PSYCHOSOCIAL DETERMINANTS OF INTENTIONAL AND UNINTENTIONAL NONADHERENCE IN PATIENTS UNDERGOING AUTOMATED PERITONEAL DIALYSIS AND CONTINUOUS AMBULATORY PERITONEAL DIALYSIS YU ZHENLI A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF SOCIAL SCIENCES NATIONAL UNIVERSITY OF SINGAPORE 2011 Acknowledgments I would like to express my deep thanks to my supervisor Dr. Konstantina Griva for offering me the opportunity to work with her and enlightening me on my research. Her dedication to health research energizes me and helps me find my sense of direction in my life. I am also very grateful to Jo-an, Augustine and Zhihui for spending much of their precious time on proofreading this thesis. In addition, my sincere thanks go to Ivy for providing me great assistance in the recruitment process. Finally, I owe my sincere thanks to my good friends, Anastasia, Jannah and Jean for working together with me and being there for me at the most needed times. i Table of Contents Acknowledgments ............................................................................................................... i Table of Contents ................................................................................................................ ii Summary ............................................................................................................................ iv List of Tables ...................................................................................................................... v List of Figures .................................................................................................................... vi List of Appendices ............................................................................................................ vii Chapter One ........................................................................................................................ 1 Introduction ......................................................................................................... 1 End Stage Renal Disease ................................................................................. 1 Health Beliefs .................................................................................................. 9 Emotional Distress......................................................................................... 11 Quality of Life ............................................................................................... 14 Nonadherence ................................................................................................ 17 Determinants of Nonadherence ..................................................................... 26 Limitations of Previous Studies..................................................................... 30 Study Objectives ............................................................................................ 31 Study Hypotheses .......................................................................................... 31 Chapter Two ..................................................................................................................... 33 Methodology ..................................................................................................... 33 Participants .................................................................................................... 33 Study Instruments .......................................................................................... 35 Study Languages............................................................................................ 43 Data Analysis................................................................................................. 44 Ethics ............................................................................................................. 48 Chapter Three ................................................................................................................... 49 ii Results ............................................................................................................... 49 Demographics ................................................................................................ 49 Health Beliefs ................................................................................................ 52 Emotional Distress......................................................................................... 54 Quality of Life ............................................................................................... 57 Prevalence of Nonadherence ......................................................................... 59 Factors Associated With Self-reported Nonadherence.................................. 65 Factors Associated With Nonadherence Based on Biochemical Markers .... 81 Chapter Four ..................................................................................................................... 86 Discussion ......................................................................................................... 86 Overview ....................................................................................................... 86 Health Beliefs ................................................................................................ 86 Emotional Distress......................................................................................... 88 Quality of Life ............................................................................................... 92 Prevalence of Nonadherence ......................................................................... 96 Determinants of Nonadherence ................................................................... 102 Clinical Recommendations .......................................................................... 114 Study Strengths and Limitations.................................................................. 116 Future Studies .............................................................................................. 118 Chapter Five .................................................................................................................... 120 Conclusion ....................................................................................................... 120 References ....................................................................................................................... 121 iii Summary The purpose of this study was to evaluate the prevalence of nonadherence and associated psychosocial factors (i.e., health beliefs, emotional distress and quality of life) in patients undergoing peritoneal dialysis (PD). Intentional (e.g., reasoned decision) and unintentional (e.g., forgetting) nonadherence were investigated as separate constructs and compared. It was a cross-sectional survey conducted with 144 PD patients. Nonadherence to all three main components of the therapeutic regimen, i.e., performing dialysis exchanges, taking medication as instructed and restricting dietary intake, was common, with dietary guidelines the most difficult to adhere to. Intentional nonadherence occurred more frequently than unintentional nonadherence for dialysis and diet; intentional and unintentional nonadherence to medication were equivalent. Nonadherence was strongly affected by psychosocial factors. More specifically, patient satisfaction was the most important predictor of intentional nonadherence to dialysis, whereas environment quality of life was the strongest predictor of unintentional nonadherence to dialysis. Self-efficacy was the strongest predictor of intentional and unintentional nonadherence to both medication and diet. iv List of Tables Table 1: Nonadherence Rates Documented in PD Patients .............................................. 20  Table 2: Number of Items, Reliability Coefficients and Concepts Measured by the KDQOL-SF Domains ....................................................................................................... 40  Table 3: Number of Items, Reliability Coefficients and Concepts Measured by the WHOQOL-BREF Domains .............................................................................................. 41  Table 4: Distribution of Missing Values........................................................................... 44  Table 5: Demographical Characteristics of APD and CAPD Patients ............................. 50  Table 6: Clinical Characteristics of APD and CAPD Patients ......................................... 51  Table 7: Health Beliefs Results in APD and CAPD Patients ........................................... 53  Table 8: Emotional Distress Results in APD and CAPD Patients .................................... 55  Table 9: Quality of Life Results in APD and CAPD Patients .......................................... 58  Table 10: Self-reported Nonadherence Outcomes in APD and CAPD Patients ............... 63  Table 11: Nonadherence Based on Biochemical Markers in APD and CAPD Patients ... 64  Table 12: Comparisons of Self-reported Nonadherence Between Different Subgroups .. 67  Table 13: Spearman Rank Correlations Between Self-reported Nonadherence and Selected Variables............................................................................................................. 69  Table 14: Multivariate Correlates of Self-reported Nonadherence to Dialysis Guidelines ........................................................................................................................................... 72  Table 15: Multivariate Correlates of Self-reported Nonadherence to Medication Guidelines ......................................................................................................................... 73  Table 16: Multivariate Correlates of Self-reported Nonadherence to Dietary Guidelines 75  Table 17: Spearman Rank Correlations Between Emotional Distress and Health Beliefs76  Table 18: Factors Affecting Nonadherence Based on Biochemical Markers in Univariate Analyses ............................................................................................................................ 82 Table 19: Factors Affecting Nonadherence Based on Biochemical Markers in Multivariate Analyses ....................................................................................................... 85 Table 20: WHO Identified Categories Affecting Nonadherence and Significant Predictors of Nonadherence in Our Study ....................................................................................... 103  v List of Figures Figure 1. Illustration of Hemodialysis Procedure ............................................................... 3  Figure 2. Illustration of Peritoneal Dialysis Procedure ....................................................... 5  Figure 3. Flowchart of the Recruitment Process. .............................................................. 34  Figure 4. A Simple Mediation Model. .............................................................................. 46  Figure 5. Anxiety Score Distribution in All Patients. ....................................................... 54  Figure 6. Depression Score Distribution in All Patients ................................................... 55  Figure 7. Loneliness Score Distribution in All Patients ................................................... 56  Figure 8. Distribution of All Patients’ Frequencies of Overall Deviation From Different Aspects of the Therapeutic Regimen ................................................................................ 61  Figure 9. Distribution of All Patients’ Frequencies of Intentional and Unintentional Deviation From Different Aspects of the Therapeutic Regimen. ..................................... 62  Figure 10. Health Belief Mediators of the Relationship Between Anxiety and Intentional Nonadherence to Medication. ........................................................................................... 77  Figure 11. Health Belief Mediators of the Relationship Between Depression and Intentional Nonadherence to Medication .......................................................................... 78  Figure 12. Health Belief Mediators of the Relationship Between Anxiety and Unintentional Nonadherence to Diet ................................................................................ 79  Figure 13. Health Belief Mediators of the Relationship Between Depression and Unintentional Nonadherence to Diet ................................................................................ 80  Figure 14. Health Belief Mediator of the Relationship Between Anxiety and Intentional Nonadherence to Diet. ...................................................................................................... 80  vi List of Appendices Appendix A: Permission to Use Figure 1 and Figure 2 .................................................. 149  Appendix B: Participation Information Sheet................................................................. 152  Appendix C: Consent Form ............................................................................................ 156  Appendix D: Demographics Questionnaire .................................................................... 157 Appendix E: Medical Form ............................................................................................ 159  Appendix F: Permission to Use Table 2 ......................................................................... 161  Appendix G: Permission to Use Table 3......................................................................... 163  Appendix H: Nonadherence Measures ........................................................................... 165  Appendix I: Research Ethics Approval........................................................................... 167  vii CHAPTER ONE Introduction End Stage Renal Disease End stage renal disease (ESRD) is the final stage of chronic kidney disease (CKD), marked by a glomerular filtration rate (GFR) of less than 15 ml/min/1.73 m2. Patients with ESRD have kidneys failing to effectively remove wastes, keep appropriate levels of electrolytes (e.g., sodium, potassium, calcium, magnesium) and reabsorb glucose, blood proteins (e.g., albumin) and other small molecules (Kaazempur-Mofrad, Vacanti, Krebs, & Borenstein, 2004). Reported incident rates of ESRD varied from 13 in Bangladesh to 557 per million population in Morelos and reported prevalent rates of ESRD varied from 110 in Philippines to 2311 per million population in Taiwan in 2008 (United States Renal Data System [USRDS], 2010). Diabetes mellitus is a major cause of ESRD, accounting for more than 40% of ESRD incident cases in most countries (USRDS, 2010). Other important causes of ESRD include glomerulonephritis and high blood pressure. Symptoms common in ESRD patients include fatigue/tiredness, pruritus, constipation, anorexia, pain, sleep disturbance, anxiety, dyspnea, nausea, restless legs and depression (Murtagh, Addington-Hall, & Higginson, 2007). Transplant and dialysis are renal replacement therapy (RRT) choices for ESRD patients to partially restore their kidney functions and sustain life. Transplant Transplant is the most ideal form of treatment for ESRD patients. Prevalent rates of functioning grafts worldwide varied from 29 in Romania to 572 per million population in Norway in 2008 (USRDS, 2010). The corresponding rate in Singapore was 344.5 1 (Singapore Renal Registry [SRR], 2010). The 1- and 5-year survival rates for transplant patients in Singapore were 97.5% and 91.5% respectively (SRR, 2010). In transplant, a healthy kidney donated by a relative or others is placed in the body to take over the work of the old, dysfunctional kidney. Patients are required to continuously take immunosuppressants to prevent the body from rejecting the new kidney after transplantation. Apart from this, transplant patients live a relatively normal life, with much less fluid and dietary restrictions and clinical visits when compared to dialysis patients (Christensen & Ehlers, 2002). Despite the good clinical and psychological outcomes, kidney transplantation remains underutilized mainly due to a shortage of kidney donors. Only 23% of the treated ESRD patients worldwide were living with a functioning transplanted kidney at the end of 2004 (Grassmann, Gioberge, Moeller, & Brown, 2005). Furthermore, kidney transplantation is only an option for a select group of patients. Medical contraindications and high comorbidity burden limit the patient pool. Legislation may also preclude elderly patients, the fastest rising segment of the renal population (SRR, 2010), from transplant candidacy. In Singapore, patients over 60 are not eligible transplant candidates (Vathsala & Chow, 2009). National data indicates that most transplant kidneys (68%) are from deceased donors (SRR, 2010) and patients need to wait for a median of 9.44 years for deceased-donor renal transplants (Vathsala & Chow, 2009). Thus most ESRD patients need dialysis to sustain life. Hemodialysis Hemodialysis (HD) is the predominant dialysis modality used in most parts of the world (USRDS, 2010). Approximately 89% of dialysis patients were undergoing HD and 2 mann et al., 2005). In 11% underggoing peritooneal dialysiis (PD) at thhe end of 20004 (Grassm S Singapore, 85.7% of diialysis patieents were trreated with H HD and 14..3% with PD D in 2008 ( (SRR, 20100). Prevalennt rates of H HD worldwiide varied ffrom 103 inn Philippinees to 2097 p million populationn in Taiwann in 2008 (U per USRDS, 20010). The ccorrespondinng rate in S Singapore w was 908.6 (SRR, ( 20100). The 1- annd 5-year survival s ratees for HD ppatients in S Singapore w were 89.2% and 58.9% respectivelyy (SRR, 20110). Figgure 1. Illusstration of hhemodialysis proceduree. Note. Used with perm N mission from m the Nationnal Institutee of Diabetees and Digeestive and K Kidney Disseases, Natioonal Instituttes of Healtth (see Apppendix A). R Retrieved August A 16, 2 2011, from http://kidneey.niddk.nihh.gov/kudiseeases/pubs/hhemodialysiis/images/diialysis.gif. The basic HD pprocedure iss illustratedd in Figure 11. The undeerlying mechanism of H is simillar to PD. B HD Blood and ddialysate aree separated by a semippermeable m membrane. S Substances (e.g., wasstes, toxinss, excessivee water) diffuse d from m a regionn of high c concentratio on (blood) into one off lower conncentration ((dialysate) and a are subbsequently r removed froom the bodyy together with w the diaalysate. In H HD, this difffusion proceess occurs o outside the bbody relyinng on a HD machine. m HD D patients usually u go too hospitals or o dialysis c centers to aattend dialyssis sessionss and are paassive recipients of treaatment. Stanndard HD s sessions aree 4 to 5 hourrs, 3 times a week (Madduell et al., 22003). 3 Peritoneal Dialysis Peritoneal dialysis (PD) is a home-based renal therapy involving patients’ active participation. Patients or carers receive trainings about how to perform PD exchanges from PD nurses and are supposed to perform the procedures independently at home after the training period. Prevalent rates of PD worldwide varied from 2 in Bangladesh to 846 in Hong Kong per million population in 2008 (USRDS, 2010). The corresponding rate in Singapore was 163.6 (SRR, 2010). The 1- and 5-year survival rates for PD in Singapore were 81.1% and 26.7% respectively (SRR, 2010). The basic PD procedure is demonstrated in Figure 2. PD uses patient’s peritoneum as a natural semipermeable membrane. While the diffusion process occurs outside the body of a HD patient, it occurs in the abdomen of a PD patient. A catheter (soft tube) is inserted into the abdomen of a PD patient and this operation makes the patient vulnerable to peritonitis. Dialysate flows into the peritoneal cavity through the catheter, stays there (patients are ambulatory during this period) absorbing wastes, toxins and excess water from the blood and then is drained out of the body together with the wastes. Then the infusion process begins again and the procedure repeats. The draining and infusion process is called an exchange, taking 30 to 60 minutes depending on the patient’s health status. These repeated exchanges can be performed either manually by patients (continuous ambulatory peritoneal dialysis, CAPD) or automatically using a mechanical devise over night (automated peritoneal dialysis, APD). As opposed to intermittent schedules of HD, PD is a continuous treatment that is performed daily. A typical CAPD 4 p prescription n involves 4-6 manual eexchanges ddaily typicallly every 3 or o 4 hours. A APD lasts f 9 to 12 hhours while the patient sleeps at nigght. for Figurre 2. Illustraation of periitoneal dialyysis proceduure. Note. Used with perm N mission from m the Nationnal Institutee of Diabetees and Digeestive and K Kidney Disseases, Natioonal Instituttes of Healtth (see Apppendix A). R Retrieved August A 16, 2 2011, from http://kidneey.niddk.nihh.gov/kudiseeases/pubs/pperitoneal /images/Perittoneal.gif. D Dialysis vs. Transplantt Diallysis is not as effectivee as transplaant. Projecteed years of life are 3 too 17 years l longer for trransplant paatients than those on thhe waiting llist (Wolfe eet al., 1999)). Dialysis p patients also have worrse outcomes than trannsplant patiients in term ms of qualiity of life ( (Basok et aal., 2009; L Lee, Morgann, Conway, & Currie, 2005; Maglakelidze, P Pantsulaia, T Tchokhonel lidze, Manaagadze, & C Chkhotua, 22011; Ogutm men et al., 22006; Panaagopoulou, H Hardalias, B Berati, & F Fourtounas, 2009), althhough thesee results shoould be vieewed with c caution as transplant patients p tennd to be yoounger and healthier tthan dialysiis patients ( (Bakewell, H Higgins, & Edmunds, 22001; Niu & Li, 2005; O Ogutmen ett al., 2006). 5 PD vs. HD Clinical outcomes, namely mortality, morbidity and hospitalization, are comparable between PD and HD (Harris, Lamping, Brown, & Constantinovici, 2002; Keshaviah, Collins, Ma, Churchill, & Thorpe, 2002; Selgas et al., 2001). A study in Singapore found higher mortality in diabetic patients on PD than HD but not in nondiabetic patients (Noshad, Sadreddini, Nezami, Salekzamani, & Ardalan, 2009). But the generalization of study is called into question as it only recruited 60 PD and 60 HD patients. There is good evidence indicating that PD enables better preservation of residual renal function (Moist et al., 2000; Oreopoulos, Ossareh, & Thodis, 2008) and is associated with less cognitive decline (Conde et al., 2010). A limiting factor of PD use is peritonitis (Bender, Bernardini, & Piraino, 2006). Peritonitis is probably the most important reason for PD technique failure and drop-out from PD programs, contributing to approximately16% death in PD patients (Davenport, 2009; Kawaguchi et al., 2003; Li et al., 2010). There is a misconception that PD patients are more vulnerable to infections than HD patients due to peritonitis. In fact, the overall risks of infection are similar for PD and HD, although infection types are different (Aslam, Bernardini, Fried, Burr, & Piraino, 2006). PD and HD populations have distinct characteristics. PD population tend to be younger, married, healthier and more educated (Ahlmen, Carlsson, & Schonborg, 1993; Little, Irwin, Marshall, Rayner, & Smith, 2001; Marron et al., 2005; Ponz Clemente et al., 2010; Stack, 2002). There is an increasing emphasis on expanding the penetration and utilization rate of PD. In Hong Kong, 79.4% patients are on CAPD (USRDS, 2010). Jalisco and Morelos are two places in Mexico where more than half of the dialysis 6 patients are on PD as well (USRDS, 2010). Since PD is less expensive, increasing the use of PD has significant effect on government budget. It has been estimated that if PD utilization increases to 40% in Singapore, savings to government will be around $ 25 million per year (Walker, Chen, & Bhattacharyya, 2007). Although PD tends to be favored by the younger patients who are fully ambulatory and independent, it is important to recognize that currently there is a shift in recommending and placing less independent patients on PD regimes (Dimkovic & Oreopoulos, 2008). Offering PD to older patients can be supported by informal or formal carers who oversee dialysis procedures or by assisted PD schemes that involve daily visits by community nurses at patients’ home to initiate PD exchanges (Jassal & Watson, 2011). APD vs. CAPD Research to date has focused predominantly on comparing medical endpoints between APD and CAPD, leaving the psychosocial outcomes poorly understood (Guney et al., 2010). Clinical studies cannot confirm a clear superiority of one modality over another, with most studies documenting equivalent outcomes in the two groups in terms of survival, technical failure, hospitalization, peritonitis, dialysis adequacy, clearance, hernias rates and the decline of residual renal function (Balasubramanian, McKitty, & Fan, 2011; Mehrotra, 2009; Mehrotra, Chiu, Kalantar-Zadeh, & Vonesh, 2009; Michels, Verduijn, Boeschoten, Dekker, & Krediet, 2009; Tang & Lai, 2007). APD may be more beneficial for certain groups, such as high transporters whose peritoneal membrane allows for rapid solute transport (Johnson et al., 2010). In addition, peritonitis risk is reduced in APD with Luer connections compared with CAPD with a disconnect system, as shown in two randomized controlled trials (Piraino & Sheth, 2010). 7 Another advantage of APD is its ability to reduce intra-abdominal pressure (Enoch, Aslam, & Piraino, 2002) which is frequently intolerable for some elderly. Known concerns regarding APD involve inadequate removal of sodium and poor hypertension control (Ortega et al., 2001; Rodriguez-Carmona, Perez-Fontan, Garca-Naveiro, Villaverde, & Peteiro, 2004). Individualization of APD based on patient characteristics may modify these risk factors (Brunkhorst, 2005). The percentage of APD in PD users is rising steadily in recent years in many parts of the world, like Canada, US, Singapore and Switzerland (Blake, 1999; Dell'Aquila, Berlingo, Pellanda, & Contestabile, 2009; Mehrotra, 2009) In Singapore, incident rate of PD patients choosing APD had increased from 3% in 1999 to 50.5% in 2008 (SRR, 2010). Patients are motivated to choose APD mainly due to the autonomy it provides instead of medical considerations (Mehrotra, 2009). APD is especially appealing to young and independent patients (Badve et al., 2008; Balasubramanian et al., 2011; Fine & Ho, 2002; Johnson et al., 2010). A main factor limiting APD use is the high cost associated with the machine. APD on average costs 20% more than CAPD (Dell'Aquila et al., 2009). Since APD is performed at night and less onerous, it is especially favorable for two groups of patients. The first group includes students and employers whose day time activities demand minimum disruptions (Dell'Aquila et al., 2009; Liakopoulos & Dombros, 2009). With the use of APD, patients do not need to interrupt their study or work several times a day to perform exchanges. APD also allows patients to avoid the embarrassing experience of performing exchanges in front of peers and thus is more appealing to patients. 