measurement of barriers.108,116,120 Furthermore, there are no studies specifically on
barriers to screening for depression. This chapter describes the comprehensive methodology
used to identify and categorize all possible patient-related barriers to screening for depression
and critically reviews currently available scales. This preliminary work supports the study
design used later in this thesis by (1) identifying all potential barriers related to the dialysis
patients that are required to be measured, and (2) critically reviewing the available barrier
scales.
3.2. IDENTIFICATION OF POSSIBLE BARRIERS
Theory, literature, and expert opinion were used to identify barriers to an SPD that would be
relevant in an ESRD population.129 Patients’ perceived barriers to mental health care
utilization were identified from the literature. A conceptual framework was developed and
additional possible barriers proposed by a group of experts were added. This resulted in a
comprehensive list of potential barriers to participation in an SPD.
24
3.2.1. Conceptualizing participation in a screening program for depression
An essential basis for understanding barriers to mental health care is a conceptual framework
to explain how patients decide on whether to participate in a health care program. This
framework allows exploration of the concept of barriers, helps determine hypothetical
constructs, and allows possible barriers to be categorized into hypothetical constructs.
Several models have been developed to explain human behaviour and used as a guide for
health care research.130-132 These models and their relationship with the concept of barriers
are briefly described in Appendix B.130-138 I decided to base my conceptual framework on
Health Belief Model (HBM) because it appropriately recognizes the interplay between
patients, the disease, and the intervention of interest. This model has been revisited by
Henshaw and Freedman-Doan131 for the context of mental health care utilization.
Additionally, having been primarily proposed to explain preventive health behaviour138 HBM
suggests the most appropriate conceptual framework to explain participation of individuals at
risk of depression in an SPD. To ensure all components in the pathway to making decisions
by health care users were addressed in the conceptual framework, HBM was complemented
by other proposed models, including self-regulatory model,137 theory of planned
behaviour,133 help-seeking model,136 sociobehavioural model,134 and social cognitive
theory.130
Figure 3.1 demonstrates the conceptual framework for patient’s decision process about
participation in an SPD. Briefly, the HBM hypothesizes that individuals are likely to engage
in a health-related behaviour if they believe that (1) there is a real risk of contracting an
illness and the disease is serious in terms of its medical and non-medical consequences, (2)
the health behaviour of interest is beneficial in reducing the threat of the health condition and
there is no perceived negative consequence of the action, and (3) there is no barriers to take
the action.131 These three constructs are named as perceived threat, perceived benefit, and
perceived barrier, respectively. Different aspects of perception of threat are described in the
literature as perceived susceptibility and severity (HBM model),131 illness representation
(self-regulatory model),137 and incentive (theory of planned behaviour).133 Perceived barriers
can be psychological, social, and practical factors. These are factors that make one unwilling
25
to take part in the SPD even if they perceive the threat of depression and believe in the
benefits of the SPD.131 Different aspects of psychological barriers such as self-concealment,
perceived control, and self-efficacy are elaborated in the help-seeking model and theory of
planned behaviour.133,136 Social barriers include stigma and lack of social support, which
have been explained by the sociobehavioural model, help-seeking model, and the theory of
planned behaviour.133,134,136 All these variables are affected by internal predisposing factors,
external predisposing factors, and satisfaction with previous experiences.
Some authors suggest that the perceived barriers construct in the HBM be further categorized
into psychological and practical barriers.131 Furthermore, since social barriers such as stigma
and lack of social support are addressed as major components in other theories and models,
the distinction between psychological factors and social factors may allow for a better
understanding of these barriers. This is supported by studies on barriers to mental health care
utilization that identify distinctive psychological and social themes for barriers.108,120
Therefore, for the purposes of this thesis work, 5 major hypothetical constructs around the
decision to participate, or not, in a SPD were identified. These barrier constructs were
named (Table 3.1):
• Perceiving no threat
• Perceiving no benefit
• Psychological barriers
• Social barriers
• Practical barriers
Extensively reviewing the conceptual framework and theories, possible factors related to
each barrier construct were determined and categorized into the components of the SPD
(Table 3.2). Below is a brief review of these constructs and how they relate to possible
barriers to SPD.
Perceiving no threat
The two elements of perceived threat in the HBM are perceived susceptibility and perceived
severity (Appendix B, Figure 1). The first step to consider a health care action is to
26
understand that one is at risk of contracting a disease, and if affected, to accept the diagnosis.
The next step is to perceive that the condition may be severe enough to require treatment.
Hence, barriers related to perceived susceptibility and severity can be a lack of recognition of
the depression or the risk of becoming depressed as well as an underestimation of the
severity of the problem. Considering the self-regulatory model (Appendix B, Figure 4),
barriers can also be related to how a patient labels the illness, attributes symptoms, and thinks
about the duration, consequences, causes, and controllability of a health care problem;
accordingly, these barriers were identified: considering the depressive symptoms as normal
sadness, attributing its symptoms to physical illness, believing that the condition is transient
and will spontaneously improve, blaming oneself as the cause of depressive symptoms, and
personal or cultural explanations for symptoms. In addition, the self-regulatory model
describes that patients continuously appraise the symptoms and distress induced by the
disease when they try to cope with the problem.135 If patients assess their disease as not
severe enough to seek help, they continue their attempt to handle the problem on their own.
This can play a role as a barrier to refuse help when needed.
The concept of ‘incentive’ (Appendix B, Figure 2) is an independent component in social
cognitive theory. It describes the value an individual gives to the improvement of a certain
health condition. Incentive is different from the concept of perceiving benefit, as it is
irrespective of the benefits or harms of a certain health behaviour of interest, ie, one might
believe that removing a disease is important (incentive), but the current treatment is not
effective (benefit). However, expecting no incentive can overlap in meaning with perceiving
no threat, depending on the reason why one gives a low value to the possibility of reduction
of depression despite recognizing it. This can be due to low perceived severity, or a result of
prioritizing other health conditions. Thus, if patients on dialysis give a higher value to
treatment of ESRD or other physical problems, they are less likely to accept participating in
an intervention that targets depression. It should be noted that expecting no incentive may
also be a psychological barrier when the reason behind giving a low value to participation is
a lack of motivation for seeking treatment.
27
Perceiving no benefit
Patient concerns about the helpfulness of the screening for depression, the subsequent testing
or assessments, and the effectiveness of treatments influence participation. They may see no
benefit in being regularly screened and prefer to seek help based on when they judge their
symptoms to be severe enough. In addition, the patient’s perception of the competency of the
physician; the mental health care team involved in the referral, diagnosis and treatment; and
their faith in the health care system may influence participation (Appendix B, Figure 6).134
Side effects of medications is an example of a perceived barrier.132 Since the harm arising
from an intervention can be compared to its benefit, the perceived harm from participation is
included in the construct of perceiving no benefit. This is consistent with Ajzen’s concept of
“attitude towards a behaviour and its consequences133”, which supports the idea of looking at
the benefits and harms together as positive and negative consequences of behaviour. This
allows classification of concerns about side effects of antidepressant medications as a barrier
related to the construct of perceiving no benefit rather than a practical barrier.
Psychological barriers
Perceived psychological barriers are those factors that impede health care utilization in spite
of the patient acknowledging the threat of a disease and the benefit from a certain
intervention. Examples include the patient feeling that the intervention may be upsetting or
inconvenient, barriers arising from personality characteristics, and phobic reactions to
diagnosis and treatment.132,138 Becker and Maiman have described 3 reasons for not
accepting the diagnosis: powerful health beliefs that conflict with physician’s assessments,
incidents that reduce the patient’s confidence about the diagnosis, and a denial reaction to
being given a diagnosis of a serious illness.139 Specifically, denial is an aspect of not
accepting the diagnosis which can be a psychological reaction rather than lack of cognition
(perceiving no threat).
Other theoretical models propose self-concealment and self-efficacy as an important
behavioural modifier.133,136 Self-concealing individuals (an internal psychological factor in
the framework) are more likely to have negative attitudes towards treatment and less likely to
28
recognize distress.136 Self-efficacy has been considered as a perceived expectation
independent from perceived barriers.132 However, lack of self-efficacy can also be considered
a psychological barrier; the psychological concept of confidence in the ability to perform a
behaviour (perceived behavioural control) explains the link between self-efficacy and
psychological barriers.133 Patients on dialysis are usually overwhelmed with questionnaire
assessments, clinical visits, and load of medications, and therefore feel they are not able to
undertake more diagnostic or therapeutic procedures. Also, they might feel that they cannot
change themselves through psychotherapy. They might have poor motivation, which further
lessens when interventions such as cognitive behavioural therapy fail. Social cognitive theory 132 suggests that lack of motivation (eg, due to depression itself) is a psychological barrier
that may influence the perceived value of treatment outcomes.
Social barriers
Social norms and social support are important concepts in socio-behavioural models
(external factors in the framework).133,134,136 Therefore, it is reasonable to differentiate
barriers related to social structure with psychological barriers, even though there are
substantial overlaps between these sets of barriers.
Social stigma occurs when an individual or group of individuals are identified, labelled and
linked with negative attributes (stereotyping). The consequences are isolation, losing status,
and facing discrimination.140 Individuals diagnosed with depression may fear stereotyping.
Patients may fear that others would hear of them being diagnosed with depression and judge
them poorly. They may fear stigmatization by other patients, their families or friends and also
by the health care team. Even if they do not perceive social pressure, their own belief about
mental disorders that is formed under the influence of social norms may reflect in their
feeling of seeing depression as a sign of weakness.
Another social barrier is the lack of social support (described by Cramer and Andersen134,136;
Appendix B, Figures 5 and 6). Patients need the support of their family and physician.
Patients may perceive that their family will refuse to help them, or that their physician is
disinterested.
29
Practical barriers
Barriers such as organizational, financial, logistic, and physical limitations are categorized as
practical or “structural” barriers.115,122,131 Examples include time constraints due to work and
family responsibilities, lack of services in the local health authority, and health insurance
restrictions. Although these factors are conceptually different, they are often grouped
together. Some of these barriers may not apply to the SPD in a dialysis setting particularly
when appropriate care can be provided in the dialysis unit. Physical limitations may still play
an important role. For example, therapy sessions may be held concomitantly with dialysis
treatments times, or they may have difficulty completing the questionnaires due to visual
impairment or due to the severe fatigue.
3.2.2. Literature review for identification of barriers
An extensive search of MEDLINE, EMBASE, and PsychINFO was carried out to identify
studies that report barriers to utilization of mental health care and help seeking (for any
mental health problem and specifically for treatment of depression). Overall, 22 studies were
identified (Table 3.3), from which 271 possible barriers were identified.
3.2.3. Expert opinion on possible barriers
Three experts in the fields of nephrology and psychiatry were identified. They were asked to
review the conceptual framework and identify possible barriers to SPD. Consensus on a list
of barriers was achieved through iterative discussions around the conceptual framework and
concerns specific to dialysis patients.
3.2.4. Summarizing and categorizing barriers
A total of 271 barriers were identified from the literature and conceptual framework. As there
was a substantial degree of overlap between the barrier concepts, the 271 items could be
summarized into 65 distinctive barriers to mental health care (Table 3.2). An additional 5
barriers based only on the expert panel opinions were added. Thus, a total of 70 items were
identified through consensus as possible barriers to participation in the SPD among dialysis
patients (Table 3.4). These barriers were categorized into the five identified barrier
30
constructs based on the conceptual model (perceiving no threat, perceiving no benefit,
psychological barriers, social barriers, and practical barriers).
3.3. CRITICAL REVIEW OF AVAILABLE BARRIER SCALES
Three studies have used a systematically developed tool to measure barriers to mental health
care (Features of these three scales are summarised in Table 3.5)108,116,120: (1) Endo et al
developed the 36-item Barriers to Psychological Care Questionnaire for use in patients with
cancer. Each item was rated using a 5-point Likert scale ranging from 1 (do not agree at all)
to 5 (agree very much). Patients’ responses of 3 to 5 for each item were considered as
perceiving a barrier. (2) Pepin et al developed the 56-item Barriers to Mental Health Services
Scale (BMHSS), to examine intrinsic and extrinsic barriers for use in the general population.
Items were rated from 1 (strongly disagree) to 5 (strongly agree), and the subscale scores for
10 barrier constructs (continuous variable) were used for comparison of subgroups of
participants. (3) Mohr et al developed the 27-item Perceived Barriers to Psychological
Treatment (PBPT) questionnaire to measure barriers to counseling among primary care
patients. The level of difficulty a certain issue would cause in the way of seeing a counselor
was rated from zero (not difficult at all) to 4 (impossible) in each item. A score of 3 and 4 for
each item indicated that the barrier was present.
Direct contact with the authors resulted in two of the questionnaires (BMHSS and PBPT)
being available. Contact with Endo et al failed. Thus, I briefly review the BMHSS and the
PBPT with special attention to the measure of interest:
Target population. Neither of the scales is developed specifically for patients with chronic
physical illnesses. Thus, special concerns and considerations that respondents might have in
the context of ESRD are not addressed by the BMHSS or the PBPT.
Intervention of interest. Both BMHSS and PBPT measure barriers to psychotherapy.
Therefore, barriers related to other types of treatment, including pharmacotherapy, are not
covered by these questionnaires. Barriers to the screening and diagnosis processes are also
not addressed in these scales either.
31
Sensibility. The BMHSS measures barriers using response rates of agreement-disagreement
for statements describing certain barriers. The meaning of some items is unclear and thus
prone to misinterpretation. For example, the item “I would see a psychotherapist (counselor)
if it were free” is rated from “strongly disagree” to “strongly disagree.” It is not clear if the
respondent were to disagree with it, whether they meant they would not see a psychotherapist
even if it was without cost or if they would see a psychotherapist even if it were not free. An
additional criticism is that the questionnaire may include redundant items as item reduction
techniques were not applied during development of the scale. As a result, the questionnaire is
too long (56 items), which may exhaust respondents and increase non-participation.
In contrast, the PBPT questionnaire underwent exploratory and confirmatory factor analysis
methods on split samples; item-total correlation coefficients for item reduction; and
concurrent validity and internal consistency testing. The PBPT’s formulation of the items is
more direct by asking the level of difficulty a possible barrier may cause. This allows
interpretation of the score as the level of the severity of a barrier and dichotomization of the
responses. While advantageous the questionnaire items can be difficult to answer. In a
number of items, two components are combined in the question, one addressing an attitude,
and one asking the extent to which the attitude acts as a barrier. As a result the response
options are not appropriate. For example, the item “I wouldn’t expect counseling to be
helpful,” is hard to rate in terms of level of difficulty if the respondent believes that
counseling would be helpful.
Reliability and validity. The internal consistency of BMHSS was not favourable for any of
the subscales. The Cronbach alpha ranged from 0.48 to 0.90 for the 10 subscales. The scale
was not validated against a gold standard measure such as eventual participation in
psychotherapy. The PBPT’s internal consistency was adequate for all subscales (Cronbach
alpha ranged from 0.71 to 0.89). The authors assessed concurrent and predictive validity of
PBPT and reported that current use of psychotherapy was associated with some, but not all
subscales of barriers. The use of psychotherapy within 1 year from the study correlated with
the total PBPT score.
32
Table 3.1. Constructs of a conceptual framework for health care utilization
Hypothetical Constructs Description Perceiving threat Perception of one’s risk of being depressed or being susceptible to the
problem; perception of symptoms of depression and their cause, duration, medical and non-medical consequences, and controllability (illness representation); and assessment of the value of improvement of depressive symptoms (incentive)
Perceiving benefits Believing that the screening program will be effective and safe in reducing depressive symptoms and that the health care system and health care professionals are able to provide the appropriate care
Psychological barriers Factors related to one’s psychological characteristics or emotional state that may impede participation in the screening program
Social barriers Factors related to normative beliefs, social pressure, and support of others (family, friends, doctors, etc) that may impede participation in the screening program
Practical barriers External factors such as time constraint, limited access, and physical problems that make it difficult to participate in the screening program
33
Table 3.2. Possible barriers arising from each construct of mental health care utilization
Possible Barriers Constructs Screening Diagnosis/Referral Treatment
Perceiving no threat
• Lack of recognition of depression or its risk
• Underestimation of problem • Considering sadness as normal • Hoping to get better soon • Self-reliance • Prioritizing other problems
• Lack of recognition of depression or its risk • Underestimation of problem • Considering sadness as normal • Personal explanation for depression • Connecting the problem to physical problems • Hoping to get better soon • Self-reliance • Prioritizing other problems
• Lack of recognition of problem or its risk • Underestimation of problem • Considering sadness as normal • Personal explanation for depression • Connecting the problem to physical problems • Hoping to get better soon • Self-reliance • Prioritizing other problems
Perceiving no benefit
• Lack of faith in accuracy and effectiveness of screening
• Lack of faith in care providers • Lack of faith in healthcare system • One’s bad previous experience • Others’ bad previous experience
• Lack of faith in the intervention • Lack of faith in care providers • Lack of faith in health care system • Concerns about adverse effects • One’s bad previous experience • Others’ bad previous experience
Psychological barriers
• Self-concealment • Being overwhelmed by tests • Being overwhelmed by visits • Fear of screening results • Lack of motivation
• Self-concealment • Being overwhelmed by tests • Being overwhelmed by visits • Fear of diagnosis • Lack of motivation
• Self-concealment • Lack of confidence about one’s ability to take action • Being overwhelmed by visits • Being overwhelmed by medications • Lack of motivation
Social barriers • Fear of others finding about one’s mental problem
• Fear of being judged by others • Fear of being judged by physician • Negative attributes such as weakness
• Fear of others finding about one’s mental problem • Fear of being judged by others • Fear of being judged by physician • Negative attributes such as weakness • Lack of family/friends support • Lack of support by physician
• Fear of others finding about one’s mental problem • Fear of being judged by others • Fear of being judged by physician • Lack of family/friends support • Lack of support from physician
Practical barriers
• Limitations due to physical problems • Lack of information about access to care • Cost of care and health insurance limits • Transportation problems • Time constraints • Responsibilities • Limitations with physical problems
• Lack of access to treatment • Lack of information about access to care • Cost of care and health insurance limits • Transportation problems • Time constraints • Responsibilities • Limitations with physical problems
34
Table 3.3. Studies on barriers to mental health care utilization
Study Disorder Population Measurement Items Aromaa et al112 Depression 5160 Finnish participants from general
population Attitudes towards people with depression, antidepressants, and desire for social distance
Different sets of scales
Blumenthal et al113
Depression 101 individuals with a history of major depression disorder
Barriers to seeking help 43 possible barriers based on literature review, expert opinion, and lay persons’ opinion
Conner et al114 Depression 37 depressed African-American older adults
Qualitative study of the experience of depression and barriers to seeking help
...
Ell115 Depression Depressed elderly Not applicable (review article) ... Endo et al116 Psychological
problems 100 patients with lung cancer Barriers to psychological care A 36-item questionnaire developed
for the study Goodman117 Perinatal depression 509 pregnant women Barriers to treatment of depression Not explained in detail Johnson and Dwyer110
Depression 103 dialysis patients Barriers to treatment of depression 14 barriers based on a previous studies
Kravitz et al118 Depression 15 people with a history of depression in themselves or their family
Qualitative study of barriers to depression help-seeking in primary care setting
...
Lee et al119 Mental health disorder
211 Chinese patients Barriers to help seeking 8 structural and attitudinal barriers, no details on source of items
Mohr et al108,128 Mental health disorder
2 studies on 290 and 600 primary care patients
Barriers to psychotherapy A 27-item questionnaire developed for the study
Mojtabai et al109 Mental health disorder
1350 patients with mental health disorder who had not used health services in 12 months
Barriers to seeking help A list of barriers, no details on source of items
Nutting et al106 Depression 18 physicians and nurses with 60 patients with depression
Qualitative and quantitative studies of barriers to diagnosis and treatment
...
35
Table 3.3. Studies on barriers to mental health care utilization. Cont’d.
Study Disorder Population Measurement Items Pepin et al120 Mental health
disorder 70 students and 80 older adults Barriers prohibiting individuals from
seeking mental health services A 56-item questionnaire developed for the study
Rochlen et al121 Depression 45 men with a history of depression Qualitative study of interaction between male characteristics and depression
...
Sareen et al122 Mental health disorder
General populations from the United States, the Netherlands , and Canada (Ontario) with a history of mental health problem
Reasons for not seeking help if one had been feeling the need for help because of a mental problem
A list of barriers, no details on items source
Wang et al123 Mental health disorder
4094 Canadians with a history of mental health problem
Reasons for not seeking help if one had been feeling the need for help because of a mental problem
13 barriers, no details on items source
Ward et al124 Mental health disorder
15 African-American women Qualitative stud of beliefs about mental illness, coping behaviors, barriers to treatment seeking
...
Weinberger et al125
Depression Cancer patients Not applicable (review article) ...
Wong et al126 Mental or emotional problem
490 Cambodian refugees Factors that prevents one from getting help
9 items based on literature and expert opinion
Woodall et al127 Psychosis 26 individuals with a first psychosis episode
Qualitative study of reasons to participate or decline participating in a research
...
Wuerth et al111 Depression 320 peritoneal dialysis patients screened positive for depression
Asking reasons for not being willing to see health care professionals after screening
...
36
Table 3.4. Barriers to mental health care identified through review of literature, theory, and expert opinion
Item Inclusion in PBPT* Construct
Believing that one's problems are not severe enough for seeking help Yes Threat Believing that attending mental health services is too self-indulgent Yes Threat Believing that one's problem is not a disease No Threat Feeling that one has caused problems by oneself No Threat Believing that the problem is caused by the physical illness No Threat Preference to handle the problem on one's own No Threat Believing that the problem would get better soon No Threat Having other problems which are more important No Threat Not feeling to get depressed No Threat Feeling that sadness is normal No Threat Not feeling the need to have treatment No Threat Believing that sadness is normal among dialysis people† No Threat Preference to decide when one needs help† No Threat Not expecting treatment to be helpful Yes Benefit Having had or having heard about unsatisfactory experiences with treatment Yes Benefit Not expecting medication to be helpful No Benefit Not expecting psychotherapy to be helpful No Benefit Believing that current health care system would not be effective No Benefit Concerns about dependence on antidepressant medications No Benefit Concerns about antidepressant medications side effects No Benefit Concerns about competence of mental health professionals No Benefit Physician's concerns about treatment of depression No Benefit Feeling that questionnaire assessments about depression are not helpful† No Benefit Discomfort with being seen when becoming emotional Yes Psychological Feeling that talking about upsetting issues makes them worse and that Yes Psychological Lack of energy or motivation to make an appointment and then go yes Psychological Distrust of mental health professional yes Psychological Having to talk to someone unknown about personal issues yes Psychological Anxiety about going far from one's home yes Psychological Concerns about having upsetting feelings yes Psychological Difficulty motivating oneself to do anything at all yes Psychological Fear of visits in psychiatric clinic No Psychological Fear of hospitalization against one's will No Psychological Feeling guilty about having treatment for depression No Psychological Negative feelings about antidepressant medications No Psychological
37
Table 3.4. Cont’d
Item Inclusion in PBPT* Construct
Overwhelmed with load of medications and visits No Psychological Concerns about one's pride being wounded No Psychological Concern about one's relation with my physician being affected No Psychological Feeling one can't change No Psychological Not willing to take more questionnaires or have more visits† No Psychological Being afraid of screening test results† No Psychological Not expecting that mental health professionals truly care about one's issues yes Social Having a medical or insurance record of mental health care services yes Social Concerns about being judged by the mental health professional yes Social Having family and friends know one was going to mental health services yes Social Thinking that using mental health services is a sign of weakness yes Social Physician's discrimination against the elderly No Social Health care discrimination No Social Physician's priorities regarding other problems No Social A previous experience of stigma No Social Feeling shame No Social Concerns about losing one's job No Social One's family wouldn't want them to have treatment for depression No Social The lack of available mental health care services yes Practical The cost of treatment yes Practical Not knowing how to find a good mental health professional yes Practical The responsibility of caring for loved ones yes Practical Daily responsibilities and activities yes Practical Getting time off work yes Practical Problems with transportation yes Practical Physical symptoms (fatigue, pain, breathing difficulties, etc.) yes Practical A serious illness which requires one to stay close to home yes Practical Physical problems, such as difficulties walking or getting around yes Practical Concerns about waiting time to make appointment No Practical Insurrance coverage limitations No Practical Cost of medications No Practical Costs of visits or psychotherapy No Practical Not knowing about possible treatments No Practical Language problems No Practical Work responsibilities No Practical
*PBPT indicates Perceived Barriers to Psychological Treatment questionnaire. †Based on conceptual framework and expert opinion and not from the literature.