8 The second group of patients who can gain great benefits from APD includes those who are highly dependent on others for their treatment, such as children and the elderly (Dell'Aquila et al., 2009; Liakopoulos & Dombros, 2009). Since only two connections are required for APD each day (vs. multiple connections in CAPD), it is easier for employed carers to perform these exchanges without major life disruptions. For elderly patients without good social support but still want to receive treatment at home, it is less expensive for APD patients to hire nurses to visit their house since APD requires fewer visits. Elderly patients on APD in nursing homes have more time to take part in day time activities and this greatly facilitates their rehabilitations (Dimkovic & Oreopoulos, 2008). Health Beliefs Past studies on patients’ health behaviors tend to depict patients as passive recipients of medical advice which is given to patients’ best interest (Donovan, 1995). Doctors feel frustrated about patients’ inability to stick to treatment plans which give rise to various adverse outcomes such as elevated hospitalization, morbidity and mortality (Vermeire, Hearnshaw, Van Royen, & Denekens, 2001). These disappointing outcomes motivate researchers to reexamine patients’ involvement in their therapy decisions and to take into consideration patients’ beliefs. Different models have been proposed to explain how patients’ beliefs affect their health decisions, such as the Health Belief Model (HBM) (Rosenstock, 1974), Theory of Planned Behaviour (TPB) (Ajzen & Fishbein, 1980), Common Sense Model (CSM) (Leventhal, Diefenbach, & Leventhal, 1992) and Medication Adherence Model (MAM) (Johnson, 2002). This study focused on two 9 models that receive significant research attention, namely Horne and Weinman’s (1999) necessity-concerns model and Bandura’s (1977) self-efficacy theory. Beliefs About Medicines Horne and Weinman (1999) proposed a necessity-concerns model to understand medication-taking behaviors among patients with chronic illnesses. The basic idea is that patients do not follow doctors’ advice without questioning, but perform elaborate calculations based on their beliefs about medicines. Main considerations include perceived usefulness/necessity of the prescribed medication and perceived disruptive effects/injuries caused by the medication. Patients are inclined to take their prescribed medication when the perceived necessity exceeds perceived concerns. Otherwise patients may adjust or skip medication to suit their needs. Necessity or concern beliefs can also work independently. For instance, if patients experience dangerous drug interactions as a result of taking their prescribed medication, they tend to adjust medication doses no matter how important they perceive the medication to be. Self-efficacy Self-efficacy is a key concept in Bandura’s social learning theory and is defined as “people’s beliefs in their capabilities to exercise control over their own level of functioning and over events that affect their lives” (Bandura, 1991, p. 257). The effect of self-efficacy in determining health behaviors is well-established in literature and this theory has been successfully applied in different settings such as smoking relapse prevention, pain management, weight control and rehabilitation from myocardial infarction (O'Leary, 1985). Dialysis patients experience various stressors caused by their disease and treatment. The top five stressors among dialysis patients are limitation of 10 physical activity, decrease in social life, uncertainty about the future, fatigue and muscle cramps (Lok, 1996). Patients’ with high self-efficacy are very likely to face these stressors directly, actively search for effective ways to minimize the influence of these stressors, set high goals for their behaviors to maintain health and remain resilient when confronted with physiological or social barriers (Bandura, 1977). Emotional Distress Emotional distress has received considerable attention in dialysis patients. Patients on dialysis have been reported to spend about six hours of their day (not including sleeping hours) in negative affective states (Song et al., 2011). A lot of factors, such as the intrusive nature of renal treatment (Griva, Davenport, Harrison, & Newman, 2010), high burden of symptoms (Murtagh et al., 2007), impaired daily functioning (Cook & Jassal, 2008) and severe sleep problems (Guney et al., 2010) may give rise to high emotional distress in this group. Emotional distress can be further worsened by patients’ reluctance to get evaluation and treatment for it (Wuerth, Finkelstein, & Finkelstein, 2005), which may be especially pronounced in Singapore where there is a high level of stigma attached to mental illness (Lai, Hong, & Chee, 2001) Depression Typical major depression symptoms include “depressed mood, anhedonia, appetite or weight change, sleep disturbance, fatigue, psychomotor disturbance, feelings of worthlessness or guilt, impaired concentration and suicidal thoughts” (Koenig, George, Peterson, & Pieper, 1997, p. 1378). The golden standard for diagnosis of clinical depression is the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition 11 (DSM-IV) criteria. The Beck Depression Inventory (BDI) and Hospital Anxiety and Depression Scale (HADS) are the most widely used self-administered questionnaires for screening depression. Around 80% of patients classified as depressed based on BDI or HADS meet DSM-IV criteria (Atalay et al., 2010; Herrero et al., 2003; Wuerth et al., 2005). Depression is the most common psychiatric disorder seen in dialysis patients (Ibrahim & El Salamony, 2008). The prevalence of major depression in patients on renal replacement therapy has been estimated to be from 20% to 30%, considerably higher than general populations (Atalay et al., 2010; Fukunishi et al., 2002). Since many symptoms of depression, such as appetite loss, fatigue and impaired concentration, may be renal disease manifestations, physicians may overlook the presence of depression among dialysis patients. Age, gender, smoking, comorbidities, serum IL-6, albumin, perceived illness effects and social support are useful in predicting depression (Hung et al., 2011; Kimmel et al., 1995; Osthus et al., 2010; Patel, Shah, Peterson, & Kimmel, 2002). Depression is associated with various adverse outcomes such as death, hospitalization and peritonitis (Diefenthaeler, Wagner, Poli-de-Figueiredo, Zimmermann, & Saitovitch, 2008; Hedayati et al., 2008; Troidle et al., 2003). For instance, it has been recognized that patients who scored 16 or higher on BDI had a 2.7-fold increased risk of mortality than those who scored lower on BDI (Chilcot, Davenport, Wellsted, Firth, & Farrington, 2011). It is unclear why depression is associated with poor outcomes, probably through impairing immune functioning, nutritional status and self-care abilities (Kimmel, Weihs, & Peterson, 1993). In addition, depressed patients are more likely to have poor sleep quality 12 (Guney et al., 2008) and sexual dysfunction (Lew-Starowicz & Gellert, 2009). Depression may also affect patients’ intention to withdraw from dialysis treatment (Christensen & Ehlers, 2002). Anxiety Anxiety is characterized by feelings of uncertainty, tension, helplessness, inadequacy, self-consciousness, concentration difficulties, feeling flushed, perspiring, damp hands, irregular breathing, racing heartbeat, and dry mouth (Endler, Parker, Bagby, & Cox, 1991). In contrast to the prosperity of studies on depression, anxiety in dialysis patients receives little attention, even though anxiety is also associated with poor outcomes such as lowered quality of life and increased likelihood of sexual dysfunction (Sayin, Mutluay, & Sindel, 2007; Steele et al., 1996; Vasilieva, 2006; Vazquez et al., 2005). Conventional measures of anxiety include the Hospital Anxiety and Depression Scale (HADS) and the State-Trait Anxiety Inventory (Spielberger, 1985). The prevalence of anxiety was reported to be 18.6% in a recent study involving 97 adult HD and PD patients (Partridge & Robertson, 2011). Around one third of the HD patients were diagnosed with anxiety in the study of Taskapan et al. (2005). Few studies investigated the prevalence of anxiety in PD patients. A study in Singapore reported 13% depression and 50% anxiety in 30 CAPD patients (Lye, Chan, Leong, & van der Straaten, 1997). Loneliness Loneliness is defined as "to the extent that a person's network of social relationships is smaller or less satisfying than the person desires" (Peplau & Perlman, 1979, p. 101). Perceptions of loneliness have rarely been recorded in dialysis patients, although it is reasonable to expect its high prevalence in this group, since dialysis patients 13 tend to have decreased social life and frequently face family problems and marriage malaises due to dialysis (Lok, 1996; Rapisarda et al., 2006). One phenomenological study identified loneliness as an important theme in patients’ experience with dialysis (Herlin & Wann-Hansson, 2010). Another study assessed loneliness among CAPD patients together with their carers and detected low level of loneliness in both groups (Asti, Kara, Ipek, & Erci, 2006). But this study used a relatively young patient sample (mean age around 45) and it may not be appropriate to generalize findings in this sample to the elderly patients who tend to have decreased economic and social resources (Buemi et al., 2008). Emotional Distress on APD vs. CAPD It is unclearly whether APD offers patients better psychological adjustment than CAPD, since only three studies with small sample sizes compared emotional outcomes in APD and CAPD patients and reported mixed results. One study reported equivalent depression rates in the two groups (Guney et al., 2010). Another study reported lower rate of depression in APD than CAPD patients (Griva et al., 2010). Similarly, de Wit, Merkus, Krediet and de Charro (2001) reported that APD patients were less depressed and anxious than CAPD patients. Quality of Life Quality of life is defined by the World Health Organization (WHO) as “individuals’ perception of their position in life in the context of the culture and the value systems in which they live and in relation to their goals, expectations, standards and concerns” (Harper & Power, 1998, p. 551). 14 Quality of life measures can be divided into generic and disease-specific. Generic instruments measure concepts that are relevant to everyone’s well-being (e.g., life satisfaction), can be applied in different populations and allow comparisons across different groups (Patrick & Deyo, 1989). Disease specific measures are used only in limited populations, but are more sensitive in detecting small quality of life changes associated with specific conditions (e.g., severity of disease) (Valderrabano, Jofre, & Lopez-Gomez, 2001). The 36-item Short-Form Health Survey (SF-36), EuroQOL 5 Dimension (EQ-5D), and World Health Organization Quality of Life Instrument, Short Form (WHOQOL-BREF) are the top three used generic quality of life measures, whereas the Kidney Disease Quality of Life instrument (KDQOL) and its shortened versions (KDQOL-SF, KDQOL-36) are the most commonly used measures for assessing disease specific quality of life in ESRD patients (Glover, Banks, Carson, Martin, & Duffy, 2011). The current study measured both generic and disease specific quality of life, as recommended in literature (Valderrabano et al., 2001). Quality of life impairment is predominant in dialysis patients when compared with healthy individuals (Maglakelidze et al., 2011; Osthus et al., 2010; Sayin et al., 2007) , especially in the physical health domain (Bohlke et al., 2008; Brown et al., 2010; de Wit et al., 2001). Psychological factors, such as health beliefs and social support/deprivation, have been found to be important in predicting quality of life (Bakewell, Higgins, & Edmunds, 2002; Theofilou, 2011; Wight et al., 1998). Nonpsychological factors, such as age, gender, hospitalization, number of comorbid diseases, primary kidney disease, nutritional status and dialysis adequacy may also influence 15 quality of life (de Wit et al., 2001; Fructuoso, Castro, Oliveira, Prata, & Morgado, 2011; Senol, Sipahioglu, Ozturk, Argun, & Utas, 2010). The importance of quality of life is increasingly appreciated, not only because it is inversely correlated with hazards of hospitalization and mortality (DeOreo, 1997; Valderrabano et al., 2001), but also because it evaluates the effectiveness of treatment based on patients’ subjective feelings (Fructuoso et al., 2011). Dialysis patients are willing to trade less living time for better quality of life (Jhamb et al., 2011; Tsevat et al., 1998), corresponding to Socrates’ adage that “The really important thing is not to live, but to live well”. Nephrologists also place more weight on quality of life than mortality and morbidity in recommending dialysis modalities (Mendelssohn, Mullaney, Jung, Blake, & Mehta, 2001). Various efforts have been initiated to improve patients’ quality of life, such as adjusting dialysis prescription, controlling comorbidities, treating anemia and alleviating depression (Ross, Hollen, & Fitzgerald, 2006). Quality of Life on APD vs. CAPD Although APD is expected to offer patients better quality of life due to its less onerous nature (Balasubramanian et al., 2011), this hypothesis is not well-supported in literature. Five studies compared quality of life outcomes between APD and CAPD patients, with four of them suggesting equivalent quality of life between the two groups based on SF- 36 scores (Balasubramanian et al., 2011; Bro et al., 1999; de Wit et al., 2001; Guney et al., 2010) and one suggesting worse physical but better mental quality of life in APD patients (Diaz-Buxo, Lowrie, Lew, Zhang, & Lazarus, 2000). The last study (Diaz-Buxo et al., 2000) did not control for critical covariates (e.g., comorbidity, time on dialysis) and thus its results should be viewed with caution. A recent longitudinal study 16 examined quality of life in 119 APD and 105 CAPD patients with SF-36 and reported worse baseline quality of life in CAPD patients, but no differences were found after a year (Balasubramanian et al., 2011). Thus it is quite possible that PD modality per se is not a significant predictor of quality of life after patients adapt to it. Nonadherence Definition and Measurement of Nonadherence Any deviation from doctor’s treatment instructions was viewed as nonadherence in this study. In contrast to abundance of research on quality of life and emotional adjustment, there has been little focus on treatment nonadherence. Nonadherence to treatment regime is a key contributor to poor survival in patients treated with dialysis, probably in the same order of importance as medical indicators (Bander & Walters, 1998). Dialysis regimen is extremely complicated and time-consuming, involving regular clinical visits, attending dialysis sessions, taking a variety of medications, limiting water intake and paying great attention to food choices. As treatment complexity has been cited as the most important reason affecting patients’ nonadherence (Donovan, 1995), it is not unexpected that it is easy for dialysis patients to be nonadherent. Nonadherence studies in dialysis patients are greatly hindered by a lack of consistent standards for measuring nonadherence. Common measures of nonadherence include: (a) report from patients or medical staff, (b) biological and biochemical markers, (c) electronic monitoring and (d) checking medication refill status and inspection of dialysate delivery records. Each method has its own drawbacks. The most widely used method is self-report, a cost-effective way of measuring nonadherence (George, 17 Mackinnon, Kong, & Stewart, 2006), though there are doubts about the accuracy of selfreport (Horne & Weinman, 1999; Vlaminck, Maes, Jacobs, Reyntjens, & Evers, 2001). For instance, Haynes et al. (1980) reported that patients underestimated their nonadherence by 17% in self-reports when compared with pill count. Common biological and biochemical markers used to measure nonadherence include interdialytic weight, phosphorus, potassium and albumin levels (Karamanidou, Clatworthy, Weinman, & Horne, 2008; Kugler, Maeding, & Russell, 2011). The validity of using biochemical markers to indicate nonadherence is challenged by using arbitrary, instead of theory supported, cut-off values to divide patients into adherence and nonadherence groups and by factors irrelevant with nonadherence such as residual renal function, dialysis prescriptions, disease conditions and demographic characteristics (Denhaerynck et al., 2007). Electronic devices such as the Medication Event Monitoring System (MEMS, Aardex, Switzerland) and the Home-Choice Pro card (Baxter Healthcare Corporation, Deerfield, Illinois, USA) are capable of providing reliable measures of nonadherence (Chua & Warady, 2011; Sevick et al., 1999), but the associated high costs limit their wide application. Medication refill rates (Gincherman, Moloney, McKee, & Coyne, 2010) and dialysate delivery records (Fine, 1997) cannot reveal whether patients actually utilize the medication or dialysate (e.g., patients may give the medication or dialysate to others) or whether patients use them correctly. The current study used the combination of selfreport (used in a non-threatening way) and biochemical markers (i.e., serum potassium, phosphate and albumin levels) to detect nonadherence and this design has been suggested as powerful at detecting nonadherence (George et al., 2006; Inui, Carter, & Pecoraro, 1981). 18 Another key problem in past studies is treating nonadherence as an unidimensional concept, reflected by using a composite score to indicate overall nonadherence (Lin & Liang, 1997; Pakpour et al., 2010; Sayin et al., 2007; Vives et al., 1999). However, patients do not perceive different aspects of the therapeutic regimen (e.g., dialysis, medication, diet, fluid) as equally important and have differing levels of difficulty in managing separate treatment components (Smith et al., 2010; Stack et al., 2010). In studies measuring nonadherence to several components of the therapeutic regimen simultaneously, nonadherence rates were found to be different and affected by different factors (Karamanidou, Clatworthy, et al., 2008; Sensky, Leger, & Gilmour, 1996). Therefore, it is more reasonable to assess nonadherence to different aspects of the therapeutic regimen as separate constructs. Studies reporting nonadherence rates to three main components of the therapeutic regimen (i.e., dialysis, medication and diet) among PD patients are listed out in Table 1. 19 20 Note. NA = nonadherence. USA Warren et al., 1994 121 CAPD USA 64 PD 191 PD 19 APD Italy Italy 173 CAPD Hong Kong Russo et al., 2006 119 HD 51 PD USA 42 APD 75 HD, 14 CAPD USA Juergensen et al., 2004 39 HD, 15 PD 67 / 26 CAPD 30 PD Israel USA Holley et al., 2006 Katzir et al., 2010 Kutner et al., 2002 Lam et al., 2010 Neri et al., 2002 Nolph et al., 1995 Canada Fine, 1997 35 PD China Brazil 92 PD USA Figueiredo et al., 2005 49 PD USA 51 pediatric APD 15 CAPD 5 APD USA USA 50 PD Sample Italy Area Chua et al., 2011 Amici et al., 1996 Bernardini et al., 1997 Bernardini et al., 1998 Bernardini et al., 2000 Chen et al., 2006 Author, year of publication Performing less than 90% of prescribed exchanges Excessive dietary protein intake (DPI) < 0.8 g/kg/d or > 1.2 g/kg/d Following a prescription variable (duration/ number of cycles/number of sessions/dialysate volume) < 95% of the time. Performing less than 90% of prescribed exchanges Performing less than 90% of prescribed exchanges Performing less than 90% of prescribed exchanges Definition of NA Not clear Laboratory data Home visit Laboratory data Creatinine excretion (CrEx) > 1.24 Dialysis: not compliant with the exchange protocol procedures Medication: incongruence between drugs at home and prescribed in the clinical file Creatinine excretion (CrEx) > 1.24 Baxter PD Link software Missing PD sessions in 90 days Self-report / laboratory Missing at least one treatment / shortening at least one data treatment / PO4 > 7.5mg/dl Mild, moderate, severe or very severe deviation from Self-report therapeutic regimen Self-report Self-report / records of dialysate delivery and Using less than 90% of prescribed dialysate pick-up Not to fill a prescription / Not to take specific Self-report medications Baxter Home-Choice Delivered dialysis volume/prescribed dialysis PRO Card volume)*100 < 95% Self-report Baxter Home-Choice PRO Card Laboratory data Self-report Self-report Self-report PD Adequest software Software model Method Table 1Equation 1 Nonadherence Rates Documented in PD Patients 26% 23% Male / Female: 17% / 5% 53% 7% HD:19% / 31% / 19% PD:30% / 4% / 10% 17% 12% / 5% 30% 4%/8%/18%/22% 30% 35% 40% 22% Dialysis   25% 17% HD / CAPD: 27% / 2% 30% / 21% Medication % NA     20 62% 77.1% Diet Nonadherence to Dialysis Procedures It is life-threatening if dialysis patients fail to perform dialysis exchanges as prescribed, since wastes, toxins and excess water may accumulate in the body and disturb its hemodynamic status. Despite its importance, nonadherence to dialysis prescriptions is a common problem. Reported rates of nonadherence to dialysis prescriptions in PD varied from 4% to 53% (Amici et al., 1996; Bernardini, Nagy, & Piraino, 2000; Bernardini & Piraino, 1997, 1998; Chua & Warady, 2011; Figueiredo, Santos, & Creutzberg, 2005; Fine, 1997; Kutner, Zhang, McClellan, & Cole, 2002; Lam, Twinn, & Chan, 2010; Neri, Viglino, Cappelletti, Gandolfo, & Barbieri, 2002; Nolph et al., 1995; Russo et al., 2006; Warren & Brandes, 1994). The corresponding rates in HD varied from 0% to 32% (Bleyer et al., 1999; Block, Hulbert-Shearon, Levin, & Port, 1998; DeOreo, 1997; Hecking et al., 2004; Kutner et al., 2002; Leggat et al., 1998; Sherman, Cody, Matera, Rogers, & Solanchick, 1994; Taskapan et al., 2005). Factors associated with nonadherence to dialysis prescriptions include smoking, younger age and ethnicity with blacks reporting more nonadherence than whites (Kimmel et al., 1995; Kutner et al., 2002; Leggat et al., 1998; Unruh, Evans, Fink, Powe, & Meyer, 2005). Perceived negative effects of treatment on daily life and less control perception over future health were identified as predictors of shortening behaviors in one study (Kutner et al., 2002). Nonadherence to dialysis has been found to be associated with higher mortality and lower likelihood of kidney transplantation in HD (Unruh et al., 2005). Data on PD patients showed that nonadherence to dialysis is associated with technique failure, inadequate dialysis, increased peritonitis rates and hospitalizations (Bernardini et al., 2000; Bernardini & Piraino, 1998). 21 Nonadherence to Medication In addition to performing exchanges regularly, dialysis patients are expected to take multiple tablets to control their phosphate levels (dialysis procedure is unable to remove phosphate from the body adequately) and manage symptoms and comorbid diseases. Dialysis pill burden is ranked as one of the highest among chronic illnesses (Chiu et al., 2009). Patients take 10 to 12 different types of medications and one fourth of dialysis patients take more than 25 pills per day (Chiu et al., 2009; Manley et al., 2004). Medications have distinct requirements for mode, timing and amount of intake. The complexity of medication regimen significantly increases patients’ likelihood of nonadherence (Chiu et al., 2009). A total of 2% to 30% PD patients fail to take their prescribed medication as instructed (Holley & DeVore, 2006; Katzir et al., 2010; Lam et al., 2010; Russo et al., 2006), whereas 17% to 99% HD patients do not adhere to their prescribed medication (Curtin, Svarstad, & Keller, 1999; Lin & Liang, 1997). Age, pill burden, health literacy, health beliefs, personality, social support, and patient satisfaction have been cited as important factors affecting medication nonadherence (Browne & Merighi, 2010; Karamanidou, Clatworthy, et al., 2008). Important barriers to medication adherence include non user-friendly drug compound, feeling of discomfort, forgetfulness, polypharmacy and patient ignorance (Lindberg & Lindberg, 2008). Inadequate control of phosphorus level is linked with several risk factors for cardiovascular disease, such as elevated blood pressure, hyperkinetic circulation, increased cardiac work, and high arterial tensile stress (Marchais, Metivier, Guerin, & London, 1999). While hyperkalemia (high potassium) is a common problem among HD 22 patients, hypokalemia (low potassium) is profound among PD patients because there is greater filtration of potassium from blood to dialysate during the dialysis process in PD than HD (Factor, 2007; Khan, Bernardini, Johnston, & Piraino, 1996). Potassium < 3.5 mmol/l is associated with increased mortality, risk of peritonitis and poor nutritional status (Chuang, Shu, Yu, Cheng, & Chen, 2009; Szeto et al., 2005) and can be managed effectively with potassium supplements or increasing dietary potassium intake. No studies have investigated nonadherence to potassium supplements among PD patients. Nonadherence to Diet Because dialysis does not restore functioning levels comparable to a health kidney, dietary restrictions are often used together with medications to prevent the increment of certain elements, such as sodium, phosphorus and protein, in the body. Dietary restriction is the most distressing part of dialysis regimen (Durose, Holdsworth, Watson, & Przygrodzka, 2004; Lam et al., 2010), probably because it involves profound alterations to individuals’ lifestyles. Only two studies investigated dietary nonadherence in PD patients and found that 62% to 77.1% of patients did not follow their dietary guidelines (Chen, Lu, & Wang, 2006; Lam et al., 2010). The prevalent rates of dietary nonadherence in HD patients varied from 24% to 81.4% (Kara, Caglar, & Kilic, 2007; Kugler, Vlaminck, Haverich, & Maes, 2005; Lin & Liang, 1997; Vlaminck et al., 2001). Factors affecting dietary nonadherence have not been adequately examined. One study in Hispanic patients identified knowledge of diet, language, food consumption frequency, socioeconomic status, family support and attitudes toward the renal diet as important factors related to dietary nonadherence (Morales Lopez, Burrowes, Gizis, & Brommage, 2007). Another study revealed that younger male patients and smokers were 23 more likely to be nonadherent to diet and fluid restrictions (Kugler et al., 2005). Since poor understanding with dietary contents was an important reason for nonadherence, using menu suggestion was found effective in reducing dietary nonadherence (Chen et al., 2006). Nonadherence with salt and fluid was a critical reason for PD drop-out (Kawaguchi et al., 2003). Intentional and Unintentional Nonadherence Increasing knowledge is a standard way employed to reduce nonadherence in intervention programmes. However, nonadherence is problematic even among those with good knowledge (Lee & Molassiotis, 2002; Nerbass et al., 2010). Clarifying causes of nonadherence and intervening accordingly may be more cost-effective. Two broad categories have been proposed to classify causes of nonadherence: intentional and unintentional (Clifford, Barber, & Horne, 2008). This “intentional-unintentional” typology is predominantly used in investigating medication-taking behaviors (Daleboudt, Broadbent, McQueen, & Kaptein, 2010; Unni & Farris, 2011). Both forms of nonadherence have been observed in dialysis patients (McCarthy, Cook, Fairweather, Shaban, & Martin-McDonald, 2009; Nerbass et al., 2010; Polaschek, 2007), although no attempt has been made to document their prevalent rates. Unintentional nonadherence is a passive process, like forgetting to take medication or failing to recognize what are contained in food when eating. Patients are usually not aware of their deviation from treatment guidelines when it occurs. Factors such as complexity of treatment and disease severity contribute to unintentional nonadherence (Schuz et al., 2011). Intentional nonadherence is an active, decision-making process. Patients deliberately adjust their regimen to suit their needs, like forgoing medications to avoid side effects. Intentional 24 nonadherence is especially likely to occur if patients experience consequences as a result of adherence (e.g., dangerous drug interactions) or if patients are not well-informed and henceforth feel uncertain about the effectiveness of treatment (Schuz et al., 2011). There is good evidence suggesting that unintentional nonadherence occurs more often than intentional nonadherence in other populations (Rees, Leong, Crowston, & Lamoureux, 2010; Sewitch et al., 2003; Unni & Farris, 2011). A phenomenological study in another type of chronic illness identified forgetfulness, accidentally overdose and the unavailability of medication as reasons for unintentional nonadherence and intentional nonadherence was mainly caused by side effects, social activities, eating out, drinking alcohol or traveling (Eliasson, Clifford, Barber, & Marin, 2011). To the best of our knowledge, no studies concerning ESRD patients have distinguished intentional and unintentional nonadherence. Nonadherence on APD vs. CAPD Only one known study compared nonadherence difference between APD and CAPD patients (Bernardini et al., 2000). Home visit supply inventories were used to evaluate nonadherence to dialysis exchanges in this study and PD modality was identified as an independent predictor of nonadherence, with more nonadherence reported in CAPD than APD patients. Limitations of Previous Studies on Nonadherence in PD Patients As can be seen from Table 1, nonadherence levels in PD patients are relatively understudied compared to outcomes such as quality of life. Seventeen studies could be retrieved but these present several limitations. Most of these studies focused on nonadherence to dialysis and medication, overlooking nonadherence in relation to dietary 25 recommendations for PD patients. Although changes in diet are essential for management of conditions to ensure good clinical outcomes, little is known on rates of nonadherence with respect to dietary recommendation in this population. Recruited study samples were very small, which limits generalizability of findings. Only four studies had sample sizes above 100. Methodological criteria to define nonadherence in some studies are questionable. For instance, two studies defined nonadherence as creatinine excretion (CrEx) > 1.24 and this was later found to be an unreliable marker of nonadherence (Blake, Spanner, McMurray, Lindsay, & Ferguson, 1996). No previous studies used traditional biochemical markers (e.g., potassium, phosphate) to measure nonadherence in PD patients. Moreover, no studies have compared nonadherence outcomes in APD and CAPD. The majority of studies were forced to merge between APD and CAPD groups due to small sample sizes or were only based on CAPD patients. Nonadherence rates in APD patients were hence either not assessed or reported together with CAPD patients, so the question of which PD modality may be associated with less nonadherence remains largely unanswered. Lastly, no studies in PD patients have looked at intentional and unintentional nonadherence despite their important implications for intervention. Determinants of Nonadherence Previous studies on determinants of nonadherence tend to focus on demographical and clinical variables, overlooking the effects of psychosocial variables (Karamanidou, Clatworthy, et al., 2008; Russell, Knowles, & Peace, 2007). However, identifying demographical and clinical associates of nonadherence is of limited use in clinical applications as these factors are usually not modifiable (Sensky et al., 1996). Moreover, it 26 has been suggested that psychosocial factors are stronger determinants of nonadherence than demographical and clinical variables (Karamanidou, Clatworthy, et al., 2008). Therefore, it is imperative to examine the effects of psychosocial factors on nonadherence. This study focuses on three psychosocial variables: health beliefs, emotional distress and quality of life. Health Beliefs Little data is available about the effects of beliefs about medicines on nonadherence in ESRD patients. However, the association between beliefs about medicines and nonadherence to medication is well-supported in other populations (Daleboudt et al., 2010; Horne & Weinman, 1999; Schuz et al., 2011; Unni & Farris, 2011). In a study involving 324 patients from different chronic conditions, the difference between perceived necessity of the prescribed medication and perceived concerns about the medication (e.g., side effects, long-term dependence) was found to be an independent predictor of medication nonadherence, accounting for a good portion (19%) of the variance (Horne & Weinman, 1999). Concern beliefs and necessity beliefs appear to have different roles in determining intentional and unintentional nonadherence. Concern beliefs have been reported to affect both intentional and unintentional nonadherence (Daleboudt et al., 2010; Unni & Farris, 2011), whereas necessity beliefs were mainly associated with intentional nonadherence (Schuz et al., 2011; Unni & Farris, 2011). There is ample evidence suggesting that self-efficacy is correlated with nonadherence to fluid, medication and diet in dialysis patients (Brady, Tucker, Alfino, Tarrant, & Finlayson, 1997; Christensen, Wiebe, Benotsch, & Lawton, 1996; Eitel, Friend, Griffin, & Wadhwa, 1998; Lindberg & Fernandes, 2010; Oka & Chaboyer, 2001; 27 Rosenbaum & Ben-Ari Smira, 1986; Schneider, Friend, Whitaker, & Wadhwa, 1991; Zrinyi et al., 2003). The impact of self-efficacy on nonadherence to dialysis exchanges is however, yet to be determined. A recent study consisting of 133 HD patients reported significantly less fluid intake in patients with high self-efficacy than patients with low self-efficacy (Lindberg, Wikstrom, & Lindberg, 2010). Similarly, another study involving a large group of HD patients associated self-efficacy with dietary nonadherence based on self-report and biochemical markers (Zrinyi et al., 2003). Emotional Distress Depression has been found to be associated with nonadherence (Brownbridge & Fielding, 1994; Cukor, Rosenthal, Jindal, Brown, & Kimmel, 2009; De-Nour & Czaczkes, 1976). Depression may have a direct effect on nonadherence as symptoms may manifest as reduced appetite, excessive fatigue and a lack of energy which limit patients’ ability to adhere (McCarthy et al., 2009). Depression may also have an indirect effect on nonadherence through beliefs and cognitions. Depressed patients tend to have negative thoughts and feel hopeless and despair about self, world and future (Dekker et al., 2011). It is very likely that depressed patients devaluate their ability to cope with their diseases (i.e., having low self-efficacy), underestimate the effectiveness of their treatment and hold exaggerated concerns about possible disruptive effects of their treatment and thus show nonadherence behaviors (DiMatteo, Lepper, & Croghan, 2000). Given that previous intervention programmes targeting depression tend to have low response rates (Wuerth et al., 2005), identifying mediators between depression and nonadherence has the potential to find a more direct and appealing way to help depressed patients with elevated nonadherence. 