38
Table 3.5. Scales for measurement of perceived barriers to mental health care utilization
Characteristics Endo et al116 Pepin et al120 Mohr et al108 Scale BPCQ BMHSS PBPT Intervention of Interest Psychological Care Psychotherapy Psychotherapy Target Population Cancer patients General population Primary care patients Number of Items 36 56 27 Scoring system 5-point Likert on agreement (1 to 5) 5-point Likert on agreement (1 to 5) 5-point level of difficulty (0 to 4)
Summary score 4 subscores 10 subscores Total score 8 subscores
Cutoff point > 3 for each item None ≥ 3 for each item Factors (hypothetical or based on factor analysis)
Emotional communication with physician; Psychiatric consultation; Psychotropic medications; and Counseling
Help-seeking attitudes; Stigma; Knowledge and fear of psychotherapy; Belief about inability to find a psychotherapist; Belief that depressive symptoms are normal; Insurance and payment concerns; Ageism; Concerns about psychotherapist’s qualifications; Physician referral; and Transportation concerns
Stigma; Lack of motivation; Emotional concerns; Negative evaluation of therapy; Misfit of therapy to needs; Time construction; Participation restrictions; and Availability of services
*BPCQ indicates Barriers to Psychological Care Questionnaire; BMHSS, Barriers to Mental Health Services Scale; and PBPT, Perceived Barriers to Psychological Treatment.
39
Figure 3.1. Conceptual framework for participation in a screening program for depression based on the Health Belief Model,131 complemented by other social and behavioural models. The Health Belief Model suggests the 3 main elements of perceived threat, perceived benefits, and barriers. The latter is further categorized into psychological, social, and practical factors based on suggestions on the use of Health Belief Model in mental health care as well as other conceptual models on health care behaviours. The Self-regulatory Model further explores perception of the patients about their symptoms and their coping behaviour. Other models add the concepts of feedback, incentive, and external factors to this framework. See the figures in Appendix B for details of each conceptual model.
40
4. METHODS
The purpose of this chapter is:
• To detail the design of the study • To define the outcome measures and explanatory factors • To describe the research procedure and data management • To summarise the statistical methods used for the analysis of data
4.1. STUDY DESIGN, SETTING, AND PARTICIPANTS
This research project was a questionnaire-based cross-sectional observational study of
outpatient hemodialysis patients. Participants were recruited from in-centre hemodialysis
units of the University Health Network (UHN) and Sunnybrook Health Sciences Centre
(SHSC). All patients undergoing chronic outpatient in-centre hemodialysis during the period
of July to September 2012 were eligible. The inclusion criteria were as follows:
a. age ≥ 18 years old;
b. comprehension of written English at grade 6 level;
c. ability to read print materials with large print (Arial font 16 points in size); and
d. undergoing chronic hemodialysis treatment for ≥ 30 days at one of the study sites.
The exclusion criteria were:
a. documentation of clinical dementia or cognitive impairment;
b. acute inpatient status; and
c. inability or unwillingness to provide informed consent.
4.2. MAIN OUTCOME MEASURE
4.2.1. Definitions
The primary outcome was defined as patient-perceived barriers to participation in an SPD.
The SPD was used to mean a program incorporating routine questionnaire assessments,
41
referral, and treatment for depression. Treatment was considered to be any of the currently
available treatment options for depression, including psychotherapy and pharmacotherapy.
4.2.2. Measurement tool
For the purpose of this Master’s thesis, an adapted version of the Perceived Barriers to
Psychological Treatment (PBPT) questionnaire was used as a measurement tool of patient-
perceived barriers. Mohr et al108 developed a 27-item PBPT in 2010 to measure perceived
barriers to psychotherapy among primary care patients. Each item asked participants to rate
the degree to which different kinds of problems would get in the way of seeing a counselor or
a therapist. Response options were “not difficult at all,” “slightly difficult,” “moderately
difficult,” “extremely difficult,” and “impossible,” rated from zero to 4, respectively.
After critically appraising questionnaires that have been used (Chapter 4; Table 3.5),108,116,120
we selected the PBPT as the most appropriate despite two main limitations–a focus limited to
psychotherapy treatments and the absence of barriers that may be more common in a
medically unwell population. The PBPT was selected because (1) it directly addresses the
concept of barriers, (2) it is shorter than the other available scales, (3) it allows
dichotomization of the responses (presence versus absence of barriers), and (4) it is a
validated barriers questionnaire against concurrent and future use of mental health care.
4.2.3. Adaptation of perceived barriers to psychological treatment
In order to make the scale appropriate for this study, “counselling” was replaced by
“screening program for depression” throughout the questionnaire. The definition and purpose
of the SPD, written to a readability grade 6 level, was added to the instructions (Appendix
C):
Some centres suggest that we routinely use a screening questionnaire about
depression in dialysis patients. Screening helps the dialysis team to identify people
with depressive symptoms, so that they can help if needed. For example, they may
adjust treatments, refer to a mental health specialist, or prescribe medications, if
needed.
42
For the purpose of this questionnaire, assume that a screening program for
depression would involve completing a questionnaire about your mood and
feelings, and if needed being referred to a mental health specialist for further
assessment, counseling or medications.
To ensure comprehensiveness of the scale, we compared the PBPT against a summary list of
possible barriers to participation in the SPD generated from the literature (described in
Chapter 4; Table 3.4). Of the 70 identified barriers, 27 were included in the original PBPT.
Based on consensus among the expert team of the project, we decided to add 11 items from
among the remaining 43 identified items. These items were deemed to be essential and most
relevant to the dialysis patients (Table 4.1). Thus, the adapted PBPT comprised 38 items.
4.2.4. Content validity
The adapted PBPT was peer reviewed by three content experts in the fields of nephrology,
psychiatry, and dialysis care nursing. Final revisions of the questionnaire were carried out
based on the experts’ comments on the instructions and the added items. No new items were
added and none of the items were removed.
4.2.5. Subscales
Scoring was based on the original PBPT. The original PBPT consisted of 2 items related to
perceiving no threat, 2 items related to perceiving no benefit, 8 items related to psychological
barriers, 5 items related to social barriers, and 10 items related to practical barriers. The 11
additional items in the adapted PBPT were also categorized into the five hypothetical
constructs (Table 3.1). Thus, the adapted PBPT consisted of 8 items related to perceiving no
threat, 4 items related to perceiving no benefit, 11 items related to psychological barriers, 5
items related to social barriers, and 10 items related to practical barriers (Appendix C;
scoring system).
43
4.3. EXPLANATORY FACTORS
4.3.1. Depression scale
The presence of depressive symptoms influences a patient’s decision to seek help. It may
also alter their perception of the barriers to mental health care.108 Since those with depressive
symptoms are actually the target of the SPD, we measured depressive symptoms to
understand the effect of this variable on the burden and pattern of barriers. We used the
Patient Health Questionnaire (PHQ-2), a depression scale with 2 questions about two core
depressive symptoms (Appendix D).2,141 The two items of the PHQ-2 are questions about the
number of days one feels depressed and lack of interest in doing things in the past 2 weeks.
These items are the two core symptoms of depression based on the DSM-IV definition of
MDD.2 The PHQ-2 is validated in the primary care setting (sensitivity of 62% and a
specificity of 95%; scores ≥ 3),141 and has been used in some large-scale studies on dialysis
patients.17,142 Alternative depression scales used in ESRD patients were considered. These
include the BDI and CESD,79 as they are validated in this population with adjusted cutoff
points and acceptable diagnostic accuracy. Gyamlani et al demonstrated that the PHQ-2
significantly correlated with CESD.142 PHQ-2, and CESD identified 24% and 30% of the
CKD patients to have depressive symptoms, respectively.142 A positive PHQ-2 serves to alert
the clinician that further clinical evaluation may be appropriate.141
4.3.2. Covariates
Data were collected on the main baseline demographic factors and clinical characteristics that
may correlate with depression and patient-related barriers to mental health
care.18,54,108,112,121,123,143 Data were collected from the patients and their medical charts. These
included age, gender, education level, marital status, cause of end-stage renal disease,
comorbidities listed in the Charlson Comorbidity Index,144 time on RRT, and history of
diagnosis and/or treatment of depression. Education level was recorded as the number of
years of studying since primary school. Marital status was categorized as married (or living
with partner), single, divorced, and widowed. Causes of ESRD included diabetes mellitus,
hypertension and vascular diseases, glomerulonephritis and interstitial diseases, hereditary
disorders, other causes, and unknown.
44
4.4. RESEARCH PROCEDURE
4.4.1. Informed consent
The informed consent forms were prepared and revised according to the recommendations of
the Research Ethic Boards of the UHN and SHSC (Appendix E).
4.4.2. Non-participation
To identify sources of bias due to non-participation, we asked eligible patients who opted not
to participate in the study to consent to limited chart review. Data collected included age,
gender, cause of ESRD, and time on RRT.
4.4.3. Study visits
Patients who consented to full participation in the study were asked to participate in two
study visits. The date and time of visits were arranged at the convenience of the patients, and
interference with the routine clinical care of the patients was avoided.
Visit 1. The first visit was usually immediately after obtaining the consent. Participants were
asked to answer questions about their demographics and medical history. Medical charts
were reviewed to confirm and complete data collection. Participants were also asked to
complete the PHQ depression scale. Assistance in reading or filling out the questionnaire was
provided upon request.
Visit 2. Within 1-2 weeks, participants were asked to fill out the study questionnaire designed
to measure perceived barriers to participation in an SPD (PBPT). Assistance in reading or
filling out the questionnaire was provided upon request. Upon their request, a copy of the
signed consent form was provided to the participants at the end of this visit.
4.4.4. Data management
Collected data were stored in a locked file cabinet in the administrative area designated to the
research project at the Division of Nephrology, UHN. Signed consent forms and the
enrollment sheets were stored separately in another locked file cabinet. Data entry was
carried out on a weekly basis using the Microsoft Excel. The Excel spread sheet file was
45
saved on a secured network directory of the UHN and was password protected. The dataset
was double-checked and cleaned through randomly picking different segments to look for
incorrectly entered data and routine monitoring of descriptive data (eg, minimum and
maximum values). No interim analysis was planned.
4.5. SAMPLE SIZE
Based on the literature on perceived barriers to mental health care in primary care setting and
among dialysis patients,108,110 it was estimated that 50-70% of the patients would have one or
more barriers to the SPD. Based on a normal approximation to the binomial theorem to
evaluate the 95% CI for the estimate of the proportion and assuming an average proportion of
60% and a 95% CI of 15%, a total of 159 hemodialysis patients were required. A sample size
of 159 patients allowed for a CI of ± 7.0-7.7% with the estimated range of proportion (50-
70%) and 8-11 variables in multivariable models for the dichotomized outcome variable
(presence of barriers) and 15 variables in models for the continuous outcome variable (barrier
score). This confidence range was considered to be clinically justifiable because of the
estimated high proportion (Appendix F).
4.6. FEASIBILITY
Intermittent hemodialysis patients attend dialysis sessions 3 times per week. Thus,
participants were restricted to in-centre hemodialysis patients because of their accessibility.
Patients were approached during 4-hour dialysis sessions. A total of 530 patients undergo
regular chronic hemodialysis at the UHN and SHSC. It was anticipated that 60% would be
eligible (n = 318) and of those eligible, 50% would consent to participate (n = 159).
4.7. DATA ANALYSIS
4.7.1. Psychometrics of the adapted PBPT
The adapted scale was tested for internal consistency using the Cronbach α coefficient, which
is an index of reliability associated with the variation accounted for by the true score of the
46
construct measured. A value greater than 0.70 for the scale and subscales was considered
appropriate.145 The corrected item-total correlations were used to test the homogeneity of the
subscores. This method measures the correlation of an individual item score with the total
score when that item is omitted. Each item should correlate with the total score above 0.2.
Therefore, items with a coefficient less than 0.2 would be discarded from the analyses of
each subscale of barriers.129
Assessment of the criterion validity of the PBPT is not possible because of the lack of a gold
standard for measurement of the barriers. However, the literature shows that patients with the
experience of diagnosed and treated clinical depression are less likely to refuse intervention
for depression.143 As a result, convergent validity was tested by comparing the total score and
subscores against a past history of the diagnosis of depression, hypothesizing that these
patients will have lower barrier scores. Logistic regression analysis was used to assess the
association between the total barriers score (dependent variable) and the past history of
depression, adjusted for other independent variables.
4.7.2. Descriptive analysis of patients with barriers (Objective 1)
An item score ≥ 3 was considered as presence of a barrier to the SPD.108 Data of the
proportion of participants who perceived ≥ one barrier were demonstrated as frequency and
percentage (95% CI). The total barrier score was demonstrated as the median value
(interquartile range). These statistics were also reported for the subscales of each barrier
construct, including perceiving no threat, perceiving no benefit, psychological barriers, social
barriers, and practical barriers.
4.7.3. Descriptive analysis of barriers (Objective 2)
The proportions of patients who perceived each barrier (item score ≥ 3) were demonstrated as
frequencies and percentages. The cutoff point for dichotomization of data was based on the
methodology used for the original PBPT.108
4.7.4. Sensitivity analysis: comparison with the original barrier questionnaire
The original PBPT scale (27 items) was compared to the adapted version with the additional
items, in terms of the percentage of participants with barriers. The most common barriers
based on the original and adapted PBPT scales were compared to each other. The agreement
47
between the original PBPT and the adapted PBPT regarding the presence of at least one
barrier was tested using the kappa coefficient of agreement.
4.7.5. Non-participants’ data
Age, sex, cause of ESRD, and time on RRT were compared between participants and non-
participants to identify potential sources of bias. Between-group comparisons of continuous
variables were done using the independent Student t test. The Shapiro-Wilk test was used to
assess normality of distributions and comparisons of skewed data were done using the
Wilcoxon rank sum test. Comparisons of categorical data were done using the chi-square
test. If the number of expected values was less than 5 in more than 25% of the cells in the
contingency tables, the Fisher exact test was used.
4.7.6. Patient characteristics associated with barriers (Objective 3)
Univariable analysis. The associations between each of the five barrier subscores and the
patient characteristics were examined. The included patient characteristics were age, sex,
education level, marital status (dichotomized to married or living with a partner versus not
living with a partner), time on RRT, causes of ESRD (diabetes mellitus, hypertension, and
glomerulonephritis, each versus other causes), Charlson comorbidity score, PHQ-2 score, and
history of depression. Analyses were performed using the Pearson correlation coefficient and
the independent Student t test or the non-parametric counterparts (Spearman rho test and
Wilcoxon rank sum test, respectively). Normality of the distributions was tested using the
Shapiro-Wilk test.
Multivariable analysis. Multiple regression models were constructed for each of the 5
subscores. Linear regression analyses were applied to assess the association between each of
the barrier subscore and the PHQ-2 score. The included covariates were age, sex, education
level, marital status, time on RRT, diabetes mellitus as a cause of ESRD versus other causes,
hypertension as a cause of ESRD versus other causes, glomerulonephritis as a cause of
ESRD versus other causes, Charlson score, and history of depression. Since the assumption
of normality of residuals was not met in the models, natural logarithmic (log) transformation
of the dependent variables (subscores) was applied. The final model goodness of fit was
48
assessed using the R-squared index. Multicollinearity of the independent variables was tested
on for the using the variance inflation factor (> 4) and the collinearity index (condition index
> 100), and exclusion of the collinear variables, if any, was decided based on their clinical
importance. The overall significance of the model was tested using the omnibus F test, and
the interpretation of significant variables in the model was conditional to the significance of
the F test. The normality of the residuals in the linear regression was tested by the Shapiro-
Wilk test. Outlier observations in the model with a studentized residual absolute value greater
than 2 or a DIFFITS (standard influence of observation on predicted value) absolute value
greater than 2 were considered as potential influential outliers. The DIFFITS measures the
standardized change in the dependent variable if an observation is removed. The model was
tested after removal of these observation and the changes in estimates were assessed
subjectively. If changes were small, the final model was reported without excluding the
outlier observations. Finally, residuals were plotted to assess homoscedasticity and straight
line relationship of the exposures and the outcome.
If log transformation of the independent variables did not result in normality of residuals, two
alternative models were used to validate findings from the linear regression model: (1)
Categorical logistic regression model with the dependent variables categorized into four
groups of 1 to 4 based on the 1st quartile, median, and 3rd quartile values, and (2) Cox
proportional hazard model with the subscore values treated as time to event and assuming a
value of 1 as the event for all observations.146 The proportionality of the hazards assumption
was tested using the proportionality test when the interaction term between the subscore and
the PHQ-2 score was included in the model (A P value less than .05 rejects the null
hypothesis of proportionality of hazards).
4.7.7. Correction of significance level for multiple testing
Using five separate models may cause an inflated type I error. Therefore, correction for
multiple testing was applied using the Benjamini-Hochberg correction procedure.147 This is
to control for the false discovery rate and offers a more powerful method than the Bonferroni
correction. The adjusted P values for each model were calculated using the PROC
MULTTEST in the SAS statistical software.
49
4.7.8. Analysis software
We used the SAS (Statistical Analysis System, version 9.2, SAS, Cary, NC, USA) for the
statistical analyses and considered a two-tailed type I error rate of .05 as the threshold for
statistical significance.
4.8. ETHICS
The protocol, questionnaires, data collection forms, and consent forms were submitted to the
Office of Research Ethics of the University of Toronto and the local hospital Research Ethics
Boards. Consent forms for the UHN and SHSC are included in the Appendix E.
As a safety measure, patients with a PHQ-2 result indicative of depressive symptoms were
identified and reported to the primary care nephrologist. Further assessments, referral, or
treatment was left to the discretion of the clinical team.
50
Table 4.1. Added items to PBPT
Construct Item
Perceiving no threat I would prefer to handle it on my own if I was depressed.
Perceiving no threat Having other problems that are more important
Perceiving no threat I would prefer to decide when I need help for depression on my own.
Perceiving no threat I do not think I will get depressed.
Perceiving no threat I think sadness is normal among people on dialysis.
Perceiving no threat I think better treatment of the kidney problem would improve depression.
Perceiving no benefit I would be concerned about side effects of medications for depression, if needed.
Perceiving no benefit I wouldn’t expect questionnaires for depression to be helpful.
Psychological barrier Having to fill out additional questionnaires
Psychological barrier Having to take more medications
Psychological barrier I would be afraid of screening results for depression.
51
5. RESULTS
The purpose of this chapter is:
• To demonstrate the reliability and validity of the adapted PBPT scale in dialysis patients • To describe the barriers perceived by dialysis patients to participation in screening program
for depression • To identify the characteristics of the patients who are more likely to have barriers to
screening for depression • To examine the association of depressive scores of the patients with their barriers subscores
through multivariable analyses controlling for other covariates
5.1. BASELINE DATA
5.1.1. Participation
Overall, 373 of 488 patients maintained on chronic in-centre hemodialysis were assessed for
eligibility at the hemodialysis units of the UHN and SHSC. All of the 279 patients at UHN
were assessed for eligibility. Given the higher-than-expected participation rate from this site,
patients at the SHSC were approached randomly, and the assessments were stopped after
recruitment of 94 of the 209 patients as we had reached the a priori determined sample size
required for the study.
Two hundred and forty-two patients were eligible (65%), of whom 169 consented to
participate in all parts of the study (70%). Of the non-participants (n = 73), 17 patients (23%)
consented to basic data collection (Part A of the study). Of the 169 participants, 160 (94.7%)
completed the study; therefore, overall, 66% of the eligible patients (43% of all patients)
participated in and completed the study (Figure 5.1).
5.1.2. Baseline characteristics
Table 5.1 summarizes the baseline characteristics of the patients. The mean age of the
participants was 57.1 ± 17.0 years old (range 21 to 92 years) and 61% were men. The
52
participants had been receiving RRT for a median time of 2 years (range 1 to 420 months).
Non-participants’ characteristics were not significantly different from the characteristics of
the participants on these variables.
5.1.3. Depressive symptoms
Figure 5.2 demonstrates PHQ-2 scores of the patients. Twenty-seven patients (16.0%) had
depressive symptoms (PHQ-2 score ≥ 3). Of these patients, 3 (11.1%) were concurrently on
treatment for depression. Sixteen patients who screened positive for depressive symptoms
had a past history of diagnosis of depression, of whom 2 had been diagnosed within the past
6 months.
5.1.4. History of depression
A clinical history of depression was identified in 37 patients (21.9%). The time since the
diagnosis of depression was made ranged from 1 to 40 years prior to the time of this study. In
21 of the 37 patients (56.8%), the diagnosis had been made after the initiation of RRT. The
diagnosis of depression was documented in the medical charts of only 9 patients (5.3%).
Twenty-six patients gave a prior history of treatment for depression. However, only 6
patients (3.6%) were on any treatment for depression at the time of recruitment. Overall,
treatments included pharmacotherapy in 21 patients, psychotherapy in 10 patients, and
exercise therapy in 1 patient. Pharmacotherapy included selective serotonin-reuptake
inhibitors in 7 patients (fluoxetine in 4 and citalopram in 3); mirtazapine in 3, lithium in 2,
amitriptyline in 1, and unknown in 8 (Figure 5.3).
5.2. RELIABILITY AND VALIDITY OF THE ADAPTED PBPT
Of the 169 participants, 160 (94.7%) completed the barriers questionnaire. Three patients
were transferred or admitted to hospital before the second study visit and 6 refused to fill out
the questionnaire. Of the respondents, 147 answered all the questions. Response rates to the
items ranged between 96.9% and 100%.
53
The Cronbach alpha coefficient was 0.95 for the barriers questionnaire and 0.75 to 0.89 for
the questionnaire’s subscales (Table 5.2). The item-total correlations were all greater than
0.2 (Table 5.3).
The logistic regression model for assessment of convergent validity of the adapted PBPT
failed to show a significant association between a history of diagnosis of depression and the
barriers questionnaire total score (PBPT score as the independent variable; P = .52). The
model was adjusted for age, sex, marital status, education level, RRT time, and the Charlson
score.
5.3. BARRIERS
5.3.1. Barriers scores and dichotomized results
Table 5.4 summarizes the barrier scores and percentage of participants with one or more
barriers (score ≥ 3 for one or more questionnaire items). Overall, 117 participants (73.1%;
95% CI, 66.2% to 80%) perceived barriers to participating in an SPD. They had a median of
6 barriers to the SPD (range, 1 to 30; quartile range, 2 to 10). Half of the participants
perceived at least one psychological barrier, and 51.3% perceived at least one practical
barrier. Figure 5.4 shows the histogram of barriers scores and the number of barriers
perceived by the participants.
The most frequent reasons patients gave for not being willing to participate in the SPD were
their concerns about anti-depressant medications (concerns about the side effects and
difficulty taking additional medications) and perceiving no threat (feeling that their problem
is not severe or that they are not at risk of becoming depressed). Table 5.5 summarizes the
most frequent barriers (a complete list of barriers is available in Appendix G).
5.3.2. Sensitivity analysis: barriers using the PBPT without additional items
Considering responses to the 27 items of the original PBPT, 66.3% of the participants had
barriers to the SPD (95% CI, 58.9% to 73.6%; Table 5.6). The agreement between the
dichotomized responses to the original and adapted PBPT was high (kappa, 0.83). However,
54
the original PBPT did not capture 4 of the barriers identified as the ten most common ones
based on the adapted PBPT; concerns about antidepressant medications and having more
medications were the two most common barriers captured by the additional items in the
adapted PBPT version (Table 5.5).
5.3.3. Relationship between barriers and PHQ-2 scores
Patients with depressive symptoms (PHQ-2 score ≥ 3) were more likely to perceive barriers
to SPD (96% versus 68.9%, P = .005; Figure 5.5). They had significantly higher scores for
perceiving no benefit, psychological barriers, and practical barriers (Table 5.7). The most
frequent barriers were concerns about side effects of medications for depression and practical
barriers related to the physical illness and costs of treatment, if needed (Table 5.8). Concerns
about medication side effects were considerably more frequent among patients with
depressive symptoms as compared to those without depressive symptoms. Figure 5.6
compares frequency of the most common barriers between patients with and without
depressive symptoms.
5.4. UNIVARIABLE ANALYSES
The associations between 11 independent variables and the barriers subscores were assessed
using univariable analysis methods. Since none of the barrier subscores had a normal
distribution, univariable analyses were done using non-parametric tests, including the
Wilcoxon rank-sum test and the Spearman rho correlation coefficient test. The PHQ-2 scores
significantly correlated with 4 of 5 barrier subscores of perceiving no benefit, psychological,
social, and practical barrier scores. A history of diagnosis of depression was another
covariate associated with practical barriers. Interestingly, none of the variables were linked
with perceiving no threat scores (Table 5.9 to Table 5.13).
5.5. MULTIVARIABLE ANALYSES
Multiple linear regression models were built for each barrier subscore as the dependent
variable. The PHQ-2 score, as well as 10 covariates were included in the models. Covariates
55
were age, sex, education level, marital status (married or living with a partner versus not
living with a partner), time on RRT, causes of ESRD (diabetes mellitus, hypertension, and
glomerulonephritis, each versus other causes), Charlson comorbidity score, and history of
depression.
Perceiving no threat. Multiple linear regression failed to show overall significant results for
the subscore of perceiving no threat (omnibus F test, 0.78; P = .66).
Perceiving no benefit. Depressive symptom scores and time on RRT were linked with the
subscore of perceiving no benefit, when treated as natural log-tranformed continuous variable
for multiple linear regression analysis (Table 5.14). The Benjamini-Hochberg correction for
multiple testing did not change the results (P = .02 and P = .03, respectively). The normality
of residuals assumptions was not met for the linear model (Shapiro-Wilk test, P = .004), but
the results were consistent with the logistic regression for categorized subscores. The Cox
model’s assumption of proportional hazards was not met.
Psychological barriers. Depressive symptom scores and age were linked with the log-
transformed subscore of psychological barriers; however, after correction for multiple
testing, only depressive symptoms were associated with psychological barriers (P = .01;
Table 5.15). The distribution of the residuals in the linear regression model was normal
(Shapiro-Wilk test, P = .08).
Social barriers. Depressive symptom scores and time on RRT were linked with the log-
transformed subscore of social barriers, when controlling for the other covariates in the
multiple linear regression model. The associations remained significant after corrections for
multiple testing (P = .048 for both variables; Table 5.16). The model, however, was not fit
because of non-normal distribution of residuals (Shapiro-Wilk test, P = .001), but the results
were confirmed in the logistic model for categorical social barriers variable (The Cox
model’s assumptions were not met).
Practical barriers. When controlled for other factors in a linear regression model, log-
transformed practical barriers subscore was associated with age, Charlson score, and PHQ-2
scores; however, correction for multiple testing demonstrated that only greater depressive
56
symptom scores were linked with higher practical barriers scores (Table 5.17). The model’s
assumption of normality of residuals was met (Shapiro-Wilk test, P = .08).
57
Table 5.1. Baseline characteristics of participants and non-participants*
Parameter Participants Non-participants† P Number 169 17 … Site …
TGH 115 16 SHSC 54 1
Mean age, y 57.1 ± 17.0 52.3 ± 14.3 .21 Male sex 103 (60.9) 12 (70.6) .44 Marital status …
Married/living with partner 81 (47.9) …
Single 48 (28.4) … Divorced 30 (17.8) … Widowed 10 (5.9) …
Mean education level, y 13.6 ± 3.4 … … Median RRT time, mo 48 (18 – 102) 66 (24 – 96) .48 ESRD cause .86
DM 40 (23.7) 5 (29.4) HTN 29 (17.2) 3 (16.7) GN 41 (24.3) 6 (35.3) Hereditary 27 (16.0) 2 (11.8) Others 20 (11.8) 1 (5.9) Unknown 12 (7.1) 0
Median Charlson score 4 (2 – 5) … … History of depression 37 (21.9) … … Treatment of depression …
Current 6 (3.6) … Previous 29 (17.2) …
PHQ … Median total score 0 (0 – 2) … Positive (>= 3) 27 (16.0) …
*Values in parentheses are percentages for frequencies and the 1st and 3rd quartiles for the median values.
†Basic data of these patients were collected after obtaining consent. 56 (35%) did not participate and did not provide consent to basic data collection.
58
Table 5.2. Internal consistency of the modified PBPT subscale subscores
Barrier Subscale Items Cronbach α Perceiving no threat 8 0.89 Perceiving no benefit 4 0.75 Psychological barriers 11 0.89 Social barriers 5 0.84 Practical barriers 10 0.82 All barriers 38 0.95
59
Table 5.3. Corrected item-total correlations (Pearson correlation coefficient) and internal consistency of subscales with each item when that item is omitted.
Barrier item* Correlation with Total
Cronbach α with Deleted Item
Barrier item* Correlation with Total
Cronbach α with Deleted Item
Perceiving no threat Social barriers p 0.47 0.89 gg 0.67 0.80 q 0.66 0.87 ii 0.67 0.80 r 0.59 0.88 jj 0.60 0.82 s 0.69 0.87 kk 0.61 0.82 bb 0.73 0.86 ll 0.66 0.80 cc 0.69 0.87 Practical barriers dd 0.79 0.86 a 0.50 0.80 ee 0.62 0.87 b 0.46 0.80
Perceiving no benefit c 0.36 0.81 k 0.48 0.73 d 0.59 0.79 m 0.62 0.65 e 0.56 0.79 n 0.51 0.71 f 0.56 0.79 o 0.58 0.67 g 0.25 0.82
Psychological barriers h 0.54 0.80 l 0.40 0.90 i 0.59 0.79 t 0.56 0.89 j 0.57 0.79 u 0.53 0.89 v 0.67 0.88 w 0.75 0.88 x 0.66 0.88 y 0.71 0.88 z 0.65 0.88 aa 0.68 0.88 ff 0.61 0.88 hh 0.58 0.89
*Letters refer to the individual questions or items used. These have been named alphabetically from “a” to “ll” in the questionnaire. See Appendix C for barrier items.
60
Table 5.4. Median subscores and percentage of participants with barriers for each barrier subscale
Score Barrier Positive
Barrier Subscale (Possible Score Range) Median 1st - 3rd
Quartiles Range n % (95% CI)
Perceiving no threat (0 – 32) 5 1 – 11 0 – 32 69 43.1 (35.4 – 50.8)
Perceiving no benefit (0 – 16) 3 1 – 5 0 – 12 68 42.5 (34.8 – 50.2)
Psychological barriers (0 – 44) 7 3 – 14 0 – 40 80 50.0 (42.3 – 57.7)
Social barriers (0 – 20) 3 0 – 7 0 – 20 48 30.0 (22.9 – 37.1)
Practical barriers (0 – 40) 8 3 – 13 0 – 31 82 51.3 (43.5 – 59.0)
Overall (0 – 152) 30 14 – 51 0 – 114 117 73.1 (66.2 – 80.0)
61
Table 5.5. Ten most frequently perceived barriers*
Subscale Barrier Responses (%) n (%) Benefit Concerns about side effects of medications† 159 (99.4) 63 (39.6) Psychological Having to take more medications† 160 (100) 51 (31.9) Threat My problems are not severe enough 160 (100) 37 (23.1) Threat I do not think I will get depressed† 158 (98.8) 36 (22.8) Practical The cost of treatment, if needed 159 (99.4) 33 (20.8)
Practical A serious illness which requires me to stay close to home 160 (100) 30 (18.8)
Threat Having other problems that are more important† 155 (96.9) 29 (18.7)
Practical Physical problems, such as difficulties walking or getting around 160 (100) 27 (16.9)
Psychological Anxiety about going far from my home 160 (100) 26 (16.3)
Social Having a medical or insurance record of mental health services 160 (100) 26 (16.3)
*See Appendix G for the complete list of barriers. †These items were among the additional items to the original PBPT scale.
62
Table 5.6. Percentage of participants with barriers for each subscale based on the original PBPT
Score Barrier Positive
Barrier Subscale (Possible Score Range) Median 1st - 3rd
Quartiles Range n % (95% CI)
Perceiving no threat (0 – 8) 1 0 – 3 0 – 8 40 25.0 (18.3 – 32.5)
Perceiving no benefit (0 – 8) 0 0 – 2 0 – 6 20 12.5 (7.4 – 18.6)
Psychological barriers (0 – 32) 4 2 – 10 0 – 29 63 39.4 (31.8 – 47.0)
Social barriers (0 – 20) 3 0 – 7 0 – 20 48 30.0 (22.9 – 37.1)
Practical barriers (0 – 40) 8 3 – 13 0 – 31 82 51.3 (43.5 – 59.0)
Overall (0 – 108) 18.5 9 – 34 0 – 81 106 66.3 (58.9 – 73.6)
63
Table 5.7. Median barrier scores (1st – 3rd quartiles) in patients with (PHQ-2 ≥ 3) and without depressive symptoms (PHQ-2 < 3)
PHQ-2
Barrier (Possible Range) < 3 ≥ 3 P
Perceiving no threat (0 – 32) 5 (1 – 11) 6 (2 – 10) .90
Perceiving no benefit (0 – 16) 3 (0 – 5) 5 (3 – 7) .02
Psychological barriers (0 – 44) 7 (2 – 13) 12 (6 – 18) .03
Social barriers (0 – 20) 2 (0 – 6) 4 (1 – 9) .23
Practical barriers (0 – 40) 7 (3 – 12) 12 (8 – 16) .005
Overall (0 – 152) 26 (13 – 50) 38 (25 – 52) .05
64
Table 5.8. Ten most frequent barriers perceived by the patients with depressive symptoms (PHQ-2 ≥ 3)
Subscale Barrier Responses (%) n (%)
Benefit Concerns about side effects of medications 25 (100) 14 (56.0)
Practical A serious illness which requires me to stay close to home 25 (100) 9 (36.0)
Practical The cost of treatment, if needed 25 (100) 8 (32.0)
Practical Physical problems, such as difficulties walking or getting around 25 (100) 7 (28.0)
Practical Physical symptoms (fatigue, pain, breathing difficulties, etc) 25 (100) 7 (28.0)
Psychological Having to take more medications 25 (100) 7 (28.0)
Threat Having other problems that are more important 22 (88.0) 6 (27.3)
Psychological Anxiety about going far from my home 25 (100) 6 (24.0)
Psychological Lack of energy or motivation to make an appointment and then go 25 (100) 5 (20.0)
Threat My problems are not severe enough 25 (100) 5 (20.0)
65
Table 5.9. Univariable analysis of the association between the subscore of perceiving no threat and covariates
Perceiving No Threat
Factor Median Subscore (1st – 3rd quartiles)
Correlation Coefficient P
Age 0.13 .08 Sex Male 5 (1 – 11) Female 6 (2 – 11) .41 Married/living with partner Yes 5 (2 – 11) No 5 (1 – 11) .44 Education 0.08 .31 Diabetes as ESRD cause Yes 6 (1 – 11) No 5 (2 – 11) .81 Hypertension as ESRD cause Yes 6 (2 – 10) No 5 (1 – 11) .81 GN as ESRD cause Yes 6 (2 – 11) No 5 (1 – 11) > .99 RRT time 0.03 .66 Charlson score 0.06 .45 PHQ-2 score 0.08 .27 History of depression Yes 6 (1 – 10) No 5 (2 – 11) .97
66
Table 5.10. Univariable analysis of the association between the subscore of perceiving no benefit and covariates
Perceiving No Benefit
Factor Median Subscore (1st – 3rd quartiles)
Correlation Coefficient P
Age -0.03 .74 Sex Male 3 (1 – 6) Female 3 (0 – 5) .74 Married/living with partner Yes 3 (1 – 5) No 3 (1 – 6) .84 Education 0.13 .10 Diabetes as ESRD cause Yes 3 (0 – 5) No 3 (1 – 6) .24 Hypertension as ESRD cause Yes 3 (1 – 4) No 3 (0.5 – 6) .28 GN as ESRD cause Yes 4 (0 – 6) No 3 (1 – 5) .47 RRT time 0.09 .25 Charlson score -0.001 .99 PHQ-2 score 0.19 .01 History of depression Yes 4 (1 – 6) No 3 (1 – 5) .64
67
Table 5.11. Univariable analysis of the association between the subscore of psychological barriers and covariates
Psychological Barriers
Factor Median Subscore (1st – 3rd quartiles)
Correlation Coefficient P
Age -0.02 .81 Sex Male 7 (2 – 13) Female 10 (4 – 18) .08 Married/living with partner Yes 7 (3 – 14) No 7 (2 – 16) .87 Education 0.12 .15 Diabetes as ESRD cause Yes 7 (2 – 13) No 7 (3 – 15.5) .59 Hypertension as ESRD cause Yes 6 (1.5 – 10.5) No 7.5 (3 – 16) .15 GN as ESRD cause Yes 9 (4 – 19) No 7 (2 – 13) .17 RRT time 0.08 .29 Charlson score 0.08 .31 PHQ-2 score 0.26 < .001 History of depression Yes 10 (4 – 16) No 7 (3 – 13) .24
68
Table 5.12. Univariable analysis of the association between the subscore of social barriers and covariates
Social Barriers
Factor Median Subscore (1st – 3rd quartiles)
Correlation Coefficient P
Age -0.05 .55 Sex Male 3 (0 – 7) Female 2 (0 – 6) .67 Married/living with partner Yes 4 (1 – 8) No 2 (0 – 6) .06 Education ... 0.12 .15 Diabetes as ESRD cause Yes 2 (0 – 4) No 3 (0 – 7) .12 Hypertension as ESRD cause Yes 6 (1.5 – 10.5) No 2.5 (0 – 4.5) .38 GN as ESRD cause Yes 4 (0 – 7) No 2 (0 – 6) .60 RRT time 0.12 .12 Charlson score -0.03 .72 PHQ-2 score 0.17 .03 History of depression Yes 3 (0 – 7) No 3 (0 – 7) .99
69
Table 5.13. Univariable analysis of the association between the subscore of practical barriers and covariates
Practical Barriers
Factor Median Subscore (1st – 3rd quartiles)
Correlation Coefficient P
Age -0.08 .32 Sex Male 7 (3 – 12) Female 8 (4 – 14) .29 Married/living with partner Yes 7 (3 – 12) No 8 (3 – 13) .85 Education 0.07 .38 Diabetes as ESRD cause Yes 7 (3 – 12.5) No 8 (3.5 – 13) .76 Hypertension as ESRD cause Yes 7 (3 – 10.5) No 8 (3 – 14) .33 GN as ESRD cause Yes 11 (4 – 15) No 7 (3 – 12) .16 RRT time -0.07 .41 Charlson score 0.13 .10 PHQ-2 score 0.38 < .001 History of depression Yes 11 (5 – 16) No 7 (3 – 12) .046
70
Table 5.14. Multivariable regression model for the subscore of perceiving no benefit (log transformed)*
Variable Adjusted Beta (95% CI) Test Statistic P Adjusted P†
Omnibus F statistic (ndf, ddf) … 2.39 (11, 142) .009 ...
Intercept 1.791 (1.438, 2.144) 0.179 < .001 …
Age 0.001 (-0.002, 0.004) 0.002 .56 .84
Male sex 0.012 (-0.101, 0.125) 0.057 .83 .91
Married or partnered -0.024 (-0.085, 0.038) 0.031 .45 .82
Education level 0.012 (-0.007, 0.030) 0.009 .21 .50
Time on RRT 0.001 (0.0002, 0.001) 0.0003 .01 .03
Diabetes as ESRD cause -0.103 (-0.270, 0.065) 0.084 .23 .50
Hypertension as ESRD cause -0.111 (-0.273, 0.050) 0.082 .17 .50
GN as ESRD cause -0.037 (-0.180, 0.106) 0.072 .61 .84
Charlson score -0.001 (-0.034, 0.031) 0.017 .93 .93
History of depression -0.025 (-0.164, 0.115) 0.071 .73 .89
PHQ-2 score 0.063 (0.025, 0.102) 0.020 .001 .02 *R2 = 0.16 †Corrected P values for multiple testing.
71
Table 5.15. Multivariable regression model for the subscore of psychological barriers (log transformed)*
Variable Adjusted Beta (95% CI) Test Statistic P Adjusted P†
Omnibus F statistic (ndf, ddf) … 3.01 (11, 144) .001 ...
Intercept 2.058 (1.524, 2.593) 0.270 < .001 …
Age -0.0001 (-0.005, 0.005) 0.003 .98 .99
Male sex -0.209 (-0.382, -0.036) 0.087 .02 .10
Married or partnered -0.040 (-0.134, 0.053) 0.047 .40 .54
Education level 0.018 (-0.010, 0.045) 0.014 .21 .46
Time on RRT 0.001 (-0.0002, 0.002) 0.0005 .11 .39
Diabetes as ESRD cause 0.002 (-0.254, 0.259) 0.130 .99 .99
Hypertension as ESRD cause -0.123 (-0.378, 0.132) 0.129 .34 .53
GN as ESRD cause 0.128 (-0.095, 0.350) 0.112 .26 .47
Charlson score 0.038 (-0.013, 0.089) 0.026 .14 .39
History of depression 0.005 (-0.211, 0.220) 0.109 .97 .99
PHQ-2 score 0.104 (0.044, 0.164) 0.030 < .001 .01 *R2 = 0.19 †Corrected P values for multiple testing.
72
Table 5.16. Multivariable regression model for the subscore of social barriers (log transformed)*
Variable Adjusted Beta (95% CI) Test Statistic P Adjusted P†
Omnibus F statistic (ndf, ddf) … 2.30 (11, 143) .01 ...
Intercept 1.922 (1.493, 2.351) 0.217 < .001 …
Age 0.001 (-0.004, 0.005) 0.002 .80 .88
Male sex 0.075 (-0.067, 0.217) 0.072 .30 .47
Married or partnered -0.062 (-0.138, 0.015) 0.039 .11 .32
Education level 0.011 (-0.011, 0.033) 0.011 .34 .47
Time on RRT 0.001 (0.0003, 0.002) 0.0004 .007 .048
Diabetes as ESRD cause -0.055 (-0.264, 0.154) 0.106 .60 .74
Hypertension as ESRD cause -0.162 (-0.364, 0.041) 0.102 .12 .32
GN as ESRD cause -0.104 (-0.285, 0.077) 0.092 .26 .47
Charlson score -0.027 (-0.070, 0.015) 0.021 .21 .45
History of depression 0.008 (-0.166, 0.182) 0.088 .93 .93
PHQ-2 score 0.065 (0.017, 0.114) 0.025 .009 .048 *R2 = 0.15 †Corrected P values for multiple testing.
73
Table 5.17. Multivariable regression model for the subscore of practical barriers (log transformed)*
Variable Adjusted Beta (95% CI) Test Statistic P Adjusted P†
Omnibus F statistic (ndf, ddf) … 3.15 (11, 145) < .001 ...
Intercept 2.307 (1.832, 2.781) 0.240 < .001 … Age -0.005 (-0.010, -0.0003) 0.002 .04 .14 Male sex -0.102 (-0.254, 0.049) 0.077 .18 .51 Married or partnered -0.002 (-0.085, 0.080) 0.042 .95 .95 Education level 0.012 (-0.013, 0.036) 0.012 .34 .76 Time on RRT 0.0002 (-0.001, 0.001) 0.0004 .63 .91 Diabetes as ESRD cause -0.032 (-0.258, 0.194) 0.114 .78 .91 Hypertension as ESRD cause -0.024 (-0.248, 0.199) 0.113 .83 .91 GN as ESRD cause 0.078 (-0.118, 0.275) 0.100 .43 .79 Charlson score 0.055 (0.010, 0.100) 0.023 .02 .09 History of depression 0.041 (-0.149, 0.231) 0.096 .67 .91 PHQ-2 score 0.101 (0.048, 0.154) 0.027 < .001 .003 *R2 = 0.19 †Corrected P values for multiple testing.
74
Figure 5.1. Recruitment of hemodialysis patients
75
Figure 5.2. PHQ-2 score of the participants
0
10
20
30
40
50
60
0 1 2 3 4 5 6
Patie
nts,
%
PHQ-2 Score
Positive for Depressive Symptoms
76
Figure 5.3. History of diagnosis and treatment of depression
77
A B
Figure 5.4. A: (Diagrams on the left side) Histograms showing the number of individuals (frequency) across all the possible barrier subscores. B: (Diagrams on the right side) Graphs showing the number of participants (frequency) plotted against the number of barriers perceived by that individual (ie, item score ≥ 3). These are reported for each of the five barrier constructs.
78
A B
Figure 5.4. cont’d.
79
Figure 5.5. Percentage of participants who perceived one barrier or more by PHQ-2 results
0
10
20
30
40
50
60
70
80
90
100
Perceiving no Threat
Perceiving no Benefit
Psychological Barriers
Social Barriers Practical Barriers
All Barriers
Perc
enta
ge o
f Pat
ient
s with
Bar
riers
PHQ < 3
PHQ ≥ 3
80
Figure 5.6. Most frequently perceived barriers to screening for depression among participants without depressive symptoms (PHQ-2 < 3), as compared to those in participants with depressive symptoms (PHQ-2 ≥ 3)
81
6. DISCUSSION
The purpose of this chapter is:
• To summarize finding of this study • To discuss limitations of the study, their impact on the interpretation of data, and measures
to address them • To outline conclusions made based on the findings and implications and future directions
6.1. COMMENTARY
6.1.1. Summary of Study Results
This cross-sectional questionnaire study is the first to report specifically, from the patient’s
perspective, barriers that impact how successful an SPD aimed at reducing depression would
be. It shows a substantial proportion of the target population of ESRD patients would have at
least one reason to find participation in the SPD extremely difficult or impossible. Most
commonly, patients had concerns about the side effects of antidepressant medications and
being prescribed more medications. Many patients did not feel they were at risk of
depression or felt that their symptoms were unlikely to warrant intervention. Perhaps most
concerningly, the results showed that those at highest risk of having undetected depression
were those with greatest number of barriers to the SPD.
6.1.2. Proportion of patients with barriers to the screening program for depression
Although there are no other studies asking whether an SPD can be implemented in the ESRD
population, our results are consistent with much of the literature in the general population
and other special populations looking at barriers to treatment of depression.122,123 The
proportion of hemodialysis patients in this study who reported to have barriers was similar to
that reported by two surveys of dialysis and cancer patients designed to address barriers to
treatment of (and not screening for) depression,110,116 supporting the accuracy of the results
shown here.
82
The proportion of hemodialysis patients who reported to have barriers was higher in this
study than in the one reported by Mohr et al, which used the original PBPT questionnaire
with the primary care patients.108 This can be explained by the observation that poor health
status increases the likelihood of perceiving barriers to psychological treatment. Moreover,
since the adapted PBPT questionnaire used in the present study had 11 additional items,
covering a broader intervention of interest (the SPD), it is expected that patients would
identify more barriers.
6.1.3. Barriers to the screening program for depression
The findings of this study support our hypothesis that patients undergoing hemodialysis have
barriers to the SPD mostly because of concerns about antidepressant medications (perceiving
no benefit) and failing to accept the risk of becoming depressed (perceiving no threat).
Similarly, a considerable proportion of peritoneal dialysis patients has been shown to reject
medical treatment for depression when diagnosed .111 Several other studies have addressed
patients’ concerns about antidepressant medications in different populations such as cancer
patients and pregnant women.116,117
Perceiving no threat is another major barrier to treatment for depression in dialysis patients
and other populations. In two studies on patients undergoing dialysis, many of those who
screened positive for depression did not feel depressed and did not feel the need for
help.110,111 Similar findings have been documented in the general population; for example,
the most common reason for the Ontarians who did not seek help for their mental health
problems was their expectation that the problem would get better by itself (not perceiving the
severity of the disease).122
Although most of the health care expenses of dialysis patients are paid by the government,
the cost of treatment was the next most frequent barrier among the participants of the present
study. The costs of mental health care are less of a barrier in Canada compared to the United
States.122 However, regarding the high prevalence of depression among dialysis patients,
potential costs of the treatment is still a factor that may influence patient’s decision about
participation in the SPD.
83
The role of other practical barriers, especially those related to physical problems, was
considerable in our cohort. These barriers are not addressed by most of the other
studies.110,116,122,123 In the primary care setting, less than one-fifth of the patients perceive
barriers related to physical problems.108 The relatively higher rates of barriers due to physical
problems were expected in our cohort because of the medical status and higher average age
of dialysis patients. Interestingly, unadjusted analyses of the findings of this study showed
that older hemodialysis patients and those with more comorbidities were less likely to
participate in the SPD because of their physical problems.
6.1.4. Correlation between depressive symptoms and barriers
Similar to the finding of this study, primary care patients with depressive symptoms are more
likely to perceive barriers to psychotherapy.108 However, this association was not found in
cancer patients.116 In addition to methodological limitations, such as type II error in the latter
study, differences in characteristics of dialysis and cancer patients may explain their
differences in association between depression and perceiving barriers to mental health care.
The data presented here are the first to explore characteristics of the barriers in those with
and without depressive symptoms. Concerns about side effects of the medications were the
most frequently perceived barrier among PHQ-positive patients, but barriers related to
perceiving no threat ranked lower in this subgroup. It is speculated that patients who reported
to have depressive symptoms had already accepted the presence of the threat of depression.
However, lack of perception of the severity of depressive symptoms seemed to be a barrier to
treatment of depression, as 27% the patients with depressive symptoms still believed that
their other problems were more important.
6.1.5. Correlation between time on renal replacement therapy and barriers
The duration of RRT was found to be predictive of barriers. This relationship was noted for
multiple (albeit not all) barrier constructs, suggesting that patients on dialysis for a longer
time may be at higher risk of not participating in the SPD, when they might need it. This may
be because patients who have been on dialysis for a longer time have experienced a higher
burden of disease over time and are exhausted by their health care experiences, while patients
84
who start chronic dialysis are more willing to seek help for their problems. The former group
have adapted to their condition and may be less interested in receiving new treatments. Over
time, patients may have more physical impairment and face physical barriers due to disability
in addition to other barriers.148 This can make these patients sensitive to their disabilities,
explaining their social barriers to mental health care.
6.1.6. Correlation between socio-demographic factors and barriers
Although the analyses initially demonstrated significant associations between some of the
barrier subscores and age, sex, and marital status, multivariable analyses failed to show these
associations. These socio-demographic factors are associated with depression in many
populations, including dialysis patients,18,54,108,143 and are expected to explain some variations
in perceiving barriers.122,112 However, the literature has failed to demonstrate the link
between these factors and barriers. This might be because barriers constitute several different
concerns, each of which needs be examined separately rather than together with other barrier
items as subscales.