28 Health beliefs appear to be important mediators between depression and nonadherence in other populations (Chao, Nau, Aikens, & Taylor, 2005; Sacco et al., 2007; Sacco et al., 2005; Schoenthaler, Ogedegbe, & Allegrante, 2009). For instance, a study involving 445 patients with diabetes reported that depression affected nonadherence mainly via perceived side effects, perceived general barriers, and selfefficacy (Chao et al., 2005). Given that depression is significantly associated with health beliefs (Devins et al., 1982; Tsay & Healstead, 2002) and health beliefs, as stated above, are commonly associated with nonadherence. It is highly probable that self-efficacy also mediates the depression-nonadherence relationship in dialysis patients. This hypothesis was tested in this study. Anxiety was also found to be associated with nonadherence (Brownbridge & Fielding, 1994). The relationships among anxiety, health beliefs and nonadherence are rarely explored in literature and were also examined in this study. Quality of Life Only two studies associated quality of life with nonadherence (DeOreo, 1997; Pakpour et al., 2010) and the direction of this association is not clear. DeOreo (1997) studied 1000 HD patients and revealed that physical quality of life was higher, but mental quality of life was lower in patients who skipped more than two treatments per month than other patients. In contrast, Pakpour et al. (2010) reported a positive association between physical quality of life and adherence and no association between mental quality of life and adherence in a group of 250 HD patients. 29 Limitations of Previous Studies Previous studies have shed some light on outcomes related to PD yet present with several shortcomings that limit generalizability of findings to other PD populations. The majority of studies comparing outcomes in APD and CAPD have focused exclusively on clinical endpoints such as peritonitis, morbidity and mortality. Psychological and behavioral outcomes, such as emotional distress, quality of life, and nonadherence, have largely been overlooked. Studies that explored psychological outcomes have very small sample sizes (not exceeding 70) and some merge across APD and CAPD into one group. There are also conceptual and method limitations in the measures of nonadherence such as using unreliable biochemical markers (e.g., creatinine excretion). Most of the studies on PD patients have focused on nonadherence yo dialytic prescriptions (e.g., missing exchanges, shortening dialysis time) or prescribed medication, overlooking the lifestyle aspects of treatment such as dietary recommendations. No studies have explored intentional and unintentional nonadherence. The factors that may explain/predict nadherence difficulties and self-care behaviors in patients maintained on PD are also not well understood. Psychosocial factors, which are more proximal predictors of nonadherence and more amendable to interventions, receive little attention in literature than demographical and clinical variables. Furthermore, to the best of our knowledge, there have been no studies on psychosocial outcomes in PD patients in Singapore. It is hard to extrapolate or generalize 30 experience gained in other countries to Singapore due to the local variation in the important factors (e.g., sociopolitical, economic, cultural). Study Objectives  To assess the effect of PD modality (APD or CAPD) on health beliefs, emotional distress and quality of life  To document overall, intentional and unintentional nonadherence rates to different aspects of the therapeutic regimen (i.e., dialysis, medication and diet)  To identify psychosocial determinants of nonadherence Study Hypotheses Two main hypotheses were suggested for this study based on previous findings: Hypothesis 1 focuses on the prevalence of nonadherence among PD patients. Based on what has been discussed above, we proposed Hypothesis 1a that nonadherence to dietary restrictions is higher than nonadherence to medical aspects of treatment regimes, namely dialysis exchanges and medication. The exact differences between the three components could not be predicted, since past studies used different definitions and measures of nonadherence. Hypothesis 1b is that unintentional nonadherence is expected to occur more than intentional nonadherence, as explained above. Hypothesis 2 is regarding determinants of nonadherence. Hypothesis 2a is that positive health beliefs (i.e., high necessity beliefs, low concern beliefs and high selfefficacy) are expected to be associated with less nonadherence; Hypothesis 2b that more emotional distress is expected to be associated with more nonadherence. 31 No apriori directional hypotheses were formulated regarding psychosocial (i.e., health beliefs, emotional distress, quality of life and nonadherence) differences between APD and CAPD groups as well as the association between quality of life and nonadherence, as there is no clear theoretical or empirical evidence. 32 CHAPTER TWO Methodology Participants This was a cross-sectional study. Participants were recruited from the Peritoneal Dialysis Center, Singapore General Hospital (SGH) from October 2010 to June 2011. Inclusion Criteria Patients were recruited if they met the following criteria: 1. In the PD programme for a minimum of 3 months; 2. Aged 21 years or over; 3. Able to communicate verbally with research assistants; 4. Able to provide informed consent. Recruitment Process Patients were randomly approached by research assistants while awaiting consultations with nephrologists at the PD center. All approached patients were given a short introduction about the purpose and procedure of our study. After a brief screening, patients failing to meet the inclusion criteria were excluded from participation. The rest of the eligible patients were given an information sheet listing out details of the study (see Appendix B). It was made clear to the patients that participation was strictly voluntary and confidential. Any additional questions were answered appropriately before patients made their decisions concerning participation. The overall recruitment process is shown in Figure 3. 33 Patients approached n = 199 Eligible n = 183 Agreed to participate n = 146 Included n = 144 Excluded n = 16 Declined n = 37 Dropped-out n=2 Figure 3. Flowchart of the recruitment process. One hundred and ninety nine patients were approached at the PD center. Of the 199, 16 failed to meet the inclusion criteria and were excluded. Of the eligible 183, 37 declined invitation to participate. Of the 146 who agreed to participate, n = 2 patients dropped out due to sudden decline in health status. Thus, our final sample consisted of 144 PD patients (response rate = 79%). Written informed consent (see Appendix C) was obtained from patients prior to questionnaire administration. Standard to research conducted in busy clinical settings, study assessments were scheduled based on patients’ preference, availability and convenience. Questionnaires were therefore completed and returned to research in the following ways (data only available for n = 132 patients): 1. Completing the questionnaires at the hospital while awaiting consultations with nephrologists or medicines (n = 33). 34 2. Bringing the study information sheet and questionnaires back to further consider about it and posting their questionnaires back to us if they wanted to (n = 37). 3. Indicating a preferred time and allowing a research assistant to administer a home visit (n = 62). Study Instruments Demographics Demographics (age, gender, marital status, ethnicity, education, employment status, housing condition and time on dialysis) were collected with self-designed questions (see Appendix D). Patients were also inquired about who were taking a more active role (self vs. carer) in managing their disease. Medical Information Medical information (prescribed medicines, comorbid diseases, primary cause of ESRD, PD modality, creatinine, potassium, phosphate, albumin, hemoglobin and Kt/V, see Appendix E) was abstracted from medical records by nursing staff in the participating clinic. Charlson Comorbidity Index (CCI). CCI score was calculated using the validated method suggested by Beddhu et al. (2002) for PD patients. Included comorbid diseases were coronary artery disease, heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, connective tissue disorder, peptic ulcer disease, diabetes, hemiplegia, moderate or severe renal disease, tumor, leukemia, lymphoma, liver disease, metastatic solid tumor and AIDS. One point was added to the CCI score for each decade of > 40 years of age. Patients were classified into low, 35 moderate and high comorbidity groups based on being positioned before the 33rd, between 33rd and 66th, and after the 66th percentiles of CCI scores respectively. Health Beliefs Measures The Beliefs About Medicines Questionnaire (BMQ).Patients’ medicine beliefs were assessed with the BMQ (Horne, Weinman, & Hankins, 1999). It consists of a necessity subscale and a concerns subscale (5 items for each), measuring perceived necessity/usefulness about taking the prescribed medication and perceived disruptive effects/dependence effects caused by the medication, respectively. Aggregate scores (range = 5−25) were used, with higher scores indicating more necessity or concern beliefs. The difference between the necessity score and the concerns score was calculated for each individual. If this value was positive, it indicated more necessity beliefs over concern beliefs. Otherwise, it indicated necessity beliefs equal to or lower than concern beliefs. Both subscales showed high reliability in our study. Necessity subscale Cronbach’s α = .81 and concerns subscale Cronbach’s α = .77. Self-efficacy. It was assessed with questions adapted from the Self Efficacy for Managing Chronic Disease Scale developed by Lorig et al. (2001). It consists of a set of 13 items, asking patients to rate their confidence in accomplishing various tasks involved in managing their illness on a 10-point Likert scale (e.g., “How confident are you that you can perform your PD exchanges as instructed by your doctor?” 1 = not at all confident, 10 = totally confident). The first six items (Cronbach’s α = .89) were from the original scale, measuring generic self-efficacy. We also designed additional seven items, one item measuring dialysis self-efficacy, two measuring medication self-efficacy (Cronbach's α = .61) and four measuring dietary self-efficacy (Cronbach's α = .87). Mean 36 scores (range = 1−10) were used, with higher scores denoting higher levels of selfefficacy. Emotional Distress Measures The Hospital Anxiety and Depression Scale (HADS). Symptoms of depression and anxiety were assessed using the HADS (Zigmond & Snaith, 1983). HADS is a widely used instrument, consisting of an anxiety subscale (7 items) and a depression subscale (7 items). Patients were asked to rate their degree of anxiety and depression in the past 14 days on a 4-point Likert scale (0 = yes, definitely, 3 = no, not at all). Aggregate score for each subscale was derived, ranging from 0 to 21. A higher score signifies higher level of anxiety or depression. Both the depression and anxiety subscales demonstrated satisfactory reliability. Anxiety subscale Cronbach’s α = .83. Depression subscale Cronbach’s α = .76. The aggregate scores were classified into three categories: normal (0−7), borderline (8−10) and abnormal (11−21) (Thompson et al., 2000). HADS has been validated and applied successfully in measuring anxiety and depression among ESRD patients (Loosman, Siegert, Korzec, & Honig, 2010; Riezebos, Nauta, Honig, Dekker, & Siegert, 2010). This measure was preferred over other instruments as it precludes somatic symptoms such as fatigue, anorexia, and weight loss (Fernandes et al., 2010) which may confound with symptoms related to renal condition and uremia. The Revised UCLA Loneliness Scale. Loneliness was assessed with the Revised UCLA Loneliness Scale (Russell, Peplau, & Cutrona, 1980). It includes 20 items. Patients responded to the questions on a 4-point Likert scale. The aggregate score (range = 20−80) was calculated, with higher score indicating a higher degree of loneliness. This scale had high reliability on the study (Cronbach’s α =.85). The loneliness scores were 37 classified into four categories: low (20−34), moderate (35−49), moderately high (50−64) and high (65−80) (Perry, 1990). Quality of Life Measures Three measures were used to capture all aspects of quality of life: generic health related quality of life, disease specific quality of life and subjective global quality of life. The Kidney Disease Quality of Life Short Form (KDQOL-SF). Disease specific quality of life was assessed with the KDQOL-SF (Hays et al., 1997), a special instrument targeting ESRD patients. The original questionnaire includes the 36-item Short-Form Health Survey (SF-36), supplemented by kidney disease-specific questions. SF-36 was replaced by its shorter version, the 12-item Short-Form Health Survey (SF-12) in our study to reduce burden of completion for respondents. SF-12 is a generic health related quality of life measure. Two summary scores, physical component summary score (PCS) and mental component summary score (MCS), were calculated using the standard way (Ware, Kosinski, Keller, & Institute, 1998). The eight domain scores used in SF-36 were not calculated, as suggested in the SF-12 manual (Ware et al., 1998). Conveniently, Singapore’s SF-36 PCS and MCS norms are available (Thumboo et al., 2001) and a calculator is provided online (http://www.singhealth.com.sg/Research/ HealthServicesResearch/OurServices/OutcomesResearch/Documents/HSRSingHealthSF 36calculator-v3-beta.xls) to calculate PCS and MCS norms after adjusting for age, gender and ethnicity. It has been justified to use SF-36 norms to interpret SF-12 results (Brown et al., 2010). The kidney disease specific part includes 31 items to measure disease specific quality of life, encompassing 6 domains: symptoms, effects of kidney disease, burden of kidney disease, patient satisfaction, staff encouragement and social support. All 38 6 domain scores were transformed to be varying from 0 to 100. Higher scores signify better quality of life for all KDQOL-SF domains. Table 2 provides a summary for the KDQOL domains regarding their number of items, reliability coefficients on the study and interpretations. All KDQOL-SF subscales used in this study showed acceptable reliability (range = .66 to. 85). KDQOL-SF has been verified in 1180 Singapore HD patients and showed satisfactory psychometric properties (Joshi, Mooppil, & Lim, 2010). The World Health Organization Quality of Life Instrument, Short Form (WHOQOL-BREF). Subjective global quality of life was measured with the WHOQOLBREF (Herrman et al., 1998). This instrument includes 26 items. Twenty four out of the 26 items were used to assess patients' quality of life in four domains: physical health, psychological health, social relations, and environment. Each domain score ranged from 4 to 20, with higher score signifying better quality of life. Table 3 displays the number of items, reliability coefficients and interpretations for the four domains. All domains showed acceptable reliability, with Cronbach’s α ranging from .61 to .81. WHOQOLBREF has been validated in 23 countries consisting of 11,830 sick and well adults (Skevington, Lotfy, & O'Connell, 2004) and the norms reported in this article were compared with our WHOQOL-BREF results. 39 Table 2: Number of Items, Reliability Coefficients and Concepts Measured by KDQOL-SF Domains Number of Items, Reliability Coefficients and Concepts Measured by the KDQOL-SF Domains No. of items Cronbach’s α Concept measured SF-12 PCS 6 .66 Physical component summary score SF-12 MCS 6 .72 Mental component summary score Symptom list 12 .85 Extent that patients are bothered by dialysis-related symptoms such as chest pain, cramps and itchy skin Effects of kidney 8 .85 disease Extent that patients are bothered by daily life issues such as fluid restriction, dietary restriction, inability to dependence on medical staff Burden of kidney 4 .79 travel and Extent that patients are bothered by time disease consumed by dialysis, its intrusiveness and Patient satisfaction burden on family Patients’ satisfaction with care received for 1 dialysis Staff 2 .84 patients to be independent and support patients in dealing with their kidney disease encouragement Social support Extent to which renal staff encourage 4 .71 Patients’ satisfaction with togetherness and support from family and friends Note. KDQOL-SF = Kidney Disease Quality of Life Short Form. SF-12 = 12-item ShortForm Health Survey. PCS = physical component summary score. MCS = mental component summary score. Cited from “Psychosocial predictors of non-compliance in haemodialysis and peritoneal dialysis patients,” by N. G. Kutner, R. Zhang, W. M. McClellan, & S. A .Cole, 2002, Nephrology, Dialysis, Transplantation, 17, p. 95. Used with permission from the Oxford University Press (see Appendix F). ("WHOQOL-BREF Introduction, Administration, Scoring and Generic Version of the Assessment, Field Trial Version," 1996 ) 40 Table 3: Number of Items, Reliability Coefficients and Concepts Measured by WHOQOL-BREF Domains Number of Items, Reliability Coefficients and Concepts Measured by the WHOQOLBREF Domains Domain Concepts measured Physical health Activities of daily living 7 items, Dependence on medicinal substances and medical aids Cronbach’s α = .80 Energy and fatigue Mobility Pain and discomfort Sleep and rest Work Capacity Psychological health Bodily image and appearance 6 items, Negative feelings Cronbach’s α = .81 Positive feelings Self-esteem Spirituality / Religion / Personal beliefs Thinking, learning, memory and concentration Social relations Personal relationships 3 items Social support Cronbach’s α = .61 Sexual activity Environment Financial resources 8 items Freedom, physical safety and security Cronbach’s α = .76 Health and social care: accessibility and quality Home environment Opportunities for acquiring new information and skills Participation in and opportunities for recreation / leisure activities Physical environment (pollution / noise / traffic / climate) Transport Note. WHOQOL-BREF = World Health Organization Quality of Life Instrument, Short Form. Cited from “WHOQOL-BREF Introduction, Administration, Scoring and Generic Version of the Assessment, Field Trial Version,” by M. Rapley, 1996, World health Organization, p. 7. Used with permission from the World Health Organization (see Appendix G). 41 Nonadherence Measure No single method of measuring nonadherence is perfect. For this study, we elected to use self-report as the primary outcome and serum concentrations of potassium, phosphate and albumin as a secondary outcome. Self-reported nonadherence. A dialysis specific measure (see Appendix H) was designed for the purpose of this study to assess self-reported nonadherence to the PD regimens. The scale contains 9 items, measuring nonadherence to prescribed dialysis exchanges (3 items), medication (3 items) and diet (3 items). The three questions used to assess medication nonadherence were adapted from Horne and Weinman (1999): 1. “How often do you follow this regime?” (overall nonadherence) 2. “Some people forget to take their medicines. Overall, how often does this happen to you?” (unintentional / accidental nonadherence) 3. “Some people decide to miss out a dose of their medication or adjust it to suit their own needs. Overall, how often do you do this?” (intentional / deliberate nonadherence) Similar formats were used for measuring nonadherence to dialysis and diet. But for nonadherence to diet, one more item was added before the three questions: “Have you been asked to follow a diet?” If patients answered “no” to this question, they did not need to answer the rest questions regarding nonadherence to diet. Potential responses to each question are from 1 (= never) to 5 (= very often). The score for the first item (“How often do you follow this regime?”) was reversed. Thus higher scores indicated greater nonadherence for all questions. The overall nonadherence scale showed good reliability in this study (Cronbach's α = .77). Nonadherence was defined as at least rarely deviating 42 from treatment regimen. Two additional questions were developed in accordance with the criteria defined in the USRDS (1997) survey of dialysis morbidity and mortality to assess patients' behavioral nonadherence to dialysis guidelines: 1. “During the last 4 weeks, how many times have you skipped/missed one of your PD sessions/exchanges?” 2. “During the past 4 weeks, how many times have you shortened your PD session by 10 minutes?” These two items were treated as stand alone items for the analyses, with higher scores signifying more frequent nonadherence to dialysis exchanges/regimen. Nonadherence based on biochemical markers. Three biochemical markers (i.e., potassium, phosphate and albumin) were used to estimate nonadherence to diet and medication. Nonadherence was defined as levels outside clinical cut-offs/ targets: potassium < 3.5 mmol/l (signifying nonadherence to dietary guidelines or taking potassium supplements), phosphate > 1.78 mmol/l (signifying nonadherence to dietary guidelines or phosphate binders), and albumin < 3.5 g/dl (signifying malnutrition and thereby dietary nonadherence). These cut-off values have been widely used in previous studies (Moe et al., 2005; Plantinga et al., 2004; Szeto et al., 2005) and were confirmed by the key PD consultants in patient care as clinical targets for the patient population. Study Languages Patients were free to choose their preferred language in this study. All study instruments were available in English and Chinese versions. The HADS, UCLA loneliness Scale, KDQOL-SF and WHOQOL-BREF have been translated into Chinese 43 and validated in other studies (Fang, Hao, & Li, 1999; Leung, Ho, Kan, Hung, & Chen, 1993; Li, 1998; Liu, 1993). The BMQ and demographical, self-efficacy and adherence questionnaires have no standardized Chinese versions available. They were translated by a research assistant proficient in both English and Chinese. All patients who could only speak Chinese dialects (e.g., Hokkien) or Malay were interviewed with assistance from their family members or by research assistants. Data Analysis Missing Data Table 4 shows the distribution of missing items. As missing items were small in number and randomly distributed, mean imputations were used if at least half of the items in the same domain were answered by the participants (missing Kt/V values were not imputed). Table 4: Distribution of Missing Values Distribution of Missing Values Frequency Percent Kt/V 4 2.8 BMQ 6 0.4 UCLA loneliness 1 0.03 SF-12 2 0.1 WHOQOL-BREF 3 0.08 Nonadherence measures 1 0.08 Note. BMQ = Beliefs about Medicines Questionnaire. UCLA loneliness = Revised UCLA Loneliness Scale. SF-12 = 12-item Short-Form Health Survey. WHOQOL-BREF = World Health Organization Quality of Life Instrument, Short Form. 44 Statistical Analysis Descriptives. Descriptive statistics included medians, means and standard deviations for continuous variables and frequencies and percentages for categorical variables. Univariate analyses. Kolmogorov-Smirnov tests were used to assess the normality of variable distribution. Means were compared with independent t tests, paired t tests, ANOVA or ANCOVA and percentages were compared with χ2 tests unless otherwise specified. When data were not normally distributed, non-parametric tests (Mann-Whitney tests, Wilcoxon signed-rank tests, Kruskal-Wallis tests or Fisher's exact tests) were used instead and noted. Spearman rank correlation coefficients were calculated to denote the correlations among variables. Multivariate analyses. Hierarchical stepwise multiple linear regressions were used to identify important predictors of self-reported nonadherence levels. Hierarchical stepwise logistic regressions were used to identify significant predictors of nonadherence based on biochemical markers. Entry and removal criteria were fixed at p = .05 and p = .10 respectively. The large number of variables measured increased the risk of type I error. Therefore, the first stage of model selection involved initial screening using univariate tests of associations (e.g., ANOVAs, Mann–Whitney tests, correlations or χ2 tests as appropriate) to identify demographical, clinical and psychosocial variables associated with nonadherence indices. A forward stepwise procedure was used to select variables from those that were significant at .05 on initial screening. Mediation analyses. Figure 4 is used to illustrate how to test the existence of mediation effect. X represents the predictor, Y represents the dependent variable and M 45 represents the mediator. Based on the suggestions of Barron and Kenney (1986), the X-Y relationship is totally mediated by M if (1) there is a significant relationship between X and Y before controlling for M (total effect, represented by c path in Figure 4); (2) M is significantly associated with X (a path); (3) the influence of M on Y (b path) is significant; (4) the effect of X on Y becomes zero after controlling for M (direct effect, represented by cʹ path). Indirect effect is the difference between total effect and direct effect and it estimates the effect of X on Y via M. c Y X M b a Y X cʹ Figure 4. A simple mediation model. We tested mediation effects with the use of a SPSS macro (available at www.quantpsy.org ) provided by Preacher and Hayes (2008). It performs the four steps recommended by Baron and Kenny (1986) simultaneously. The significance of indirect effect can be estimated through a nonparametric method (bootstrapping) contained in this Macro. It produces a 95% confidence interval (CI) for indirect effect. If this 95% CI does not include zero, it indicates that indirect effect is significantly different from zero. This statistic method (bootstrapping) is highly recommended in literature (Shrout & Bolger, 2002), because it does not require normal distribution, reduces the probability of Type I 46 error, and allows for multiple mediators to be tested simultaneously after controlling for covariates (Buffardi & Campbell, 2008). Parameters were estimated based on 1000 bootstrap samples in the current study. Specifically, we created and tested all possible mediation effects of health belief variables (i.e., medicine beliefs and/or self-efficacy) on the relationship between emotional distress (i.e., depression, anxiety or loneliness) and nonadherence in three steps, similar to previous studies (Buffardi & Campbell, 2008; Sebire, Standage, & Vansteenkiste, 2009). Firstly, if an emotional factor significantly correlated with one of the six nonadherence measures (intentional/unintentional nonadherence to dialysis/medication/diet) after controlling for covariates, this emotional factor was chosen as X (the predictor) and the corresponding nonadherence measure was chosen as Y (the dependent variable). Secondly, if a health belief variable significantly correlated with both X and Y after controlling for covariates, it was chosen as M (the mediator). If there was more than one health belief variable meeting this criteria, they were entered into the same model and tested simultaneously in order to reduce number of tests and type I error. Thus a mediation model in our study contained only one predictor and one dependent variable, but allowed for multiple mediators. Finally, we used the SPSS macro provided by Preacher and Hayes (2008) to examine the significance of the total effect, direct effect and indirect effect of emotional distress (X) on nonadherence (Y). Statistical significance level was set at .05 for all procedures. Bonferroni αadjustment was applied for all post-hoc comparisons. All the procedures were performed using the Statistical Package for Social Sciences (SPSS) version 17.0 (SPSS Inc, Chicago, IL, USA). 47 Required sample size. Required sample sizes were calculated with G*power 3.1 (Faul, Erdfelder, Buchner, & Lang, 2009) corresponding to two primary study goals: (a) comparing APD and CAPD outcomes (using independent t tests or Mann-Whitney tests) and (b) identifying predictors of intentional and unintentional nonadherence (using multiple linear regressions). A sample size of 64 for each group is required to detect a medium effect size of d = .5,  = .05, power (1) = .8 for two-tailed independent t test. A sample size of 67 for each group is required to detect a medium effect size of d = .5,  = .05, power (1) = .8 for two-tailed Mann-Whitney test. For linear regression with 17 predictors, a sample size of 146 is required to detect a medium effect size of 2 = .15,  = .05, power (1) = .8. Ethics This study was approved by Centralized Institutional Review Board, SingHealth Research Facilities (see Appendix I). ("Seventh Report of the Singapore Renal Registry, 2007/2008," 2010; "USRDS 2010 Annual Data Report," 2010; "The USRDS Dialysis Morbidity and Mortality Study: Wave 2," 1997) 48 CHAPTER THREE Results Demographics Fifty eight automated peritoneal dialysis (APD) and 86 continuous ambulatory peritoneal dialysis (CAPD) patients participated in the study (response rate = 79%). Table 5 summarizes the patients' demographic characteristics. The median age of the patients was 59 years (range = 21−89 years), close to the median age of 60 years for Singapore prevalent PD patients (SRR, 2010). Forty three percent of the patients were male and most patients (72%) were married. Comparisons between PD subgroups indicated that APD patients had been on PD for less time, χ2(2, N = 144) = 13.90, p = .001, and were significantly more likely to be employed, χ2(1, N = 144) = 6.66, p = .01 compared to CAPD respondents. No differences were found for other demographics (age, gender, marital status, ethnicity, education level, employment status and housing condition) between the two groups. These case-mix differences (i.e., time on dialysis and employment status) were entered as covariates in ANCOVA tests conducted to compare APD and CAPD patients, as long as they were associated with the outcome in question. 49 Table 5: Demographical Characteristics of APD and CAPD Patients Demographical Characteristics of APD and CAPD Patients Total APD CAPD (n =144) (n = 58) (n = 86) 58.0 ± 12.9 55.6 ± 14.5 59.6 ± 11.5 .08 62 (43) 26(45) 36(42) .74 103 (72) 37(64) 66 (77) .13 41 (28) 21(36) 20 (23) Chinese 102 (71) 42 (72) 60 (70) Malay/Indian/Other 42 (29) 16 (28) 26 (30) Low 54 (38) 20(34) 34 (39) Intermediate 54 (38) 19(33) 35 (41) High 36 (24) 19(33) 17 (20) Employed 38 (26) 22(38) 16 (19) Unemployed 106 (74) 36(62) 70 (81) 1−4 rooms flats 88 (61) 31(53) 57 (66) 5 rooms flats 35 (24) 16(28) 19 (22) Condominiums or above 21 (15) 11(19) 10 (12) Less than 1 year 32 (22) 21(36) 11 (13) 1 to 2 years 20 (14) 10(17) 10 (12) More than 2 years 92 (64) 27(47) 65 (75) Age in years Male p Marital status Married Widowed/divorced/single/ other Ethnicity .85 Education Level .23 Employment status .01 Housing condition .29 Time on dialysis .001 Note. Data expressed as M ± SD or n (%). APD = Automated Peritoneal Dialysis. CAPD = Continuous Ambulatory Peritoneal Dialysis. 50 Table 6: Clinical Characteristics of APD and CAPD Patients Clinical Characteristics of APD and CAPD Patients Total APD CAPD (n =144) (n = 58) (n = 86) 36 (25) 16 (28) 20 (23) .70 Pill burden 10.3 ± 2.6 10.2 ± 2.7 10.3 ± 2.5 .77a CCI 5.6 ± 2.0 5.3 ± 2.2 5.7 ± 1.9 .33a Low ( ≤ 5) 72(50) 30(52) 42(49) .88 Moderate ( 6) 30(21) 11(19) 19(22) High ( ≥ 7) 42(29) 17(29) 25(29) Diabetes 62(43) 28(48) 34(39) Hypertension 28(19) 8(14) 20(23) Glomerulonephritis 38(27) 16(28) 22(26) Other 16(11) 6(10) 10(12) Creatinine (mmol/l) 0.87 ± 0.31 0.89 ± 0.33 0.85 ± 0.29 .39 Potassium (mmol/l) 4.1 ± 0.7 4.1 ± 0.7 4.1 ± 0.6 .76 Phosphate (mmol/l) 1.60 ± 0.54 1.68 ± 0.57 1.55 ± 0.51 .18 2.9 ± 0.5 2.9 ± 0.6 2.9 ± 0.5 .74 Hemoglobin (g/dl) 10.76 ± 1.69 10.83 ± 1.58 10.72 ± 1.77 .72 Kt/V 2.33 ± 0.99 2.39 ± 1.26 2.29 ± 0.76 .44a Dependence on carer p Primary cause of ESRD Albumin (g/dl) .52 Note. Data expressed as M ± SD or n (%).APD = Automated Peritoneal Dialysis. CAPD = Continuous Ambulatory Peritoneal Dialysis. Dependence on carer = carer accomplishing most treatment tasks (e.g., performing dialysis). Pill burden = number of prescribed medicines. CCI = Charlson Comorbidity Index. ESRD = End Stage Renal Disease. a Mann-Whitney test. 51 Clinical Characteristics Table 6 displays patients' clinical characteristics. Seventy five percent of the patients were relatively independent, requiring minimum or no assistance in their treatment activities. Twenty five percent relied on their carers to accomplish most or all of the treatment tasks (e.g., performing dialysis). A median of 10 types of medicines were prescribed to the patients (M = 10.3, SD = 2.6, range = 5−17). Based on 33rd (≤ 5) and 67th (> 6) percentiles of the CCI scores, 50% patients had moderate to high comorbidity. The most common causes of ESRD were diabetes (43%), glomerulonephritis (27%) and hypertension (19%). Patients were adequately dialyzed but malnourished, as suggested by a mean Kt/V level of 2.33 (SD = 0.99) and a mean albumin level of 2.9 (SD = 0.5) g/dl, respectively. There were no significant differences between APD and CAPD patients on any of the clinical parameters recorded (see Table 6). Health Beliefs BMQ Patients expressed strong beliefs in the necessity of their prescribed medication (M = 20.5, SD = 3.9) and a moderate level of concerns about the disruptive effects of the medication (M = 16.1, SD = 5.1) (Table 7). The necessity-concerns difference score was calculated through subtracting the concerns score from the necessity score for each individual. This difference score was positive (more positive evaluation of necessity beliefs vs. concern beliefs; necessity outweighing any concerns) in 73%, zero (equal necessity and concern beliefs) in 12% and negative (negative evaluation of need of medication vs. concerns; concerns outweighing 52 perceived need) in 15% of the patients. Mann-Whitney tests indicated no differences between APD and CAPD patients on necessity and concern beliefs. Table 7: Health Beliefs Results in APD and CAPD Patients CAPD Patients Health Beliefs Results in APD and CAPD Patients Total APD CAPD (n =144) (n = 58) (n = 86) Necessity 20.5 ± 3.9 20.1 ± 4.0 20.7 ± 3.9 .21a Concerns 16.1 ± 5.1 16.7 ± 4.7 15.7 ± 5.4 .40a Generic SE 5.9 ±2.0 6.0± 2.3 5.8 ± 1.8 .76 Dialysis SE 8.9 ± 1.7 8.7 ± 2.1 9.1 ± 1.3 .53a Medication SE 7.8 ± 1.9 7.2 ± 2.2 8.2 ± 1.6 .01a Dietary SE 7.3 ± 2.1 7.1 ± 2.2 7.5 ± 2.0 .23a p BMQ SE Note. Data expressed as M ± SD. Higher value indicates more necessity/concern beliefs or higher self-efficacy. APD = Automated Peritoneal Dialysis. CAPD = Continuous Ambulatory Peritoneal Dialysis. BMQ = Beliefs about Medicines Questionnaire. SE = self-efficacy. a Mann-Whitney test. Self-efficacy Patients reported high self-efficacy in following dialysis (M = 8.9 out of possible 10, SD = 1.7), medication (M = 7.8, SD = 1.9) and dietary guidelines (M = 7.3, SD = 2.1) and moderate generic self-efficacy in managing their illness (M = 5.9, SD = 2.0). Mann-Whitney tests indicated that mean medication self-efficacy was higher in CAPD (Mdn = 8.5) patients than APD (Mdn = 7.2) patients, U = 1873.0, p = .01. This was further confirmed by a ANCOVA test after controlling for time on dialysis and employment status, F(1, 139) = 10.00, p = .002. Generic, dialysis and dietary self- 53 e efficacy on the other haand were noot significanntly differennt between thhe two PD ssubgroups ( (Table 7). Emotioonal Distresss H HADS HAD DS scores are a displayedd in Table 88. Mean anxxiety score was w 7.0 (SD D = 5.0), in t normal range (0−77). Mean deepression sccore was 9..1 (SD = 4.6), in the bborderline the r range (8−100) based on the classificcation schem me of Thom mpson et al. ((2000). Usinng 8 as cut--off value, possible annxiety was detected d in 41% (n = 59) of all p patients (Figgure 5). Annxiety rates among a APD D and CAPD D groups weere 41% (n = 24) and 4 41% (n = 35), respectively (Tablee 8). Possiblle depressioon was detected in 62% % (n = 90) o all patiennts (Figure 6). of 6 Depressiion rates am mong APD aand CAPD ggroups weree 64% (n = 3 and 62% 37) % (n = 53), respectively r y (Table 8). 26% 15% Normal (0--7) Boderrline (8-10) 59% Abnormal ((11-21) Figgure 5. Anxiiety score ddistribution iin all patientts. 54 38% 40% 22% N Normal (0-7) Boderline (8-10) Abbnormal (11-221) Figuure 6. Depression scoree distributioon in all patiients. T Table 8: Em motional Disstress Resullts in APD aand CAPD P Patients E Emotional D Distress Ressults in APD D and CAPD D Patients Total D APD C CAPD (n =144) (n = 558) (nn = 86) H HADS Anxxiety 7.0 ± 5.0 7.0 ± 4.8 7.00 ± 5.0 0.98a Normal (0-7) 85 (59) 34 (559) 51 (59) 1.00b Borderlinee (8-10) 22 (15) 9 (15) 13 (15) Abnormal (11-21) 37 (26) 15 (226) 22 (26) 9.1 ± 4.6 9.3 ± 4.5 9.00 ± 4.7 0.59a Normal (0-7) 54 (38) 21 (336) 33 (38) 0.59b Borderlinee (8-10) 32 (22) 11 (119) 21 (24) Abnormal (11-21) 58 (40) 26 (445) 32 (38) 36.5 ± 10.1 36.7 ± 10.2 36.33 ± 10.1 0.86c Low (20-334) 62 (43) 27 (447) 35 (41) 0.75d Moderate ((35-49) 67 (47) 24 (441) 43 (50) Moderatelyy high (50-664) 13 (9) 6 (10) 7 (8) High (65-880) 2 (1) 1 (22) 1 (1) H HADS Deprression U UCLA Loneeliness p Note. Data eexpressed aas M ± SD oor n (%). AP N PD = Autom mated Peritooneal Dialyssis. CAPD = Continuouus Ambulattory Peritoneeal Dialysiss. HADS = H Hospital Annxiety and Depression D S Scale. UCL LA lonelinesss = Revisedd UCLA Lonneliness Scaale. a Mann-Whiitney test. bχ2test. cindeppendent t tesst. dFisher’ss exact test. 55 Therre were no differencess in either the levels of depression or anxieety (using c continuous ween APD HADS scores) or freqquencies of depression or anxiety cases betw a and CAPD D patients. These com mparisons were w perform med with Mann-Whittney tests ( (supplemen nted with AN NCOVA tests) using continuous c H HADS scorres and χ2 tests t using c categorized HADS scorres (see Tabble 8). R Revised UC CLA Loneliness Scale Baseed on the classification c n scheme oof Perry (19990), the m mean lonelinness score 3 36.5 (SD = 10.1) was inn moderate range (35− −49). Amongg all patientts, about hallf (n = 67) r reported mooderate loneeliness and additional 10% (n = 115) reportedd moderatelly high or v very high degree d of lonneliness (Fiigure 7). Fiffty three peercent of AP PD (n = 31)) and 59% C CAPD patieents (n = 511) reported at least mooderate degrree of loneliiness (Tablee 8). Both t mean looneliness scoore (tested w the with independent t test)) and the freequency of loneliness c cases (testedd with Fisheer’s exact teest) did not ddiffer betweeen APD annd CAPD paatients. % 9%1% 43% 47% Low (20-344) Moderatee (35-49) M Moderately highh (50-64) H High (65-80) Figure 7. L Loneliness sccore distribuution in all ppatients. 56 Quality of Life KDQOL-SF Table 9 reports quality of life levels for the total PD sample as well as APD and CAPD subgroups. Results on the generic health related quality of life measure, i.e. SF-12, indicated that patients' mean PCS score of 35.7 (SD = 9.2) was significantly lower than the mean MCS score of 43.6 (SD = 11.0), t(143) = −7.16, p < .001. Both the mean PCS and MCS scores were significantly lower than the age-, gender- and ethnicity-adjusted Singapore norms (Thumboo et al., 2001) in one sample t tests (PCS norm 51.2, t[143] = −20.16, p < .001; MCS norm 50.3, t[143] = −7.25, p < .001), indicating that patients' quality of life was impaired in our sample, especially in the physical domain/well-being. Among kidney disease specific quality of life subscales, staff encouragement (M = 72.2, SD = 30.7) and social support (M =71.6, SD = 21.6) scored the highest (indicative of better quality of life scores), while on the other hand burden of kidney disease (M = 32.4, SD = 26.5) scored the lowest. Thus, patients perceived a relatively high level of support from medical staff, family members and friends. They also felt severely bothered by time spent on dialysis, its intrusiveness and burden on family. Mann-Whitney or t tests were performed to compare KDQOL differences between APD and CAPD. APD (Mdn = 79.2) and CAPD (Mdn = 70.8) patients only scored differently on symptoms score, with marginally significantly less symptoms reported by APD patients, Mann-Whitney test U = 2005.5, p = .05. However, ANCOVA tests showed no quality of life differences (including symptoms) between the two groups after controlling for covariates. 57 Table 9: Quality of Life Outcomes in APD and CAPD Patients Quality of Life Results in APD and CAPD Patients Total APD CAPD (n =144) (n = 58) (n = 86) SF-12 PCS 35.7 ± 9.2 36.7 ± 9.9 34.9 ± 8.7 .26 SF-12 MCS 43.6 ± 11.0 44.0 ± 10.3 43.4 ± 11.6 .73 Symptoms 68.8 ± 19.4 72.0 ± 19.7 66.7 ± 18.9 .05a Effects of kidney disease 67.0 ± 22.9 69.1 ± 22.9 65.6 ± 23.0 .37 Burden of kidney disease 32.4 ± 26.5 34.2 ± 27.2 31.2 ± 26.1 .57a Patient satisfaction 65.7 ± 22.1 63.5 ± 21.0 67.2 ± 22.8 .46a Staff encouragement 72.2 ± 30.7 74.6 ± 26.7 70.6 ± 33.1 .83a Social support 71.6 ± 21.6 70.3 ± 22.4 72.5 ± 21.2 .43a Physical health 11.7 ± 3.2 11.7 ± 3.1 11.6 ± 3.3 .93a Psychological health 13.0 ± 3.2 12.7 ± 3.4 13.3 ± 3.0 .28 Social relations 13.2 ± 3.4 13.0 ± 3.6 13.4 ± 3.3 .40a Environment 13.6 ± 2.5 13.4 ± 2.9 13.8 ± 2.3 .32 p KDQOL-SF WHOQOL-BREF Note. Data expressed as M ± SD. Higher values indicates better quality of life. APD = Automated Peritoneal Dialysis. CAPD = Continuous Ambulatory Peritoneal Dialysis. KDQOL-SF = Kidney Disease Quality of Life Short Form. SF-12 = 12-item Short-Form Health Survey. PCS = physical component summary score. MCS = mental component summary score. WHOQOL-BREF = World Health Organization Quality of Life Instrument, Short Form. a Mann-Whitney test. WHOQOL-BREF The international mean scores (Skevington et al., 2004) for the physical health, psychological health, social relations and environment domains were 16.2 (SD = 2.9), 15.0 (SD = 2.8), 14.3 (SD = 3.2) and 13.5 (SD = 2.6). The corresponding four domain scores in our study were 11.7 (SD = 3.2), 13.0 (SD = 3.2), 13.2 (SD = 3.4) and 13.6 (SD = 58 2.5), with the first three scores significantly lower than the norms (t[143] = −16.88, p < .001; t[143] = −7.51, p < .001; t[143] = −3.80, p < .001 respectively) and the last score equivalent to the normative mean (t[143] = 0.55, p = .58). Thus WHOQOL-BREF results indicated that patients’ physical, psychological and social relations, but not environment, quality of life was poorer than international norms. APD and CAPD patients did not score differently on any of the WHOQOL-BREF domains (Mann-Whitney, t tests or ANCOVA tests). Prevalence of Nonadherence Self-reported Nonadherence The percentages of patients reporting “never”, “rarely”, “sometimes”, “often” or “very often” deviating from their treatment recommendations for dialysis, medication and diet are depicted in Figure 8. A total of n = 12 patients reported no dietary restrictions, so dietary nonadherence outcomes were based on the rest 132 patients. Twenty percent, 47% and 75% of the patients reported at least rarely deviating from dialysis, medication and diet guidelines respectively. Mean dialysis nonadherence score was the lowest (M = 1.3, SD = 0.6), followed by medication (M = 1.6, SD = 0.8) and diet (M = 2.2, SD = 0.9) (Table 10). The percentages of patients reporting intentionally and unintentionally deviating from guidelines regarding dialysis exchanges/schedule, medication and diet (“never”, “rarely”, “sometimes”, “often” or “very often”) are shown in Figure 9. Nonadherence was defined as at least rarely deviating from treatment guidelines. According to this criterion, 28%, 58% and 81% of the patients were classified as 59 intentional nonadherent to dialysis, medication and diet, whereas 19%, 71% and 73% were classified as unintentional nonadherent to dialysis, medication and diet, respectively. Wilcoxon signed-rank tests showed that patients reported more intentional (Mdn = 1.0) than unintentional (Mdn = 1.0) nonadherence to dialysis, z = 3.07, p = .002; patients also reported more intentional (Mdn = 3.0) than unintentional (Mdn = 2.0) nonadherence to diet, z = 3.48, p < .001; no significant difference was found between intentional (Mdn = 2.0) and unintentional (Mdn = 2.0) nonadherence to medication, z = 1.73, p = .08. Mann-Whitney tests showed that APD and CAPD patients reported similar nonadherence levels (overall, intentional or unintentional) in all three domains (Table 10). These results were confirmed by ANCOVA tests after controlling for covariates. χ2 test indicated that a significantly higher percentage of CAPD patients (23%) reported shortening treatment sessions when compared with APD patients (2%), χ2(1, N = 144) = 12.89, p < .001. 60 60% 40% 20% 0% Percentages of patients 80% 60% 40% 20% 0% 80% 60% 40% 20% 0% Never Rarely Dialysis Sometimes Medication Often Very often Diet Figure 8. Distribution of all patients’ frequencies of overall deviation from different aspects of the therapeutic regimen. 61 60% 40% 20% 0% Percentages of patients 80% 60% 40% 20% 0% 80% 60% 40% 20% 0% Dialysis Medication Diet Figure 9. Distribution all patients’ frequencies of intentional and unintentional deviation from different aspects of the therapeutic regimen. Note. Intentional nonadherence on the left and unintentional nonadherence on the right of each stack. 62 Table 10: Self-reported Nonadherence Outcomes in APD and CAPD Patients Self-reported Nonadherence Results in APD and CAPD Patients Total APD CAPD (n = 144) (n = 58) (n = 86) Dialysis 1.3 ± 0.6 1.2 ± 0.6 1.3 ± 0.6 .76 Medication 1.6 ± 0.8 1.7 ± 0.8 1.6 ± 0.8 .14 Dietb 2.2 ± 0.9 2.4 ± 0.9 2.2 ± 0.9 .27 Dialysis 1.4 ± 0.8 1.3 ± 0.6 1.5 ± 0.8 .07 Medication 2.0 ± 1.0 2.0 ± 1.0 2.0 ± 1.1 .72 Dietc 2.6 ± 1.1 2.7 ± 1.1 2.6 ± 1.1 .58 Dialysis 1.2 ± 0.5 1.2 ± 0.4 1.2 ± 0.5 .91 Medication 2.1 ± 0.9 2.2 ± 0.9 2.0 ± 0.9 .20 Dietb 2.4 ± 1.1 2.5 ± 1.1 2.3 ± 1.1 .32 Skipping PD sessions 13 (9) 3 (5) 10 (12) .24 Shortening PD sessions 21 (15) 1 (2) 20 (23) < .001 pa Overall nonadherence Intentional nonadherence Unintentional nonadherence Note. Data expressed as M ± SD or n (%). Higher value indicates more nonadherence. APD = Automated Peritoneal Dialysis. CAPD = Continuous Ambulatory Peritoneal Dialysis. a Mann-Whitney or χ2 test. bn = 132 because 12 patients reported no dietary restrictions. cn = 131 because 12 patients reported no dietary restrictions and one did not answer this question. Nonadherence Based on Biochemical Markers Three biochemical markers, potassium, phosphate and albumin, were used to assess clinical nonadherence in this study. Table 11 includes reference values for the three biochemical markers, their clinical meanings and nonadherence rates (i.e. percentages of patients with values outside these targets). 63 Twenty seven (19%) patients had potassium level < 3.5 mmol/l, indicating nonadherence to diet. This might also reflect nonadherence to the prescribed medication as n = 20 out of the 27 low in potassium patients were prescribed with potassium supplements. Forty eight (33%) patients had phosphate level > 1.78 mmol/l, which reflects a combination of poor adherence to diet and phosphate binders. Based on medical records, all respondents but one were prescribed with phosphate binders. One hundred and twenty seven (88%) patients did not achieve target albumin levels. Low albumin reflects poor nutritional status, suggesting that most patients were not eating appropriately. Table 11: Nonadherence Based on Biochemical Markers in APD and CAPD Patients Nonadherence Based on Biochemical Markers in APD and CAPD Patients Target levels Potassium ≥ 3.5 mmol/l Phosphate ≤ 1.78mmol/l Albumin ≥ 3.5 g/dl NA, n (%) Out-target values signify NA to… dietary guidelines/ potassium supplements dietary guidelines/ phosphate binders dietary guidelines Total APD CAPD p 27 (19) 12 (21) 15 (17) .67 48 (33) 23 (40) 25 (29) .21 127 (88) 50 (86) 77 (90) .60 Note. APD = Automated Peritoneal Dialysis. CAPD = Continuous Ambulatory Peritoneal Dialysis. NA = nonadherence. Nonadherence rates based on biochemical markers were equivalent between APD and CAPD patients (see Table 11). Interestingly, n = 12 patients reported that they were not given any dietary recommendations as part of their treatment. There were however no differences in biochemical levels/outcomes (i.e., potassium, phosphate and albumin) 64 between this group and the group of patients who reported having been given dietary recommendations. Factors Associated With Self-reported Nonadherence Univariate Analyses Self-reported nonadherence comparison between different subgroups. Kruskal– Wallis and Mann-Whitney tests were used to compare nonadherence levels in subgroups with different demographical and clinical profiles (gender, marital status, ethnicity, educational level, employment status, housing, time on dialysis, dependence on carer and primary cause of ESRD). Results indicated significant differences for ethnicity, employment status, education, time on dialysis and primary cause of ESRD (Table 12). The Chinese were less likely to forget taking medications (Mdn = 2.0) than nonChinese (Mdn = 3.0), U = 1689.0, p = .04. Employed patients (Mdn = 3.0) were significantly more likely to intentionally deviate from their dietary demands than unemployed patients (Mdn = 2.5), U = 1259.5, p = .05. Three nonadherence measures, namely intentional nonadherence to medication and diet and unintentional nonadherence to medication, varied with education level (H[2] = 11.24, p = .003; H[2] = 6.59, p = .04; H[2] = 6.84, p = .03, respectively). Bonferronicorrected Mann-Whitney tests (adjusted α = .05/3 = .017) followed up the results and revealed that patients with intermediate education showed more both intentional and unintentional nonadherence to medication than patients with low education (U = 994.5, p 65 = .002; U = 1089.0, p = .017, respectively), whereas patients with high education only showed more intentional nonadherence to medication than patients with low education (U = 669.0, p =. 007). Time on dialysis had significant impact on intentional nonadherence to dialysis, H(2) = 8.58, p = .01. When this was followed up by Bonferroni-corrected Mann-Whitney tests (adjusted α = .05/3 = .017), both patients on dialysis for more than two years and patients on dialysis between one to two years were found to be significantly more likely to intentionally violate their dialysis guidelines (U = 1143.0, p = .015; U = 205.5, p = .004 respectively) relative to patients who had been on PD for less than one year. Primary cause of ESRD was closely linked with intentional nonadherence to medication and diet (H[3] = 12.78, p = .005; H[3] = 15.63, p = .001, respectively). Bonferroni-corrected Mann-Whitney tests (adjusted α = .05/6 = .008) were used to make pairwise comparisons. Patients with other cause of ESRD were significantly more likely to be intentionally nonadherent to medication and dietary guidelines than patients with hypertension as primary cause of ESRD (U = 96.5, p = .001; U = 61.0, p = .001 respectively) and patients with diabetes as primary cause of ESRD (U = 281.5, p = .005; U = 190.5, p = .003 respectively). In addition, patients with glomerulonephritis as primary cause of ESRD showed significantly more intentional nonadherence to dietary guidelines than patients with hypertension as primary cause of ESRD (U = 261.0, p = .004). 66 Table 12: Comparisons of Self- reported Nonadherence Between Different Subgroups Comparisons of Self-reported Nonadherence Between Different Subgroups Dependent variables (n = 144) Intentional NA to dialysis Intentional NA to medication Unintentional NA to medication Intentional NA to dietc Factors Time on dialysis Less than 1 year 1 to 2 years More than 2 years Education Level Low Intermediate High Primary cause of ESRD Diabetes Hypertension Glomerulonephritis Other Ethnicity Chinese Non-Chinese Education Level Low Intermediate High Employment status Employed Unemployed Education Level Low Intermediate High Primary cause of ESRD Diabetes Hypertension Glomerulonephritis Other n NA score, M ± SD 32 20 92 1.1 ± 0.4 1.6 ± 0.9 1.5 ± 0.8 .01a 54 54 36 1.6 ± 1.0 2.1 ± 1.0 2.2 ± 1.1 .003a 62 28 38 16 1.9 ± 1.0 1.6 ± 0.8 2.2 ± 1.1 2.6 ± 1.0 .005a 102 42 2.0 ± 0.8 2.4 ± 1.1 .04b 54 54 36 1.9 ± 0.9 2.3 ± 0.9 2.2 ± 0.8 .03a 33 98 2.9 ± 0.9 2.6 ± 1.1 .05b 49 51 31 2.3 ± 1.1 2.8 ± 1.1 2.9 ± 1.1 .04a 58 26 34 13 2.5 ± 1.2 2.2 ± 1.0 2.9 ± 1.0 3.5 ± 0.8 .001a p Note. APD = Automated Peritoneal Dialysis. CAPD = Continuous Ambulatory Peritoneal Dialysis. NA = nonadherence. ESRD = end stage renal disease. a Kruskal–Wallis test. bMann-Whitney test. cn = 131 because 12 patients reported no dietary restrictions and one did not answer this question. 67 Correlates of self-reported nonadherence measures. Spearman rank correlation coefficients were calculated to investigate the influence of continuous variables (age, pill burden, CCI, biochemical marker values, BMQ, self-efficacy, HADS anxiety and depression , UCLA loneliness, KDQOL-SF and WHOQOL-BREF scores) on selfreported nonadherence measures. Significant correlates of self-reported nonadherence measures are presented in Table 13. Older age was associated with less nonadherence as indexed by inverse associations with unintentional nonadherence to medication and diet (rs = −.29, p < .001; rs = −.19, p = .03, respectively) and intentional dietary nonadherence (rs = −.24, p = .007). Comorbidity (assessed with CCI) was associated with less nonadherence in some domains (see Table 13). Patients’ cognitions were significantly associated with nonadherence in the expected direction. Strong necessity beliefs were associate with less intentional nonadherence to medication (rs = −.19, p = .03), whereas higher concern beliefs were associated with nonadherence on almost all measures (all ps < .05; see Table 13 for full results). Higher self-efficacy scores were associated with less nonadherence in most domains as indexed with inverse correlation with nonadherence scores ( see Table 13). Emotional distress (greater symptoms of depression or anxiety) increased nonadherence. Quality of life levels (as measured with KDQOL-SF symptoms, effects of kidney disease and patient satisfaction, WHOQOL-BREF environment) were also significantly associated with nonadherence measures (significant rs ranging from −.17 to −32, all ps < ,05, see Table 13 for full results); the overall pattern of results showed that 68 better quality of life was associated with less nonadherence to treatment recommendations. Table 13: Correlates of Self-reported Nonadherence in Spearman Rank Correlations Spearman Rank Correlations Between Self-reported Nonadherence and Selected Variables Intentional NA Unintentional NA Dialysis Medication Diet Dialysis Medication Diet ** Age −.16 −.14 −.24 −.13 −.29*** −.19* CCI −.08 −.13 −.20* −.09 −.19* −.16 Necessity −.06 −.19* −.06 −.04 −.15 −.08 Concerns −.20* −.30*** −.27** −.16 -.21* −.38*** Dialysis −.19* −.16 −.07 −.25** −.20* Medication −.12 −.35*** −.27** −.04 −.32*** −.40*** Diet −.13 −.37*** −.38*** −.14 −.26** −.52*** Anxiety −.12 −.21* −.23** −.12 -.16 −.32*** Depression −.12 −.20* −.16 −.10 -.11 −.21* −.28** −.18* −.12 −.19* −.14 −.14 −.17* −.19* −.18* −.17* −.08 −.21* −.32*** −.16 −.10 −.26** −.18* −.26** −.30*** −.16 −.18* −.28** −.16 −.25** BMQ Self-efficacy −.08 HADS KDQOL-SF Symptoms Effects of kidney t disease Patient satisfaction WHOQOL-BREF Environment Note. Spearman rank correlation. NA = nonadherence. CCI = Charlson Comorbidity Index. BMQ = Beliefs about Medicines Questionnaire. HADS = Hospital Anxiety and Depression Scale. KDQOL-SF = Kidney Disease Quality of Life–Short Form. WHOQOL-BREF = World Health Organization Quality of Life Instrument, Short Form. * p < .05. **p < .01. ***p < .001. 69 Multivariate Analyses Six hierarchical stepwise multiple linear regressions were performed to identify important demographical, clinical and psychosocial multivariate correlates of selfreported intentional and unintentional nonadherence to different aspects of therapeutic regimen (dialysis intentional and unintentional, medication intentional and unintentional and diet intentional and unintentional). Only variables associated with self-reported nonadherence measures in univariate analyses were included in the models. In all regression analyses, demographics (age, ethnicity, education level and employment status) were entered on the first step, followed by medical variables (modality, time on dialysis, CCI and primary cause of ESRD) on the second step and psychosocial variables (BMQ necessity and concerns, self-efficacy scores, HADS depression and anxiety scores, KDQOL-SF symptoms, effects of kidney disease, patient satisfaction and WHOQOLBREF environment) on the final step. Categorical variables were coded as dummy variables as appropriate. If one dummy variable was selected by the regression model, all related dummy variables were kept in the final model. Forward selection method was used for entry of variable into respective blocks. The problem of multicollinearity (strong linear relationship among two or more predictors in regression model) was assessed with the Variance Inflation Factor (VIF) value. VIF value more than 10 indicates multicollinearity (Myers, 1990). VIF values in the current study varied from 1.00 to 2.79, indicating no multicollinearity problem. The final regression models explained low to moderate proportions of variance in self-report nonadherence scores (cumulative R2 = .11 to .31). Psychosocial variables (ΔR2 = .07 to .31) independently accounted for more variance in self-reported nonadherence 70 scores relative to demographical (ΔR2 = .00 to .09) and clinical variables (ΔR2 = .00 to .08) (see Table 14 to Table 16) Intentional nonadherence to dialysis exchanges. Education, modality, KDQOLSF patients satisfaction and WHOQOL-BREF environment were significant predictors of intentional nonadherence to dialysis exchanges in the final regression model, accounting for R2 = .19, R2adj = .16 of the variance, F(5, 138) = 6.32, p < .001. High education (vs. low education) (β = .22, p = .01, ΔR2 = .03) and CAPD (β = .22, p = .006, ΔR2 = .03) were positively associated with intentional nonadherence to dialysis guidelines, whereas patient satisfaction (β = −.21, p = .02, ΔR2 = .09) and WHOQOL environment (β = −.21, p = .02, ΔR2 = .04) were inversely associated with intentional nonadherence to dialysis guidelines (Table 14). Unintentional nonadherence to dialysis. The final regression model (F[3, 140] = 5.92, p = .001) explained R2 = .11, R2adj = .09 of the variance in unintentional nonadherence to dialysis, with time on dialysis and WHOQOL environment quality of life being significant at the final step. More than 2 years on dialysis (vs. less than 1 year) (β = .22, p = .02, ΔR2 = .04) and lower environment quality of life score (β = −.27, p = .001, ΔR2 = .07) were associated with greater unintentional nonadherence ( i.e,.forgetfullness) to dialysis exchanges (Table 14). 71 Table 14: Multivariate Correlates of Self-reported Nonadherence to Dialysis Guidelines Multivariate Correlates of Self-reported Nonadherence to Dialysis Guidelines Dependent variable: intentional nonadherence to dialysis guidelines Step 1 Predictor R2 R2adj ΔR2 ΔF p .03 .02 .03 2.48 .09 −0.34 −.22** .06 .05 .03 4.74 .03 10.23 B β Intermediate education −0.04 −.03 (vs. low education) High education −0.39 −.22* (vs. low education) 2 3 CAPD (vs. APD) KDQOL-SF patient * −0.007 −.21 .15 .13 .09 −0.06 −.21* .19 .16 .04 satisfaction WHOQOL-BREF < .001 environment Dependent variable: unintentional nonadherence to dialysis guidelines Step 2 Predictor R2 R2adj ΔR2 ΔF p .04 .03 .04 2.89 .06 −0.05 −.27** .11 .09 .07 11.56 .001 B β 1 to 2 years on dialysis −0.05 −.04 (vs. less than 1year) More than 2 years on −0.21 −.22* dialysis (vs. less than 1 year) 3 WHOQOL-BREF environment Note. Hierarchical stepwise multiple linear regressions were used. APD = Automated Peritoneal Dialysis. CAPD = Continuous Ambulatory Peritoneal Dialysis. KDQOL-SF = Kidney Disease Quality of Life–Short Form. WHOQOL-BREF = World Health Organization Quality of Life Instrument, Short Form. * p < .05. **p < .01. 