6.1.7. Correlation between comorbidities and barriers
Consistent with other patient populations,116 the data of this study did not demonstrate any
relationship between comorbidity and barriers. However, our findings contrast with the work
done in the general Canadian population, demonstrating that comorbidities increase the
likelihood of perceiving barriers to mental health services use.123 One likely explanation is
that cohorts of patients with serious chronic illnesses such as ESRD and cancer are more
homogenous with regards to comorbidity scores compared to the general population.
Additionally, the role of comorbidities is masked by the greater influence of the primary
disease on their perception of their psychological health status. To further explore the
predictive effects of comorbidities, a larger sample size might be needed for studies of such
populations.
85
6.2. LIMITATIONS
6.2.1. Study population
The findings of this study are limited to the population recruited; thus, patients of other
cultures or those with visual or cognitive issues who were excluded may have barriers not
identified in this survey. Non-participation in the survey was another major challenge to this
study. One speculates that those who were not interested in participation in this study were
likely to have depression and were less motivated to participate in research. Several strategies
were used to reduce non-participation. These included strategies such as allowing patients to
get familiar with the researcher, introducing the study as a ‘survey focusing on the opinion of
patients,’ and restricting the number of questions about mood. Although these strategies were
successful in minimising non-participation, the findings are based on a limited sample of
ESRD patients who were less likely to have barriers (volunteer bias). More than 70% of the
study cohort perceived barriers, and we speculate that even more patients will be reluctant to
participate in an SPD.
Some barriers such as those related to concerns about medications, or physical limitations
might be the common ones among different ESRD populations, while some others such as
attitudes towards depressive symptoms and benefits of treatment may vary among different
groups. For example, it has been shown that different ethnicities perceive mental health
disorders differently.119,124 Addressing barriers in other groups will require studies with
tailored barriers questionnaires to the target population.
6.2.2. Measurement of primary outcome
Critically reviewing the available scales,108,116,120 I adapted the newly developed PBPT
scale108 that measures barriers to psychological treatment in the primary care setting.
Sensibility: The content and face validity of the PBPT imposed some limitations to the study.
I added items relevant to the concept of screening and the clinical condition of the dialysis
patients. Content validity of the adapted version was assessed by three content experts to
ensure all items are relevant, all possible barriers are covered, and there is balance between
barrier constructs in the scale. When applied to the dialysis population, the following
86
comments were made by patients: (1) Cultural issues that might serve as barriers to mental
health care were pointed out by two patients. In my search of possible barriers, I identified
barriers particularly related to the cultural and religious beliefs. These barriers were
addressed in studies focusing on specific groups such as ethnic minorities and
refugees.114,119,124,126 I decided not to address these types of barriers because of the diversity
among the patients approached. It was not possible to include all possible barriers related to
various cultural and religious beliefs in one questionnaire. (2) Some items were confusing to
the participants because of the complexity of the sentences. For example, one might have
possessed the idea that “I would prefer to decide when I need help for depression on my
own,” but eventually would not have a difficulty with participation in the SPD. Such items
combined a question about an attitude with which a respondent would agree or disagree and a
question on the extent to which that attitude would act as a barrier to the SPD. This problem
had been discussed by the content experts involving in the content validity of the
questionnaire and was later pointed out by a few patients. However, we opted not to change
the structure of the original PBPT, in order not to impact the validity of the questionnaire.
The use of the other available scale120 would have prevented this problem, but the length of
that questionnaire and its more subjective approach to the concept of barriers were its major
limitations.
The added items to the original PBPT did not considerably change the overall proportion of
patients with barriers, but allowed a better understanding of concerns of the patients about
screening. The sensitivity analysis demonstrated that the percentage of patients with barriers
did not reduce dramatically (66% versus 73%). This shows that the adapted PBPT captured
the same group of patients as compared to the original PBPT. An additional 11 patients were
identified only by the modified questionnaire. The added items in the adapted version were
among the most important barriers of the dialysis patients, which would have not been
identified by the original PBPT.
Adding items limited my use of the constructs suggested for the original PBPT using this
method. I defined five hypothetical constructs for the adapted barriers. These construct were
meticulously identified based on theory. The items were categorized based on these
constructs. Similarly, Pepin et al used a priori subdomains for their barriers questionnaire.120
87
However, this categorization is dependent on the viewpoints of the researcher and items
might be misclassified. I performed item-total correlation analysis, which showed acceptable
correlation of the each item with the sum of other items in each construct. This indicates that
items categorized to each construct were addressing a common concept.
The adapted PBPT was successfully applied in the dialysis population, and only 6 patients
refused to complete the questionnaire. The large print and limiting the number of items to six
in each page were deemed to be helpful. Negative reactions to the questions were observed in
a few cases, mainly because the respondents believed they were not depressed or at risk of
depression. Nonetheless, the relatively high participation rate indicates that the questionnaire
was overall acceptable by the patients.
Reliability: I assessed the internal consistency of the questionnaire and its subscales. Internal
consistency is an index of reliability that examines whether those items supposed to measure
one specific construct produce similar results. Coefficients less than 0.70 indicate
heterogeneity of items and those greater than 0.90 are indicative of redundancy of items.129
The Cronbach α for the barrier subscales ranged from 0.75 to 0.89 for subscales. The overall
coefficient for the adapted PBPT was 0.95, indicating that some items measure the same
concept and thus are redundant. The original PBPT’s Cronbach α coefficient was quite high,
too (0.92),108 and generally, adding items to a scale increases its Cronbach α coefficient. This
may cause inflation of the scores, but I assume that its effect on the percentage of patients
with barriers was minimal, because of the definition I used for having barriers; response rates
≥ 3 for one or more items that address a same concept (redundancy) does not change the
result, as a respondent is categorized as barrier-positive with one or more such responses.
One must assess the validity of a scale to ensure it measures the construct it is supposed to
measure. Validation against a gold standard measure, if available, is the ideal approach
(criterion validity). It was impossible to assess criterion validity of the adapted PBPT,
because of lack of a gold standard for measurement of barriers. Concurrent diagnosis and
treatment of depression was considered gold standard by Mohr et al108; however, it is not a
reliable measure in the hemodialysis population, because of the high rate of under-diagnosis
and under-treatment of depression in ESRD patients. As an alternative, I planned to evaluate
88
the convergent validity of the adapted PBPT using the past history of a diagnosis or treatment
of depression. Convergent validity is evaluation of the correlation between the scale of
interest and a measure with which it is assumed to be associated. The data, however, failed to
show a significant association between the barrier total score and subscores and a positive
history of depression. This might be because of the wide range of the time since diagnosis or
treatment of depression, which imposes two problems: first, data were prone to recall bias
and dependent on the memory of the patients, and second, a history of depression before the
start of RRT (which was the case in 43%) does not necessarily correlate with the barriers
perceived by that patient after developing ESRD.
A better variable for evaluation of the convergent validity is depressive scores. Like Mohr et
al, this study showed the strong correlation of depressive symptoms with perceiving barriers.
However, since depressive score was my main exposure variable, I did not use it a priori as
my variable of choice for convergent validity.
6.2.3. Measurement of covariates
By having the patients fill out a depression questionnaire, it was possible that they would
become sensitive to the issue of depression. A heightened awareness of depression could
contaminate patients’ responses to the barrier questions; they might have tried to support
their responses to a depression scale throughout the barrier questionnaire. For example, those
who believed that they were not depressed would emphasize that they do not need to be
screened or vice versa. To avoid contamination, the PHQ-2 was used, which is an ultra-short
scale. Asking only two questions about mood problem, the PHQ-2 does not further explore
depressive symptoms, and thus, minimises sensitisation to the issue of depression as a
disease.
The PHQ-2 has not been validated in the dialysis population; however, two large studies on
hemodialysis patients used the 2 items of feeling depressed and anhedonia of the Short Form-
36 questionnaire of quality of life as a depression scale and reported that their 2-item scale
was associated with comorbidities and mortality of dialysis patients (convergent validity).18,54
Asking about two depressive symptoms, the PHQ-2 might lack sensitivity and its score might
not well represent the severity of depressive symptoms. The specificity of the PHQ-2,
89
however, is considerably high (95%) in the general population. This can be explained by the
fact that the questions in PHQ-2 are about feeling depressed and lack of interest, which tackle
the two core symptoms of depression without which diagnosis of MDD and minor depression
is not made. We can expect that those identified to have depressive symptoms in our cohort
are highly likely to be depressed. This subgroup is in fact our target for screening; thus, we
have explored barriers in a group of dialysis patients who are very likely to have depression.
6.2.4. Analysis of data
The concept of the main outcome measure in this study (perceived barriers to an SPD) is
complex and thus prone to lack of precision and misinterpretation. The barrier data was
dichotomized to identify patients with barriers. However, participants who perceived less
difficulty in taking part in the SPD (slightly or moderately difficult) will not necessarily
participate in the SPD, as they might have several mild to moderate barriers that impact their
decision cumulatively. In other words, a minimum of one barrier was considered as being
positive for barriers, while one could perceive some degrees of difficulty participating in the
SPD (‘slightly’ or ‘moderately difficult’) because of several issues listed in the questionnaire,
but ultimately be categorized as not having barriers.
The use of barrier subscores allowed a more powerful assessment of the associations between
barriers and covariates than the use of dichotomized barrier data. The use of scores is a better
estimate of the severity of barriers, because patients perceiving several barriers have higher
barrier subscores even if they do not choose response rates that would categorize them as
barrier positive (dichotomized data). However, the skewed nature of the barrier subscores
imposed natural logarithmic transformation of the outcome variable. This prevents
interpretation of the estimates provided by the regression analysis, because the unit changes
in estimates are not clinically meaningful when log transformation is applied.
6.2.5. Interpretation of data
Measurement of barriers using a self-report scale does not necessarily provide us with the
eventual reasons behind the decisions of not taking part in an SPD when offered. One’s
behaviour can be predicted by determination of a particular constellation of one’s beliefs
90
only when these beliefs are assumed to remain unchanged prior to, at the time of, and after
the behaviour.138 The cognitive dissonance theory describes that when individuals have
conflicting judgements about an issue, they try to reduce dissonance by altering their existing
attitudes, adding new ones to create a consistent belief system, or reducing the importance of
any one of the dissonant beliefs.149 In the case of taking part in an SPD, patients may have
conflicting ideas about the risk of becoming depressed, the benefits and side effects of
treatment, and the psychological and social barriers to the use of mental health services.
When asked about these, they try to alter their viewpoints and give sound reasons about their
decision in a hypothetical situation. When asked to participate in an SPD in a real situation,
their answers might not be those they considered before as their reasoning may alter because
of the continuing conflicts between their cognitions. Even, the decision to accept or reject
screening for depression in a real-life situation may itself modify one’s perception of the
barriers they had considered before making the decision.138 Thus, the identified barriers allow
us to understand areas of patients’ concerns rather than finding the exact reasons behind
patients’ decisions about participation in the SPD.
6.3. IMPLICATIONS
Through non-participation, the SPD may fail to identify those who are in need of further
assessments and treatment, and even if screening occurs, it may not lead to diagnosis and
treatment. The findings in this thesis highlight the need for revisions in the recommendations
about screening for depression. Currently, screening for depression is recommended not only
in high-risk groups, but also in the general population.100-102 These recommendations are
criticized because of the challenges arising from the accuracy of screening tools and health
policy issues.99,103 Our study adds patient-related barriers to these concerns. These barriers
can be a major challenge because they are extremely common even in a high-risk group of
patients with the highest rate of depression. More concerningly, this study demonstrated that
the targets of the SPD, those with depressive symptoms, are the ones who are less likely to
participate in the SPD.
91
We identified several reasons because of which individuals may not consider taking part in
the SPD. The possibility of reducing these barriers should be studied by designing and
evaluating educational programs around the seriousness of depression and various
treatments. In the case of dialysis patients, the strong association of depression and mortality
warrants programs aimed at reducing the burden of depression.
6.4. FUTURE DIRECTIONS
The validity and reliability of the adapted PBPT requires further investigation in an
independent population. However, the considerably high burden of barriers shown in this
study warrant devising strategies to reduce possible barriers to participation in an SPD. It
may be plausible to develop a number of solutions aimed at education of those delivering
health care, as well as patients and their families; improved awareness of the clinical
significance of depressive symptoms; and through routine use of interventions such as
support groups or counselling. Further solutions may be identified, for example through
focus groups interviewing patients or clinicians with expertise in the field. All such strategies
would need to be studied for their effectiveness in increasing patient participation prior to
their introduction into a clinical setting.
6.5. CONCLUSIONS
This study addressed the patient-perceived barriers to an SPD aimed at preventing and
treating depression among hemodialysis patients. It was shown that there were significant
perceived beliefs and limitations regarding participation in an SPD for depression and that
those with a high burden of depression symptoms were more likely to perceive these barriers
than those without. The relationship between the presence of depressive symptoms and
perceived barriers to SPD suggest that implementation of an SPD for depression is likely to
systematically miss those individuals it is meant to benefit because of non-participation.
92
7. REFERENCES
1. Williams JW, Jr., Noel PH, Cordes JA, Ramirez G, Pignone M. Is this patient clinically
depressed? JAMA. 2002;287:1160-70.
2. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders -
Text Revision (DSMIV-TR). 4th ed. Arlington, VA: American Psychiatric Association; 2000.
3. Vasiliadis HM, Lesage A, Adair C, Wang PS, Kessler RC. Do Canada and the United States
differ in prevalence of depression and utilization of services? Psychiatr Serv. 2007;58:63-71.
4. Wang J, Williams J, Lavorato D, Schmitz N, Dewa C, Patten SB. The incidence of major
depression in Canada: the National Population Health Survey. J Affect Disord. 2010;123:158-
63.
5. Cuijpers P, Smit F. Excess mortality in depression: a meta-analysis of community studies. J
Affect Disord. 2002;72:227-36.
6. Patten SB, Williams JV, Lavorato D, Li Wang J, Khaled S, Bulloch AG. Mortality associated
with major depression in a Canadian community cohort. Can J Psychiatry. 2011;56:658-66.
7. Egede LE. Major depression in individuals with chronic medical disorders: prevalence,
correlates and association with health resource utilization, lost productivity and functional
disability. Gen Hosp Psychiatry. 2007;29:409-16.
8. Sullivan MD, O'Connor P, Feeney P, et al. Depression Predicts All-Cause Mortality:
Epidemiological evaluation from the ACCORD HRQL substudy. Diabetes Care.
2012;35:1708-15.
9. Damen NL, Versteeg H, Boersma E, et al. Depression is independently associated with 7-year
mortality in patients treated with percutaneous coronary intervention: Results from the
RESEARCH registry. Int J Cardiol. 2012.
10. Barth J, Schumacher M, Herrmann-Lingen C. Depression as a risk factor for mortality in
patients with coronary heart disease: a meta-analysis. Psychosom Med. 2004;66:802-13.
11. Farrokhi F, Abedi N, Beyene J, Kurdyak P, Jassal SV. Depression as a risk factor for mortality
in patients on dialysis: a systematic review and meta-analysis. Paper presented at: American
Sociaty of Nephrology Kidney Week; 30 October, 2012, 2012; San Diego.
93
12. Levey AS, Coresh J, Balk E, et al. National Kidney Foundation practice guidelines for chronic
kidney disease: evaluation, classification, and stratification. Ann Intern Med. 2003;139:137-
47.
13. Canadian Institute for Health Information. 2011 CORR Report—Treatment of End-Stage
Organ Failure in Canada, 2000 to 2009: Canadian Institute for Health Information; 2011.
14. Cukor D, Cohen SD, Peterson RA, Kimmel PL. Psychosocial aspects of chronic disease:
ESRD as a paradigmatic illness. J Am Soc Nephrol. 2007;18:3042-55.
15. Cohen SD, Norris L, Acquaviva K, Peterson RA, Kimmel PL. Screening, diagnosis, and
treatment of depression in patients with end-stage renal disease. Clin J Am Soc Nephrol.
2007;2:1332-42.
16. Kimmel PL, Cohen SD, Peterson RA. Depression in patients with chronic renal disease: where
are we going? J Ren Nutr. 2008;18:99-103.
17. Kutner NG, Zhang R, Huang Y, Johansen KL. Depressed mood, usual activity level, and
continued employment after starting dialysis. Clin J Am Soc Nephrol. 2010;5:2040-5.
18. Lopes AA, Bragg J, Young E, et al. Depression as a predictor of mortality and hospitalization
among hemodialysis patients in the United States and Europe. Kidney Int. 2002;62:199-207.
19. Murtagh FE, Addington-Hall J, Higginson IJ. The prevalence of symptoms in end-stage renal
disease: a systematic review. Adv Chronic Kidney Dis. 2007;14:82-99.
20. Katon WJ. Clinical and health services relationships between major depression, depressive
symptoms, and general medical illness. Biol Psychiatry. 2003;54:216-26.
21. Hedayati SS, Minhajuddin AT, Afshar M, Toto RD, Trivedi MH, Rush AJ. Association
between major depressive episodes in patients with chronic kidney disease and initiation of
dialysis, hospitalization, or death. JAMA. 2010;303:1946-53.
22. Tsai YC, Chiu YW, Hung CC, et al. Association of symptoms of depression with progression
of CKD. Am J Kidney Dis. 2012;60:54-61.
23. Bonnet F, Irving K, Terra JL, Nony P, Berthezene F, Moulin P. Anxiety and depression are
associated with unhealthy lifestyle in patients at risk of cardiovascular disease.
Atherosclerosis. 2005;178:339-44.
24. Cukor D, Rosenthal DS, Jindal RM, Brown CD, Kimmel PL. Depression is an important
contributor to low medication adherence in hemodialyzed patients and transplant recipients.
Kidney Int. 2009;75:1223-9.
94
25. DiMatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with
medical treatment: meta-analysis of the effects of anxiety and depression on patient adherence.
Arch Intern Med. 2000;160:2101-7.
26. Everett KD, Brantley PJ, Sletten C, Jones GN, McKnight GT. The relation of stress and
depression to interdialytic weight gain in hemodialysis patients. Behav Med. 1995;21:25-30.
27. Golden SH, Lazo M, Carnethon M, et al. Examining a bidirectional association between
depressive symptoms and diabetes. JAMA. 2008;299:2751-9.
28. Safdar N, Baakza H, Kumar H, Naqvi SA. Non-compliance to diet and fluid restrictions in
haemodialysis patients. J Pak Med Assoc. 1995;45:293-5.
29. Ziegelstein RC, Fauerbach JA, Stevens SS, Romanelli J, Richter DP, Bush DE. Patients with
depression are less likely to follow recommendations to reduce cardiac risk during recovery
from a myocardial infarction. Arch Intern Med. 2000;160:1818-23.
30. Friend R, Hatchett L, Wadhwa NK, Suh H. Serum albumin and depression in end-stage renal
disease. Adv Perit Dial. 1997;13:155-7.
31. Koo JR, Yoon JW, Kim SG, et al. Association of depression with malnutrition in chronic
hemodialysis patients. Am J Kidney Dis. 2003;41:1037-42.
32. Empana JP, Sykes DH, Luc G, et al. Contributions of depressive mood and circulating
inflammatory markers to coronary heart disease in healthy European men: the Prospective
Epidemiological Study of Myocardial Infarction (PRIME). Circulation. 2005;111:2299-305.
33. Howren MB, Lamkin DM, Suls J. Associations of depression with C-reactive protein, IL-1,
and IL-6: a meta-analysis. Psychosom Med. 2009;71:171-86.
34. Irwin MR, Miller AH. Depressive disorders and immunity: 20 years of progress and
discovery. Brain Behav Immun. 2007;21:374-83.
35. Rief W, Hennings A, Riemer S, Euteneuer F. Psychobiological differences between depression
and somatization. J Psychosom Res. 2010;68:495-502.
36. Serebruany VL, Glassman AH, Malinin AI, et al. Platelet/endothelial biomarkers in depressed
patients treated with the selective serotonin reuptake inhibitor sertraline after acute coronary
events: the Sertraline AntiDepressant Heart Attack Randomized Trial (SADHART) Platelet
Substudy. Circulation. 2003;108:939-44.
37. Barefoot JC, Burg MM, Carney RM, et al. Aspects of social support associated with
depression at hospitalization and follow-up assessment among cardiac patients. J Cardiopulm
Rehabil. 2003;23:404-12.
95
38. Suzuki R, Rygh LJ, Dickenson AH. Bad news from the brain: descending 5-HT pathways that
control spinal pain processing. Trends Pharmacol Sci. 2004;25:613-7.
39. Katon W, Sullivan M, Walker E. Medical symptoms without identified pathology: relationship
to psychiatric disorders, childhood and adult trauma, and personality traits. Ann Intern Med.
2001;134:917-25.
40. Cukor D, Coplan J, Brown C, Peterson RA, Kimmel PL. Course of depression and anxiety
diagnosis in patients treated with hemodialysis: a 16-month follow-up. Clin J Am Soc
Nephrol. 2008;3:1752-8.
41. Walters BA, Hays RD, Spritzer KL, Fridman M, Carter WB. Health-related quality of life,
depressive symptoms, anemia, and malnutrition at hemodialysis initiation. Am J Kidney Dis.
2002;40:1185-94.
42. Halen NV, Cukor D, Constantiner M, Kimmel PL. Depression and mortality in end-stage renal
disease. Curr Psychiatry Rep. 2012;14:36-44.
43. McEwen BS. Protective and damaging effects of stress mediators. N Engl J Med.
1998;338:171-9.
44. Schiepers OJ, Wichers MC, Maes M. Cytokines and major depression. Prog
Neuropsychopharmacol Biol Psychiatry. 2005;29:201-17.
45. Ranjit N, Diez-Roux AV, Shea S, et al. Psychosocial factors and inflammation in the multi-
ethnic study of atherosclerosis. Arch Intern Med. 2007;167:174-81.
46. Hedayati SS, Finkelstein FO. Epidemiology, diagnosis, and management of depression in
patients with CKD. Am J Kidney Dis. 2009;54:741-52.
47. Wilson B, Spittal J, Heidenheim P, et al. Screening for depression in chronic hemodialysis
patients: comparison of the Beck Depression Inventory, primary nurse, and nephrology team.
Hemodial Int. 2006;10:35-41.
48. Balogun RA, Balogun SA, Kepple AL, et al. GDS-15 as a predictor of mortality in elderly
hemodialysis patients. Am J Kidney Dis. 2011;57:A23.
49. Boulware LE, Liu Y, Fink NE, et al. Temporal relation among depression symptoms,
cardiovascular disease events, and mortality in end-stage renal disease: contribution of reverse
causality. Clin J Am Soc Nephrol. 2006;1:496-504.
50. Diefenthaeler EC, Wagner MB, Poli-de-Figueiredo CE, Zimmermann PR, Saitovitch D. Is
depression a risk factor for mortality in chronic hemodialysis patients? Rev Bras Psiquiatr.
2008;30:99-103.
96
51. Drayer RA, Piraino B, Reynolds CF, 3rd, et al. Characteristics of depression in hemodialysis
patients: symptoms, quality of life and mortality risk. Gen Hosp Psychiatry. 2006;28:306-12.
52. Einwohner R, Bernardini J, Fried L, Piraino B. The effect of depressive symptoms on survival
in peritoneal dialysis patients. Perit Dial Int. 2004;24:256-63.
53. Kojima M, Hayano J, Suzuki S, et al. Depression, alexithymia and long-term mortality in
chronic hemodialysis patients. Psychother Psychosom. 2010;79:303-11.
54. Lacson E, Jr., Li NC, Guerra-Dean S, Lazarus M, Hakim R, Finkelstein FO. Depressive
symptoms associate with high mortality risk and dialysis withdrawal in incident hemodialysis
patients. Nephrol Dial Transplant. 2012;27:2921-8.
55. Lopes AA, Albert JM, Young EW, et al. Screening for depression in hemodialysis patients:
associations with diagnosis, treatment, and outcomes in the DOPPS. Kidney Int.
2004;66:2047-53.
56. Mahajan S, Tiwari SC, Kalra V, et al. Analysis of depression and its effect on outcome among
adult Indian peritoneal dialysis patients. Perit Dial Int. 2007;27:94-6.
57. Riezebos RK, Nauta KJ, Honig A, Dekker FW, Siegert CE. The association of depressive
symptoms with survival in a Dutch cohort of patients with end-stage renal disease. Nephrol
Dial Transplant. 2010;25:231-6.
58. Santos PR. Evaluation of objective and subjective indicators of death in a period of one year in
a sample of prevalent patients under regular hemodialysis. BMC Res Notes. 2012;5:24.
59. Shulman R, Price JD, Spinelli J, Shulman R, Price JD, Spinelli J. Biopsychosocial aspects of
long-term survival on end-stage renal failure therapy. Psychol Med. 1989;19:945-54.
60. van den Beukel TO, Verdujin M, Boeschoten EW, Krediet RT, Dekker FW, Van Dijk S. The
Association of Baseline, New-Onset, and Persistent Depressive Symptoms with Mortality in
ESRD Patients. Paper presented at: American Society of Nephrology Kidney Week, 2010.
61. Craven JL, Rodin GM, Littlefield C. The Beck Depression Inventory as a screening device for
major depression in renal dialysis patients. Int J Psychiatry Med. 1988;18:365-74.
62. Hedayati SS, Bosworth HB, Kuchibhatla M, Kimmel PL, Szczech LA. The predictive value of
self-report scales compared with physician diagnosis of depression in hemodialysis patients.
Kidney Int. 2006;69:1662-8.
63. Watnick S, Wang PL, Demadura T, Ganzini L. Validation of 2 depression screening tools in
dialysis patients. Am J Kidney Dis. 2005;46:919-24.
64. Hedayati SS, Minhajuddin AT, Toto RD, Morris DW, Rush AJ. Validation of depression
screening scales in patients with CKD. Am J Kidney Dis. 2009;54:433-9.
97
65. Butt AA, Skanderson M, McGinnis KA, et al. Impact of hepatitis C virus infection and other
comorbidities on survival in patients on dialysis. J Viral Hepat. 2007;14:688-96.
66. Genestier S, Meyer N, Chantrel F, et al. Prognostic survival factors in elderly renal failure
patients treated with peritoneal dialysis: A nine-year retrospective study. Perit Dial Int.
2010;30.
67. Hedayati SS, Grambow SC, Szczech LA, Stechuchak KM, Allen AS, Bosworth HB.
Physician-diagnosed depression as a correlate of hospitalizations in patients receiving long-
term hemodialysis. Am J Kidney Dis. 2005;46:642-9.
68. Miskulin D, Bragg-Gresham J, Gillespie BW, et al. Key comorbid conditions that are
predictive of survival among hemodialysis patients. Clin J Am Soc Nephrol. 2009;4:1818-26.
69. Soucie JM, McClellan WM, Soucie JM, McClellan WM. Early death in dialysis patients: risk
factors and impact on incidence and mortality rates. J Am Soc Nephrol. 1996;7:2169-75.
70. Tsai CJ, Loh EW, Lin CH, Yu TM, Chan CH, Lan TH. The correlation of antidepressive
agents and the mortality of end-stage renal disease: a Health Insurance Research Database
Study in Taiwan. Nephrology (Carlton). 2012;17:390-7.
71. Hedayati SS, Bosworth HB, Briley LP, et al. Death or hospitalization of patients on chronic
hemodialysis is associated with a physician-based diagnosis of depression. Kidney Int.
2008;74:930-6.
72. Finkelstein FO, Wuerth D, Finkelstein SH. An approach to addressing depression in patients
with chronic kidney disease. Blood Purif. 2010;29:121-4.
73. Kurella M, Kimmel PL, Young BS, Chertow GM. Suicide in the United States end-stage renal
disease program. J Am Soc Nephrol. 2005;16:774-81.
74. McDade-Montez EA, Christensen AJ, Cvengros JA, Lawton WJ. The role of depression
symptoms in dialysis withdrawal. Health Psychol. 2006;25:198-204.
75. Christensen AJ, Wiebe JS, Smith TW, et al. Predictors of survival among hemodialysis
patients: effect of perceived family support. Health Psychol. 1994;13:521-5.
76. Kutner NG, Lin LS, Fielding B, et al. Continued survival of older hemodialysis patients:
investigation of psychosocial predictors. Am J Kidney Dis. 1994;24:42-9.
77. Peterson RA, Kimmel PL, Sacks CR, et al. Depression, perception of illness and mortality in
patients with end-stage renal disease. Int J Psychiatry Med. 1991;21:343-54.
78. Kimmel PL, Peterson RA, Weihs KL, et al. Multiple measurements of depression predict
mortality in a longitudinal study of chronic hemodialysis outpatients. Kidney Int.
2000;57:2093-8.
98
79. Hedayati SS, Yalamanchili V, Finkelstein FO. A practical approach to the treatment of
depression in patients with chronic kidney disease and end-stage renal disease. Kidney Int.
2012;81:247-55.
80. Chilcot J, Wellsted D, Farrington K. Screening for depression while patients dialyse: an
evaluation. Nephrol Dial Transplant. 2008;23:2653-9.
81. National Collaborating Centre for Mental Health. Depression in adults with a chronic physical
health problem: the treatment and management London: National Institute for Health and
Clinical Excellence; 2009.
82. Simon GE, VonKorff M. Somatization and psychiatric disorder in the NIMH Epidemiologic
Catchment Area study. Am J Psychiatry. 1991;148:1494-500.
83. Atalay H, Solak Y, Biyik M, et al. Sertraline treatment is associated with an improvement in
depression and health-related quality of life in chronic peritoneal dialysis patients. Int Urol
Nephrol. 2010;42:527-36.
84. Blumenfield M, Levy NB, Spinowitz B, et al. Fluoxetine in depressed patients on dialysis. Int
J Psychiatry Med. 1997;27:71-80.
85. Doyle GD, Laher M, Kelly JG, Byrne MM, Clarkson A, Zussman BD. The pharmacokinetics
of paroxetine in renal impairment. Acta Psychiatr Scand Suppl. 1989;350:89-90.
86. Kalender B, Ozdemir AC, Yalug I, Dervisoglu E. Antidepressant treatment increases quality
of life in patients with chronic renal failure. Ren Fail. 2007;29:817-22.
87. Kamo T, Horikawa N, Tsuruta Y, Miyasita M, Hatakeyama H, Maebashi Y. Efficacy and
pharmacokinetics of fluvoxamine maleate in patients with mild depression undergoing
hemodialysis. Psychiatry Clin Neurosci. 2004;58:133-7.
88. Kennedy SH, Craven JL, Rodin GM. Major depression in renal dialysis patients: an open trial
of antidepressant therapy. J Clin Psychiatry. 1989;50:60-3.
89. Koo JR, Yoon JY, Joo MH, et al. Treatment of depression and effect of antidepression
treatment on nutritional status in chronic hemodialysis patients. Am J Med Sci. 2005;329:1-5.
90. Levy NB, Blumenfield M, Beasley CM, Jr., et al. Fluoxetine in depressed patients with renal
failure and in depressed patients with normal kidney function. Gen Hosp Psychiatry.
1996;18:8-13.
91. Spigset O, Hagg S, Stegmayr B, Dahlqvist R. Citalopram pharmacokinetics in patients with
chronic renal failure and the effect of haemodialysis. Eur J Clin Pharmacol. 2000;56:699-703.
92. Turk S, Atalay H, Altintepe L, et al. Treatment with antidepressive drugs improved quality of
life in chronic hemodialysis patients. Clin Nephrol. 2006;65:113-8.
99
93. Wuerth D, Finkelstein SH, Finkelstein FO. The identification and treatment of depression in
patients maintained on dialysis. Semin Dial. 2005;18:142-6.
94. Hosseini SH, Espahbodi F, Mirzadeh Goudarzi SMM. Citalopram versus psychological
training for depression and anxiety symptoms in hemodialysis patients. Iran J Kidney Dis.
2012;6:446-51.
95. Duarte PS, Miyazaki MC, Blay SL, Sesso R. Cognitive-behavioral group therapy is an
effective treatment for major depression in hemodialysis patients. Kidney Int. 2009;76:414-21.
96. Ouzouni S, Kouidi E, Sioulis A, Grekas D, Deligiannis A. Effects of intradialytic exercise
training on health-related quality of life indices in haemodialysis patients. Clin Rehabil.
2009;23:53-63.
97. Watnick S, Kirwin P, Mahnensmith R, Concato J. The prevalence and treatment of depression
among patients starting dialysis. Am J Kidney Dis. 2003;41:105-10.
98. Peckham CS, Dezateux C. Issues underlying the evaluation of screening programmes. Br Med
Bull. 1998;54:767-78.
99. Thombs BD, Coyne JC, Cuijpers P, et al. Rethinking recommendations for screening for
depression in primary care. CMAJ. 2012;184:413-8.
100. MacMillan HL, Patterson CJ, Wathen CN, et al. Screening for depression in primary care:
recommendation statement from the Canadian Task Force on Preventive Health Care. CMAJ.
2005;172:33-5.
101. Pignone MP, Gaynes BN, Rushton JL, et al. Screening for depression in adults: a summary of
the evidence for the U.S. Preventive Services Task Force. Ann Intern Med. 2002;136:765-76.
102. National Collaborating Centre for Mental Health. Depression: the treatment and management
of depression in adults. London: The British Psychological Society & the Royal College of
Psychiatrists; 2010.
103. Palmer SC, Coyne JC. Screening for depression in medical care: pitfalls, alternatives, and
revised priorities. J Psychosom Res. 2003;54:279-87.
104. Baas KD, Wittkampf KA, van Weert HC, et al. Screening for depression in high-risk groups:
prospective cohort study in general practice. Br J Psychiatry. 2009;194:399-403.
105. Saver BG, Van-Nguyen V, Keppel G, Doescher MP. A qualitative study of depression in
primary care: missed opportunities for diagnosis and education. J Am Board Fam Med.
2007;20:28-35.
106. Nutting PA, Rost K, Dickinson M, et al. Barriers to initiating depression treatment in primary
care practice. J Gen Intern Med. 2002;17:103-11.
100
107. Hollon SD, Munoz RF, Barlow DH, et al. Psychosocial intervention development for the
prevention and treatment of depression: promoting innovation and increasing access. Biol
Psychiatry. 2002;52:610-30.
108. Mohr DC, Ho J, Duffecy J, et al. Perceived barriers to psychological treatments and their
relationship to depression. J Clin Psychol. 2010;66:394-409.
109. Mojtabai R, Olfson M, Sampson NA, et al. Barriers to mental health treatment: results from
the National Comorbidity Survey Replication. Psychol Med. 2011;41:1751-61.
110. Johnson S, Dwyer A. Patient perceived barriers to treatment of depression and anxiety in
hemodialysis patients. Clin Nephrol. 2008;69:201-6.
111. Wuerth D, Finkelstein SH, Kliger AS, Finkelstein FO. Chronic peritoneal dialysis patients
diagnosed with clinical depression: results of pharmacologic therapy. Semin Dial.
2003;16:424-7.
112. Aromaa E, Tolvanen A, Tuulari J, Wahlbeck K. Personal stigma and use of mental health
services among people with depression in a general population in Finland. BMC Psychiatry.
2011;11:52.
113. Blumenthal R, Endicott J. Barriers to seeking treatment for major depression. Depress
Anxiety. 1996;4:273-8.
114. Conner KO, Copeland VC, Grote NK, et al. Barriers to treatment and culturally endorsed
coping strategies among depressed African-American older adults. Aging Ment Health.
2010;14:971-83.
115. Ell K. Depression care for the elderly: reducing barriers to evidence-based practice. Home
Health Care Serv Q. 2006;25:115-48.
116. Endo C, Akechi T, Okuyama T, et al. Patient-perceived barriers to the psychological care of
Japanese patients with lung cancer. Jpn J Clin Oncol. 2008;38:653-60.
117. Goodman JH. Women's attitudes, preferences, and perceived barriers to treatment for perinatal
depression. Birth. 2009;36:60-9.
118. Kravitz RL, Paterniti DA, Epstein RM, et al. Relational barriers to depression help-seeking in
primary care. Patient Educ Couns. 2011;82:207-13.
119. Lee S, Guo WJ, Tsang A, et al. Perceived barriers to mental health treatment in metropolitan
China. Psychiatr Serv. 2010;61:1260-2.
120. Pepin R, Segal DL, Coolidge FL. Intrinsic and extrinsic barriers to mental health care among
community-dwelling younger and older adults. Aging Ment Health. 2009;13:769-77.
101
121. Rochlen AB, Paterniti DA, Epstein RM, Duberstein P, Willeford L, Kravitz RL. Barriers in
diagnosing and treating men with depression: a focus group report. Am J Mens Health.
2010;4:167-75.
122. Sareen J, Jagdeo A, Cox BJ, et al. Perceived barriers to mental health service utilization in the
United States, Ontario, and the Netherlands. Psychiatr Serv. 2007;58:357-64.
123. Wang J. Perceived barriers to mental health service use among individuals with mental
disorders in the Canadian general population. Med Care. 2006;44:192-5.
124. Ward EC, Clark le O, Heidrich S. African American Women's beliefs, coping behaviors, and
barriers to seeking mental health services. Qual Health Res. 2009;19:1589-601.
125. Weinberger MI, Bruce ML, Roth AJ, Breitbart W, Nelson CJ. Depression and barriers to
mental health care in older cancer patients. Int J Geriatr Psychiatry. 2011;26:21-6.
126. Wong EC, Marshall GN, Schell TL, et al. Barriers to mental health care utilization for U.S.
Cambodian refugees. J Consult Clin Psychol. 2006;74:1116-20.
127. Woodall A, Howard L, Morgan C. Barriers to participation in mental health research: findings
from the Genetics and Psychosis (GAP) Study. Int Rev Psychiatry. 2011;23:31-40.
128. Mohr DC, Hart SL, Howard I, et al. Barriers to psychotherapy among depressed and
nondepressed primary care patients. Ann Behav Med. 2006;32:254-8.
129. Streiner DL, Norman GR. Health measurement scales: a practical guide to their development
and use. 4th ed. New York: Oxford University Press; 2008.
130. Bandura A. Social Foundations of Thought and Action. Englewood Cliffs, NJ: Prentice Hall;
1986.
131. Henshaw EJ, Freedman-Doan CR. Conceptualizing mental health care utilization using the
Health Belief Model. Clin Psychol Sci Prac. 2009;16:420-39.
132. Rosenstock IM, Strecher VJ, Becker MH. Social learning theory and the Health Belief Model.
Health Educ Q. 1988;15:175-83.
133. Ajzen I. Perceived behavioral control, self-efficacy, locus of control, and the theory of planned
behavior. J Appl Soc Psychol. 2002;32:665-83.
134. Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J
Health Soc Behav. 1995;36:1-10.
135. Cameron L, Leventhal EA, Leventhal H. Symptom representations and affect as determinants
of care seeking in a community-dwelling, adult sample population. Health Psychol.
1993;12:171-9.
102
136. Cramer KM. Psychological antecedents to help-seeking behavior: A reanalysis using path
modeling structures. J Couns Psychol. 1999;46:381-7.
137. Leventhal H, Meyer D, Nerenz D. The common-sense representation of illness danger. In:
Rachman S, ed. Medical psychology. Vol 2. New York: Pergamon; 1980:7-30.
138. Rosenstock IM. Why people use health services. Milbank Mem Fund Q. 1966;44:Suppl:94-
127.
139. Becker MH, Maiman LA. Strategies for enhancing patient compliance. J Community Health.
1980;6:113-35.
140. Link BG, Phelan JC. Conceptualizing stigma. Annu Rev Sociol. 2001;27:363-85.
141. Kroenke K, Spitzer RL, Williams JB. The Patient Health Questionnaire-2: validity of a two-
item depression screener. Med Care. 2003;41:1284-92.
142. Gyamlani G, Basu A, Geraci S, et al. Depression, screening and quality of life in chronic
kidney disease. Am J Med Sci. 2011;342:186-91.
143. National Institute for Clinical Excellence. Depression: core interventions in the management
of depression in primary and secondary care. London: Her Majesty's Stationary Office; 2004.
144. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic
comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373-
83.
145. Hsueh IP, Lin JH, Jeng JS, Hsieh CL. Comparison of the psychometric characteristics of the
functional independence measure, 5 item Barthel index, and 10 item Barthel index in patients
with stroke. J Neurol Neurosurg Psychiatry. 2002;73:188-90.
146. Basu A, Manning WG, Mullahy J. Comparing alternative models: log vs Cox proportional
hazard? Health Econ. 2004;13:749-65.
147. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful
approach to multiple testing. J R Stat Soc Series B Stat Methodol. . 1995;57:289-300.
148. Kurella Tamura M, Covinsky KE, Chertow GM, Yaffe K, Landefeld CS, McCulloch CE.
Functional status of elderly adults before and after initiation of dialysis. N Engl J Med.
2009;361:1539-47.
149. Festinger L. A theory of cognitive dissonance. Stanford, CA: Staford University Press; 1957.
103
8. APPENDICES
Appendix A. The systematic review manuscript on the association of depression with mortality among dialysis patients
TITLE Depression as a Risk Factor for Mortality in Patients on Dialysis: A Systematic Review and Meta-analysis AUTHORS Farhat Farrokhi, MD, MSc (candidate)1, Neda Abedi, MD2, Joseph Beyene, PhD1,3, Paul Kurdyak, MD, PhD1,4,5, Sarbjit Vanita Jassal, MD, MSc1,6
1Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Canada 2Department of Psychiatry, University of Saskatchewan, Saskatoon, Canada 3Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Canada 4Institute for Clinical Evaluative Sciences, University of Toronto, Toronto, Ontario, Canada 5Centre for Addiction and Mental Health, Toronto, Ontario, Canada 6Division of Nephrology, Department of Medicine, University Health Network, University of Toronto, Toronto, Canada ABSTRACT Background: An estimated 20-40% of dialysis patients suffer from depressive symptoms. We aimed to systematically review and analyze the association between depression and mortality risk in adult patients on chronic maintenance dialysis. Methods: Searching MEDLINE, EMBASE, PsychINFO, we identified studies examining the relationship between depression, measured as depressive symptoms or clinical diagnosis, and mortality among patients receiving chronic renal dialysis. Quality appraisal was done using the Newcastle-Ottawa Scale. The inverse variance method and random effects model were used to summarize the effect sizes, and the trim-and-fill method to adjust for potential publication bias. Results: Fifteen of 31 included studies showed a significant association between depression and mortality, including 5 of 6 studies with more than 6000 participants. A significant link was established between presence of depressive symptoms and mortality (HR, 1.51; 95% CI, 1.35 to 1.69; I2=40%), based on 12 studies reporting depressive symptoms using depression scales (n=21055; mean age, 57.6 years). After adjusting for publication bias, presence of depressive symptoms remained a significant predictor of mortality (HR, 1.45; 95% CI, 1.27 to 1.65). In addition, combining across 6 studies reporting per unit change of depression score (n=7857) resulted in a significant effect (HR, 1.04 per unit change of score; 95% CI,
104
1.01 to 1.06; I2=74%). Based on 3 retrospective studies reporting odds ratios, the association of depression (diagnosis in medical charts) with mortality was nonsignificant. Conclusions: There is considerable between-study heterogeneity in the reports of depressive symptoms among dialysis patients, likely caused by high variability in the way depressive symptoms are measured. However, the overall significant independent effect of depressive symptoms on survival of dialysis patients warrants studying the underlying mechanisms of this relationship and the potential benefits of interventions to improve depression on the outcomes. BACKGROUND Over the recent years, there has been a paradigm shift in the treatment of end-stage renal disease (ESRD), with more attention being paid to nonrenal symptoms.1 Depression is one such symptom.1,2 It is estimated that 27% (range, 5-58%) of ESRD patients have depressive symptoms.3 Several factors contribute to development of depressive symptoms, such as loss of the primary role in the family, inability to continue working, decreased physical function, medications, dietary restrictions, and tethering to “lifesaving” dialysis treatments.2,4,5 Depressive symptoms, accompanied by a high burden of physical symptoms, are associated with poor adherence to treatment and loss of wellbeing in ESRD patients.1,2 Accordingly, depression has been suggested as an independent risk factor for hard outcomes, including mortality. Earlier studies linking depression to mortality risk in ESRD patients were inconclusive; while recent large studies have demonstrated an independent association between depression and mortality.6-9 Nonetheless, there is considerable variation in the reported findings, in part due to differences in study design, statistical methodology, and the method used to ascertain depression (eg, self-report depressive symptoms versus physician-diagnosed depression). A systematic review of the literature to estimate the strength of the relationship between depression and death in dialysis patients would help inform if there is any possible survival benefit from developing interventional strategies. The objective of this review is to evaluate the association between depression, measured either as depressive symptoms using depression scales or clinical diagnosis, and mortality of adult patients on long-term dialysis. METHODS Criteria for selection of studies Type of studies: All observational cohort studies, case-control studies, and longitudinal studies published in either abstract or full form that included an assessment of the ability of depressive symptoms or clinical depression to predict mortality were included. Letters to the editor were included if contained relevant data. Non-English articles were considered for inclusion provided that an abstract in English was available. Type of participants: Adult patients (>18 years) receiving dialysis (home or hospital-based hemodialysis and peritoneal dialysis modalities) as a long-term renal replacement therapy. Types of exposure measures: Depression was defined as documentation of clinical depression (major depression, minor depression, or dysthymia) or depressive symptoms in any of the following ways: (1) a diagnosis of depression based on structured or semi-structured clinical interviews validated against the Diagnostic and Statistical Manual of Mental Disorders or the International Classification of Diseases criteria, (2) any clinical record of the diagnosis of depression, (3) measurement of depressive symptoms using a
105
depression scale, and (4) a measurement of depressive symptoms by subscales of other questionnaires such as quality of life scales, if recognized and validated as an indicator of depressive symptoms. Studies that measured depression using personality trait scales were excluded. Studies that measured depression prior to renal dialysis initiation were also excluded. Types of outcome: The primary outcome of interest was all-cause mortality after the start of dialysis therapy. Studies with assessment of the outcome <3 months or >10 years after depression measurement were not included. The window of observation was selected based on the assumption that mortality occurring either very early or late after the screening is unlikely to be depression-related. Composite outcomes were not considered unless authors could provide analyses on mortality alone. Search methods for identification of studies Electronic searches: A comprehensive search strategy was applied with the help of an expert librarian. Three online databases of MEDLINE (1948-August 2012), EMBASE (1947-August 2012), and PsychINFO (1806-August 2012) were searched using text words of “dialysis” OR “hemodialysis,” “depression” OR “depressive,” and “mortality” OR “survival,” OR “death” as well as the vocabulary terms specific to each database. No filters for language, publication status, or study design were applied. Other resources: In order to reduce publication bias, the following resources were also searched: bibliographic information of pertinent review articles; proceedings of international conferences (World Congress of Nephrology, American Society of Nephrology Renal Week, and European Renal Association-European Dialysis and Transplant Association Congress; 2006-August 2012); and dissertations (Proquest; 1637-August 2012). Authors of abstracts were contacted for detailed data where possible. Data collection and analysis Selection of studies: Search results were imported into Endnote X for Windows (Thomson Reuters, New York, NY), and duplications excluded. After inclusive screening of the titles by one review author (by FF) to exclude irrelevant records two authors independently reviewed the refined list of records (FF, NA) for eligibility. A third review author (SVJ) was involved to achieve consensus. In all data extraction discrepancies were resolved by consensus. Data extraction and management: Two review authors (FF, NA) independently extracted study characteristics and effect estimates. Double data entry into the RevMan version 5.1 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark) was applied. In case of missing data, the investigators contacted the authors. Assessment of risk of bias in included studies: Data quality was appraised independently by three review authors (FF, NA, SVJ). A modified version of the Newcastle-Ottawa Scale (NOS)10 for cohort studies was used for quality appraisal. We considered the clinical structured interview of all participants for diagnosis of depression as the highest level of ascertainment of exposure with identification of depressive symptoms using a standardized depression scale applied to all participants as acceptable. Studies with documentation of depression or depressive symptoms without assessment in selected groups did not meet ascertainment of exposure quality standard. Clinically important determinants for mortality for the NOS comparability tool included age and the presence or absence of both diabetes and cardiac disease in adjusted analyses. Since
106
depression-related mechanisms are most likely chronic (with the rare exception of suicide and dialysis withdrawal), the authors agreed on at least 1-year followup for quality appraisal. A maximum loss-to-followup rate of <10% was acceptable. In addition to assessment using the NOS, we dichotomized the studies based on low and high risk of bias. Studies that met criteria for representativeness of the exposed cohort ≥3 criteria for selection, ≥1 for comparability, and ≥2 in the outcome sections were considered low risk. Measures of exposure effect: For data presented as a dichotomous variable (presence or absence of clinical depression and depressive symptoms scores above or below a cutoff), crude and adjusted hazard ratios (HRs) and/or odds ratios (ORs) were extracted. Studies reporting data presented as a continuous variable had the HRs and ORs for each unit change in scores extracted. When risk estimates were not reported, crude ORs were calculated with 95% confidence intervals based on the two-by-two table of exposure and outcome, if possible. Standard errors of the risk estimates were calculated using standard methods. Assessment of heterogeneity: Between-study heterogeneity was investigated by the chi-square test (P <0.1), and the I2 statistic was used to quantify its impact.11 Data synthesis: Eligible studies for quantitative data synthesis were imported into RevMan 5.1. Data were categorized based on depression measurement method and analyses were done separately for each category. Meta-analysis was done to estimate a summary measure of the ORs and HRs for binary and time-to-event outcomes, respectively. The inverse variance method was used to test the overall effect for reports of crude and adjusted HRs and ORs. Since we anticipated significant between-study heterogeneity, we used the random effects model as a conservative approach to summarize the findings. Assessment of publication bias: The funnel plot was used to visualize potential publication bias. We used the trim-and-fill method to adjust the calculated effect sizes for publication bias. The trim-and-fill method uses a nonparametric technique to identify studies in the asymmetric part of the funnel plot. These studies are trimmed off, and the symmetric remainder is used to estimate the true center of the funnel. Then, the trimmed studies and their missing counterparts are replaced around the center to return the final estimate and its variance based on the filled funnel plot.12 The R statistical software package version 2.15.1 (R Development Core Team, Vienna, Austria) was used. Subgroup analysis and investigation of heterogeneity: Subgroup analyses were planned a priori for the following possible sources of heterogeneity: (1) followup time (<1 year, 1 to 3 years, and >3 years); (2) time of measurement of depressive symptoms in relation to dialysis start (incident versus prevalent or mixed prevalent and incident dialysis patients); (3) country (the United States versus other countries); and (4) single versus repeated measurements of depression. Sensitivity analysis: The following sensitivity analyses were planned a priori: (1) exclusion of studies with high risk of bias defined using NOS (2) exclusion of studies with <100 participants. RESULTS Results of the search The search yielded 2528 records, of which 63 were potentially relevant. The kappa index for agreement between the two reviewers was 0.96. The full texts and abstracts of the selected reports were screened for eligibility for inclusion. Authors were contacted for additional information; attempts were successful in 7 cases with additional data provided by authors,
107
successful in 2 cases but authors were unable to provide additional data, and unsuccessful in 13 cases. Thirty-two studies were excluded (9 with duplicate data, 9 with nonstandard measurements of depression, 6 with insufficient data or unavailable full texts, 5 with mixed study populations with chronic kidney disease or transplant patients, and 3 with outcomes of interest other than all-cause mortality). Thus, a total of 31 articles were included for qualitative synthesis, and data required for quantitative synthesis were available for 25 studies (Figure 1). Qualitative Analyses Characteristics of included studies: Table 1 is a summary of the characteristics of 31 studies included in the systematic review (n=67075; mean age, 60.4 years; male, 54.4%).6,8,9,13-40 Sample sizes varied from 40 to 16965. Eight studies were on samples smaller than 100, while 6 were on samples larger than 6000 patients (Table 1). Publication dates ranged from 1989 to 2012. Studies were of prospective (n=25) or retrospective (n=6) cohort design. Fifteen studies were from the United States, Eighteen were limited to hemodialysis patients, 4 to peritoneal dialysis patients, and 9 included both. Most were representative of the dialysis population, except for one study that was limited to only men,22 one that predominantly included African-Americans,8 and 3 that were limited to older dialysis patients.13,20,24 In 7 studies, patients were incident cases at the start of followup, while in the remaining studies, they were either prevalent or a mix of prevalent and incident dialysis patients. The followup duration was up to 1 year in 7 studies, 1 to 3 years in 14 studies, and ≥3 years in 10. In 25 studies, depression was measured using a screening scale at baseline (22 studies) or repeatedly (3 studies). Eleven of these studies reported Beck Depression Inventory scores, with various cutoffs to identify screen-positive patients for depression. The remaining 14 studies used 8 different scales (Table 1). In a second group of studies (n=8), clinical depression (physician-diagnosed) was identified from medical records. None of these studies assessed all patients systematically. Characteristics of excluded studies: Table 2 summarizes the excluded studies.7,41-71 Three studies were exclude because they only reported the (not recognized specifically for depression). A total of 7 studies were excluded because they did not determine depression using a disease-specific-screening tool (mental health component of the SF-36, n=3; depression subscale of Personality Trait Inventories, n=4; single-item questionnaire, n=1; and unspecified, n=1). Only data from the most complete and updated reports from large studies, such as those by the Dialysis Outcomes and Practice Patterns Study (DOPPS) 49,57,64 and the Netherlands Co-operative Study Adequacy of Dialysis study 65, were included to limit duplication. Risk of bias in included studies: None of the included studies provided clinical structured interview of all participants. Overall, 7 of the 31 studies were considered to be prone to a high risk of bias (Table 3). The main sources of bias were high or unreported rate of loss-to-followup and suboptimal method of ascertainment of the exposure. Effects of exposure: Depression was reported in three ways: presence/absence of depressive symptoms (based on depression scale cutoff), depression score (continuous data), and physician diagnosis. Fourteen studies reported depression using depression scale cutoffs. Depression was estimated at 29.7% patients (point prevalence range 8.1% to 65.4%, n=21146). Physician-diagnosed depression was reported in 16.2% of 48465 patients in 7 studies (point prevalence range 4.4% to 27.7%).