72 Intentional nonadherence to medication. In the final regression model, three variables, namely primary cause of ESRD, medication self-efficacy and BMQ concerns, emerged as significant multivariate correlates of intentional nonadherence to medication, explaining R2 = .24, R2adj =.21of the variance, F(5, 138) = 8.55, p < .001 (Table 15). Diabetes (β = −.38, p = .003, ΔR2 = .01) and hypertension (β = −.30, p = .008, ΔR2 = .06) as primary causes of ESRD (vs. other cause of ESRD), as well as medication selfefficacy (β = −.32, p < .001, ΔR2 = .13) were negatively associated with intentional nonadherence to medication. Concern beliefs (β = .19, p = .02, ΔR2 = .03) were positively correlated with intentional nonadherence to medication. Unintentional nonadherence to medication. Two demographic variables (age and ethnicity) and one psychosocial factor (medication self-efficacy) remained significant in predicting unintentional nonadherence to medication (i.e., medication forgetfulness) at the final step of the regression, F(3, 140) = 9.58, p < .001, R2 = .17. R2adj = .15 (Table 15). All of them were negatively correlated with unintentional nonadherence to medication (age β = −.20, p = .01, ΔR2 = .05; Chinese β = −.18, p = .02, ΔR2 = .04; self-efficacy β = −.28, p < .001, ΔR2 = .08). Intentional nonadherence to dietary guidelines. The final regression model only included age and dietary self-efficacy, F(2, 128) = 12.83, p < .001, R2 = .17, R2adj = .15 (Table 16). Both age (β = −.18, p = .03, ΔR2 = .04) and dietary self-efficacy (β = −.36, p = 1.78 mmol/l) were significantly more likely to be younger (Mann-Whitney U = 1400.0, p < .001), female (χ2[1, N = 144] = 7.49, p = .007), employed (χ2[1, N = 144] = 6.45, p = .02), and having less comorbidity (MannWhitney U = 1650.0, p = .005). Patients nonadherent to albumin (< 3.5 g/dl) were significantly more likely to be older (Mann-Whitney U = 705.0, p = .02), unemployed (Fisher’s exact test, p = .02), living in poorer housing conditions (Fisher’s exact test, p = .01), having more comorbidity (Mann-Whitney U = 680.5, p = .01), worse physical quality of life (SF-12 PCS t[142] = −2.13, p = .04) and worse mental quality of life (SF-12 MCS MannWhitney U = 740.0, p = .04). 81 Table 18: Factors Affecting Nonadherence Based on Biochemical Markers in Univariate Analyses Factors Affecting Nonadherence Based on Biochemical Markers in Univariate Analyses Adherent Nonadherent p Potassium Potassium ≥ 3.5 mmol/l < 3.5 mmol/l (n = 117) (n = 27) 25 (21) 11 (41) .05 Pill purden 10.1 ± 2.7 11.1 ± 1.8 .02a CCI 5.4 ± 2.0 6.4 ± 1.9 .01a Phosphate Phosphate ≤ 1.78 mmol/l > 1.78 mmol/l (n = 96) (n = 48) Age 60.8 ± 12.0 52.4 ± 13.0 Male 49 (51) 13 (27) .007 Employed 19 (20) 19 (40) .02 5.8 ± 1.8 5.0 ± 2.3 Albumin Albumin ≥ 3.5 g/dl < 3.5g/dl (n = 17) (n = 127) 48.5 ± 17.6 59.3 ± 11.7 .02a 9 (53) 29 (23) .02b 1−4 rooms flats 5 (29) 83 (65) .01b 5-rooms flats 7 (42) 28 (22) Condominum or above 5 (29) 16 (13) 4.4 ± 2.1 5.7 ± 1.9 .01a SF-12 PCS 40.1 ± 11.1 35.1 ± 8.8 .04 SF-12 MCS 48.4 ± 8.8 43.0 ± 11.2 .04a Dependence on carer, CCI Age Employed < .001a .005a Housing condition CCI Note. Data expressed as M ± SD or n (%). ESRD = End Stage Renal Disease. CCI = Charlson Comorbidity Index. SF-12 = 12-item Short-Form Health Survey. PCS = physical component summary score. MCS = mental component summary score. a Mann-Whitney test. bFisher’s exact test. 82 Multivariate Analyses Clinical nonadherence indices (based on potassium, phosphate and albumin clinical targets) were subsequently regressed to demographic, clinical and psychosocial variables in three separate hierarchical stepwise logistic regression models. Only variables associated with nonadherence based on biochemical markers in univariate analyses (reported earlier) were entered as potential predictors. In all regression analyses, demographics (age, gender, employment status and housing condition) were entered on the first step, followed by medical variables (dependence on carer, pill burden and CCI) on the second step and psychosocial variables (SF-12 PCS and SF-12 MCS) on the final step. Nagelkerke R2, similar to R2 in linear regression, was calculated to assess how much variance in nonadherence (based on biochemical markers) could be explained by the regression models. Forward: likelihood ratio method was used for variable selection. Table 19 presents the final regression models. Potassium < 3.5 mmol/l. The only variable that was significant in predicting potassium < 3.5 mmol/l at the final step of the logistic regression was CCI, accounting for Nagelkerke R2 = .07 of the variance (Omnibus χ2[1, N = 144] = 6.05, p = .01). One score increment on CCI was associated with 30% more likelihood of potassium < 3.5 mmol/l (OR = 1.30, 95% CI = [1.05, 1.62], p = .02). This model correctly classified potassium status in 81% of the cases. Phosphate > 1.78 mmol/l. Only age and gender were significant in predicting phosphate > 1.78 mmol/l at the final step of the logistic regression, accounting for Nagelkerke R2 = .18 of the variance, Omnibus χ2(2, N = 144) = 19.63, p < .001. One year increment of age was associated with 5% less likelihood of phosphate > 1.78 mmol/l (OR 83 = 0.95, 95% CI = [0.92, 0.98], p = .001). Female was 1.5 times more likely than men to have phosphate > 1.78 mmol/l (OR = 2.53, 95% CI = [1.15, 5.55], p = .02). The final model correctly classified phosphate status in 70% of the cases. Albumin < 3.5 g/dl. Albumin level < 3.5 g/dl was explained by age and housing condition, Omnibus χ2(3, N = 144) = 20.87, p < .001, Nagelkerke R2 = .26. One increment on age increased the likelihood of albumin < 3.5 g/dl by 8% (OR = 1.08, 95% CI = [1.03, 1.12], p = .001). Patients living in 5-rooms flats and Condominiums or above were 85% less likely to have albumin level < 3.5 g/dl when compared to patients living in 1-4 rooms flats (OR = 0.15, 95% CI = [0.04, 0.57], p = .006; OR = 0.15, 95% CI = [0.03, 0.65], p = .01). The final model correctly classified albumin status in 90% of the cases. 84 Table 19: Predictors of Nonadherence Based on Biochemical Markers in Multivariate Analyses Factors Affecting Nonadherence Based on Biochemical Markers in Multivariate Analyses Dependent variable: potassium < 3.5 mmol/l Step 2 Predictor B −.26* CCI OR Nagelkerke R2 χ2 p 1.30 .07 6.05 .01 Dependent variable: phosphate > 1.78 mmol/l Step 1 B OR Nagelkerke R2 χ2 p Age −0.05** 0.95 .13 19.63 < .001 Female (vs. male) −0.93* 2.53 .18 Predictor Dependent variable: albumin < 3.5 g/dl Step 1 B OR Nagelkerke R2 χ2 p Age −0.07** 1.08 .13 20.87 1.78 mmol/l, comparable to previous findings (Hecking et al., 2004; Leggat et al., 1998; O'Connor, Jardine, & Millar, 2008). Both potassium and phosphate levels are traditional biochemical markers used to indicate nonadherence to medication and diet (Denhaerynck et al., 2007). Nonadherence rates estimated by biochemical values were lower than self-reports which were 47% for medication and 75% for diet. This was 96 in line with the notion by Kugler et al. (2011) that self-report, when used in an unthreatening way, was more capable of detecting minor nonadherence than biochemical markers. The rate of nonadherence to albumin was 88%, remarkably higher than previous studies (Russell et al., 2008; Sehgal, Leon, & Soinski, 1998). This is a cause for concern, given that serum albumin level is a robust associate of high mortality and morbidity in patients with and without renal diseases (Goldwasser & Feldman, 1997; Owen, Lew, Liu, Lowrie, & Lazarus, 1993). Nonadherence based on biochemical values were not associated with all selfreport nonadherence scores, consistent with previous studies (Chiu et al., 2009; Kaveh & Kimmel, 2001; Lee & Molassiotis, 2002). One reason for the lack of concordance between the two methods is likely to be due to differences in time measurement. It is also important to note that biochemical levels are likely to be influenced by a variety of factors such as comorbidity, dialysis adequacy, residual renal function, time at which the tests were performed and acid-based and hormonal status (Denhaerynck et al., 2007). In addition, biochemical values may be affected by “white coat adherence” noted in different populations which means patients tend to have good adherence several days preceding a clinic visit and much less satisfactory adherence in other days (Podsadecki, Vrijens, Tousset, Rode, & Hanna, 2008). Comparison of Nonadherence to Dialysis, Medication and Diet Nonadherence to dialysis exchanges was high. This is most probably supported by patients’ high awareness about the importance of dialysis, family enforcement, and the comfort of doing dialysis at home. PD patients do not need to be concerned about transportation, a common cause of nonadherence in HD patients (Latham, 1998). 97 Moreover, some patients are more motivated to adhere to dialysis because they hold the misconception that dialysis can compensate for their violations in other domains of the therapeutic regimen (Smith et al., 2010). Nonadherence to medication fell between nonadherence to dialysis and diet. As mentioned above, patients in our study had elevated level of concerns about the disruptive effects (e.g., side effects, drug dependence) of their prescribed medication when compared to other populations (Horne & Weinman, 1999; Rees et al., 2010).This might be an important contributor of medication nonadherence in this group. In addition, patients may be discouraged to take medication when they hold the belief that dialysis is enough to keep their health, as medicines produce no immediate symptomatic relief and they frequently receive information regarding the effects of medicines from alternative therapists (e.g., practitioner of Chinese medicine), other patients and friends that conflicts with physicians’ suggestions (Donovan & Blake, 1992; Karamanidou, Clatworthy, et al., 2008). Dietary nonadherence was the greatest of the three, consistent with previous findings (Lam et al., 2010; Leggat et al., 1998). This supports the notion that nonadherence is very likely to occur when it requires major changes in lifestyle (Arenas et al., 2010). Following dietary guidelines is especially challenging for patients in countries where there is a huge discrepancy between traditional eating preferences and dietary needs for dialysis patients (Lam et al., 2010; Park, Choi-Kwon, Sim, & Kim, 2008). Eating out is common in Singapore, as testified by the abundance of food courts and restaurants in the country. However, commercially prepared food usually contain high fat, cholesterol and salt (Ang & Foo, 2002) and is considered unhealthy for dialysis 98 patients. Thus, patients need to make food choices cautiously when they eat out because the alternative would be to eat unpalatable food at home. This restricted diet can be very distressing for patients. Nonadherence is especially likely to occur among patients with low frustration tolerance (De-Nour & Czaczkes, 1972). Even for patients who succeed in adjusting to renal dietary plans, dietary abuse can appear in social occasions because the collectivist culture in Singapore expects people to cater to the needs of others and suppress their own needs, such as in the case of “Guanxi” (McCarthy et al., 2009; Triandis, 2001). For patients with unsuccessful dietary adherence, guiding them to cope with frustration and to be more assertive may be helpful. Intentional and Unintentional Nonadherence Intentional nonadherence to each domain of the treatment regimen was considerable, 28%, 58% and 81% for dialysis, medication and diet respectively. Unintentional nonadherence rates were also high, 19%, 71% and 73% for dialysis, medication and diet, respectively. Hypothesis 1b that unintentional nonadherence is more common than intentional nonadherence was not supported by our data based on mean scores, in contrast with findings in other chronic illness conditions (Rees et al., 2010; Sewitch et al., 2003; Unni & Farris, 2011). Part of the discrepancy may be related to mode of administration. The questionnaire developed by Morisky, Green and Levine (1986) was used to assess intentional and unintentional nonadherence in other studies. It asks patients about nonadherence directly (e.g., “Do you ever forget to take your medicines?”). Patients may feel more comfortable to cite forgetfulness to justify their nonadherence rather than admitting deliberate deviations since the former is less blamable (Unni & Farris, 2011). 99 We have opted for patient self-report administration to an independent researcher not involved with patients’ care in a confidential setting outside the clinic to minimize social desirability bias. All items were carefully phrased in a nonjudgmental manner to validate experience and facilitate disclosure (e.g., “Some people forget to take their medicines. Overall, how often does this happen to you?”). Such an approach has been recommended as effective in reducing response bias in self-report (Vlaminck et al., 2001). As there is no previous work on intention/unintentional nonadherence in ESRD patients, further work is needed to examine if the discrepancy may be somehow related to this population per se rather than the methodology. It is noteworthy that although intentional nonadherence seems puzzling to health providers, it may be more understandable when viewed from patients’ perspectives. Intentional nonadherence can be positively reinforced in daily life. For instance, Eliasson et al. (2011) interviewed a group of chronic myeloid leukemia patients and reported that patients’ intentional nonadherence to medication could be encouraged by the reduction of symptoms following altering drugs’ doses and by perceived doctors’ unintentional reassurance that occasional deviations would not lead to detrimental consequences. Moreover, nonadherence to HD regimen and remaining sick had been found helpful for patients to gain attention from others and to resolve family problems (De-Nour & Czaczkes, 1972). On the other hand, adherence can be negatively reinforced. Patients may experience no adverse effects when reducing or stopping therapy for some time, but may feel tiredness, itching and headaches caused by the treatment (Polaschek, 2007). Renal nurses commented in the study of McCarthy et al. (2009) that PD patients were at greater risk of being marginalized by their social network if they followed their treatment 100 guidelines as instructed. For instance, if patients kept their rooms very clean, it might discourage others from visiting due to concerns about causing infections to patients. In addition, PD patients have more chances to intentionally modify some aspects of their therapy such as starting time and the length of dialysis, since they are not dependent on the fixed schedule in dialysis centers (Polaschek, 2007). Patients’ attributes of foresight and foreplanning have been proven to be an important predictor of unintentional nonadherence (Daleboudt et al., 2010). However, this may not play a major role in this study, as even though cognitive decline is common among older dialysis patients (Tyrrell, Paturel, Cadec, Capezzali, & Poussin, 2005), they did not report more forgetfulness than younger patients. In addition, unintentional nonadherence could be caused by patients’ poor comprehension about their treatment and language barriers (Browne & Merighi, 2010). Education from pharmacists has been found to be effective in reducing patients’ difficulty about remembering medicines (Sathvika, Naraharib, Gurudevb, & Parthasarathia, 2009). Comparison of Nonadherence Between APD and CAPD Patients APD patients reported lower level of nonadherence than CAPD patients. More specifically, APD was associated with less intentional nonadherence to dialysis exchanges than CAPD in multivariate analysis and it was less common for APD patients to shorten their treatment in univariate analysis. Bernardini et al. (2000) also evaluated nonadherence to dialysis prescriptions in these two groups and noted elevated nonadherence in CAPD patients than APD patients. This result is not surprising, given the distinct CAPD and APD demands. CAPD involves performing several exchanges manually during the day time, whereas APD is performed at night by a mechanical 101 device when patients are sleeping. Thus CAPD patients may face more barriers in terms of time (e.g., going out for no more than 4 hours), space (e.g., employers searching for space to perform exchanges) and physiological status (e.g., continuous inter-abdominal pressure) than APD patients (Liakopoulos & Dombros, 2009). Determinants of Nonadherence Our results show that nonadherence is a multi-faceted phenomenon, concurrently determined by the interplay of five categories identified by WHO (Sabaté, 2003) as critical for nonadherence. They are: social-economic factors, health care system related factors, condition-related factors, therapy-related factors and patient-related factors. The relationship between these five categories and predictors spotted in our study is displayed in Table 20. Demographics and Nonadherence Our study identified a set of demographical variables (age, gender, ethnicity, education and housing condition) associated with nonadherence based on multivariate analyses. They generally made small contributions to the prediction of nonadherence outcomes, except when nonadherence was based on phosphate and albumin levels. Several systematic reviews report that demographic variables (excluding age and smoking) are inconsistent correlates of nonadherence (Karamanidou, Clatworthy, et al., 2008; Russell et al., 2007). 102 Table 20:WHO Identified Categories Affecting Nonadherence and Significant Predictors in Our Study WHO Identified Categories Affecting Nonadherence and Significant Predictors of Nonadherence in Our Study Five categories affecting nonadherence Social-economic factors Significant predictors of nonadherence in our study  Demographics (age, gender, ethnicity, education level and housing condition)  WHOQOL-BREF environment Health care system-related factors  KDQOL-SF patient satisfaction Condition-related factors  Comorbidity  Primary cause of ESRD  PD modality  Time on dialysis  BMQ concerns  Self-efficacy  HADS depression  HADS anxiety Therapy-related factors Patient-related factors Age. Patients who were older had less unintentional nonadherence to medication, less intentional nonadherence to dietary demands and less likelihood of phosphate > 1.78 mmol/l than younger patients in the current sample. Older age is a consistent predictor of less nonadherence to all aspects of the medical regimen in other studies (Ismail, Hakim, Oreopoulos, & Patrikarea, 1993; Kugler et al., 2005; Lam et al., 2010). The association between younger age and nonadherence may reflect the practical and logistic challenges in sustaining care and following treatment as the younger patients are more likely to be occupationally and socially active (van der Mei et al., 2007). Younger patients also find it difficult coming into terms with their sickness and tend to have poor tolerance about 103 limitations associated with treatment (Dimkovic & Oreopoulos, 2000; GonsalvesEbrahim, Sterin, Gulledge, Gipson, & Rodgers, 1987; Herlin & Wann-Hansson, 2010). Nonadherence can help younger patients to deny their dependence on dialysis and the severity of their condition (De-Nour & Czaczkes, 1972). This can be further exacerbated by younger patients’ perception of a lack of support from family members, doctors, nurses, and technicians (Oka & Chaboyer, 1999). Older age was also found to be associated with reduced albumin level in our study, although due in part to albumin level tending to decrease steadily with age (Greenblatt, 1979; Salive et al., 1992). It is also plausible that older patients do not take in adequate amount of protein due to reduced appetite (Park et al., 2008). Gender. Females were 1.5 times more likely to have phosphate > 1.78 mmol/l than males, indicating possible dietary or medication violations. Another study also reported a higher level of fluid nonadherence among females than males (Chilcot, Wellsted, & Farrington, 2010). In contrast, other studies suggested that males are more nonadherent to all aspects of renal regimen than females (Bame, Petersen, & Wray, 1993; Lam et al., 2010; O'Connor et al., 2008; Setoguchi, Choudhry, Levin, Shrank, & Winkelmayer, 2010). Asefzadeh, Asefzadeh and Javadi (2005) revealed that female patients were less likely to be adherent than male patients because of forgetfulness, doubts about the importance and necessity of medicines and religious considerations. In the current study, females did not report more nonadherence in survey but showed elevated phosphate levels. Chen et al. (2006) suggested that some patients adhere to their medical regimen wrongly, such as when patients think that they are duly following doctors’ instruction but in fact their methods are not correct. This is probably the case 104 with some female patients in our study. Interestingly, females are also more responsive to educational interventions about nonadherence (Barnett, Li Yoong, Pinikahana, & Si-Yen, 2008). Another reason to explain the gender difference on self-report and phosphate level might be that females feel more reluctant about reporting nonadherence than males, since females tend to have more concerns about others’ evaluation of them than males (Hoffman, 1977). Ethnicity. The Chinese were less likely to forget taking their prescribed medication in our study than non-Chinese. The cohesion of Chinese families might play an important role in this. Closer ties amongst family members can facilitate patients’ medication adherence through providing financial resources, increasing motivation, educating patients about the dosing and effects of medicines, setting reminders or directly administrating medications (Fredriksen-Goldsen et al., 2011). Education level. Patients with high education (post-secondary school) reported more intentional deviation from dialysis guidelines than patients with low educational level (no or primary school) in our study, whereas the difference between patients with intermediate education (secondary school) and patients with low education was not significant. Our finding is in contrast with other studies such as Lam et al. (2010) which reported a negative association between education level and nonadherence to dialysis, fluid and dietary restrictions. Comparing patients on different educational levels revealed that 39% patients with high education were still working while only 15% patients with low education were still working. Patients with high educational level were also younger than patients with low education level (data not shown). Thus patients with high education level might be occupationally and socially more active and their daily activities 105 might have more conflicts with the rigorous demands of their medical regimen. Patients with busy or unpredictable schedules are found to be less likely to be adherent in other studies (Browne & Merighi, 2010). Housing condition. Patients living in more expensive property (5-rooms flats/condominiums or above) had a lower risk of malnutrition (indicated by albumin level) than patients living in less expensive property (1-4 rooms flats). Housing condition was used as a rough indicator of patients’ financial status in our study, since there were many missing values for monthly family income. Poor financial status is a barrier to adequate protein nutrition amongst dialysis patients, since renal diet can be expensive to maintain (Sehgal et al., 1998; Soinski, Kelly, & Komaransky, 1993). Clinical Variables and Nonadherence The association between PD modality and nonadherence has been discussed above. Other factors affecting nonadherence involved time on dialysis, comorbidity and primary cause of ESRD. Clinical factors were also found to be poor predictors of nonadherence in literature (Karamanidou, Clatworthy, et al., 2008). Time on dialysis. Patients on PD for more than two years were more likely to intentionally deviate from their dialysis guidelines than patients treated with PD for less than one year. Difficulty in sustained motivation in long-term dialysis patients is commonly documented (Lam et al., 2010). It has also been suggested that patients follow doctors’ advice initially when they are unfamiliar with their disease and treatment. As they become more knowledgeable about their treatment gradually through their observations and body reactions, they intentionally adjust certain aspects of their treatment to feel better or to regress back to their previous lifestyle (Polaschek, 2007). 106 Comorbidity. Comorbidity was the only factor associated with potassium level below clinical target. A positive relationship between comorbidity and nonadherence has also been observed in other studies (Pang, Ip, & Chang, 2001; Setoguchi et al., 2010). Patients with higher number of comorbid diseases may have more physical or emotional difficulties about being adherent to potassium. Dependence on carers for shopping and cooking also predispose patients to nonadherence in literature (Lee & Molassiotis, 2002; Sehgal et al., 1998), but our study could not confirm these results. Primary cause of ESRD. Patients with diabetes and hypertension as primary cause of ESRD showed less intentional nonadherence to medication than patients with other causes of ESRD (e.g, systemic lupus erythematous). Diabetic status was also found associated with less nonadherence in another study (Christensen, Benotsch, Wiebe, & Lawton, 1995). It has been observed in different populations that patients stop taking medicines as soon as there is a reduction of symptoms (Addington, 1979; Britten, 1994). Given that both diabetes and hypertension are asymptomatic, it is very likely that patients with these conditions are more aware about the importance of being adherent to medication regimen regardless of symptoms due to previous experience with diabetic and hypertension regimen. It is noteworthy that PD patients in our study had a very high pill burden, taking a median of 10 types of medication per day. Contrary to two previous studies (Chiu et al., 2009; Neri et al., 2011), our study detected no association between pill burden and medication nonadherence. This discrepancy may be due to the fact that the other two studies used the exact number of tablets (dosages per day), whereas we only evaluated the number of prescribed medicines and this parameter might be less sensitive. 107 Psychosocial Variables and Nonadherence Health beliefs. Hypothesis 2a that positive health beliefs (i.e., high necessity beliefs, low concern beliefs and high self-efficacy) are associated with less nonadherence was partially confirmed. Self-efficacy and concern beliefs had a major impact on nonadherence to diet and medication, but not on nonadherence to dialysis exchanges. The lack of association between dialysis nonadherence and health beliefs may be due to the fact that dialysis presents a more immediate gratification of their illness, thus involving less rational decision making. In our study higher level of concern beliefs were associated with more intentional nonadherence to medication. This is in line with studies in other patient populations (Daleboudt et al., 2010; Lowry, Dudley, Oddone, & Bosworth, 2005). It is understandable that patients’ intentional nonadherence to medication was affected by concern beliefs in our sample. Dialysis patients may become more sensitive to the costs associated with taking medication (e.g., side effects, drug dependence) after they have already experienced multiple losses such as functional decline, unemployment and reduced social life (Chan, Brooks, Erlich, Chow, & Suranyi, 2009). Thus the need to prevent further losses takes over and leads to intentional deviations from medication guidelines. The observed inverse association between concern beliefs and unintentional nonadherence to diet is noteworthy as unintentional nonadherence is not typically thought of as an active process (Clifford et al., 2008). Other studies (Daleboudt et al., 2010; Unni & Farris, 2011) also reported that concern beliefs played an important role in unintentional nonadherence to medication. It seems that concerns may somehow lead to 108 patients attaching low priority to treatment which may lead to more forgetfulness. More work using qualitative method is needed to explore these relationships. Some concerns may be fueled by a lack of understanding of how the drugs work, it is also important to remember that these concerns may be well validated by patients’ subjective experience and not solely founded by a lack of understanding or poor knowledge (Schuz et al., 2011). The finding of the role on concerns beliefs driving nonadherence highlights the need to correct patients’ misperceptions and worries about side effects. Corrective actions need to be taken either in the form of educating patients about drug effects or when the situation calls, changing prescribed drug (there may for instance be another hypertensive agent that the health care professions could prescribe if one is not that well tolerated) so as to reduce risk of intentional deviation from prescribed medication regimen. Self-efficacy influenced most nonadherence measures (intentional/unintentional nonadherence to medication/diet) in our study. This is in line with the finding of Smith et al. (2010) that sense of self-responsibility and perceived capability of self-restriction are critical facilitators of adherence. There is overwhelming theoretical and empirical evidence on the role of self-efficacy in guiding adherence and self-care management in populations with different chronic illnesses (Lindberg & Fernandes, 2010; Mo & Mak, 2009; Taal, Rasker, Seydel, & Wiegman, 1993; Zrinyi et al., 2003). Self-efficacy training in HD patients has been found to be effective in reducing weight gains (Tsay, 2003). Incorporating self-efficacy into a self-management programme for PD patient achieved successful effects in terms of enhancing volume status, quality of life, rehabilitation status, and reducing nonadherence to dialysis and dietary guidelines (Su et al., 2009). 109 Dialysis patients face a range of adherence barriers such as debilitating symptoms, a lack of motivation and conflicts between therapy and other life activities (Polaschek, 2007). The importance of self-efficacy is that it “determines whether coping behavior will be initiated, how much effort will be expended, and how long it will be sustained in the face of obstacles and aversive experiences” (Bandura, 1977, p. 191). Thus patients with stronger self-efficacy are very likely to view their treatment demands as challenging tasks to be mastered, actively participate in solving their problems and remain resilient when their efforts do not achieve desirable effects immediately. On the other hand, patients with low self-efficacy are very likely to judge treatment tasks as too difficult for them and thus give up quickly when they encounter barriers in managing their treatment or when they do not see effects of adherence in short term (Bandura, 1989). Bandura (1977) proposed four sources of self-efficacy: mastery experience, vicarious experience, verbal persuasion and physiological states. Strategies corresponding to these postulates can be effective in enhancing self-efficacy in PD patients, such as: (a) providing adequate training to ensure that patients master necessary knowledge and skills and helping patients achieve successes at the initiation of dialysis; (b) using social model and asking successful patients to share their adherence experiences and tips so other patients can learn from them; (c) reassuring patients that they are capable of accomplishing the tasks; (d) helping patients to reduce stress and view emotional and physiological arousals as facilitators instead of barriers to adherence (Bandura, 2006; Locke & Latham, 2002). Given the importance of cognitive factors (i.e., beliefs about medicines and selfefficacy), it is imperative to take them into account in interventions aimed at reducing 110 nonadherence. Little success is found in previous educational programmes which increased patients’ understanding about treatment but failed to change patients’ health beliefs and self-efficacy (Cummings, Becker, Kirscht, & Levin, 1981; Karamanidou, Weinman, & Horne, 2008; Tanner et al., 1998). Emotional distress and nonadherence. Hypothesis 2b that more emotional distress is associated with more nonadherence was confirmed in univariate analyses, but negated in multivariate analyses after controlling for covariates. The relationship between depression and nonadherence is well supported in different studies (Brownbridge & Fielding, 1994; Cukor et al., 2009; De-Nour & Czaczkes, 1976). A rigorous review reported a considerable relationship between depression and nonadherence across different chronic illnesses, with depressed patients three times as likely as non-depressed patients to be nonadherent (OR = 3.03, 95% CI = [1.96, 4.89]) (DiMatteo et al., 2000). However, the association between anxiety and nonadherence is ambiguous in current literature (DiMatteo et al., 2000). The association between anxiety and nonadherence in dialysis patients was confirmed in one study (Brownbridge & Fielding, 1994) but negated in three other studies (Christensen, Moran, Lawton, Stallman, & Voigts, 1997; Katz et al., 1998; Schneider et al., 1991). A significant finding of our study is that emotional distress only affects nonadherence indirectly through cognitions. More specifically, depression and anxiety affect nonadherence through increasing patients’ concerns about the adverse effects of medicines and lowering patients’ self-efficacy about achieving treatment goals. This result has important clinical implications, since it suggests that modifying beliefs may be a feasible way to reduce nonadherence in depressed patients. 111 Quality of life and nonadherence. Two quality of life parameters turned out to be important in predicting nonadherence outcomes: KDQOL-SF patient satisfaction and WHOQOL-BREF environment. Patient satisfaction was the most important predictor of intentional nonadherence to dialysis in our study. Patients’ satisfaction with their nephrologists has also been found associated with better attendance at HD sessions (Kovac, Patel, Peterson, & Kimmel, 2002). Low satisfaction with care can fuel patients’ nonadherence. A tense doctor-patient relationship lowers the perceived authority of health providers. Patients view those bythe-book treatments given by the doctors as not to their best interest and have a tendency to use their subjective experience to modify their treatments (Allen, Wainwright, & Hutchinson, 2011). For instance, patients experiment with their medication to find out minimum dosage workable for them and to control side effects instead of following doctor’s advice blindly (Donovan, Blake, & Fleming, 1989; Vermeire et al., 2001). In addition, nonadherence could be a way of “acting out”, used by passive-aggressive patients to express their underlying hostility towards the medical team who did not treat them well (De-Nour & Czaczkes, 1972). Furthermore, patients are apt to conceal their nonadherence behaviors from health providers, and when they do so, it blocks out necessary discussion regarding the appropriateness of such behaviors (Allen et al., 2011; Donovan & Blake, 1992). The importance of patients’ faith in their doctors is testified by reports of healthy physician-patient relationship related to adhering to treatment plans even when facing severe physical and social stressors (Eliasson et al., 2011). Good environment quality of life decreased nonadherence to dialysis guidelines. Patients with better environmental quality of life are more likely to have adequate 112 resources (e.g., financial, informational) to secure their needs, gain easy accessibility to health care when problems arise, have better chances to participate in recreational activities to alleviate stress and a more favorable home environment (e.g., safety, ease of transportation, low noise). All these factors can help patients establish a positive attitude towards dialysis which has been found to be important in motivating patients to perform exchanges. This is well illustrated by the comment of a patient in the study of Polascheck (2007): “If life is going well, then dialysis is not a problem.” Comparison of Determinants of Intentional and Unintentional Nonadherence There is theoretical and empirical evidence to show that intentional nonadherence to medication is largely driven by patients’ motivation and beliefs (Branin, 2001; George, Kong, & Stewart, 2007). These are factors largely amenable to change. In contrast, unintentional nonadherence to medication is thought to be the result of a passive process that is less strongly associated with individuals' beliefs and more closely related to patients’ skills or their abilities to follow their medication regimen (e.g. age, manual dexterity, cognitive impairment, health literacy) or treatment factors (e.g. dose, complexity) (Clifford et al., 2008). However, predictors of intentional and unintentional nonadherence were not clearly separated in our study. Intentional nonadherence was determined by the interplay of age, education, PD modality, primary cause of ESRD, patient satisfaction, environment quality of life, concern beliefs and self-efficacy, whereas unintentional nonadherence was affected by age, ethnicity, time on dialysis, environment quality of life, concern beliefs and self-efficacy. Thus, age, environment quality of life, concern beliefs and self-efficacy were all predictors of both intentional and unintentional nonadherence. 113 One possible reason could be that there was moderate to high overlap between patients reporting intentional nonadherence and unintentional nonadherence (Spearman correlation, rs = .35 for dialysis, rs = .56 for medication, rs = .67 for diet, all ps < .001). This is not unexpected since factors such as a busy life schedule can simultaneously contribute to intentional and unintentional nonadherence. Nonadherence is not a dichotomy of intentionality. For instance, patients forgetting to take medication at lunch may intentionally increase dosage at dinner (Eliasson et al., 2011). Clinical Recommendations It has been advocated that views concerning nonadherence should be expanded from merely focusing on patients as cause of nonadherence to considering the interactions between patient, health provider and the wider health care system (Kammerer, Garry, Hartigan, Carter, & Erlich, 2007). We agree that health providers are essential in preventing patients’ nonadherence. In order to do this, there are several suggestions. Firstly, the extensiveness of nonadherence should be noted, especially regarding dietary restrictions. Few patients discuss their nonadherence behaviors with physicians without prompting (Allen et al., 2011; Donovan & Blake, 1992). This may give physicians the wrong impression that few patients are nonadherent (McHorney, 2009). Patients also feel frustrated if doctors do not check upon them regarding their therapy (Polaschek, 2007). Thus it is crucial for physicians to become aware of this issue and provide consultations for patients claiming difficulties about following doctors’ instructions. As shown in our results, patients are willing to acknowledge nonadherence when asked in a nonthreatening, non-authoritative way. Secondly, intentional and unintentional nonadherence 114 are different phenomena. The intentionality of nonadherence should be considered in designing interventional programmes to reduce nonadherence. It should be pointed out that it is common to suggest behavioral strategies such as timer-alarms and medication boxes to patients reporting forgetfulness about therapy (Unni & Farris, 2011). Our results showed that these strategies alone are not likely to be effective without considering patients’ beliefs (self-efficacy and beliefs about medicines). Thirdly, it is notable that nonadherence is determined by a complex interplay of social-economic factors (age, gender, ethnicity, education level, housing condition and environment quality of life), health care team related factors (patient satisfaction), condition-related factors (comorbidity and primary cause of ESRD), therapy-related factors (PD modality and time on dialysis) and patient-related factors (medicine beliefs, self-efficacy, depression and anxiety). Social-economic, condition and therapy related factors are less modifiable than other factors. Patients with younger age, female gender, non-Chinese ethnicity, high education, poor housing condition, poor environment quality, high comorbidity, primary cause of ESRD not being diabetes or hypertension, and longer time on dialysis should be monitored for risk of nonadherence. For patients consistently showing nonadherent behaviors while on CAPD, switching to APD may be a solution (Raj, 2002). Patients’ cognitions, emotions and satisfaction with care are also critical determinants of nonadherence and they are more modifiable in interventions. Health providers can play an important role in reducing nonadherence through ways such as opening up communication regarding nonadherence, addressing patients’ concerns about medication and providing assurance that patients have adequate self-care abilities. A specific strategy suggested in literature that can be quite useful in this population is regular phone calls 115 from health providers (physicians, nurses or pharmacist) (Eliasson et al., 2011). Phone calls serve as reminders for patients to follow treatment guidelines and at the same time enhance patients’ relationship with medical staff and reduce patients’ uncertainty regarding their disease and treatment. Implementing this strategy has been proven to be effective in improving quality of life, resolving patients’ concerns and reducing nonadherence among PD patients and patients with other chronic conditions (Chow & Wong, 2010; Clifford, Barber, Elliott, Hartley, & Horne, 2006; Eliasson et al., 2011). Another more time-consuming strategy is negotiated care in which patients take an active role making decisions regarding their prescriptions. This strategy produced marked outcomes in a group of diabetic PD patients, reducing patient nonadherence to dietary salt and fluid restriction from 80.5% to 23.8% (Quan et al., 2006). Study Strengths and Limitations This study has several important strengths. Firstly, our study sample was highly representative of Singapore PD patients. It comprised almost one quarter of all local PD patients and had similar age, gender, ethnicity and primary cause of ESRD as the overall PD population in Singapore. Selection bias was minimized in our study. Participants were not limited to those who could read and completed the questionnaires by themselves. Patients could participate in our study through different ways based on their preferences for time, location and language. This prevented loss of patients due to poor health condition, illiteracy, visual problems and language barriers. Secondly, this is the first systematic study to investigate prevalence and predictors of intentional and unintentional nonadherence in PD patients. A series of determinants of intentional and unintentional 116 nonadherence were identified, thus offering multiple choices for targeting nonadherence. In addition, our study provided critical information regarding health beliefs, emotional distress, quality of life and nonadherence for patients on APD and CAPD. This can facilitate better management of PD programme and help patients’ selection of PD modality. Limitations of our study should be noted. Regression models in our study only explained low to moderate variance in nonadherence outcomes. More than half of the variance was not tapped in our study. Our study failed to assess some important associates of nonadherence such as cognitive decline (Daleboudt et al., 2010) and smoking status (Kugler et al., 2011; Russell et al., 2007) as well as factors beyond patients’ control such as cost of medicines (Hirth, Greer, Albert, Young, & Piette, 2008). In addition, we did not assess patients who did not show up for their regular clinics. Since patients who miss their clinic appointments are also more likely to be nonadherent with other aspects of the therapeutic regimen (Daleboudt et al., 2010), nonadherence rates in our study might be underestimated. Moreover, we used adapted questionnaires in assessing dialysis, medication and dietary self-efficacy and nonadherence. Since the psychometric properties of these two questionnaires have not been tested in other studies, findings regarding nonadherence and self-efficacy should be interpreted with caution. It should also be noted that about half of the participants were interviewed by research assistants. Interviewed patients may tend to give favorable responses due to social desirability issue. However, all analyses did not control for patients’ participation method because there were 12 missing items for this variable. These missing items would lead to loss of power and increase type II error in data analyses. Our current sample size 117 (n = 144) is powerful enough to detect medium effect sizes but underpowered to detect small effect sizes based on G*power 3.1 (Faul et al., 2009) calculations. Post-hoc power (1) = .83, .81, .79 for independent t, Mann-Whitney tests and regression analyses respectively when fixing  = .05 and effect sizes to be medium. A sample size of 394 for each group is required to detect a small effect size of d = .2,  = .05, power (1) = .8 for two-tailed independent t test. A sample size of 412 for each group is required to detect a small effect size of d = .2,  = .05, power (1) = .8 for two-tailed Mann-Whitney test. For linear regression with 17 predictors, a sample size of 1000 is required to detect a small effect size of 2 = .02,  = .05, power (1) = .8. Thus, it is necessary to continue the research with more participants in order to detect small effect sizes. Future Studies Our study makes the first attempt to distinguish intentional and unintentional nonadherence in ESRD patients and to identify associated factors. More studies are required to test whether findings in our study can be repeated in PD patients from other regions or be generalized to other ESRD populations (e.g., HD, transplant patients). Moreover, since this is an observational study, we can only evaluate factors affecting absolute nonadherence. It has been recognized that factors maintaining nonadherence are different from factors changing nonadherence (Friend, Hatchett, Schneider, & Wadhwa, 1997). Longitudinal studies are needed to describe the dynamics of nonadherence as the renal disease progresses and identify relevant predictors. Our study reported high prevalence of intentional nonadherence. However, intentional nonadherence does not always appear to be related with detrimental clinical 118 consequences. Some studies paradoxically identified a positive relationship between mortality and adherence (Park et al., 2008). This was interpreted using “reasoned adherence”, that is, patients decide to follow treatment guidelines only to a certain point to balance their treatment needs and their daily life needs (Nevins, 2002). This kind of behavior may reduce prescriptions errors by physicians, free patients from the helpless role and hence provide survival advantages (Leggat, 2005; O'Brien, 1990). More studies are required to ascertain the relationship between intentional nonadherence and clinical outcomes. Another direction worth mentioning is regarding the influence of carers on patients’ outcomes. Since PD is a home based therapy, carers are closely involved in patients’ treatment such as moving dialysis fluids, performing exchanges, and making critical medical decisions (Fan, Sathick, McKitty, & Punzalan, 2008). Carers of PD patients are vulnerable to poor mental health and have more diminished quality of life than those of HD patients (Belasco, Barbosa, Bettencourt, Diccini, & Sesso, 2006; Fan et al., 2008; Shimoyama et al., 2003). The implication of this on patients’ outcomes has been completely overlooked. It is possible that caring for PD patients may give rise to carer burnout which may result in suboptimal care or switching to in-center HD. 119 CHAPTER FIVE Conclusion This study assessed health beliefs, emotional distress, quality of life and nonadherence in a representative cohort of APD and CAPD patients. High rates of nonadherence were observed for all three aspects of the medical regimen, i.e., dialysis, medication and diet, with dietary nonadherence being the most remarkable. Intentional nonadherence occurred more commonly than unintentional nonadherence, In addition to nonadherence, PD patients also exhibited unfavorable attitudes toward treatment regimen, poor emotional outcomes, and compromised quality of life. APD patients were more adherent than CAPD patients and CAPD patients reported higher self-efficacy about managing their medicines. No other differences between APD and CAPD patients could be demonstrated. More importantly, this study identified important determinants of nonadherence. The strongest predictors of self-reported nonadherence tend to be psychosocial factors instead of demographical and clinical factors. Health belief factors were found to be mediators in the relationship between emotional distress and nonadherence. These significant findings have important implications for the successful management of PD programme. Efforts aimed at improving patients’ communication with medical providers and addressing patients’ concerns about therapy or doubts about their own self-management capability are very likely to be effective in terms of reducing nonadherence and maximizing treatment effects in PD populations. 120 References Addington, W. W. (1979). Patient compliance: the most serious remaining problem in the control of tuberculosis in the United States. Chest, 76(Suppl. 6), S741-S743. Agganis, B. T., Weiner, D. E., Giang, L. M., Scott, T., Tighiouart, H., Griffith, J. L., & Sarnak, M. J. (2010). Depression and cognitive function in maintenance hemodialysis patients. American Journal of Kidney Diseases, 56(4), 704-712. Ahlmen, J., Carlsson, L., & Schonborg, C. (1993). Well-informed patients with end-stage renal disease prefer peritoneal dialysis to hemodialysis. Peritoneal Dialysis International, 13 (Suppl. 2), S196-S198. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall. Allen, D., Wainwright, M., & Hutchinson, T. (2011). 'Non-compliance' as illness management: Hemodialysis patients' descriptions of adversarial patient-clinician interactions. Social Science and Medicine, 73(1), 129-134. Amici, G., Viglino, G., Virga, G., Gandolfo, C., Da Rin, G., Bocci, C., & Cavalli, P. L. (1996). Compliance study in peritoneal dialysis using PD Adequest software. Peritoneal Dialysis International, 16 (Suppl. 1), S176-S178. Andrykowski, M. A., & Hunt, J. W. (1993). Positive psychosocial adjustment in potential bone marrow transplant recipients: Cancer as a psychosocial transition. Psycho Oncology, 2(4), 261-276. Ang, K. L., & Foo, S. (2002). An exploratory study of eating patterns of Singapore children and teenagers. Health Education, 102(5), 239-248. Arenas, M. D., Malek, T., Gil, M. T., Moledous, A., Alvarez-Ude, F., & Reig-Ferrer, A. (2010). Challenge of phosphorus control in hemodialysis patients: a problem of adherence? Journal of Nephrology, 23(5), 525-534. Asefzadeh, S., Asefzadeh, M., & Javadi, H. (2005). Care management: Adherence to therapies among patients at Bu-Alicina clinic, Qazvin, Iran. Journal of Research in Medical Sciences, 10(6), 343-348. Aslam, N., Bernardini, J., Fried, L., Burr, R., & Piraino, B. (2006). Comparison of infectious complications between incident hemodialysis and peritoneal dialysis patients. Clinical Journal of the American Society of Nephrology, 1(6), 1226-1233. Asti, T., Kara, M., Ipek, G., & Erci, B. (2006). The experiences of loneliness, depression, and social support of Turkish patients with continuous ambulatory peritoneal dialysis and their caregivers. Journal of Clinical Nursing, 15(4), 490-497. 121 Atalay, H., Solak, Y., Biyik, M., Biyik, Z., Yeksan, M., Uguz, F., et al. (2010). Sertraline treatment is associated with an improvement in depression and health-related quality of life in chronic peritoneal dialysis patients. International Urology and Nephrology, 42(2), 527-536. Badve, S. V., Hawley, C. M., McDonald, S. P., Mudge, D. W., Rosman, J. B., Brown, F. G., & Johnson, D. W. (2008). Automated and continuous ambulatory peritoneal dialysis have similar outcomes. Kidney International, 73(4), 480-488. Bakewell, A. B., Higgins, R. M., & Edmunds, M. E. (2001). Does ethnicity influence perceived quality of life of patients on dialysis and following renal transplant? Nephrology, Dialysis, Transplantation, 16(7), 1395-1401. Bakewell, A. B., Higgins, R. M., & Edmunds, M. E. (2002). Quality of life in peritoneal dialysis patients: decline over time and association with clinical outcomes. Kidney International, 61(1), 239-248. Balasubramanian, G., McKitty, K., & Fan, S. L. (2011). Comparing automated peritoneal dialysis with continuous ambulatory peritoneal dialysis: survival and quality of life differences? Nephrology, Dialysis, Transplantation, 26(5), 1702-1708. Bame, S. I., Petersen, N., & Wray, N. P. (1993). Variation in hemodialysis patient compliance according to demographic characteristics. Social Science and Medicine, 37(8), 1035-1043. Bander, S. J., & Walters, B. A. (1998). Hemodialysis morbidity and mortality: links to patient non-compliance. Current Opinion in Nephrology and Hypertension, 7(6), 649-653. Bandura, A. (1977). Self-efficacy: toward a unifying theory of behavioral change. Psychological Review, 84(2), 191-215. Bandura, A. (1989). Regulation of cognitive processes through perceived self-efficacy. Developmental Psychology, 25(5), 729. Bandura, A. (1991). Social cognitive theory of self-regulation. Organizational Behavior and Human Decision Processes. Bandura, A. (2006). Guide for constructing self-efficacy scales. Self-efficacy beliefs of adolescents, 5, 307-337. Barat, I., Andreasen, F., & Damsgaard, E. (2001). Drug therapy in the elderly: what doctors believe and patients actually do. British Journal of Clinical Pharmacology, 51(6), 615-622. Barnett, T., Li Yoong, T., Pinikahana, J., & Si-Yen, T. (2008). Fluid compliance among patients having haemodialysis: can an educational programme make a difference? Journal of Advanced Nursing, 61(3), 300-306. 122 Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173-1182. Basok, E. K., Atsu, N., Rifaioglu, M. M., Kantarci, G., Yildirim, A., & Tokuc, R. (2009). Assessment of female sexual function and quality of life in predialysis, peritoneal dialysis, hemodialysis, and renal transplant patients. International Urology and Nephrology, 41(3), 473-481. Beddhu, S., Zeidel, M. L., Saul, M., Seddon, P., Samore, M. H., Stoddard, G. J., & Bruns, F. J. (2002). The effects of comorbid conditions on the outcomes of patients undergoing peritoneal dialysis. American Journal of Medicine, 112(9), 696-701. Belasco, A., Barbosa, D., Bettencourt, A. R., Diccini, S., & Sesso, R. (2006). Quality of life of family caregivers of elderly patients on hemodialysis and peritoneal dialysis. American Journal of Kidney Diseases, 48(6), 955-963. Bender, F. H., Bernardini, J., & Piraino, B. (2006). Prevention of infectious complications in peritoneal dialysis: best demonstrated practices. Kidney International. Supplement(103), S44-S54. Bernardini, J., Nagy, M., & Piraino, B. (2000). Pattern of noncompliance with dialysis exchanges in peritoneal dialysis patients. American Journal of Kidney Diseases, 35(6), 1104-1110. Bernardini, J., & Piraino, B. (1997). Measuring compliance with prescribed exchanges in CAPD and CCPD patients. Peritoneal Dialysis International, 17(4), 338-342. Bernardini, J., & Piraino, B. (1998). Compliance in CAPD and CCPD patients as measured by supply inventories during home visits. American Journal of Kidney Diseases, 31(1), 101-107. Bilgic, A., Akman, B., Sezer, S., Ozisik, L., Arat, Z., Ozdemir, F. N., & Haberal, M. (2008). Predictors for quality of life in continuous ambulatory peritoneal dialysis patients. Nephrology (Carlton), 13(7), 587-592. Blake, P. G. (1999). Advantages and disadvantages of automated peritoneal dialysis compared to continuous ambulatory peritoneal dialysis. Peritoneal Dialysis International, 19 (Suppl. 2), S121-S124. Blake, P. G., Spanner, E., McMurray, S., Lindsay, R. M., & Ferguson, E. (1996). Comparison of measured and predicted creatinine excretion is an unreliable index of compliance in PD patients. Peritoneal Dialysis International, 16(2), 147-153. Bleyer, A. J., Hylander, B., Sudo, H., Nomoto, Y., de la Torre, E., Chen, R. A., & Burkart, J. M. (1999). An international study of patient compliance with hemodialysis. Journal of the American Medical Association, 281(13), 1211-1213. 123 Block, G. A., Hulbert-Shearon, T. E., Levin, N. W., & Port, F. K. (1998). Association of serum phosphorus and calcium x phosphate product with mortality risk in chronic hemodialysis patients: a national study. American Journal of Kidney Diseases, 31(4), 607-617. Bohlke, M., Nunes, D. L., Marini, S. S., Kitamura, C., Andrade, M., & Von-Gysel, M. P. (2008). Predictors of quality of life among patients on dialysis in southern Brazil. Sao Paulo Medical Journal, 126(5), 252-256. Brady, B. A., Tucker, C. M., Alfino, P. A., Tarrant, D. G., & Finlayson, G. C. (1997). An investigation of factors associated with fluid adherence among hemodialysis patients: a self-efficacy theory based approach. Annals of Behavioral Medicine, 19(4), 339-343. Branin, J. J. (2001). The role of memory strategies in medication adherence among the elderly. Home Health Care Services Quarterly, 20(2), 1-16. Britten, N. (1994). Patients' ideas about medicines: a qualitative study in a general practice population. The British Journal of General Practice, 44(387), 465-468. Bro, S., Bjorner, J. B., Tofte-Jensen, P., Klem, S., Almtoft, B., Danielsen, H., et al. (1999). A prospective, randomized multicenter study comparing APD and CAPD treatment. Peritoneal Dialysis International, 19(6), 526-533. Broekman, B. F., Nyunt, S. Z., Niti, M., Jin, A. Z., Ko, S. M., Kumar, R., et al. (2008). Differential item functioning of the Geriatric Depression Scale in an Asian population. Journal of Affective Disorders, 108(3), 285-290. Brown, E. A., Johansson, L., Farrington, K., Gallagher, H., Sensky, T., Gordon, F., et al. (2010). Broadening Options for Long-term Dialysis in the Elderly (BOLDE): differences in quality of life on peritoneal dialysis compared to haemodialysis for older patients. Nephrology, Dialysis, Transplantation, 25(11), 3755-3763. Brownbridge, G., & Fielding, D. M. (1994). Psychosocial adjustment and adherence to dialysis treatment regimes. Pediatric Nephrology, 8(6), 744-749. Browne, T., & Merighi, J. R. (2010). Barriers to adult hemodialysis patients' selfmanagement of oral medications. American Journal of Kidney Diseases, 56(3), 547-557. Brunkhorst, R. R. (2005). Individualized PD prescription: APD versus CAPD. Peritoneal Dialysis International, 25 (Suppl. 3), S92-S94. Buemi, M., Lacquaniti, A., Bolignano, D., Donato, V., Fazio, M. R., Campo, S., et al. (2008). Dialysis and the elderly: an underestimated problem. Kidney and Blood Pressure Research, 31(5), 330-336. 124 Buffardi, L. E., & Campbell, W. K. (2008). Narcissism and social networking Web sites. Personality and Social Psychology Bulletin, 34(10), 1303-1314. Cao, Z., Ong, K. C., Eng, P., Tan, W. C., & Ng, T. P. (2006). Frequent hospital readmissions for acute exacerbation of COPD and their associated factors. Respirology, 11(2), 188-195. Chan, R., Brooks, R., Erlich, J., Chow, J., & Suranyi, M. (2009). The effects of kidneydisease-related loss on long-term dialysis patients' depression and quality of life: positive affect as a mediator. Clinical Journal of the American Society of Nephrology, 4(1), 160-167. Chen, W., Lu, X. H., & Wang, T. (2006). Menu suggestion: an effective way to improve dietary compliance in peritoneal dialysis patients. Journal of Renal Nutrition, 16(2), 132-136. Chilcot, J., Davenport, A., Wellsted, D., Firth, J., & Farrington, K. (2011). An association between depressive symptoms and survival in incident dialysis patients. Nephrology, Dialysis, Transplantation, 26(5), 1628-1634. Chilcot, J., Wellsted, D., & Farrington, K. (2010). Illness representations are associated with fluid nonadherence among hemodialysis patients. Journal of Psychosomatic Research, 68(2), 203-212. Chiu, Y. W., Teitelbaum, I., Misra, M., de Leon, E. M., Adzize, T., & Mehrotra, R. (2009). Pill burden, adherence, hyperphosphatemia, and quality of life in maintenance dialysis patients. Clinical Journal of the American Society of Nephrology, 4(6), 1089-1096. Chow, S. K., & Wong, F. K. (2010). Health-related quality of life in patients undergoing peritoneal dialysis: effects of a nurse-led case management programme. Journal of Advanced Nursing, 66(8), 1780-1792. Christensen, A. J., Benotsch, E. G., Wiebe, J. S., & Lawton, W. J. (1995). Coping with treatment-related stress: effects on patient adherence in hemodialysis. Journal of Consulting and Clinical Psychology, 63(3), 454-459. Christensen, A. J., & Ehlers, S. L. (2002). Psychological factors in end-stage renal disease: an emerging context for behavioral medicine research. Journal of Consulting and Clinical Psychology, 70(3), 712-724. Christensen, A. J., Moran, P. J., Lawton, W. J., Stallman, D., & Voigts, A. L. (1997). Monitoring attentional style and medical regimen adherence in hemodialysis patients. Health Psychology, 16(3), 256-262. Christensen, A. J., Wiebe, J. S., Benotsch, E. G., & Lawton, W. J. (1996). Perceived health competence, health locus of control, and patient adherence in renal dialysis. Cognitive Therapy and Research, 20(4), 411-421. 