108
Fifteen of 31 studies found a relationship between depression and mortality after multivariable analysis (Table 1). The association was documented in 5 of 6 studies with large samples (n >6000)14,25-27,29 and all studies with repeated measurements of depression.6,8,40 Quantitative synthesis Twenty-five studies provided data appropriate for quantitative data synthesis (13 reporting a significant association with mortality and 12 without).6,8,9,13,15,17-19,21-23,25-28,30-33,35-40 Of those studies excluded from the meta-analysis, 4 did not find a statistically significant relationship between mortality and depression while 2 did. Presence of depressive symptoms. Fifteen studies reporting dichotomized results of depression scales were analysed together. Combining across studies reporting unadjusted HRs and ORs, the presence of depressive symptoms was a significant predictor of mortality (Table 4). Based on 12 studies that reported HRs adjusted for covariates (n=21055; mean age, 57.6 years; males, 53%; hemodialysis 99%),6,8,9,13,17,18,21,23,25-27,40 the presence of depressive symptoms was an independent predictor of mortality among dialysis patients, increasing the risk of death by 51% (adjusted HR, 1.51; 95% CI, 1.35 to 1.69; P <.001; I2=40%; Figure 2). The funnel plot (Figure 3) visualized the potential for publication bias and the trim-and-fill method predicted that there were 5 hypothetically missing studies and imputed them. The adjusted meta-analysis after imputation gave an adjusted HR of 1.45 (95% CI, 1.27 to 1.65) in those with depressive symptoms. Depressive score. Nine studies reported depression scale scores as a continuous variable. The unadjusted effect estimate (HR) for depressive scores showed a significant increase in mortality per unit change (Table 4). Combining across 6 studies reporting adjusted analyses (n=7857; mean age, 61.3 years; males, 53%; hemodialysis, 98%), depressive scores was significantly associated with mortality (adjusted HR, 1.04; 95% CI, 1.01 to 1.06; P=.002; Figure 4). The effect size, however, was based on heterogeneous results (I2=74%). Physician-diagnosed depression. Five studies reported physician-diagnosed depression in medical records; 2 studies reported the DOPPS I and DOPPS II results, both of which showed a significant link between depression and mortality in univariable analysis (HR, 1.42; 95% CI, 1.27 to 1.59)27 and multivariable analysis (HR, 1.23 and 1.26; 95% CI, 1.08 to 1.40 and 1.10 to 1.44, in DOPPS I and DOPPS II, respectively).26,27 The effect size for a diagnosis of depression was presented as OR in 3 studies. All of these studies failed to show a significant association between depression and death(Table 4). Hedayati et al reported an adjusted OR of 0.98 (95% CI, 0.72 to 1.34) and Soucie et al presented an adjusted OR, 1.30 (95% CI, 0.88 to 1.51).22,36 Subgroups and sensitivity analysis. Subgroup and sensitivity analyses across different population and study quality groupings showed similar results across all analyses suggesting a robust relationship between depression and mortality (Tables 5 and 6). DISCUSSION Using formal meta-analytical techniques, we have shown an independent association between depressionand subsequent increased mortality risk in the dialysis population. The magnitude of the increased risk was 1.45 times in the presence of depressive symptoms. This is seen regardless of the methods used to evaluate depression, characteristics of the population studied, and the study design, suggesting that our findings are robust. Of note studies with repeated measurements of depressive symptoms (longitudinal assessment) demonstrated a 1.66 times higher mortality risk with depressive symptoms.6,8,40 The only
109
group in which a nonsignificant relationship was found included the 3 retrospective studies where depression was determined only from documentation in medical records.22,36,39 Our results cannot determine causality. However, the characteristics consistent with causality include the finding that there is increased mortality risk per unit increase in depression scores (dose dependency), and the observation that depression can affect immune and inflammatory responses72-74 nutrition, and adherence to treatment,75,76 together with the associated risk of suicide and withdrawing dialysis (biological plausibility).77-79 Similar to our findings in the dialysis population, depression has been associated with all-cause mortality in other medically unwell populations and in population-based studies. Lemogne et al80 demonstrated in a 12-year population-based survey that depressive mood predicted natural mortality among men and women. A systematic review has confirmed this association on the community level, reporting an overall 1.8 relative risk of dying in depressed subjects.81 Higher mortality rates are seen in cancer patients with depression.82-85 Pinquart and coworkers reviewed 43 studies on cancer patients and reported a 22% higher risk of mortality among those with depression diagnosis or depressive symptoms.84 ESRD patients may be at higher risk of depression-related mortality due to concomitant comorbidity; however, our study found a relationship despite adjustment for other comorbidities. Non-ESRD patients with diabetes or cardiovascular diseases, for example, have increased mortality if depressed (relative risk, 1.8 and 2.0, respectively).86-88 Patients on dialysis are a unique group, as they suffer from several comorbidities such as diabetes and cardiovascular disease, as well as considerable functional decline.89 Given the accumulation of several factors related to depression and mortality in the setting of ESRD, the relationship between depression and patient outcome is deemed to be more complex. A large proportion of the ESRD population are aged 65 years and older, and interestingly, it has been shown that the non-demented elderly individuals have 41% higher risk of mortality if they are depressed.90 We believe depression to be a common but under-recognized comorbidity among dialysis patients. information about the outcomes with treatment for depression is limited to a few observational and uncontrolled trials that show promising results in those on dialysis ts.91-93 Both nonpharmacological and pharmacological therapies also appear promising.94,95 Nonetheless, only one-third of those diagnosed with depression receive any treatment.26,96 We propose the mortality risk reported by this systematic review provides sufficient incentive for further studies investigating the effectiveness of screening and therapy in the dialysis population. We implemented strong and effective meta-analytical methods. Our study, however, is limited by the quality and heterogeneity of the studies included. Using the NOS for quality appraisal, we identified 7 studies with a high risk of bias. In many cases, the quality of data was not optimal. One example of this is that none of the studies used structured clinical interview in the entire sample of dialysis patients, while many documented depression or depressive symptoms only from medical charts or a single self-report assessment. Our results are also limited by the heterogeneity caused by the variation in measurement methods, design, and analysis although we minimized this by conducting a variety of subgroup analyses. We found that that studies with a limited followup of dialysis patients after measurement of depression (<1 year) did not show a relationship between depression and mortality. This finding supports the decision to restrict our main analysis to those with longer followup duration.
110
Meta-analyses are prone to a variety of selection biases. We attempted to limit the impact of publication bias using the trim-and-fill method. To avoid missing studies where depression was measured as a covariate, but not explicitly reported (because it was statistically nonsignificant) we used extensive searching techniques, including review of conference proceedings, thesis dissertations and any reports investigating in the dialysis population. In conclusion, the present systematic review and meta-analysis supports the independent association between depression and mortality risk among patients on chronic maintenance dialysis. These data suggest further study evaluating if screening or case finding strategies are effective, and evaluation of the effectiveness of treatments for depression in the dialysis population through well-designed clinical trials. ACKNOWLEDGEMENTS We would like to thank Dr Prakesh Shah for his excellent guidance through completion of this review. Also we thank the authors of the reviewed publications who contributed to our work by providing additional information of their studies: Rasheed Balogun, Joseph Chilcot, Konstadina Griva, Eduardo Lacson, Rolf Peterson, Cheuk-Chun Szeto, Melissa Thong, and Tessa van den Beukel. FINANCIAL SUPPORT None. REFERENCES 1. Cukor D, Cohen SD, Peterson RA, Kimmel PL. Psychosocial aspects of chronic disease: ESRD as a paradigmatic
illness. J Am Soc Nephrol. Dec 2007;18(12):3042-3055. 2. Cohen SD, Norris L, Acquaviva K, Peterson RA, Kimmel PL. Screening, diagnosis, and treatment of depression in
patients with end-stage renal disease. Clin J Am Soc Nephrol. Nov 2007;2(6):1332-1342. 3. Murtagh FE, Addington-Hall J, Higginson IJ. The prevalence of symptoms in end-stage renal disease: a systematic
review. Adv Chronic Kidney Dis. Jan 2007;14(1):82-99. 4. Kimmel PL, Cohen SD, Peterson RA. Depression in patients with chronic renal disease: where are we going? J Ren
Nutr. Jan 2008;18(1):99-103. 5. Kutner NG, Zhang R, Huang Y, Johansen KL. Depressed Mood, Usual Activity Level, and Continued Employment
after Starting Dialysis. Clin J Am Soc Nephrol. Sep 30 2010;5(11):2040-2045. 6. Boulware LE, Liu Y, Fink NE, et al. Temporal relation among depression symptoms, cardiovascular disease events,
and mortality in end-stage renal disease: contribution of reverse causality. Clin J Am Soc Nephrol. May 2006;1(3):496-504.
7. Chilcot J, Davenport A, Wellsted D, Firth J, Farrington K. An association between depressive symptoms and survival in incident dialysis patients. Nephrol Dial Transplant. Oct 4 2010.
8. Kimmel PL, Peterson RA, Weihs KL, et al. Multiple measurements of depression predict mortality in a longitudinal study of chronic hemodialysis outpatients. Kidney Int. May 2000;57(5):2093-2098.
9. Riezebos RK, Nauta KJ, Honig A, Dekker FW, Siegert CE. The association of depressive symptoms with survival in a Dutch cohort of patients with end-stage renal disease. Nephrol Dial Transplant. Jan 2010;25(1):231-236.
10. Wells GA, Shea B, O'Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed Jul 3, 2011.
11. Higgins J, Thompson S, Deeks J, Altman D. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557. 12. Duval S, Tweedie R. Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in
meta-analysis. Biometrics. Jun 2000;56(2):455-463. 13. Balogun RA, Balogun SA, Kepple AL, et al. GDS-15 as a predictor of mortality in elderly hemodialysis patients. Am J
Kidney Dis. 2011;57:A23. 14. Butt AA, Skanderson M, McGinnis KA, et al. Impact of hepatitis C virus infection and other comorbidities on survival
in patients on dialysis. J Viral Hepat. Oct 2007;14(10):688-696. 15. Chilcot J, Wellsted D, Farrington K. Illness perceptions predict survival in haemodialysis patients. Am J Nephrol.
2011;33:358-363. 16. Christensen AJ, Wiebe JS, Smith TW, et al. Predictors of survival among hemodialysis patients: effect of perceived
family support. Health Psychol. Nov 1994;13(6):521-525.
111
17. Diefenthaeler EC, Wagner MB, Poli-de-Figueiredo CE, Zimmermann PR, Saitovitch D. Is depression a risk factor for mortality in chronic hemodialysis patients? Rev Bras Psiquiatr. Jun 2008;30(2):99-103.
18. Drayer RA, Piraino B, Reynolds CF, 3rd, et al. Characteristics of depression in hemodialysis patients: symptoms, quality of life and mortality risk. General hospital psychiatry. Jul-Aug 2006;28(4):306-312.
19. Einwohner R, Bernardini J, Fried L, Piraino B. The effect of depressive symptoms on survival in peritoneal dialysis patients. Perit Dial Int. May-Jun 2004;24(3):256-263.
20. Genestier S, Meyer N, Chantrel F, et al. Prognostic survival factors in elderly renal failure patients treated with peritoneal dialysis: A nine-year retrospective study. Perit Dial Int. 2010;30(218-26).
21. Griva K, Stygall J, Hankins M, et al. Cognitive impairment and 7-year mortality in dialysis patients. Am J Kidney Dis. Oct 2010;56(4):693-703.
22. Hedayati SS, Grambow SC, Szczech LA, Stechuchak KM, Allen AS, Bosworth HB. Physician-diagnosed depression as a correlate of hospitalizations in patients receiving long-term hemodialysis. Am J Kidney Dis. Oct 2005;46(4):642-649.
23. Kojima M, Hayano J, Suzuki S, et al. Depression, alexithymia and long-term mortality in chronic hemodialysis patients. Psychother Psychosom. 2010;79(5):303-311.
24. Kutner NG, Lin LS, Fielding B, et al. Continued survival of older hemodialysis patients: investigation of psychosocial predictors. Am J Kidney Dis. Jul 1994;24(1):42-49.
25. Lacson E, Jr., Li NC, Guerra-Dean S, Lazarus M, Hakim R, Finkelstein FO. Depressive symptoms associate with high mortality risk and dialysis withdrawal in incident hemodialysis patients. Nephrol Dial Transplant. Jan 23 2012;27(7):2921-2928.
26. Lopes AA, Albert JM, Young EW, et al. Screening for depression in hemodialysis patients: associations with diagnosis, treatment, and outcomes in the DOPPS. Kidney Int. Nov 2004;66(5):2047-2053.
27. Lopes AA, Bragg J, Young E, et al. Depression as a predictor of mortality and hospitalization among hemodialysis patients in the United States and Europe. Kidney Int. Jul 2002;62(1):199-207.
28. Mahajan S, Tiwari SC, Kalra V, et al. Analysis of depression and its effect on outcome among adult Indian peritoneal dialysis patients. Perit Dial Int. Jan-Feb 2007;27(1):94-96.
29. Miskulin D, Bragg-Gresham J, Gillespie BW, et al. Key comorbid conditions that are predictive of survival among hemodialysis patients. Clin J Am Soc Nephrol. 2009;4:1818-1826.
30. Peng YS, Chiang CK, Hung KY, et al. Are both psychological and physical dimensions in health-related quality of life associated with mortality in hemodialysis patients: a 7-year Taiwan cohort study. Blood Purification. 2010;30(2):98-105.
31. Peterson RA, Kimmel PL, Sacks CR, et al. Depression, perception of illness and mortality in patients with end-stage renal disease. Int J Psychiatry Med. 1991;21(4):343-354.
32. Rosenthal Asher D, Ver Halen N, Cukor D. Depression and nonadherence predict mortality in hemodialysis treated end-stage renal disease patients. Hemodial Int. Apr 3 2012.
33. Santos PR. Evaluation of objective and subjective indicators of death in a period of one year in a sample of prevalent patients under regular hemodialysis. BMC Res Notes. Jan 11 2012;5(1):24.
34. Shulman R, Price JD, Spinelli J, Shulman R, Price JD, Spinelli J. Biopsychosocial aspects of long-term survival on end-stage renal failure therapy. Psychol Med. Nov 1989;19(4):945-954.
35. Simic Ogrizovic S, Jovanovic D, Dopsaj V, et al. Could depression be a new branch of MIA syndrome? Clin Nephrol. Feb 2009;71(2):164-172.
36. Soucie JM, McClellan WM, Soucie JM, McClellan WM. Early death in dialysis patients: risk factors and impact on incidence and mortality rates. J Am Soc Nephrol. Oct 1996;7(10):2169-2175.
37. Szeto CC, Chow KM, Kwan BC, et al. The impact of social support on the survival of Chinese peritoneal dialysis patients. Perit Dial Int. May-Jun 2008;28(3):252-258.
38. Takaki J, Nakao M, Yano E. The relationship of quality of life and depression to mortality in hemodialysis patients. Dial Transplant. 2005;34(8):568-575.
39. Tsai CJ, Loh EW, Lin CH, Yu TM, Chan CH, Lan TH. The correlation of antidepressive agents and the mortality of end-stage renal disease: a Health Insurance Research Database Study in Taiwan. Nephrology (Carlton). Jan 19 2012;17(4):390-397.
40. van den Beukel TO, Verdujin M, Boeschoten EW, Krediet RT, Dekker FW, Van Dijk S. The Association of Baseline, New-Onset, and Persistent Depressive Symptoms with Mortality in ESRD Patients. Paper presented at: American Society of Nephrology2010.
41. Abdel-Rahman EM, Balogun SA, Kepple A, et al. Beck Depression Inventory and survival in elderly hemodialysis patients. Nephrol Dial Transplant. Mar 25 2011;26:2064-2065.
42. Bilgic A, Alasancak R, Sezer S, Kubilay Koc S, Ibis A, Nurhan Ozdemir F. The prospective effect of psychosocial factors on peritonitis and hospitalization rates in peritoneal dialysis patients. Paper presented at: ERA-EDTA2007; Barcelona, Spain.
43. Brzosko S, Hryszko T, Mysliwiec M. Depression is underrecognized in dialysis population and predicts long term survival Paper presented at: American Society of Nephrology2010.
44. Burton HJ, Kline SA, Lindsay RM, et al. The relationship of depression to survival in chronic renal failure. Psychosom Med. Mar-Apr 1986;48(3-4):261-269.
112
45. DeOreo PB, DeOreo PB. Hemodialysis patient-assessed functional health status predicts continued survival, hospitalization, and dialysis-attendance compliance. Am J Kidney Dis. Aug 1997;30(2):204-212.
46. Devins GM, Mann J, Mandin H, et al. Psychosocial predictors of survival in end-stage renal disease. J Nerv Ment Dis. Feb 1990;178(2):127-133.
47. Fischer MJ, Kimmel PL, Greene T, et al. Elevated depressive affect is associated with adverse cardiovascular outcomes among African Americans with chronic kidney disease. Kidney Int. Sep 2011;80(6):670-678.
48. Foster FG, Cohn GL, McKegney FP, Foster FG, Cohn GL, McKegney FP. Psychobiologic factors and individual survival on chronic renal hemodialysis. A two year follow-up. I. Psychosom Med. Jan-Feb 1973;35(1):64-82.
49. Fukuhara S, Green J, Albert J, et al. Symptoms of depression, prescription of benzodiazepines, and the risk of death in hemodialysis patients in Japan. Kidney Int. Nov 2006;70(10):1866-1872.
50. Hedayati SS, Bosworth HB, Briley LP, et al. Death or hospitalization of patients on chronic hemodialysis is associated with a physician-based diagnosis of depression. Kidney Int. Oct 2008;74(7):930-936.
51. Husebye DG, Westlie L, Styrvoky TJ, et al. Psychological, social, and somatic prognostic indicators in old patients undergoing long-term dialysis. Arch Intern Med. Nov 1987;147(11):1921-1924.
52. Kazama S, Kazama JJ, Ito Y, et al. Assessment of emotional condition using self-rating depression scale (SDS) predicts 2 year life prognosis in chronic hemodialysis patients ERA_EDTA. Milan, Italy 2009.
53. Kellerman QD, Christensen AJ, Baldwin AS, Lawton WJ. Association between depressive symptoms and mortality risk in chronic kidney disease. Health Psychol. Nov 2010;29(6):594-600.
54. Kimmel PL, Peterson RA, Weihs KL, et al. Dyadic relationship conflict, gender, and mortality in urban hemodialysis patients. J Am Soc Nephrol. Aug 2000;11(8):1518-1525.
55. Kimmel PL, Peterson RA, Weihs KL, et al. Psychosocial factors, behavioral compliance and survival in urban hemodialysis patients. Kidney Int. Jul 1998;54(1):245-254.
56. Koo JR, Yoon JY, Noh JW, et al. Association of the malnutrition-inflammation-depression-arteriosclerosis (mida) syndrome with adverse cardiovascular outcome in chronic hemodialysis (hd) patients: a 5-year prospective study. World Congress of Nephrology2011.
57. Lopes AA, Elder SJ, Ginsberg N, et al. Lack of appetite in haemodialysis patients--associations with patient characteristics, indicators of nutritional status and outcomes in the international DOPPS. Nephrol Dial Transplant. Dec 2007;22(12):3538-3546.
58. Manrique J, Purror C, Jesus Unzue J, Arteaga J. Association between Dependency and Depressive Symptoms with Mortality in Hemodialysis Patients. Paper presented at: American Society of Nephrology2010.
59. Numan IM, Barklind KS, Lubin B, Numan IM, Barklind KS, Lubin B. Correlates of depression in chronic dialysis patients: morbidity and mortality. Res Nurs Health. Sep 1981;4(3):295-297.
60. Parkerson GR, Jr., Gutman RA. Health-related quality of life predictors of survival and hospital utilization. Health Care Financ Rev. Spring 2000;21(3):171-184.
61. Sapilak BJ, Melon M, Hans-Wytrychowska A, et al. Chronically hemodialysed patients - Is it possible to reduce risk of their death further?. [Polish]. Family Med Prim Care Rev. 2006;8(3):756-758.
62. Sawicka A, Maciejewski M, Marczewski K. Clock drawing test and mortality in haemodialysis patients, reversed epidemiology in 5 years observation? ERA-EDTA. Munich, Germany 2010.
63. Sayana HH, Narra MB, Amin A, Tang F, Beasley BB. Mortality and hospitalization outcomes in incident dialysis patients with change in depressive symptoms – analysis of the United States Renal Data System (USRDS). Paper presented at: American Society of Nephrology2010.