125 Chua, A. N., & Warady, B. A. (2011). Adherence of pediatric patients to automated peritoneal dialysis. Pediatric Nephrology, 26(5), 789-793. Chuang, Y. W., Shu, K. H., Yu, T. M., Cheng, C. H., & Chen, C. H. (2009). Hypokalaemia: an independent risk factor of Enterobacteriaceae peritonitis in CAPD patients. Nephrology, Dialysis, Transplantation, 24(5), 1603-1608. Clifford, S., Barber, N., Elliott, R., Hartley, E., & Horne, R. (2006). Patient-centred advice is effective in improving adherence to medicines. Pharmacy World and Science, 28(3), 165-170. Clifford, S., Barber, N., & Horne, R. (2008). Understanding different beliefs held by adherers, unintentional nonadherers, and intentional nonadherers: application of the Necessity-Concerns Framework. Journal of Psychosomatic Research, 64(1), 41-46. Conde, S. A., Fernandes, N., Santos, F. R., Chouab, A., Mota, M. M., & Bastos, M. G. (2010). Cognitive decline, depression and quality of life in patients at different stages of chronic kidney disease. Journal of Brasileiro de Nefrologia, 32(3), 242248. Cook, W. L., & Jassal, S. V. (2008). Functional dependencies among the elderly on hemodialysis. Kidney International, 73(11), 1289-1295. Cukor, D., Coplan, J., Brown, C., Friedman, S., Cromwell-Smith, A., Peterson, R. A., & Kimmel, P. L. (2007). Depression and anxiety in urban hemodialysis patients. Clinical Journal of the American Society of Nephrology, 2(3), 484-490. Cukor, D., Coplan, J., Brown, C., Friedman, S., Newville, H., Safier, M., et al. (2008). Anxiety disorders in adults treated by hemodialysis: a single-center study. American Journal of Kidney Diseases, 52(1), 128-136. Cukor, D., Rosenthal, D. S., Jindal, R. M., Brown, C. D., & Kimmel, P. L. (2009). Depression is an important contributor to low medication adherence in hemodialyzed patients and transplant recipients. Kidney International, 75(11), 1223-1229. Cummings, K. M., Becker, M. H., Kirscht, J. P., & Levin, N. W. (1981). Intervention strategies to improve compliance with medical regimens by ambulatory hemodialysis patients. Journal of Behavioral Medicine, 4(1), 111-127. Curtin, R. B., Svarstad, B. L., & Keller, T. H. (1999). Hemodialysis patients' noncompliance with oral medications. ANNA Journal, 26(3), 307-316. Daleboudt, G. M., Broadbent, E., McQueen, F., & Kaptein, A. A. (2010). Intentional and unintentional treatment non-adherence in patients with systemic lupus erythematosus. Arthritis Care and Research, 63(3), 342-350. 126 Davenport, A. (2009). Peritonitis remains the major clinical complication of peritoneal dialysis: the London, UK, peritonitis audit 2002-2003. Peritoneal Dialysis International, 29(3), 297-302. De-Nour, A. K., & Czaczkes, J. W. (1972). Personality factors in chronic hemodialysis patients causing noncompliance with medical regimen. Psychosomatic Medicine, 34(4), 333-344. De-Nour, A. K., & Czaczkes, J. W. (1976). The influence of patient's personality on adjustment to chronic dialysis. Journal of Nervous and Mental Disease, 162(5), 323-333. de Wit, G. A., Merkus, M. P., Krediet, R. T., & de Charro, F. T. (2001). A comparison of quality of life of patients on automated and continuous ambulatory peritoneal dialysis. Peritoneal Dialysis International, 21(3), 306-312. Dekker, R. L., Lennie, T. A., Hall, L. A., Peden, A. R., Chung, M. L., & Moser, D. K. (2011). Developing a shortened measure of negative thinking for use in patients with heart failure. Heart and Lung, 40(3), e60-69. Dell'Aquila, R., Berlingo, G., Pellanda, M. V., & Contestabile, A. (2009). Continuous ambulatory peritoneal dialysis and automated peritoneal dialysis: are there differences in outcome? Contributions to Nephrology, 163, 292-299. Denhaerynck, K., Manhaeve, D., Dobbels, F., Garzoni, D., Nolte, C., & De Geest, S. (2007). Prevalence and consequences of nonadherence to hemodialysis regimens. American Journal of Critical Care, 16(3), 222-235. DeOreo, P. B. (1997). Hemodialysis patient-assessed functional health status predicts continued survival, hospitalization, and dialysis-attendance compliance. American Journal of Kidney Diseases, 30(2), 204-212. Diaz-Buxo, J. A., Lowrie, E. G., Lew, N. L., Zhang, H., & Lazarus, J. M. (2000). Quality-of-life evaluation using Short Form 36: comparison in hemodialysis and peritoneal dialysis patients. American Journal of Kidney Diseases, 35(2), 293-300. Diefenthaeler, E. C., Wagner, M. B., Poli-de-Figueiredo, C. E., Zimmermann, P. R., & Saitovitch, D. (2008). Is depression a risk factor for mortality in chronic hemodialysis patients? . Revista Brasileira de Psiquiatria, 30(2), 99-103. DiMatteo, M. R., Lepper, H. S., & Croghan, T. W. (2000). Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of anxiety and depression on patient adherence. Archives of Internal Medicine, 160(14), 2101-2107. Dimkovic, N., & Oreopoulos, D. G. (2000). Chronic peritoneal dialysis in the elderly: a review. Peritoneal Dialysis International, 20(3), 276-283. 127 Dimkovic, N., & Oreopoulos, D. G. (2008). Assisted peritoneal dialysis as a method of choice for elderly with end-stage renal disease. International Urology and Nephrology, 40(4), 1143-1150. Donovan, J. L. (1995). Patient decision making. The missing ingredient in compliance research. International Journal of Technology Assessment in Health Care, 11(3), 443-455. Donovan, J. L., & Blake, D. R. (1992). Patient non-compliance: deviance or reasoned decision-making? Social Science and Medicine, 34(5), 507-513. Donovan, J. L., Blake, D. R., & Fleming, W. G. (1989). The patient is not a blank sheet: lay beliefs and their relevance to patient education. British Journal of Rheumatology, 28(1), 58-61. Durose, C. L., Holdsworth, M., Watson, V., & Przygrodzka, F. (2004). Knowledge of dietary restrictions and the medical consequences of noncompliance by patients on hemodialysis are not predictive of dietary compliance. Journal of the American Dietetic Association, 104(1), 35-41. Eitel, P., Friend, R., Griffin, K. W., & Wadhwa, N. K. (1998). Cognitive control and consistency in compliance. Psychology and Health, 13, 953-973. Eliasson, L., Clifford, S., Barber, N., & Marin, D. (2011). Exploring chronic myeloid leukemia patients' reasons for not adhering to the oral anticancer drug imatinib as prescribed. Leukemia Research, 35(5), 626-630. Endler, N. S., Parker, J. D., Bagby, R. M., & Cox, B. J. (1991). Multidimensionality of state and trait anxiety: factor structure of the Endler Multidimensional Anxiety Scales. Journal of Personality and Social Psychology, 60(6), 919-926. Enoch, C., Aslam, N., & Piraino, B. (2002). Intra-abdominal pressure, peritoneal dialysis exchange volume, and tolerance in APD. Seminars in Dialysis, 15(6), 403-406. Factor, K. F. (2007). Potassium management in pediatric peritoneal dialysis patients: can a diet with increased potassium maintain a normal serum potassium without a potassium supplement? Advances in Peritoneal Dialysis, 23, 167-169. Fan, S. L., Sathick, I., McKitty, K., & Punzalan, S. (2008). Quality of life of caregivers and patients on peritoneal dialysis. Nephrology, Dialysis, Transplantation, 23(5), 1713-1719. Fang, J. Q., Hao, Y. T., & Li, C. X. (1999). Reliability and validity for Chinese version of WHO quality of life scale (in Chinese). Chinese Mental Health Journal, 13, 203205. 128 Faul, F., Erdfelder, E., Buchner, A., & Lang, A. G. (2009). Statistical power analyses using G* Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41(4), 1149-1160. Fernandes, G. V., dos Santos, R. R., Soares, W., de Lima, L. G., de Macedo, B. S., da Fonte, J. E., et al. (2010). The impact of erectile dysfunction on the quality of life of men undergoing hemodialysis and its association with depression. Journal of Sexual Medicine, 7(12), 4003-4010. Figueiredo, A. E., Santos, K. S., & Creutzberg, M. (2005). Compliance in peritoneal dialysis measured by supply inventories. Advances in Peritoneal Dialysis, 21, 7779. Fine, A. (1997). Compliance with CAPD prescription is good. Peritoneal Dialysis International, 17(4), 343-346. Fine, R. N., & Ho, M. (2002). The role of APD in the management of pediatric patients: a report of the North American Pediatric Renal Transplant Cooperative Study. Seminars in Dialysis, 15(6), 427-429. Fredriksen-Goldsen, K. I., Shiu, C. S., Starks, H., Chen, W. T., Simoni, J., Kim, H. J., et al. (2011). "You Must Take the Medications for You and for Me": Family Caregivers Promoting HIV Medication Adherence in China. AIDS Patient Care and STDs, advance online publication. Friend, R., Hatchett, L., Schneider, M. S., & Wadhwa, N. K. (1997). A comparison of attributions, health beliefs, and negative emotions as predictors of fluid adherence in renal dialysis patients: a prospective analysis. Annals of Behavioral Medicine, 19(4), 344-347. Fructuoso, M., Castro, R., Oliveira, L., Prata, C., & Morgado, T. (2011). Quality of life in chronic kidney disease. Nefrologia, 31(1), 91-96. Fukunishi, I., Kitaoka, T., Shirai, T., Kino, K., Kanematsu, E., & Sato, Y. (2002). Psychiatric disorders among patients undergoing hemodialysis therapy. Nephron, 91(2), 344-347. George, J., Kong, D. C., & Stewart, K. (2007). Adherence to disease management programs in patients with COPD. International Journal of Chronic Obstructive Pulmonary Disease, 2(3), 253-262. George, J., Mackinnon, A., Kong, D. C., & Stewart, K. (2006). Development and validation of the Beliefs and Behaviour Questionnaire (BBQ). Patient Education and Counseling, 64, 50-60. Gincherman, Y., Moloney, K., McKee, C., & Coyne, D. W. (2010). Assessment of adherence to cinacalcet by prescription refill rates in hemodialysis patients. Hemodialysis International, 14(1), 68-72. 129 Ginieri-Coccossis, M., Theofilou, P., Synodinou, C., Tomaras, V., & Soldatos, C. (2008). Quality of life, mental health and health beliefs in haemodialysis and peritoneal dialysis patients: investigating differences in early and later years of current treatment. BMC Nephrology, 9(14), 14. Glover, C., Banks, P., Carson, A., Martin, C. R., & Duffy, T. (2011). Understanding and Assessing the Impact of End-Stage Renal Disease on Quality of Life: A Systematic Review of the Content Validity of Self-Administered Instruments Used to Assess Health-Related Quality of Life in End-Stage Renal Disease. The Patient: Patient-Centered Outcomes Research, 4(1), 19-30. Goldwasser, P., & Feldman, J. (1997). Association of serum albumin and mortality risk. Journal of Clinical Epidemiology, 50(6), 693-703. Gonsalves-Ebrahim, L., Sterin, G., Gulledge, A. D., Gipson, W. T., & Rodgers, D. A. (1987). Noncompliance in younger adults on hemodialysis. Psychosomatics, 28(1), 34-41. Grassmann, A., Gioberge, S., Moeller, S., & Brown, G. (2005). ESRD patients in 2004: global overview of patient numbers, treatment modalities and associated trends. Nephrology, Dialysis, Transplantation, 20(12), 2587-2593. Greenblatt, D. J. (1979). Reduced serum albumin concentration in the elderly: a report from the Boston Collaborative Drug Surveillance Program. Journal of the American Geriatrics Society, 27(1), 20-22. Griva, K., Davenport, A., Harrison, M., & Newman, S. (2010). An evaluation of illness, treatment perceptions, and depression in hospital- vs. home-based dialysis modalities. Journal of Psychosomatic Research, 69(4), 363-370. Gudex, C. M. (1995). Health-related quality of life in endstage renal failure. Quality of Life Research, 4(4), 359-366. Guney, I., Biyik, M., Yeksan, M., Biyik, Z., Atalay, H., Solak, Y., et al. (2008). Sleep quality and depression in peritoneal dialysis patients. Renal Failure, 30(10), 1017-1022. Guney, I., Solak, Y., Atalay, H., Yazici, R., Altintepe, L., Kara, F., et al. (2010). Comparison of effects of automated peritoneal dialysis and continuous ambulatory peritoneal dialysis on health-related quality of life, sleep quality, and depression. Hemodialysis International, 14(4), 515-522. Harper, A., & Power, M. (1998). Development of the World Health Organization WHOQOL-BREF Quality of Life Assessment. Psychological Medicine, 28, 551558. 130 Harris, S. A., Lamping, D. L., Brown, E. A., & Constantinovici, N. (2002). Clinical outcomes and quality of life in elderly patients on peritoneal dialysis versus hemodialysis. Peritoneal Dialysis International, 22(4), 463-470. Haynes, R. B., Taylor, D. W., Sackett, D. L., Gibson, E. S., Bernholz, C. D., & Mukherjee, J. (1980). Can simple clinical measurements detect patient noncompliance? Hypertension, 2(6), 757-764. Hays, R. D., Kallich, J. D., Mapes, D. L., Coons, S. J., Amin, N., Carter, W., & Kamberg, C. (1997). Kidney Disease Quality of Life Short Form (KDQOL-SFTM), Version 1.3: A Manual for Use and Scoring. Santa Monica, CA: RAND. Hecking, E., Bragg-Gresham, J. L., Rayner, H. C., Pisoni, R. L., Andreucci, V. E., Combe, C., et al. (2004). Haemodialysis prescription, adherence and nutritional indicators in five European countries: results from the Dialysis Outcomes and Practice Patterns Study (DOPPS). Nephrology, Dialysis, Transplantation, 19(1), 100-107. Hedayati, S. S., Bosworth, H. B., Briley, L. P., Sloane, R. J., Pieper, C. F., Kimmel, P. L., & Szczech, L. A. (2008). Death or hospitalization of patients on chronic hemodialysis is associated with a physician-based diagnosis of depression. Kidney International, 74(7), 930-936. Herlin, C., & Wann-Hansson, C. (2010). The experience of being 30-45 years of age and depending on haemodialysis treatment: a phenomenological study. Scandinavian Journal of Caring Sciences, 24(4), 693-699. Herrero, M. J., Blanch, J., Peri, J. M., De Pablo, J., Pintor, L., & Bulbena, A. (2003). A validation study of the hospital anxiety and depression scale (HADS) in a Spanish population. General Hospital Psychiatry, 25(4), 277-283. Herrman, H., Schofield, H., Murphy, B., Metelko, Z., Szabo, S., Pibernik-Okanovic, M., et al. (1998). Development of the World Health Organization WHOQOL-BREF quality of life assessment. Psychological Medicine, 28(3), 551-558. Hirth, R. A., Greer, S. L., Albert, J. M., Young, E. W., & Piette, J. D. (2008). Out-ofpocket spending and medication adherence among dialysis patients in twelve countries. Health Affairs, 27(1), 89-102. Hoffman, M. L. (1977). Sex differences in empathy and related behaviors. Psychological Bulletin, 84(4), 712-722. Holley, J. L., & DeVore, C. C. (2006). Why all prescribed medications are not taken: results from a survey of chronic dialysis patients. Advances in Peritoneal Dialysis, 22, 162-166. 131 Horne, R., & Weinman, J. (1999). Patients' beliefs about prescribed medicines and their role in adherence to treatment in chronic physical illness. Journal of Psychosomatic Research, 47(6), 555-567. Horne, R., Weinman, J., & Hankins, M. (1999). The beliefs about medicines questionnaire: the development and evaluation of a new method for assessing the cognitive representation of medication. Psychology and Health, 14(1), 1-24. Hung, K. C., Wu, C. C., Chen, H. S., Ma, W. Y., Tseng, C. F., Yang, L. K., et al. (2011). Serum IL-6, albumin and co-morbidities are closely correlated with symptoms of depression in patients on maintenance haemodialysis. Nephrology, Dialysis, Transplantation, 26(2), 658-664. Ibrahim, S., & El Salamony, O. (2008). Depression, quality of life and malnutritioninflammation scores in hemodialysis patients. American Journal of Nephrology, 28(5), 784-791. Inui, T. S., Carter, W. B., & Pecoraro, R. E. (1981). Screening for noncompliance among patients with hypertension: is self-report the best available measure? Medical Care, 19(10), 1061-1064. Ismail, N., Hakim, R. M., Oreopoulos, D. G., & Patrikarea, A. (1993). Renal replacement therapies in the elderly: Part 1. Hemodialysis and chronic peritoneal dialysis. American Journal of Kidney Diseases, 22(6), 759-782. Jaber, B. L., Lee, Y., Collins, A. J., Hull, A. R., Kraus, M. A., McCarthy, J., et al. (2010). Effect of daily hemodialysis on depressive symptoms and postdialysis recovery time: interim report from the FREEDOM (Following Rehabilitation, Economics and Everyday-Dialysis Outcome Measurements) Study. American Journal of Kidney Diseases, 56(3), 531-539. Jassal, S. V., & Watson, D. (2011). Offering peritoneal dialysis to the older patient: medical progress or waste of time? Seminars in Nephrology, 31(2), 225-234. Jhamb, M., Tamura, M. K., Gassman, J., Garg, A. X., Lindsay, R. M., Suri, R. S., et al. (2011). Design and rationale of health-related quality of life and patient-reported outcomes assessment in the Frequent Hemodialysis Network trials. Blood Purification, 31(1-3), 151-158. Johnson, D. W., Hawley, C. M., McDonald, S. P., Brown, F. G., Rosman, J. B., Wiggins, K. J., et al. (2010). Superior survival of high transporters treated with automated versus continuous ambulatory peritoneal dialysis. Nephrology, Dialysis, Transplantation, 25(6), 1973-1979. Johnson, M. J. (2002). The Medication Adherence Model: a guide for assessing medication taking. Research and Theory for Nursing Practice, 16(3), 179-192. 132 Joshi, V. D., Mooppil, N., & Lim, J. F. (2010). Validation of the Kidney Disease Quality of Life-Short Form: a cross-sectional study of a dialysis-targeted health measure in Singapore. BMC Nephrology, 11, 36. Juergensen, P. H., Gorban-Brennan, N., & Finkelstein, F. O. (2004). Compliance with the dialysis regimen in chronic peritoneal dialysis patients: utility of the pro card and impact of patient education. Advances in Peritoneal Dialysis, 20, 90-92. Kaazempur-Mofrad, M. R., Vacanti, J. P., Krebs, N. J., & Borenstein, J. T. (2004). A MEMS-based renal replacment system. Solid-State Sensor, Actuator and Microsystems Workshop, 67-70. Kalender, B., Ozdemir, A. C., Dervisoglu, E., & Ozdemir, O. (2007). Quality of life in chronic kidney disease: effects of treatment modality, depression, malnutrition and inflammation. International Journal of Clinical Practice, 61(4), 569-576. Kammerer, J., Garry, G., Hartigan, M., Carter, B., & Erlich, L. (2007). Adherence in patients on dialysis: strategies for success. Nephrology Nursing Journal, 34(5), 479-486. Kara, B., Caglar, K., & Kilic, S. (2007). Nonadherence with diet and fluid restrictions and perceived social support in patients receiving hemodialysis. Journal of Nursing Scholarship, 39(3), 243-248. Kara, M., & Mirici, A. (2004). Loneliness, depression, and social support of Turkish patients with chronic obstructive pulmonary disease and their spouses. Journal of Nursing Scholarship, 36(4), 331-336. Karamanidou, C., Clatworthy, J., Weinman, J., & Horne, R. (2008). A systematic review of the prevalence and determinants of nonadherence to phosphate binding medication in patients with end-stage renal disease. BMC Nephrology, 9, 2. Karamanidou, C., Weinman, J., & Horne, R. (2008). Improving haemodialysis patients' understanding of phosphate-binding medication: a pilot study of a psychoeducational intervention designed to change patients' perceptions of the problem and treatment. British Journal of Health Psychology, 13(2), 205-214. Katz, R. C., Ashmore, J., Barboa, E., Trueblood, K., McLaughlin, V., & Mathews, L. (1998). Knowledge of disease and dietary compliance in patients with end-stage renal disease. Psychological Reports, 82(1), 331-336. Katzir, Z., Boaz, M., Backshi, I., Cernes, R., Barnea, Z., & Biro, A. (2010). Medication apprehension and compliance among dialysis patients--a comprehensive guidance attitude. Nephron. Clinical Practice, 114(2), 151-157. Kaveh, K., & Kimmel, P. L. (2001). Compliance in hemodialysis patients: multidimensional measures in search of a gold standard. American Journal of Kidney Diseases, 37(2), 244-266. 133 Kawaguchi, Y., Ishizaki, T., Imada, A., Oohira, S., Kuriyama, S., Nakamoto, H., et al. (2003). Searching for the reasons for drop-out from peritoneal dialysis: a nationwide survey in Japan. Peritoneal Dialysis International, 23 (Suppl. 2), S175-S177. Keshaviah, P., Collins, A. J., Ma, J. Z., Churchill, D. N., & Thorpe, K. E. (2002). Survival comparison between hemodialysis and peritoneal dialysis based on matched doses of delivered therapy. Journal of the American Society of Nephrology, 13 (Suppl. 1), S48-52. Khan, A. N., Bernardini, J., Johnston, J. R., & Piraino, B. (1996). Hypokalemia in peritoneal dialysis patients. Peritoneal Dialysis International, 16(6), 652. Kim, J. A., Lee, Y. K., Huh, W. S., Kim, Y. G., Kim, D. J., Oh, H. Y., et al. (2002). Analysis of depression in continuous ambulatory peritoneal dialysis patients. Journal of Korean Medical Science, 17(6), 790-794. Kimmel, P. L., Peterson, R. A., Weihs, K. L., Simmens, S. J., Boyle, D. H., Verme, D., et al. (1995). Behavioral compliance with dialysis prescription in hemodialysis patients. Journal of the American Society of Nephrology, 5(10), 1826-1834. Kimmel, P. L., Weihs, K., & Peterson, R. A. (1993). Survival in hemodialysis patients: the role of depression. Journal of the American Society of Nephrology, 4(1), 1227. Ko, G. J., Kim, M. G., Yu, Y. M., Jo, S. K., Cho, W. Y., & Kim, H. K. (2010). Association between depression symptoms with inflammation and cardiovascular risk factors in patients undergoing peritoneal dialysis. Nephron. Clinical Practice, 116(1), 29-35. Koenig, H. G., George, L. K., Peterson, B. L., & Pieper, C. F. (1997). Depression in medically ill hospitalized older adults: prevalence, characteristics, and course of symptoms according to six diagnostic schemes. American Journal of Psychiatry, 154(10), 1376-1383. Kovac, J. A., Patel, S. S., Peterson, R. A., & Kimmel, P. L. (2002). Patient satisfaction with care and behavioral compliance in end-stage renal disease patients treated with hemodialysis. American Journal of Kidney Diseases, 39(6), 1236-1244. Kugler, C., Maeding, I., & Russell, C. L. (2011). Non-adherence in patients on chronic hemodialysis: an international comparison study. Journal of Nephrology, 24(3), 366-375. Kugler, C., Vlaminck, H., Haverich, A., & Maes, B. (2005). Nonadherence with diet and fluid restrictions among adults having hemodialysis. Journal of Nursing Scholarship, 37(1), 25-29. 134 Kutner, N. G., Zhang, R., McClellan, W. M., & Cole, S. A. (2002). Psychosocial predictors of non-compliance in haemodialysis and peritoneal dialysis patients. Nephrology, Dialysis, Transplantation, 17(1), 93-99. Lai, K. N., Lui, S. L., & Chan, D. T. (1999). Automated peritoneal dialysis in Asia. Peritoneal Dialysis International, 19 (Suppl. 2), S125-S129. Lai, Y. M., Hong, C. P., & Chee, C. Y. (2001). Stigma of mental illness. Singapore Medical Journal, 42(3), 111-114. Lam, L. W., Twinn, S. F., & Chan, S. W. (2010). Self-reported adherence to a therapeutic regimen among patients undergoing continuous ambulatory peritoneal dialysis. Journal of Advanced Nursing, 66(4), 763-773. Latham, C. E. (1998). Obstacles to achieving adequate dialysis dose: compliance, education, transportation, and reimbursement. American Journal of Kidney Diseases, 32(Suppl. 4), S93-S95. Leary, M. R. (1990). Responses to social exclusion: Social anxiety, jealousy, loneliness, depression, and low self-esteem. Journal of Social and Clinical Psychology, 9, 221-229. Lee, A. J., Morgan, C. L., Conway, P., & Currie, C. J. (2005). Characterisation and comparison of health-related quality of life for patients with renal failure. Current Medical Research and Opinion, 21(11), 1777-1783. Lee, S. H., & Molassiotis, A. (2002). Dietary and fluid compliance in Chinese hemodialysis patients. International Journal of Nursing Studies, 39(7), 695-704. Leggat, J. E., Jr. (2005). Adherence with dialysis: a focus on mortality risk. Seminars in Dialysis, 18(2), 137-141. Leggat, J. E., Jr., Orzol, S. M., Hulbert-Shearon, T. E., Golper, T. A., Jones, C. A., Held, P. J., & Port, F. K. (1998). Noncompliance in hemodialysis: predictors and survival analysis. American Journal of Kidney Diseases, 32(1), 139-145. Leung, C. M., Ho, S., Kan, C. S., Hung, C. H., & Chen, C. N. (1993). Evaluation of the Chinese version of the Hospital Anxiety and Depression Scale. A cross-cultural perspective. International Journal of Psychosomatics, 40, 29-34. Leventhal, H., Diefenbach, M., & Leventhal, E. A. (1992). Illness cognition: using common sense to understand treatment adherence and affect cognition interactions. Cognitive Therapy and Research, 16(2), 143-163. Lew-Starowicz, M., & Gellert, R. (2009). The sexuality and quality of life of hemodialyzed patients--ASED multicenter study. Journal of Sexual Medicine, 6(4), 1062-1071. 135 Li, C. M. (1998). Development of the Hong Kong Chinese Version of the Kidney Disease Quality of Life Short Form (KDQOL-SF) Instrument. Unpublished master's thesis. The Hong Kong Poly-technic University, Hong Kong. Li, P. K., Szeto, C. C., Piraino, B., Bernardini, J., Figueiredo, A. E., Gupta, A., et al. (2010). Peritoneal dialysis-related infections recommendations: 2010 update. Peritoneal Dialysis International, 30(4), 393-423. Liakopoulos, V., & Dombros, N. (2009). Patient selection for automated peritoneal dialysis: for whom, when? Peritoneal Dialysis International, 29 (Suppl. 2), S102S107. Lin, C. C., & Liang, C. C. (1997). The relationship between health locus of control and compliance of hemodialysis patients. Kaohsiung Journal of Medical Sciences, 13(4), 243-254. Lindberg, M., & Fernandes, M. A. (2010). Self-efficacy in relation to limited fluid intake amongst Portuguese haemodialysis patients. Journal of Renal Care, 36(3), 133138. Lindberg, M., & Lindberg, P. (2008). Overcoming obstacles for adherence to phosphate binding medication in dialysis patients: a qualitative study. Pharmacy World and Science, 30(5), 571-576. Lindberg, M., Wikstrom, B., & Lindberg, P. (2010). Subgroups of haemodialysis patients in relation to fluid intake restrictions: a cluster analytical approach. Journal of Clinical Nursing, 19, 2997-3005. Little, J., Irwin, A., Marshall, T., Rayner, H., & Smith, S. (2001). Predicting a patient's choice of dialysis modality: experience in a United Kingdom renal department. American Journal of Kidney Diseases, 37(5), 981-986. Liu, P. (1993). UCLA loneliness scale. Chin Mental Health, J7, 227-230. Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation. A 35-year odyssey. American Psychologist, 57(9), 705-717. Lok, P. (1996). Stressors, coping mechanisms and quality of life among dialysis patients in Australia. Journal of Advanced Nursing, 23(5), 873-881. Loosman, W. L., Siegert, C. E., Korzec, A., & Honig, A. (2010). Validity of the Hospital Anxiety and Depression Scale and the Beck Depression Inventory for use in endstage renal disease patients. British Journal of Clinical Psychology, 49(4), 507516. Lorig, K. R., Ritter, P., Stewart, A. L., Sobel, D. S., Brown, B. W., Jr., Bandura, A., et al. (2001). Chronic disease self-management program: 2-year health status and health care utilization outcomes. Medical Care, 39(11), 1217-1223. 136 Lowrie, E. G., Curtin, R. B., LePain, N., & Schatell, D. (2003). Medical outcomes study short form-36: a consistent and powerful predictor of morbidity and mortality in dialysis patients. American Journal of Kidney Diseases, 41(6), 1286-1292. Lowry, K. P., Dudley, T. K., Oddone, E. Z., & Bosworth, H. B. (2005). Intentional and unintentional nonadherence to antihypertensive medication. Annals of Pharmacotherapy, 39(7-8), 1198-1203. Lye, W. C., Chan, P. S., Leong, S. O., & van der Straaten, J. C. (1997). Psychosocial and psychiatric morbidity in patients on CAPD. Advances in Peritoneal Dialysis, 13, 134-136. Madar, H., & Bar-Tal, Y. (2009). The experience of uncertainty among patients having peritoneal dialysis. Journal of Advanced Nursing, 65(8), 1664-1669. Maduell, F., Navarro, V., Torregrosa, E., Rius, A., Dicenta, F., Cruz, M. C., & Ferrero, J. A. (2003). Change from three times a week on-line hemodiafiltration to short daily on-line hemodiafiltration. Kidney International, 64(1), 305-313. Maglakelidze, N., Pantsulaia, T., Tchokhonelidze, I., Managadze, L., & Chkhotua, A. (2011). Assessment of health-related quality of life in renal transplant recipients and dialysis patients. Transplantation Proceedings, 43(1), 376-379. Manley, H. J., Garvin, C. G., Drayer, D. K., Reid, G. M., Bender, W. L., Neufeld, T. K., et al. (2004). Medication prescribing patterns in ambulatory haemodialysis patients: comparisons of USRDS to a large not-for-profit dialysis provider. Nephrology, Dialysis, Transplantation, 19(7), 1842-1848. Mapes, D. L., Bragg-Gresham, J. L., Bommer, J., Fukuhara, S., McKevitt, P., Wikstrom, B., & Lopes, A. A. (2004). Health-related quality of life in the Dialysis Outcomes and Practice Patterns Study (DOPPS). American Journal of Kidney Diseases, 44, 54-60. Mapes, D. L., Lopes, A. A., Satayathum, S., McCullough, K. P., Goodkin, D. A., Locatelli, F., et al. (2003). Health-related quality of life as a predictor of mortality and hospitalization: the Dialysis Outcomes and Practice Patterns Study (DOPPS). Kidney International, 64(1), 339-349. Marchais, S. J., Metivier, F., Guerin, A. P., & London, G. M. (1999). Association of hyperphosphataemia with haemodynamic disturbances in end-stage renal disease. Nephrology, Dialysis, Transplantation, 14(9), 2178-2183. Marron, B., Martinez Ocana, J. C., Salgueira, M., Barril, G., Lamas, J. M., Martin, M., et al. (2005). Analysis of patient flow into dialysis: role of education in choice of dialysis modality. Peritoneal Dialysis International, 25 (Suppl. 3), S56-S59. 137 McCarthy, A., Cook, P. S., Fairweather, C., Shaban, R., & Martin-McDonald, K. (2009). Compliance in peritoneal dialysis: a qualitative study of renal nurses. International Journal of Nursing Practice, 15(3), 219-226. McHorney, C. A. (2009). The Adherence Estimator: a brief, proximal screener for patient propensity to adhere to prescription medications for chronic disease. Current Medical Research and Opinion, 25(1), 215-238. Mehrotra, R. (2009). Long-term outcomes in automated peritoneal dialysis: similar or better than in continuous ambulatory peritoneal dialysis? Peritoneal Dialysis International, 29 (Suppl. 