64. Tentori F, Elder SJ, Thumma J, et al. Physical exercise among participants in the Dialysis Outcomes and Practice Patterns Study (DOPPS): correlates and associated outcomes. Nephrol Dial Transplant. Sep 2010;25(9):3050-3062.
65. Thong MS, Kaptein AA, Krediet RT, et al. Social support predicts survival in dialysis patients. Nephrol Dial Transplant. Mar 2007;22(3):845-850.
66. Untas A, Thumma J, Rascle N, et al. The associations of social support and other psychosocial factors with mortality and quality of life in the dialysis outcomes and practice patterns study. Clin J Am Soc Nephrol. Jan 2011;6(1):142-152.
67. Valdes C, Garcia-Mendoza M, Rebollo P, et al. Mental health at the third month of haemodialysis as a predictor of short-term survival. Nephrol Dial Transplant. Nov 2006;21(11):3223-3230.
68. Wai L, Richmond J, Burton H, et al. Influence of psychosocial factors on survival of home-dialysis patients. Lancet. Nov 21 1981;2(8256):1155-1156.
69. Young BA, Von Korff M, Heckbert SR, et al. Association of major depression and mortality in Stage 5 diabetic chronic kidney disease. General hospital psychiatry. Mar-Apr 2010;32(2):119-124.
70. Ziarnik JP, Freeman CW, Sherrard DJ, et al. Psychological correlates of survival on renal dialysis. J Nerv Ment Dis. Mar 1977;164(3):210-213.
71. Zimmermann PR, Camey SA, Mari Jde J, Zimmermann PR, Camey SA, Mari JdJ. A cohort study to assess the impact of depression on patients with kidney disease. Int J Psychiatry Med. 2006;36(4):457-468.
72. Appels A, Bar FW, Bar J, Bruggeman C, de Baets M. Inflammation, depressive symptomtology, and coronary artery disease. Psychosom Med. Sep-Oct 2000;62(5):601-605.
73. Chilcot J, Wellsted D, Da Silva-Gane M, Farrington K. Depression on dialysis. Nephron Clin Pract. 2008;108(4):c256-264.
113
74. Kiecolt-Glaser JK, Glaser R. Depression and immune function: central pathways to morbidity and mortality. J Psychosom Res. Oct 2002;53(4):873-876.
75. Bilgic A, Akgul A, Sezer S, Arat Z, Ozdemir FN, Haberal M. Nutritional status and depression, sleep disorder, and quality of life in hemodialysis patients. J Ren Nutr. Nov 2007;17(6):381-388.
76. Cukor D, Rosenthal DS, Jindal RM, Brown CD, Kimmel PL. Depression is an important contributor to low medication adherence in hemodialyzed patients and transplant recipients. Kidney Int. Jun 2009;75(11):1223-1229.
77. Kurella M, Kimmel PL, Young BS, Chertow GM. Suicide in the United States end-stage renal disease program. J Am Soc Nephrol. Mar 2005;16(3):774-781.
78. McDade-Montez EA, Christensen AJ, Cvengros JA, Lawton WJ. The role of depression symptoms in dialysis withdrawal. Health Psychol. Mar 2006;25(2):198-204.
79. Cukor D, Coplan J, Brown C, Peterson RA, Kimmel PL. Course of depression and anxiety diagnosis in patients treated with hemodialysis: a 16-month follow-up. Clin J Am Soc Nephrol. Nov 2008;3(6):1752-1758.
80. Lemogne C, Niedhammer I, Khlat M, et al. Gender differences in the association between depressive mood and mortality: a 12-year follow-up population-based study. J Affect Disord. 2012;136(3):267-275.
81. Cuijpers P, Smit F. Excess mortality in depression: a meta-analysis of community studies. J Affect Disord. 2002;72(3):227-236.
82. Onitilo AA, Nietert PJ, Egede LE. Effect of depression on all-cause mortality in adults with cancer and differential effects by cancer site. General hospital psychiatry. Sep-Oct 2006;28(5):396-402.
83. Satin JR, Linden W, Phillips MJ. Depression as a predictor of disease progression and mortality in cancer patients: a meta-analysis. Cancer. 2009;115(22):5349-5361.
84. Pinquart M, Duberstein PR. Depression and cancer mortality: a meta-analysis. Psychol Med. 2010;40(11):1797-1810. 85. Schneider S, Moyer A. Depression as a predictor of disease progression and mortality in cancer patients: a meta-
analysis. Cancer. 2010;116(13):3304; author reply 3304-3305. 86. Damen NL, Versteeg H, Boersma E, et al. Depression is independently associated with 7-year mortality in patients
treated with percutaneous coronary intervention: Results from the RESEARCH registry. Int J Cardiol. May 3 2012. 87. Sullivan MD, O'Connor P, Feeney P, et al. Depression Predicts All-Cause Mortality: Epidemiological evaluation from
the ACCORD HRQL substudy. Diabetes Care. Aug 2012;35(8):1708-1715. 88. Barth J, Schumacher M, Herrmann-Lingen C. Depression as a risk factor for mortality in patients with coronary heart
disease: a meta-analysis. Psychosom Med. 2004;66(6):802-813. 89. Halen NV, Cukor D, Constantiner M, Kimmel PL. Depression and mortality in end-stage renal disease. Curr
Psychiatry Rep. Feb 2012;14(1):36-44. 90. Schoevers RA, Geerlings MI, Deeg DJ, Holwerda TJ, Jonker C, Beekman AT. Depression and excess mortality:
evidence for a dose response relation in community living elderly. Int J Geriatr Psychiatry. 2009;24(2):169-176. 91. Blumenfield M, Levy NB, Spinowitz B, et al. Fluoxetine in depressed patients on dialysis. Int J Psychiatry Med.
1997;27(1):71-80. 92. Spigset O, Hagg S, Stegmayr B, Dahlqvist R. Citalopram pharmacokinetics in patients with chronic renal failure and
the effect of haemodialysis. European journal of clinical pharmacology. Dec 2000;56(9-10):699-703. 93. Wuerth D, Finkelstein SH, Finkelstein FO. The identification and treatment of depression in patients maintained on
dialysis. Semin Dial. Mar-Apr 2005;18(2):142-146. 94. Duarte PS, Miyazaki MC, Blay SL, Sesso R. Cognitive-behavioral group therapy is an effective treatment for major
depression in hemodialysis patients. Kidney Int. Aug 2009;76(4):414-421. 95. Ouzouni S, Kouidi E, Sioulis A, Grekas D, Deligiannis A. Effects of intradialytic exercise training on health-related
quality of life indices in haemodialysis patients. Clinical rehabilitation. Jan 2009;23(1):53-63. 96. Hedayati SS, Bosworth HB, Kuchibhatla M, Kimmel PL, Szczech LA. The predictive value of self-report scales
compared with physician diagnosis of depression in hemodialysis patients. Kidney Int. May 2006;69(9):1662-1668.
114
Table 1. Characteristics of included studies
Study Sample Characteristics Measurement, Tool Followup Adjustment Results
Balogun 201113 77 prevalent HD patients, >65 y
Depressive symptoms, GDS-15 scores ≥ 5 3 y age, race, marital status adjusted HR: 1.91 (1.05, 3.46)
Boulware 20066 917 incident HD and PD patients, >18 y
Depressive symptoms, MHI-5 scores ≤ 52 2 y
age, gender, race, marital status, education, coexistent illness, dialysis modality, antidepressant therapy, cardiovascular risk factors, blood pressure, blood chemistry
crude HR: 1.62 (1.07, 2.44) adjusted HR: 2.22 (1.36, 3.60)
Butt 200614 16965 prevalent HD and PD patients, >18 y
Clinical depression, medical records 4 y HCV, HIV, drug use, CAD, stroke, DM,
PVD, HBV, anemia significant, effect measure not reported
Chilcot 201115 223 incident HD patients, >18 y
Depressive symptoms, BDI-II scores ≥ 16 16 mo No crude HR: 1.01 (0.98, 1.05) per scores
Christensen 199416 78 incident HD patients, >18 y Depression score, BDI 3.5 y age, BUN, family support not significant, effect measure not
reported
Diefenthaeler 200817
40 incident HD patients, >15 y
Depressive symptoms, BDI scores ≥ 14 10.5 mo age, hypertension, DM crude HR: 4.5 (1.1, 17.7)
adjusted HR: 6.5 (0.8, 55.0)
Drayer 200618 62 incident and
prevalent HD patients, >18 y
Depression score, PHQ-9 2 y age, sex, race, comorbidities, albumin, KT/V adjusted HR: 4.1 (1.2, 13.8)
Einwohler 200419 66 prevalent PD patients, >18 y Depression score, Zung SDS 3.5 y albumin, comorbidity (includes age
and diabetes) crude HR: 1.06 (1.03, 1.1) per score
adjusted HR: 1.05 (1.01, 1.08) per scores
Genestier 201020 112 incident PD patients, age >75 y
Clinical depression, medical records 18 mo Charlson comorbidity, site, early
referral, polymedication not significant, effect measure not
reported
Griva 201021 145 prevalent HD and PD patients, >18 y
Depressive symptoms, BDI-II scores ≥ 16 5 y
age, employment, ESRD severity index, DM, CVD, vascular disease, SF-36, cognitive impairment
not significant, effect measure not reported
Hedayati 200522 1588 prevalent HD patients, men, >18 y
Clinical depression, medical records 2 y age, DM, HTN, CHF, cardiac disease,
liver disease, substance abuse adjusted OR: 0.98 (0.72, 1.34)
115
Table 1. Cont’d
Study Sample Characteristics Measurement, Tool Followup Adjustment Results
Kimmel 20008 295 prevalent HD
patients, 92% African-American, >18 y
Standardized depression score, BDI 3 y age, dialysis solution, severity
coefficient, serum albumin, site crude HR: 1.24 (1.05, 1.46)
adjusted HR: 1.32 (1.13, 1.55)
Kojima 201023 230 prevalent HD patients, <70 y
Depressive symptoms, BDI-II scores ≥ 14 5 y
age, sex, SF-36, education, interdialytic weight gain, comorbidity, hematocrit, serum calcium, diastolic blood pressure
adjusted HR: 2.36 (1.084, 5.15) adjusted HR: 1.05 (1.01, 1.09) per score
Kutner 199424 287 prevalent HD patients, >60 y Depression score, CESD-20 3 y
age, race, gender, education, ESRD cause, CVD, dialysis duration, exercise, functional status
not significant, effect measure not reported
Lacson 201225 6415 incident HD patients, >18 y
Depressive symptoms and depression score, 2 items of
SF-36 1 y age, race, gender, DM, SF-36,
laboratory data
crude HR: 1.24 (1.28, 1.43) adjusted HR: 1.32 (1.05, 1.79)
crude HR: 1.09 (1.03, 1.15) per score adjusted HR: 1.08 (1.01, 1.14) per score
Lopes 200227 4881 incident and
prevalent HD patients, >17 y
Depressive symptoms, 2 items of SF-36
Clinical depression, Medical records
3 y demographics, laboratory data, comorbidities, time on dialysis
crude HR: 1.39 (1.24, 1.56) adjusted HR: 1.39 (1.23, 1.57)
crude HR: 1.42 (1.27, 1.60) for medical records
adjusted HR: 1.23 (1.08, 1.40) for medical records
Lopes 200426 6987 incident and
prevalent HD patients, >17 y
Depressive symptoms, CESD10 scores ≥ 10
Clinical depression, medical records
3 y age, sex, laboratory data, KT/V, comorbidities, time on dialysis, country
adjusted HR: 1.42 (1.29, 1.57) adjusted HR: 1.26 (1.1, 1.43) for medical
records
Mahajan 200728 52 prevalent PD patients, >18 y
Depressive symptoms, BDI scores >11 2 y No crude OR: 1.38 (0.24, 7.94)
Miskulin 200929 7685 prevalent HD patients, >18 y
Clinical depression, medical records 1.3 y age, race, gender, time on dialysis adjusted HR: 1.24 (1.13, 1.37)
116
Table 1. cont’d
Study Sample Characteristics Measurement, Tool Followup Adjustment Results
Peng 201030 888 prevalent HD patients, >18 y Depression score, BDI 7 y age, gender, laboratory data,
diabetes mellitus, hepatitis C, SF-36 crude HR: 1.02 (1.01, 1.03) per score
adjusted HR: 1.00 (0.99, 1.02) per score
Peterson 199431 57 incident and
prevalent HD and PD patients, >22 y
Depression score, CDI 2 y No crude HR: 1.11 (1.00, 1.24)
Riezebos 20109 101 prevalent HD and PD patients, >18 y
Depressive symptoms, HADS scores >7 1 y age, sex, CVD, DM, time on dialysis crude HR: 3.3 (1.2, 9.6)
adjusted HR: 5.0 (1.2, 9.6)
Rosenthal Asher 201232
130 prevalent HD patients, >18 y Depression score, BDI 5 y age, DM, time on dialysis,
hospitalizations adjusted HR: 1.05 (1.01, 1.08)
Santos 201233 161 prevalent HD patients, >18 y
Depressive symptoms, CESD-10 ≥ 10 1 y No crude OR: 2.26 (0.45, 11.5)
Shulman 198934 64 prevalent HD patients, >18 y
Depressive symptoms, BDI >10 10 y No significant, effect measure not reported
Simic Ogrizovic 200935
128 prevalent HD and PD patients, >18 y Depression score, BDI 3 y age, laboratory data crude HR: 1.05 (1.02, 1.08) per score
adjusted HR: 1.33 (1.00, 1.06) per score
Soucie 199636 15245 incident HD and PD patients, >15 y
Clinical depression, medical records 3 mo age, MI, activity impairment, race,
gender, CHF, HTN, smoking crude OR: 0.79 (0.65, 0.95)
adjusted OR: 1.3 (1, 1.6)
Szeto 200837 167 prevalent PD patients, >18 y
Depressive symptoms, HADS scores >7 1 y No crude HR: 1.25 (0.58, 2.70)
Takaki 200538 490 prevalent HD patients, >18 y Depression score, HADS 2.5 y No crude HR: 1.48 (1.14, 1.92) per score
Tsai 201239 2312 incident HD and PD patients, >20 y
Clinical depression, medical records 7 y age, gender, comorbidities crude OR: 1.01 (0.65, 1.59)
Van den Beukel 201040
1078 incident HD and PD patients, >18 y
Depressive symptoms, MHI-5 ≤ 52 3.8 y
age, gender, education, marital status, Davies comorbidity index, primary kidney disease, dialysis modality, laboratory data
crude HR: 2.45 (1.87, 3.20) adjusted HR: 1.83 (1.36, 2.45)
117
Table 2. Excluded studies
Study Reason for Exclusion Abdel-Rahman 201141 Duplicated data: updated results reported in Balogun et al 2011.13 Bilgic 200742 Outcome: reported only hospitalization. Author contact unsuccessful. Brzosko 201043 Insufficient data: Author contact unsuccessful.
Burton 198644 Exposure: personality trait scale used to ascertain depression (Basic Personality Inventory)
Chilcot 20117 Duplicated data: updated results reported in Chilcot et al 2011.15 DeOreo 199745 Exposure: mental component summary of Short Form-36 used to ascertain depression. Devins 199046 Population: included both dialysis and kidney transplant patients. Fischer 201147 Population: included both dialysis and predialysis chronic kidney disease patients.
Foster 197348 Exposure: personality trait scale used to ascertain depression (Mood Adjective Check List ).
Fukuhara 200649 Duplicated data: data limited to Japanese DOPPS cohort. Patients also included in those published by Lopes et al 2004.26
Hedayati 200850 Outcome: reported only as a composite of mortality and hospitalization. Authors unable to provide mortality-only data.
Huesebye 198751 Exposure: unvalidated single item score used to evaluate depression. Kazama 200952 Insufficient data: author contact unsuccessful. Kellerman 201053 Population: included both dialysis and predialysis chronic kidney disease patients. Kimmel 199855 Duplicated data: updated results reported in Kimmel et al 2000.8 Kimmel 200054 Duplicated data: updated results reported in Kimmel et al 2000.8 Koo 201156 Outcome: reported only cardiovascular events. Author contact unsuccessful. Lopes 200757 Duplicated data: updated results reported in Lopes et al 2004.26 Manrique 201058 Insufficient data: author contact unsuccessful.
Numan 198159 Exposure: personality trait scale used to ascertain depression (Depression Adjective Check List).
Parkerson 200060 Exposure: unvalidated measure of depression. Sapilak 200661 Insufficient data: author contact unsuccessful. Sawicka 201062 Insufficient data: author contact unsuccessful. Sayana 201063 Insufficient data: : Author contact unsuccessful. Tentori 201064 Duplicated data: results reported elsewhere.26 Thong 200765 Duplicated data: updated results reported elsewhere.40 Untas 200766 Duplicated data: results reported elsewhere.26 Valdes 200667 Exposure: mental component summary of Short Form-36 used to ascertain depression. Wai 198168 Exposure: evaluation of depression not clearly described. Young 201069 Population: included both dialysis and predialysis chronic kidney disease patients.
Ziarnik 19770 Exposure: personality trait scale used to ascertain depression (Minnesota Multiphasic Personality Inventory).
Zimmermann 200671 Population: included both dialysis and transplant patients. Author contact unsuccessful.
118
Table 3. Appraisal of included studies using the NOS for quality assessment of cohort studies
Modified Newcastle-Ottawa Scale Study Selection Comparability Outcome Risks of Bias
Balogun 201113 **** * *** Low Boulware 20066 **** ** *** Low Butt 200614 *** * ** Low Chilcot 201115 **** ** *** Low Christensen 199416 ** * ** High Diefenthaeler 200817 *** ** ** High Drayer 200618 **** ** *** Low Einwohler 200419 **** ** ** Low Genestier 201020 ** * *** High Griva 201021 **** ** *** Low Hedayati 200522 *** ** *** Low Kimmel 20008 **** ** *** Low Kojima 201023 *** * ** High Kutner 199424 **** ** ** Low Lacson 201225 **** * *** Low Lopes 200227 **** ** ** Low Lopes 200426 **** ** ** Low Mahajan 200728 **** * High Miskulin 200929 *** ** ** Low Peng 201030 **** * ** Low Peterson 199431 *** ** High Riezebos 20109 **** ** ** Low Rosenthal Asher 201232 **** * ** Low Santos 201233 **** ** ** Low Shulman 198934 **** *** High Simic Ogrizovic 200935 **** * *** Low Soucie 199636 *** ** ** Low Szeto 200837 **** * *** Low Takaki 200538 **** ** ** Low Tsai 201239 *** ** *** Low Van den Beukel 201040 **** ** *** Low
119
Table 4. Non-adjusted effect sizes of the association of depression and mortality by the type of reported effect size and measurement method of depression
Effect Size Studies Sample Effect Size (95% CI) P I2 Reference Crude HR
Presence of depressive symptoms
9 12859 1.58 (1.33, 1.88) <.001 79% 6,8,9,17,21,25,27,37,40
Depressive score 7 8267 1.05 (1.02, 1.08) <.001 76% 15,19,25,30,31,35,38 Physician diagnosis 1 4881 1.42 (1.27, 1.59) <.001 ... 27
Crude OR Presence of depressive symptoms
7 559 2.33 (1.43, 3.82) <.001 5% 9,13,17-19,28,33
Physician diagnosis 3 19145 0.93 (0.73, 1.18) .53 51% 22,36,39
120
Table 5. Summary of effect sizes for subgroups of dialysis patients by followup duration, country of origin, time relative to initiation of dialysis, and number of measurements of exposure.
Subgroup Studies Sample Adjusted HR (95% CI) P I2 Reference Presence of depressive symptoms
All studies 12 21055 1.51 (1.35, 1.69) < .001 40% 6,8,9,13,17,18,21,23,25-27,40 Followup <= 1 year 3 6556 2.66 (0.85, 8.25) .09 64% 9,17,25 Followup >1 to 3 years 5 12924 1.50 (1.30, 1.72) <.001 44% 6,13,18,26,27 Followup >3years 4 1575 1.59 (1.25, 2.03) <.001 46% 8,21,23,40 Incident patients 4 8450 1.73 (1.27, 2.35) <.001 51% 6,17,25,40 Prevalent and incident patients
8 12605 1.44 (1.30, 1.60) <.001 31% 8,9,13,18,21,23,26,27
US Studies 5 7593 1.57 (1.23, 2.00) <.001 51% 6,8,13,18,25 Non-US studies 5 1594 1.95 (1.53, 2.48) <.001 0 9,17,21,23,40 Single measurement of depression
9 18938 1.48 (1.31, 1.67) <.001 31% 9,13,17,18,21,23,25-27
Repeated measurements of depression
3 2117 1.66 (1.22, 2.25) .001 70% 6,8,40
Depressive score All studies 6 7857 1.04 (1.01, 1.06) .002 74% 19,23,25,30,32,35 US Studies 3 6611 1.05 (1.03, 1.08) <.001 0 19,25,32 Non-US studies 3 1246 1.02 (1.00, 1.05) .09 73% 23,30,35
121
Table 6. Sensitivity analyses on the adjusted data for the effect sizes of depression as an independent risk factor of mortality
Subgroup Studies Sample Adjusted HR (95% CI) P I2 Reference Presence of depressive symptoms
Excluding 2 studies with a high risk of bias17,23
10 20785 1.48 (1.34, 1.65) <.001 39% 6,8,9,13,18,21,25-27,40
Excluding 4 studies with small sample size9,13,17,18
8 20775 1.44 (1.32, 1.57) <.001 25% 6,8,21,23,25-27,40
Depressive score Excluding 1 study with a high risk of bias23
5 7627 1.04 (1.01, 1.06) .007 76% 19,25,30,32,35
Excluding 1 study with small sample size 19
5 7791 1.04 (1.01, 1.06) .008 75% 23,25,30,32,35
122
Figure 1. Flow diagram of search and selection of studies
123
Figure 2. Presence of depression symptoms as a risk factor of mortality (adjusted risk estimates using hazard ratios).
124
Figure 3. Funnel plot of studies reporting hazard ratios associated with the presence of depressive symptoms for mortality. Twelve studies are included, of which 5 in the right side of the vertical line are identified as outliers in the trim-and-fill analysis.
125
Figure 4. Depression scale score as a risk factor of mortality by (adjusted risk estimates using hazard ratios per score).
126
Appendix B. Models describing human behaviour for conceptualizing barriers to mental health care utilization
Several factors may interfere with an individual’s the process of decision making. Such
factors may prevent patients with mental health problems from appropriately accepting
health care services. Pathways to care and in general to behaviour are conceptualized by
some models and theories originating from social cognitive and socio-behavioural theories.
Here, we review these models and describe how they can be applied for conceptualizing
mental health care utilization.
Health Belief Model. Henshaw and Freedman-Doan131 have elaborated mental health care
utilization using the Health Belief Model (HBM), primarily described by Rosenstock,138 and
demonstrated that this socio-cognitive model addresses key aspects of health services
utilization in mental health care context. Rosenstack developed HBM using the social
cognitive theory proposed by Bandura.130,138 According to HBM, individuals are likely to
engage in a health-related behaviour to the extent that they perceive a health problem as a
threat (perceived susceptibility to contracting a disease and perceived severity of that disease)
and the extent that they perceive benefits of and barriers to utilizing a health care behaviour
(Figure 1).131 In other words, healthy individuals are likely to act if they believe that (1) there
is a real risk of contracting an illness and they are prone to this risk, (2) the disease is serious
in terms of its medical and non-medical consequences, (3) the health behaviour of interest is
beneficial in reducing the threat of the health condition, and (4) there is no perceived
negative consequence of the action. These constructs vary among different demographic and
psychological groups. In addition, individuals’ decision to utilize a given health care is
127
sometimes triggered by incidents and experiences related to the health problem, called cues
to action.138
Primarily, an individual may decide not to take a health behaviour action because of not
perceiving the threat and not perceiving any benefit of the intervention or preventive action.
However, Rosenstock argues that an individual may even perceive a threat and believe that a
health behaviour is effective in reducing the threat, but at the same time considers the
behaviour as being “inconvenient, expensive, unpleasant, painful or upsetting.138” In other
words, barriers perceived by the patients include lack of threat perception and concerns about
benefits of a behaviour, as well as several other factors that act as barriers despite perceived
threat and benefit of a related health care behaviour. The latter are specifically called
“perceived barriers” in HBM. An example that Rosenstock makes to explain perceived
susceptibility clarifies this distinction: an individual “may deny any possibility of his
contracting a given condition,138” which is a barrier derived from lack of perceived threat,
and not from difficulties taking action (those categorized as “barriers” in HBM). Rosenstock
et al describe perceived barriers as factors that have “always had something of a catch-all
quality, including such disparate items as financial costs, phobic reactions, physical barriers,
side-effects, accessibility factors, and even personality characteristics.132”
The HBM is based on the social cognitive theory, and its constructs reflect a rational
decision-making process, and emotional and social factors are not fully addressed.131 In an
attempt to highlight emotional aspects and clarify types of the barriers, Henshaw and
Freedman-Doan classified perceived barriers into psychological and practical.131 Therefore,
they identify 4 constructs of the decision making in health care utilization, which can be
128
translated into barriers constructs when an individual decides not to take a health action:
perceiving no threat, perceiving no benefit, psychological barriers, and practical barriers.
Social Cognitive Theory. Bandura’s theory (Figure 2) holds that behaviour is determined by
expectancies and incentives.130 Expectancies are about environmental cues (beliefs about
how events are connected), consequences of one’s actions (beliefs about how a behaviour
influences outcomes), and one’s competency to perform a behaviour (self-efficacy). The first
two expectancies are similar to the perceived threat and perceived benefits in HBM, and self-
efficacy can be regarded as an aspect of psychological barriers.132
Incentive or reinforcement is defined as the value of a particular outcome understood by the
individual. In the context of health care utilization, incentive is the health motive or the value
of reduction of a perceived threat.132 Incentive is partly dependent on perception of a health
condition and its consequences if left untreated; seriousness of a disease determines the value
given by the patient to its treatment. In addition, lack of motivation in general can be
considered as a psychological barrier that impedes taking action.138
Theory of Planned Behaviour. Ajzen proposed the theory of planned behaviour as a
conceptual framework for studying human action based on the theory of reasoned action
(Figure 3).133 According to this theory, a behavioural intention is guided by the individual’s
attitudes toward a behaviour produced by beliefs about the consequences of a behaviour,
normative expectations of others, and perceived behavioural control (self-efficacy and
controllability). Perception of inability to successfully participate in health behaviour (self-
efficacy) can be considered equivalent to psychological barriers in HBM. Ajzen elaborates
perceived behavioural control by considering self-efficacy as the internal part of it that
129
addresses the concept of confidence in the ability to perform a behaviour, differentiated from
perceived controllability as the external part of perceived control (the extent to which
performance of a behaviour is up to the actor).133 The strength of theory of planned behaviour
is incorporation of social pressure and normative beliefs and addressing situations in which
people perceive lack of volitional control over the behaviour of interest, which is are not fully
addressed by the HBM.133
Self-regulatory Model. Leventhal et al focused on the process of interacting with the disease
before deciding to seek medical care and incorporated emotional factors in addition to
cognitive factors to their model (Figure 4).137 According to the self-regulatory model,
cognitive factors form illness representation (cognition) and emotional factors interact with
illness cognition and coping strategies taken by the patient. Failure of coping increases
distress, alters emotional reactions and perception of the disease, and ultimately, leads to the
decision to seek help. In addition, self-regulatory model describes illness representation in
detail by elaborating it in 5 areas of label and symptom attributes, duration, consequences,
causes, and controllability of the disease.135,137
It should be noted that self-regulatory model addresses help seeking as a health behaviour of
interest and therefore focuses on threat, and not benefits of a health care intervention or
barriers to undertaking that. Also, coping here is not the behaviour of interest, but a step
before seeking help that interacts with perceived threat (Figure 4).
Help-seeking Model. To further address emotional factors that act as intervening variables
between the recognition of psychological problems and decision to seek help, Cramer
investigated the interaction between the following factors: level of distress, attitudes toward
130
professional psychological counseling, social support, and self-concealment (Figure 5).136 He
developed a help-seeking model, in which self-concealment (“a predisposition to actively
hide personal negative information from others”) is associated with higher levels of distress,
lower social support, and negative attitudes toward counseling. Cramer showed that
individuals who tend to conceal personally distressing information are less likely to seek help
for psychological problems, and on the other hand, they have higher levels of distress that
will push them to seek help.136
Sociobehavioural Model. Andersen proposed a sociobehavioural model of health services use
with emphasis on the role of social structure and the health care system (Figure 6).134
According to Andersen, people’s use of health services is a function of predisposing
characteristics, factors that enable or impede use, and their need for care. Predisposing
characteristics include demographics, social structure, and health beliefs that together explain
perceived need for care. The need for care is influenced by one’s perception of the need and
also the evaluation made by health care professionals (evaluated need). We can also assume
that an unrecognized need by health care professional acts as a barrier to health care services
use. Andersen added two interesting factors to his primary model: health care system and
consumer satisfaction. The way health care system is presented, the information provided to
people about it, and the ease of to accessing it or finding information on access affects the
other constructs in the sociobehavioural model.134
131
132
133
134
Appendix C. Modified Perceived Barriers to Psychological Treatment
Some centres suggest that we routinely use a screening questionnaire about depression in dialysis patients. Screening helps the dialysis team to identify people with depressive symptoms, so that they can help if needed. For example, they may adjust treatments, refer to a mental health specialist, or prescribe medications, if needed.
Below is a list of issues that may make it hard for people to take part in a routine screening program for depression. We would like to learn to what extent you think these issues would make it difficult for you to take part in a screening program.
For the purpose of this questionnaire, assume that a screening program for depression would involve completing a questionnaire about your mood and feelings, and if needed being referred to a mental health specialist for further assessment, counseling or medications.
Please mark one response for each item. If a particular issue does not apply to you, please mark “not difficult at all”.
Not Difficult at All
Slightly Difficult
Moderately Difficult
Extremely Difficult Impossible
a. Problems with transportation (no car, parking problems, poor public transportation, etc.) would make it ________ for me to take part in a screening program for depression.
b. The responsibility of caring for loved ones (children, someone with an illness, etc.) would make it ________ for me to take part in a screening program for depression.
Perceived Barriers to Screening for Depression Subject ID: _____________
Date: __________________
An Opinion Survey of Hemodialysis Patients to Identify Perceived Barriers to Participation in a Screening Program for Depression
135
Not Difficult at All
Slightly Difficult
Moderately Difficult
Extremely Difficult Impossible
c. The cost of treatment, if needed, would make it ________ for me to take part in a screening program for depression.
d. My daily responsibilities and activities would make it ________ for me to take part in a screening program for depression.
e. The lack of available mental health services in my area would make it _________ for me to take part in a screening program for depression.
f. Not knowing how to find a good mental health specialist would make it ________ for me to take part in a screening program for depression.
g. Getting time off work to go for mental health services would make it ________ for me to take part in a screening program for depression.
h. Physical problems, such as difficulties walking or getting around, would make it ________ for me to take part in a screening program for depression.
i. Physical symptoms (fatigue, pain, breathing difficulties, etc.) would make it ________ for me to take part in a screening program for depression.
136
Not Difficult at All
Slightly Difficult
Moderately Difficult
Extremely Difficult Impossible
j. A serious illness which requires me to stay close to home would make it ________ for me to take part in a screening program for depression.
k. Having heard about or having had bad or unsatisfactory experiences with treatment of depression would make it ________ for me to take part in a screening program for depression.
l. Distrust of mental health specialists would make it ________ for me to take part in a screening program for depression.
m. I wouldn't expect treatment for depression to be helpful and this would make it ________ for me to take part in a screening program for depression.
n. I would be concerned about side effects of medications for depression, if I needed, and that would make it _________ for me to take part in a screening program for depression.
o. I wouldn’t expect questionnaires for depression to be helpful and that would make it _________ for me to take part in a screening program for depression.
137
Not Difficult at All
Slightly Difficult
Moderately Difficult
Extremely Difficult Impossible
p. Having treatment for depression is too self-indulgent and that would make it ________ for me to take part in a screening program for depression.
q. I would prefer to handle it on my own if I was depressed, and therefore it would be _________ for me to take part in a screening program for depression.
r. Having other problems that are more important would make it _________ for me to take part in a screening program for depression.
s. I would prefer to decide when I need help for depression on my own, and that would make it _________ for me to take part in a screening program for depression.
t. Having to fill out additional questionnaires would make it _________ for me to take part in a screening program for depression.
u. Having to take more medications would make it _________ for me to take part in a screening program for depression.
v. Anxiety about going far from my home would make it ________ for me to take part in a screening program for depression.
138
Not Difficult at All
Slightly Difficult
Moderately Difficult
Extremely Difficult Impossible
w. Concerns about having upsetting feelings would make it ________ for me to take part in a screening program for depression.
x. I feel that talking about upsetting issues makes them worse and that would make it ________ for me to take part in a screening program for depression.
y. Lack of energy or motivation to make an appointment and then go would make it ________ for me to take part in a screening program for depression.
z. Difficulty motivating myself to do anything at all would make it ________ for me to take part in a screening program for depression.
aa. Discomfort with having someone see me while I am emotional would make it ________ for me to take part in a screening program for depression.
bb. My problems are not severe enough, and therefore it would be ________ for me to take part in a screening program for depression.
139
Not Difficult at All
Slightly Difficult
Moderately Difficult
Extremely Difficult Impossible
cc. I do not think I will get depressed, and therefore it would be _________ for me to take part in a screening program for depression.
dd. I think sadness is normal among people on dialysis, and therefore it would be _________ for me to take part in a screening program for depression.
ee. I think better treatment of the kidney problem would improve depression, and therefore it would be _________ for me to take part in a screening program for depression.
ff. I would be afraid of screening results for depression and that would make it _________ for me to take part in a screening program for depression.
gg. Having family and/or friends know I was going for mental health services would make it ________ for me to take part in a screening program for depression.
hh. Having to talk to someone I do not know about personal issues would make it ________ for me to take part in a screening program for depression.
ii. My concern about being judged by health care specialists would make it _________ for me to take part in a screening program for depression.
140
Not Difficult at All
Slightly Difficult
Moderately Difficult
Extremely Difficult Impossible
jj. I just do not think mental health specialists would truly care about me and that would make it ________ for me to take part in a screening program for depression.
kk. Receiving mental health care for depression would mean I cannot solve my own problems and that would make it _________ for me to take part in a screening program for depression.
ll. Having a medical or insurance record of mental health services would make it _________ for me to take part in a screening program for depression.
Please feel free to provide other reasons that might get in the way of taking part in a screening program for depression:
_____________________________________________________________________________________________________
_____________________________________________________________________________________________________
Thank you for your time and help with this survey!
141
Scoring of the Adapted PBPT Questionnaire
1. Response Options Scores: Responses to each item will be scored as below:
Response Score Not difficult at all 0 Slightly difficult 1 Moderately difficult 2 Extremely difficult 3 Impossible 4 2. Hypothetical Constructs: Items related to each construct are as below:
Construct Original PBPT Items (n = 27) New Items (n = 11) Perceiving no threat p, bb q, r, s, cc, dd, ee Perceiving no benefit k, m n, o Psychological barriers l, v, w, x, y, z, aa, hh t, u, ff Social barriers gg, ii, jj, kk, ll Practical barriers a, b, c, d, e, f, g, h, i, j 3. Subscores: The sum of scores for each item in a construct will be calculated as the subscore for that construct:
Construct Items Subscores Perceiving no threat 8 0 – 32 Perceiving no benefit 4 0 – 16 Psychological barriers 11 0 – 44 Social barriers 5 0 – 20 Practical barriers 10 0 – 40 All 38 0 - 152 4. Dichotomized results: A perceived barrier is defined as endorsing “extremely difficult” or “impossible” for that barrier. Accordingly, the percentage of patients who perceived a barrier will be reported for each item, and the percentage of patients who perceived one or more barrier will be used to categorize patients for perception of barriers to participation in a screening program for depression.
142
Appendix D. Patient Health Questionnaire-2
143
Appendix E. Consent forms
CONSENT TO PARTICIPATE IN A RESEARCH STUDY
Title: A Patient Opinion Survey to Identify Perceived Barriers to the
Introduction of a Screening Program for Depression in a Hemodialysis Population
Investigator: Dr. SV Jassal, Toronto General Hospital
Introduction
You are being asked to take part in a research study. Please read this explanation about the study and its risks and benefits before you decide if you would like to take part. You should take as much time as you need to make your decision. You should ask the study doctor or study staff to explain anything that you do not understand and make sure that all of your questions have been answered before signing this consent form. Before you make your decision, feel free to talk about this study with anyone you wish. Participation in this study is voluntary.
Background and Purpose
Up to half of the patients on dialysis feel fatigued, or have difficulty concentrating, poor sleep, or loss of interest. Studies have shown that these symptoms are associated with a low quality of life, hospitalization, and death in dialysis patients. In many cases, the dialysis health care team are unaware of these symptoms, and therefore are unable to help support patients, adjust treatments or provide care if needed.
Some doctors suggest routinely using a set of questions to look for these symptoms in dialysis patients. This practice is called screening, and it allows the medical team to give help to those who may need it or could benefit from it. We are interested in what you think about starting such a screening program.
144
Study Design and Procedures
This study has of 2 parts. You can choose which part you would like to take part in at the end of this form.
Part A - Chart review: We will look at your medical chart and note the reason for your kidney disease, how long you have been on dialysis, your age, and gender.
Part B - Visits: You will have 2 study visits that involve questionnaires and your own assessment of your health.
First visit
This visit will take about 5 minutes of your time.
During this visit we will ask you about your health history and ask you to complete a questionnaire with 2 questions about your feelings over the last 2 weeks. This information will be made available to your kidney doctor. Further assessments and referral to the appropriate health care team will be left up to your doctor.
We will collect further information about your health and your dialysis treatment from your clinical chart. This will include any other diseases you are getting treatment for.
Second visit
This visit will take 15-20 minutes of your time.
We will give you a list of reasons that people (not on dialysis) have given for why they may or may not want to take part in a screening program like the one described here. We ask you to tell us if you feel these reasons would be important to you. The questionnaire may be done anytime during your dialysis treatment. These data will not be made available to your kidney doctor.
Risks Related to Being in the Study
There are no medical risks if you take part in this study. Sometimes answering questions about yourself or your thoughts and beliefs can make you feel uncomfortable; however, you may refuse to answer these questions.
145
Benefits to Being in the Study
You will not directly benefit from being in this study. Information learned from this study may in the future help people on dialysis. This study may help us to find areas of concern for your health that may benefit from treatment outside of this study. At your request the collected information may be passed onto your dialysis doctor.
Voluntary Participation
Your participation in this study is voluntary. You may decide not to be in this study, or to be in the study now, and then change your mind later. You may leave the study at any time without affecting your care. You may refuse to answer any question you do not want to answer.
Confidentiality
If you agree to join this study, the study doctor and his/her study team will look at your personal health information and collect only the information they need for the study. Personal health information is any information that could be used to identify you and includes your name, date of birth, and new or existing medical records, that includes types, dates and results of medical tests or procedures.
The information that is collected for the study will be kept in a locked and secure area by the study doctor for 7 years. Only the study team or the people or groups listed below will be allowed to look at your records. Your participation in this study also may be recorded in your medical record at this hospital.
Representatives of the University Health Network Research Ethics Board may look at the study records and at your personal health information to check that the information collected for the study is correct and to make sure the study followed proper laws and guidelines.
All information collected during this study, including your personal health information, will be kept confidential and will not be shared with anyone outside the study unless required by law. You will not be named in any reports, publications, or presentations that may come from this study.
If you decide to leave the study, the information about you that was collected before you left the study will still be used. No new information will be collected without your permission.
146
Expenses Associated with Participating in the Study
There will be no cost to you nor will you be paid to participate in this study.
Questions About the Study
If you have any questions about the study, please contact Farhat Farrokhi (Study Coordinator) at 416-340-4800 x 6362 or [email protected], or Dr SV Jassal at 416-340-3196.
If you have any questions about your rights as a research participant or have concerns about this study, call the Chair of the University Health Network Research Ethics Board (REB) or the Research Ethics office number at 416-581-7849. The REB is a group of people who oversee the ethical conduct of research studies. These people are not part of the study team. Everything that you discuss will be kept confidential.
147
Consent
This study has been explained to me and any questions I had have been answered. I know that I may leave the study at any time. I agree to take part in this study.
I would like to ONLY participate in Part A of the study to have my medical chart reviewed
I would like to participate in Part B to the study visits and answer the questionnaires
Print Study Participant’s Name Signature Date
(You will be given a signed copy of this consent form)
================================================================
My signature means that I have explained the study to the participant named above. I have answered all questions.
Print Name of Person Obtaining ConsentSignature Date
148
CONSENT TO PARTICIPATE IN A RESEARCH STUDY
Title: A Patient Opinion Survey to Identify Perceived Barriers to the Introduction of a Screening Program for Depression in a Hemodialysis Population
Investigator: Dr. SV Jassal, Toronto General Hospital
Introduction
You are being asked to take part in a research study. Please read this explanation about the study and its risks and benefits before you decide if you would like to take part. You should take as much time as you need to make your decision. You should ask the study doctor or study staff to explain anything that you do not understand and make sure that all of your questions have been answered before signing this consent form. Before you make your decision, feel free to talk about this study with anyone you wish. Participation in this study is voluntary.
Background and Purpose
Up to half of the patients on dialysis feel fatigued, or have difficulty concentrating, poor sleep, or loss of interest. Patients who commonly suffer from these problems may experience a low quality of life and have greater risks of hospitalization and death. In many cases, the dialysis health care team are unaware of these symptoms, and therefore are unable to help support patients, adjust treatments or provide care if needed.
Some doctors suggest routinely using a set of questions to look for these symptoms in dialysis patients. This practice is called screening, and it allows the medical team to give help to those who may need it or could benefit from it. We are interested in what you think about starting such a screening program.
149
Study Design and Procedures
This study has of 2 parts. You can choose which parts you would like to take part in at the end of this form.
Part A - Chart review: We will look at your medical chart and note the reason for your kidney disease, how long you have been on dialysis, your age, and gender.
Part B – Chart review and visits: In addition to a chart review, you will have 2 study visits that involve questionnaires and your own assessment of your health.
First visit
• This visit will take about 5 minutes of your time.
• During this visit we will ask you about your health history and ask you to complete a questionnaire with 2 questions about your feelings over the last 2 weeks. This information will be made available to your kidney doctor. Further assessments and referral to the appropriate health care team will be left up to your doctor.
• We will collect further information about your health and your dialysis treatment from your clinical chart. This will include any other diseases you are getting treatment for.
Second visit
• This visit will take 10-15 minutes of your time.
• We will give you a list of reasons that people (not on dialysis) have given for why they may or may not want to take part in a screening program like the one described here. We ask you to tell us if you feel these reasons would be important to you. The questionnaire may be done anytime during your dialysis treatment. These data will not be made available to your kidney doctor.
Risks Related to Being in the Study
There are no medical risks if you take part in this study. Sometimes answering questions about yourself or your thoughts and beliefs can make you feel uncomfortable; however, you may refuse to answer these questions.
Benefits to Being in the Study
You will not directly benefit from being in this study. Information learned from this study may in the future help people on dialysis. This study may help us to find areas of concern for your health that may benefit from treatment outside of this study. At your request the collected information may be passed onto your dialysis doctor. Voluntary Participation
150
Your participation in this study is voluntary. You may decide not to be in this study, or to be in the study now, and then change your mind later. You may leave the study at any time without affecting your care. You may refuse to answer any question you do not want to answer. You do not waive any of your legal rights by agreeing to participate in this study.
Confidentiality
If you agree to join this study, the study doctor and his/her study team will look at your personal health information and collect only the information they need for the study. Personal health information is any information that could be used to identify you and includes your name, date of birth, and new or existing medical records, that includes types, dates and results of medical tests or procedures.
The information that is collected for the study will be kept in a locked and secure area by the study doctor for 7 years. Only the study team or the people or groups listed below will be allowed to look at your records. Your participation in this study also may be recorded in your medical record at this hospital.
Representatives of the Sunnybrook Research Ethics Board may look at the study records and at your personal health information to check that the information collected for the study is correct and to make sure the study followed proper laws and guidelines.
All information collected during this study, including your personal health information, will be kept confidential and will not be shared with anyone outside the study unless required by law. You will not be named in any reports, publications, or presentations that may come from this study.
If you decide to leave the study, the information about you that was collected before you left the study will still be used. No new information will be collected without your permission.
Expenses Associated with Participating in the Study
There will be no cost to you nor will you be paid to participate in this study.
Questions About the Study
If you have any questions about the study, please contact Farhat Farrokhi (Study Coordinator) at 416-340-4800 x 6362 or [email protected], or Dr SV Jassal at 416-340-3196.
If you have any questions about your rights as a research participant or have concerns about this study, call Dr. Philip C. Hébert, Chair of the Sunnybrook Research Ethics Board (REB) at (416) 480-4276. The REB is a group of people who oversee the ethical conduct of research studies. These people are not part of the study team. Everything that you discuss will be kept confidential.
151
Consent
This study has been explained to me and any questions I had have been answered. I know that I may leave the study at any time. I agree to take part in this study.
I would like to ONLY participate in Part A of the study to have my medical chart reviewed
I would like to participate in BOTH Part A and Part B of the study to have my medical chart reviewed and answer the questionnaires during the study visits
Print Study Participant’s Name Signature Date
(You will be given a signed copy of this consent form)
================================================================
My signature means that I have explained the study to the participant named above. I have answered all questions.
Print Name of Person Obtaining ConsentSignature Date
152
Appendix F. Sample Sizes and Confidence Intervals
The confidence limits of different sample sizes by the expected range of proportion of hemodialysis patients with perceived barriers to accepting mental health care for depression
% of Patients with Barriers to Screening for Depression (Maximum number of Independent Variables for Multivariable Analysis)
Sample Size 50 60 70 80
140 41.8 - 58.2 (7 variables)
52.0 – 68.0 (8 variables)
62.5 - 77.5 (9 variables)
73.4 - 86.6 (11 variables)
159* 42.3 - 57.7 (8 variables)
52.5 - 67.5 (9 variables)
63.0 - 77.0 (11 variables)
73.9 - 86.1 (12 variables)
180 42.8 - 57.2 (9 variables)
52.9 - 67.1 (10 variables)
63.4 - 76.6 (12 variables)
74.2 - 85.8 (14 variables)
200 43.2 - 56.8 (10 variables)
53.3 - 66.7 (12 variables)
63.7 - 76.3 (14 variables)
74.5 - 85.5 (16 variables)
*A sample size of 159 hemodialysis patients will allow us to have a 95% confidence interval of 7% for an estimation for proportion of those with barriers of 70%, 7.5% for 60% and 7.7% for 50%.
153
Appendix G. Barriers perceived by the participants, sorted by prevalence
Item Barrier Response (%) Positive (%) n* Concerns about side effects of medications 159 (99.4) 63 (39.6) u* Having to take more medications 160 (100) 51 (31.9) bb My problems are not severe enough 160 (100) 37 (23.1) cc* I do not think I will get depressed 158 (98.8) 36 (22.8) c The cost of treatment, if needed 159 (99.4) 33 (20.8) j A serious illness which requires me to stay close to home 160 (100) 30 (18.8) r* Having other problems that are more important 155 (96.9) 29 (18.7) h Physical problems, such as difficulties walking or getting around 160 (100) 27 (16.9) ll Having a medical or insurance record of mental health services 160 (100) 26 (16.3) v Anxiety about going far from my home 160 (100) 26 (16.3)
kk Receiving mental health care for depression would mean I cannot solve my own problems 160 (100) 25 (15.6)
gg Having family and/or friends know I was going for mental health services 160 (100) 24 (15.0) s* I would prefer to decide when I need help for depression on my own 157 (98.1) 24 (15.3) i Distrust of mental health specialists 160 (100) 21 (13.1) l Physical symptoms (fatigue, pain, breathing difficulties, etc.) 159 (99.4) 21 (13.2) hh Having to talk to someone I do not know about personal issues 160 (100) 20 (12.5) y Lack of energy or motivation to make an appointment and then go 160 (100) 20 (12.5) jj I just do not think mental health specialists would truly care about me 159 (99.4) 18 (11.3) a Problems with transportation 160 (100) 17 (10.6) d My daily responsibilities and activities 159 (99.4) 17 (10.7) f Not knowing how to find a good mental health specialist 158 (98.8) 17 (10.8) q* I would prefer to handle it on my own if I was depressed 160 (100) 17 (10.6) g Getting time off work to go for mental health services 160 (100) 15 (9.4) ii My concern about being judged by health care specialists 159 (99.4) 15 (9.4) w Concerns about having upsetting feelings 160 (100) 15 (9.4) x I feel that talking about upsetting issues 160 (100) 15 (9.4) b The responsibility of caring for loved ones 160 (100) 14 (8.8) ff I would be afraid of screening results for depression 159 (99.4) 14 (8.8) aa Discomfort with having someone see me while I am emotional 160 (100) 13 (8.1) dd* I think sadness is normal among people on dialysis 158 (98.8) 13 (8.2) e The lack of available mental health services in my area 159 (99.4) 13 (8.2) ee* I think better treatment of the kidney problem would improve depression 158 (98.8) 13 (8.2)
k Having heard about or having had bad or unsatisfactory experiences with treatment of depression 160 (100) 13 (8.1)
t* Having to fill out additional questionnaires 160 (100) 12 (7.5) m I wouldn't expect treatment for depression to be helpful 159 (99.4) 11 (6.9) z Difficulty motivating myself to do anything at all 160 (100) 11 (6.9) o* I wouldn’t expect questionnaires for depression to be helpful 158 (98.8) 10 (6.3) p Having treatment for depression is too self-indulgent 159 (99.4) 7 (4.4)
*Additional items to the original PBPT