2), S111-S114. Mehrotra, R., Chiu, Y. W., Kalantar-Zadeh, K., & Vonesh, E. (2009). The outcomes of continuous ambulatory and automated peritoneal dialysis are similar. Kidney International, 76(1), 97-107. Mendelssohn, D. C., Mullaney, S. R., Jung, B., Blake, P. G., & Mehta, R. L. (2001). What do American nephologists think about dialysis modality selection? American Journal of Kidney Diseases, 37(1), 22-29. Michels, W. M., Verduijn, M., Boeschoten, E. W., Dekker, F. W., & Krediet, R. T. (2009). Similar survival on automated peritoneal dialysis and continuous ambulatory peritoneal dialysis in a large prospective cohort. Clinical Journal of the American Society of Nephrology, 4(5), 943-949. Miller, J. F. (1985). Assessment of loneliness and spiritual well-being in chronically ill and healthy adults. Journal of Professional Nursing, 1(2), 79-85. Mo, P. K., & Mak, W. W. (2009). Intentionality of medication non-adherence among individuals living with HIV/AIDS in Hong Kong. AIDS Care, 21(6), 785-795. Moe, S. M., Chertow, G. M., Coburn, J. W., Quarles, L. D., Goodman, W. G., Block, G. A., et al. (2005). Achieving NKF-K/DOQI bone metabolism and disease treatment goals with cinacalcet HCl. Kidney International, 67(2), 760-771. Moist, L. M., Port, F. K., Orzol, S. M., Young, E. W., Ostbye, T., Wolfe, R. A., et al. (2000). Predictors of loss of residual renal function among new dialysis patients. Journal of the American Society of Nephrology, 11(3), 556-564. Molsted, S., Prescott, L., Heaf, J., & Eidemak, I. (2007). Assessment and clinical aspects of health-related quality of life in dialysis patients and patients with chronic kidney disease. Nephron. Clinical Practice, 106(1), 24-33. Morales Lopez, C., Burrowes, J. D., Gizis, F., & Brommage, D. (2007). Dietary adherence in Hispanic patients receiving hemodialysis. Journal of Renal Nutrition, 17(2), 138-147. 138 Morisky, D. E., Green, L. W., & Levine, D. M. (1986). Concurrent and predictive validity of a self-reported measure of medication adherence. Medical Care, 24(1), 67-74. Murtagh, F. E., Addington-Hall, J., & Higginson, I. J. (2007). The prevalence of symptoms in end-stage renal disease: a systematic review. Advances in Chronic Kidney Disease, 14(1), 82-99. Myers, R. (1990). Classical and modern regression with applications (2nd ed.). Boston, MA: Duxbury. Nerbass, F. B., Morais, J. G., dos Santos, R. G., Kruger, T. S., Koene, T. T., & da Luz Filho, H. A. (2010). Adherence and knowledge about hyperphosphatemia treatment in hemodialysis patients with hyperphosphatemia. Journal of Brasileiro de Nefrologia, 32(2), 149-155. Neri, L., Martini, A., Andreucci, V. E., Gallieni, M., Rocca Rey, L. A., & Brancaccio, D. (2011). Regimen Complexity and Prescription Adherence in Dialysis Patients. American Journal of Nephrology, 34(1), 71-76. Neri, L., Viglino, G., Cappelletti, A., Gandolfo, C., & Barbieri, S. (2002). Compliance in automated peritoneal dialysis. Advances in Peritoneal Dialysis, 18, 72-74. Nevins, T. E. (2002). Non-compliance and its management in teenagers. Pediatric Transplantation, 6(6), 475-479. Ng, T. P., Niti, M., Tan, W. C., Cao, Z., Ong, K. C., & Eng, P. (2007). Depressive symptoms and chronic obstructive pulmonary disease: effect on mortality, hospital readmission, symptom burden, functional status, and quality of life. Archives of Internal Medicine, 167(1), 60-67. Niti, M., Ng, T. P., Kua, E. H., Ho, R. C., & Tan, C. H. (2007). Depression and chronic medical illnesses in Asian older adults: the role of subjective health and functional status. International Journal of Geriatric Psychiatry, 22(11), 1087-1094. Niu, S. F., & Li, I. C. (2005). Quality of life of patients having renal replacement therapy. Journal of Advanced Nursing, 51(1), 15-21. Nolph, K. D., Keshaviah, P., Emerson, P., Van Stone, J. C., Twardowski, Z. J., Khanna, R., et al. (1995). A new approach to optimizing urea clearances in hemodialysis and continuous ambulatory peritoneal dialysis. ASAIO Journal, 41(3), M446-451. Noshad, H., Sadreddini, S., Nezami, N., Salekzamani, Y., & Ardalan, M. R. (2009). Comparison of outcome and quality of life: haemodialysis versus peritoneal dialysis patients. Singapore Medical Journal, 50(2), 185-192. O'Brien, M. E. (1990). Compliance behavior and long-term maintenance dialysis. American Journal of Kidney Diseases, 15(3), 209-214. 139 O'Connor, S. M., Jardine, A. G., & Millar, K. (2008). The prediction of self-care behaviors in end-stage renal disease patients using Leventhal's Self-Regulatory Model. Journal of Psychosomatic Research, 65(2), 191-200. O'Leary, A. (1985). Self-efficacy and health. Behaviour Research and Therapy, 23(4), 437-451. Ogutmen, B., Yildirim, A., Sever, M. S., Bozfakioglu, S., Ataman, R., Erek, E., et al. (2006). Health-related quality of life after kidney transplantation in comparison intermittent hemodialysis, peritoneal dialysis, and normal controls. Transplantation Proceedings, 38(2), 419-421. Oka, M., & Chaboyer, W. (1999). Dietary behaviors and sources of support in hemodialysis patients. Clinical Nursing Research, 8(4), 302-314; discussion 314307. Oka, M., & Chaboyer, W. (2001). Influence of self-efficacy and other factors on dietary behaviours in Japanese haemodialysis patients. International Journal of Nursing Practice, 7(6), 431-439. Oreopoulos, D. G., Ossareh, S., & Thodis, E. (2008). Peritoneal dialysis: past, present, and future. Iranian Journal of Kidney Diseases, 2(4), 171-182. Ortega, O., Gallar, P., Carreno, A., Gutierrez, M., Rodriguez, I., Oliet, A., et al. (2001). Peritoneal sodium mass removal in continuous ambulatory peritoneal dialysis and automated peritoneal dialysis: influence on blood pressure control. American Journal of Nephrology, 21(3), 189-193. Osthus, T. B., Dammen, T., Sandvik, L., Bruun, C. M., Nordhus, I. H., & Os, I. (2010). Health-related quality of life and depression in dialysis patients: associations with current smoking. Scandinavian Journal of Urology and Nephrology, 44(1), 46-55. Owen, W. F., Jr., Lew, N. L., Liu, Y., Lowrie, E. G., & Lazarus, J. M. (1993). The urea reduction ratio and serum albumin concentration as predictors of mortality in patients undergoing hemodialysis. New England Journal of Medicine, 329(14), 1001-1006. Pakpour, A. H., Saffari, M., Yekaninejad, M. S., Panahi, D., Harrison, A. P., & Molsted, S. (2010). Health-related quality of life in a sample of Iranian patients on hemodialysis. Iranian Journal of Kidney Diseases, 4(1), 50-59. Panagopoulou, A., Hardalias, A., Berati, S., & Fourtounas, C. (2009). Psychosocial issues and quality of life in patients on renal replacement therapy. Saudi Journay of Kidney Diseases and Transplantation, 20(2), 212-218. Pang, S. K., Ip, W. Y., & Chang, A. M. (2001). Psychosocial correlates of fluid compliance among Chinese haemodialysis patients. Journal of Advanced Nursing, 35(5), 691-698. 140 Paniagua, R., Amato, D., Vonesh, E., Guo, A., & Mujais, S. (2005). Health-related quality of life predicts outcomes but is not affected by peritoneal clearance: The ADEMEX trial. Kidney International, 67(3), 1093-1104. Park, K. A., Choi-Kwon, S., Sim, Y. M., & Kim, S. B. (2008). Comparison of dietary compliance and dietary knowledge between older and younger Korean hemodialysis patients. Journal of Renal Nutrition, 18(5), 415-423. Partridge, K. A., & Robertson, N. (2011). Body-image disturbance in adult dialysis patients. Disability and Rehabilitation, 33(6), 504-510. Patel, S. S., Shah, V. S., Peterson, R. A., & Kimmel, P. L. (2002). Psychosocial variables, quality of life, and religious beliefs in ESRD patients treated with hemodialysis. American Journal of Kidney Diseases, 40(5), 1013-1022. Patrick, D. L., & Deyo, R. A. (1989). Generic and disease-specific measures in assessing health status and quality of life. Medical Care, 27(Suppl. 3), S217-S232. Peplau, L. A., & Perlman, D. (1979). Blueprint for a social psychological theory of loneliness. In M. Cook & G. Wilson (Eds.), Love and attraction. Oxford, England: Pergamon Press. Perry, G. R. (1990). Loneliness and coping among tertiary-level adult cancer patients in the home. Cancer Nursing, 13(5), 293-302. Piraino, B., & Sheth, H. (2010). Peritonitis - does peritoneal dialysis modality make a difference? Blood Purification, 29(2), 145-149. Plantinga, L. C., Fink, N. E., Jaar, B. G., Sadler, J. H., Coresh, J., Klag, M. J., et al. (2004). Frequency of sit-down patient care rounds, attainment of clinical performance targets, hospitalization, and mortality in hemodialysis patients. Journal of the American Society of Nephrology, 15(12), 3144-3153. Podsadecki, T. J., Vrijens, B. C., Tousset, E. P., Rode, R. A., & Hanna, G. J. (2008). "White coat compliance" limits the reliability of therapeutic drug monitoring in HIV-1-infected patients. HIV Clinical Trials, 9(4), 238-246. Polaschek, N. (2007). 'Doing dialysis at home': client attitudes towards renal therapy. Journal of Clinical Nursing, 16(3), 51-58. Ponz Clemente, E., Martinez Ocana, J. C., Marquina Parra, D., Blasco Cabanas, C., Grau Pueyo, G., Mane Buixo, N., & Garcia Garcia, M. (2010). Analysis of psychological factors influencing peritoneal dialysis selection. Nefrologia, 30(2), 195-201. Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40(3), 879-891. 141 Quan, L., Xu, Y., Luo, S. P., Wang, L., LeBlanc, D., & Wang, T. (2006). Negotiated care improves fluid status in diabetic peritoneal dialysis patients. Peritoneal Dialysis International, 26(1), 95-100. Raj, D. S. (2002). Role of APD in compliance with therapy. Seminars in Dialysis, 15(6), 434-436. Rapisarda, F., Tarantino, A., De Vecchi, A., Baggio, G., Ghezzi, F., Nicodemo, D., et al. (2006). Dialysis and kidney transplantation: similarities and differences in the psychological aspects of noncompliance. Transplantation Proceedings, 38(4), 1006-1009. Rees, G., Leong, O., Crowston, J. G., & Lamoureux, E. L. (2010). Intentional and unintentional nonadherence to ocular hypotensive treatment in patients with glaucoma. Ophthalmology, 117(5), 903-908. Riezebos, R. K., Nauta, K. J., Honig, A., Dekker, F. W., & Siegert, C. E. (2010). The association of depressive symptoms with survival in a Dutch cohort of patients with end-stage renal disease. Nephrology, Dialysis, Transplantation, 25(1), 231236. Rodriguez-Carmona, A., Perez-Fontan, M., Garca-Naveiro, R., Villaverde, P., & Peteiro, J. (2004). Compared time profiles of ultrafiltration, sodium removal, and renal function in incident CAPD and automated peritoneal dialysis patients. American Journal of Kidney Diseases, 44(1), 132-145. Rosenbaum, M., & Ben-Ari Smira, K. (1986). Cognitive and personality factors in the delay of gratification of hemodialysis patients. Journal of Personality and Social Psychology, 51(2), 357-364. Rosenstock, I. M. (1974). The health belief model and preventive health behavior. Health Education Monographs, 2, 354-385. Ross, E. A., Hollen, T. L., & Fitzgerald, B. M. (2006). Observational study of an Arts-inMedicine Program in an outpatient hemodialysis unit. American Journal of Kidney Diseases, 47(3), 462-468. Russell, C., Knowles, N., & Peace, L. (2007). Adherence in dialysis patients: A review of the literature. Journal of Nephrology Social Work, 27, 11-44. Russell, C., Whitlock, R., Knowles, N., Peace, L., Tanner, B., & Hong, B. (2008). Rates and correlates of therapy non-adherence in adult hemodialysis patients. Journal of Nephrology Social Work, 1, 11-17. Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The revised UCLA Loneliness Scale: concurrent and discriminant validity evidence. Journal of Personality and Social Psychology, 39(3), 472-480. 142 Russo, R., Manili, L., Tiraboschi, G., Amar, K., De Luca, M., Alberghini, E., et al. (2006). Patient re-training in peritoneal dialysis: why and when it is needed. Kidney International. Supplement, 103, S127-S132. Sabaté, E. (2003). Adherence to long-term therapies: evidence for action. World Health Organization, Geneva. Salive, M. E., Cornoni-Huntley, J., Phillips, C. L., Guralnik, J. M., Cohen, H. J., Ostfeld, A. M., & Wallace, R. B. (1992). Serum albumin in older persons: relationship with age and health status. Journal of Clinical Epidemiology, 45(3), 213-221. Sathvika, B. S., Naraharib, M. G., Gurudevb, K. C., & Parthasarathia, G. (2009). Impact of Clinical Pharmacist-Provided Education on Medication Adherence Behaviour in ESRD Patients on Haemodialysis. Iranian Journal of Pharmaceutical Sciences, 5(1), 21-30. Sayin, A., Mutluay, R., & Sindel, S. (2007). Quality of life in hemodialysis, peritoneal dialysis, and transplantation patients. Transplantation Proceedings, 39(10), 30473053. Schneider, M. S., Friend, R., Whitaker, P., & Wadhwa, N. K. (1991). Fluid noncompliance and symptomatology in end-stage renal disease: cognitive and emotional variables. Health Psychology, 10(3), 209-215. Schuz, B., Marx, C., Wurm, S., Warner, L. M., Ziegelmann, J. P., Schwarzer, R., & Tesch-Romer, C. (2011). Medication beliefs predict medication adherence in older adults with multiple illnesses. Journal of Psychosomatic Research, 70(2), 179-187. Sebire, S. J., Standage, M., & Vansteenkiste, M. (2009). Examining intrinsic versus extrinsic exercise goals: cognitive, affective, and behavioral outcomes. Journal of Sport and Exercise Psychology, 31(2), 189-210. Sehgal, A. R., Leon, J., & Soinski, J. A. (1998). Barriers to adequate protein nutrition among hemodialysis patients. Journal of Renal Nutrition, 8(4), 179-187. Selgas, R., Cirugeda, A., Fernandez-Perpen, A., Sanchez-Tomero, J. A., Barril, G., Alvarez, V., & Bajo, M. A. (2001). Comparisons of hemodialysis and CAPD in patients over 65 years of age: a meta-analysis. International Urology and Nephrology, 33(2), 259-264. Senol, V., Sipahioglu, M. H., Ozturk, A., Argun, M., & Utas, C. (2010). Important determinants of quality of life in a peritoneal dialysis population in Turkey. Renal Failure, 32(10), 1196-1201. Sensky, T., Leger, C., & Gilmour, S. (1996). Psychosocial and cognitive factors associated with adherence to dietary and fluid restriction regimens by people on chronic haemodialysis. Psychotherapy and Psychosomatics, 65(1), 36-42. 143 Setoguchi, S., Choudhry, N. K., Levin, R., Shrank, W. H., & Winkelmayer, W. C. (2010). Temporal trends in adherence to cardiovascular medications in elderly patients after hospitalization for heart failure. Clinical Pharmacology and Therapeutics, 88(4), 548-554. Sevick, M. A., Levine, D. W., Burkart, J. M., Rocco, M. V., Keith, J., & Cohen, S. J. (1999). Measurement of continuous ambulatory peritoneal dialysis prescription adherence using a novel approach. Peritoneal Dialysis International, 19(1), 23-30. Sewitch, M. J., Abrahamowicz, M., Barkun, A., Bitton, A., Wild, G. E., Cohen, A., & Dobkin, P. L. (2003). Patient nonadherence to medication in inflammatory bowel disease. American Journal of Gastroenterology, 98(7), 1535-1544. Sherman, R. A., Cody, R. P., Matera, J. J., Rogers, M. E., & Solanchick, J. C. (1994). Deficiencies in delivered hemodialysis therapy due to missed and shortened treatments. American Journal of Kidney Diseases, 24(6), 921-923. Shimoyama, S., Hirakawa, O., Yahiro, K., Mizumachi, T., Schreiner, A., & Kakuma, T. (2003). Health-related quality of life and caregiver burden among peritoneal dialysis patients and their family caregivers in Japan. Peritoneal Dialysis International, 23 (Suppl. 2), S200-S205. Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies: new procedures and recommendations. Psychological Methods, 7(4), 422-445. Singapore Health Promotion Board. (2010). Seventh Report of the Singapore Renal Registry, 2007/2008. Sim, K., Mahendran, R., Siris, S. G., Heckers, S., & Chong, S. A. (2004). Subjective quality of life in first episode schizophrenia spectrum disorders with comorbid depression. Psychiatry Research, 129(2), 141-147. Skevington, S. M., Lotfy, M., & O'Connell, K. A. (2004). The World Health Organization's WHOQOL-BREF quality of life assessment: psychometric properties and results of the international field trial. A report from the WHOQOL group. Quality of Life Research, 13(2), 299-310. Smith, K., Coston, M., Glock, K., Elasy, T. A., Wallston, K. A., Ikizler, T. A., & Cavanaugh, K. L. (2010). Patient perspectives on fluid management in chronic hemodialysis. Journal of Renal Nutrition, 20(5), 334-341. Soinski, J., Kelly, K., & Komaransky, Z. (1993). Determination of monthly food cost factor for a reference renal diet. Journal of Renal Nutrition, 3, 141-145. Song, M. K., Gilet, C. A., Lin, F. C., Machardy, N., Devitodabbs, A. J., Fine, J. P., et al. (2011). Characterizing daily life experience of patients on maintenance dialysis. Nephrology, Dialysis, Transplantation, 0, 1-7. 144 Spielberger, C. D. (1985). Assessment of state and trait anxiety: Conceptual and methodological issues. The Southern Psychologist, 2, 6-16. Stack, A. G. (2002). Determinants of modality selection among incident US dialysis patients: results from a national study. Journal of the American Society of Nephrology, 13(5), 1279-1287. Stack, R. J., Bundy, C. E., Elliott, R. A., New, J. P., Gibson, M., & Noyce, P. R. (2010). Intentional and unintentional non-adherence in community dwelling people with type 2 diabetes: the effect of varying numbers of medicines. The British Journal of Diabetes & Vascular Disease, 10(3), 148-152. Steele, T. E., Wuerth, D., Finkelstein, S., Juergensen, D., Juergensen, P., Kliger, A. S., & Finkelstein, F. O. (1996). Sexual experience of the chronic peritoneal dialysis patient. Journal of the American Society of Nephrology, 7(8), 1165-1168. Su, C. Y., Lu, X. H., Chen, W., & Wang, T. (2009). Promoting self-management improves the health status of patients having peritoneal dialysis. Journal of Advanced Nursing, 65(7), 1381-1389. Szeto, C. C., Chow, K. M., Kwan, B. C., Leung, C. B., Chung, K. Y., Law, M. C., & Li, P. K. (2005). Hypokalemia in Chinese peritoneal dialysis patients: prevalence and prognostic implication. American Journal of Kidney Diseases, 46(1), 128-135. Taal, E., Rasker, J. J., Seydel, E. R., & Wiegman, O. (1993). Health status, adherence with health recommendations, self-efficacy and social support in patients with rheumatoid arthritis. Patient Education and Counseling, 20(2-3), 63-76. Tang, S. C., & Lai, K. N. (2007). Does automated peritoneal dialysis provide better outcomes than continuous ambulatory peritoneal dialysis? Nature Clinical Practice. Nephrology, 3(11), 596-597. Tanner, J. L., Craig, C. B., Bartolucci, A. A., Allon, M., Fox, L. M., Geiger, B. F., & Wilson, N. P. (1998). The effect of a self-monitoring tool on self-efficacy, health beliefs, and adherence in patients receiving hemodialysis. Journal of Renal Nutrition, 8(4), 203-211. Taskapan, H., Ates, F., Kaya, B., Emul, M., Kaya, M., Taskapan, C., & Sahin, I. (2005). Psychiatric disorders and large interdialytic weight gain in patients on chronic haemodialysis. Nephrology (Carlton), 10(1), 15-20. Taylor, S. E. (1983). Adjustment to threatening events: A theory of cognitive adaptation. American Psychologist, 38(11), 1161-1173. Theofilou, P. (2011). Quality of life and mental health in hemodialysis and peritoneal dialysis patients: the role of health beliefs. International Urology and Nephrology, advance online publication. 145 Thompson, C., Kinmonth, A. L., Stevens, L., Peveler, R. C., Stevens, A., Ostler, K. J., et al. (2000). Effects of a clinical-practice guideline and practice-based education on detection and outcome of depression in primary care: Hampshire Depression Project randomised controlled trial. Lancet, 355(9199), 185-191. Thong, M. S., van Dijk, S., Noordzij, M., Boeschoten, E. W., Krediet, R. T., Dekker, F. W., & Kaptein, A. A. (2009). Symptom clusters in incident dialysis patients: associations with clinical variables and quality of life. Nephrology, Dialysis, Transplantation, 24(1), 225-230. Thumboo, J., Fong, K. Y., Machin, D., Chan, S. P., Leon, K. H., Feng, P. H., et al. (2001). A community-based study of scaling assumptions and construct validity of the English (UK) and Chinese (HK) SF-36 in Singapore. Quality of Life Research, 10(2), 175-188. Tornay, M. E., & Sanchez, A. (2003). Pediatric home peritoneal dialysis. A teaching programme. Revista de la Sociedad Espanola de Enfermeria Nefrologica, 6(4), 17-22. Triandis, H. C. (2001). Individualism-collectivism and personality. Journal of Personality, 69(6), 907-924. Troidle, L., Watnick, S., Wuerth, D. B., Gorban-Brennan, N., Kliger, A. S., & Finkelstein, F. O. (2003). Depression and its association with peritonitis in long-term peritoneal dialysis patients. American Journal of Kidney Diseases, 42(2), 350-354. Tsay, S. L. (2003). Self-efficacy training for patients with end-stage renal disease. Journal of Advanced Nursing, 43(4), 370-375. Tsevat, J., Dawson, N. V., Wu, A. W., Lynn, J., Soukup, J. R., Cook, E. F., et al. (1998). Health values of hospitalized patients 80 years or older. HELP Investigators. Hospitalized Elderly Longitudinal Project. JAMA, 279(5), 371-375. Tyrrell, J., Paturel, L., Cadec, B., Capezzali, E., & Poussin, G. (2005). Older patients undergoing dialysis treatment: cognitive functioning, depressive mood and healthrelated quality of life. Aging and Mental Health, 9(4), 374-379. United States Renal Data System (1997). The USRDS Dialysis Morbidity and Mortality Study: Wave 2. American Journal of Kidney Diseases, 30(suppl. 1), S67-S85. United States Renal Data System (2010). USRDS 2010 Annual Data Report. Retrived August 28, 2011, from http://www.usrds.org/adr.htm. Unni, E. J., & Farris, K. B. (2011). Unintentional non-adherence and belief in medicines in older adults. Patient Education and Counseling, 83(2), 265-268. 146 Unruh, M. L., Evans, I. V., Fink, N. E., Powe, N. R., & Meyer, K. B. (2005). Skipped treatments, markers of nutritional nonadherence, and survival among incident hemodialysis patients. American Journal of Kidney Diseases, 46(6), 1107-1116. Valderrabano, F., Jofre, R., & Lopez-Gomez, J. M. (2001). Quality of life in end-stage renal disease patients. American Journal of Kidney Diseases, 38(3), 443-464. van der Mei, S. F., van Sonderen, E. L., van Son, W. J., de Jong, P. E., Groothoff, J. W., & van den Heuvel, W. J. (2007). Social participation after successful kidney transplantation. Disability and Rehabilitation, 29(6), 473-483. Vasilieva, I. A. (2006). Quality of life in chronic hemodialysis patients in Russia. Hemodialysis International, 10(3), 274-278. Vathsala, A., & Chow, K. Y. (2009). Renal transplantation in Singapore. Annals of the Academy of Medicine, Singapore, 38(4), 291-299. Vazquez, I., Valderrabano, F., Fort, J., Jofre, R., Lopez-Gomez, J. M., Moreno, F., & Sanz-Guajardo, D. (2005). Psychosocial factors and health-related quality of life in hemodialysis patients. Quality of Life Research, 14(1), 179-190. Vermeire, E., Hearnshaw, H., Van Royen, P., & Denekens, J. (2001). Patient adherence to treatment: three decades of research. A comprehensive review. Journal of Clinical Pharmacy and Therapeutics, 26(5), 331-342. Vives, T., Pujolar, N., Junyent, E., Flores, I., Cordovilla, L., & Izquierdo, R. (1999). Adherence to treatment and personality in renal failure. EDTNA/ERCA Journal of Renal Care, 25(3), 13-14. Vlaminck, H., Maes, B., Jacobs, A., Reyntjens, S., & Evers, G. (2001). The dialysis diet and fluid non-adherence questionnaire: validity testing of a self-report instrument for clinical practice. Journal of Clinical Nursing, 10(5), 707-715. Walker, D. R., Chen, L., & Bhattacharyya, S. K. (2007). The impact of higher peritoneal dialysis utilization on Singapore's government expenditures, Abstract presented at the Third Asian Chapter Meeting of the International Society for Peritoneal dialysis, Hiroshima, Japan. Retrived from www.pdiconnect.com/cgi/reprint/27/Supplement_24/S62.pdf. Ware, J. E., Kosinski, M., Keller, S. D., & Institute, N. E. M. C. H. (1998). SF-12: How to score the SF-12 physical and mental health summary scales. The Health Institute, New England Medical Center. Warren, P. J., & Brandes, J. C. (1994). Compliance with the peritoneal dialysis prescription is poor. Journal of the American Society of Nephrology, 4(8), 16271629. 147 Wight, J. P., Edwards, L., Brazier, J., Walters, S., Payne, J. N., & Brown, C. B. (1998). The SF36 as an outcome measure of services for end stage renal failure. Quality in Health Care, 7(4), 209-221. Wolfe, R. A., Ashby, V. B., Milford, E. L., Ojo, A. O., Ettenger, R. E., Agodoa, L. Y., et al. (1999). Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. New England Journal of Medicine, 341(23), 1725-1730. Wong, F. K., Chow, S. K., & Chan, T. M. (2010). Evaluation of a nurse-led disease management programme for chronic kidney disease: a randomized controlled trial. International Journal of Nursing Studies, 47(3), 268-278. World Health Organization. (1996). WHOQOL-BREF Introduction, Administration, Scoring and Generic Version of the Assessment, Field Trial Version. Wuerth, D., Finkelstein, S. H., & Finkelstein, F. O. (2005). The identification and treatment of depression in patients maintained on dialysis. Seminars in Dialysis, 18(2), 142-146. Yap, K. B., Niti, M., & Ng, T. P. (2007). Nutrition screening among community-dwelling older adults in Singapore. Singapore Medical Journal, 48(10), 911-916. Zigmond, A. S., & Snaith, R. P. (1983). The hospital anxiety and depression scale. Acta Psychiatrica Scandinavica, 67(6), 361-370. Zrinyi, M., Juhasz, M., Balla, J., Katona, E., Ben, T., Kakuk, G., & Pall, D. (2003). Dietary self-efficacy: determinant of compliance behaviours and biochemical outcomes in haemodialysis patients. Nephrology, Dialysis, Transplantation, 18(9), 1869-1873. 148 Append dix A Permission n to Use Figgure 1 and Figure 2 Appeendix 1 149 150 151 Appendix B Participation Information Sheet Appendix B:Participation Information Sheet 1. Study Information Protocol Title: An evaluation of psychosocial and behavioural outcomes in patients on peritoneal dialysis regimes and their carers Principal Investigator & Contact Details: Dr. Marjorie Foo Wai Yin, Department of Renal Medicine, Singapore General Hospital, Tel: (65) 63214436 Email: marjorie.foo.w.y@sgh.com.sg 2. Purpose of the Research Study You are invited to participate in a research study. It is important to us that you first take time to read through and understand the information provided in this sheet. Nevertheless, before you take part in this research study, the study will be explained to you and you will be given the chance to ask questions. After you are properly satisfied that you understand this study, and that you wish to take part in the study, you must sign this informed consent form. You will be given a copy of this consent form to take home with you. You are invited because we are approaching patients with kidney disease who are receiving peritoneal dialysis (PD) under the care of the Renal Clinicians at Singapore General Hospital. Individuals who are over the age of 21 years, have been on PD for at least 90 days and are willing to participate are welcome. Patients judged by the senior staff members as too ill or likely to be distressed by taking part in the present study will be excluded from participating. This study is carried out to explore patients’ experience with their dialysis regime, diet and medication and to explore any difficulties they may be facing in their life and overall adjustment due to their condition and/or treatment. This study hopes to recruit 200 participants from Singapore General Hospital over a period of 24 months. 3. What procedures will be followed in this study The study will be conducted over a period of 24 months. You only need to participate once. If you decide to participate, you will need to sign a consent form and complete 8 questionnaires which assess treatment nonadherence, beliefs about medication, mood, quality of life, social isolation and sociodemographic domains. The questionnaires will either be completed independently by you (i.e. selfadministered) or administered by the research assistant in the form of a structured interview, according to your preference. This segment will span approximately 55 to 70 minutes, depending on your reading and response speed. In addition, your 152 medical records will be reviewed based on your consent. Medical data will also be retrieved from your medical records. These will include: Blood pressure ratings, dialysis adequacy delivery indices (Kt/V); albumin, hemoglobin, potassium, phosphate, and peritoneal dialysis regime/dose. Information regarding, primary cause of kidney failure and date that you started on peritoneal dialysis will also be recorded. 4. Your Responsibilities in This Study You will be required to complete 8 questionnaires asking questions about your treatment, mood and overall well-being. If you encounter difficulties in understanding the questions, you may seek assistance. Should you feel uncomfortable or unwell during the assessment process, kindly inform the investigator and the assessment will be terminated without delay. 5. Possible Risks and Side Effects We do not anticipate that answering the questionnaires entails any risk or are likely to cause any discomfort or distress to the participants. The questionnaires are selected on the basis of good psychological principles and are non-invasive. It is therefore not expected that the questionnaires would cause any harm, risk or discomfort to the respondent. 6. Possible Benefits from Participating in the Study Most participants enjoy the opportunity to express their opinions about their experience and reflect on their behaviors and/or difficulties in managing the demands of illness and treatment. We hope and anticipate that patients will find sharing their experience useful. We also hope that this will also benefit future patients and carers. Your feedback and the knowledge gained from this study will help us evaluate the care delivered and if needed design a support intervention program for PD patients and carers to assist them in better manage their condition and improve their well-being. By gaining a better insight and understanding in patients and carers’ experiences, we will be able to develop a program specific to patients and carers’ needs in the local context. 7. Costs & Payments if Participating in the Study Participation by both you and your carer will be collectively rewarded by a token fee of SGD 20. This will be covered by MOE ACRF start up research grant to Dr Konstadina Griva (Co- Investigator) and will be administered by the research assistants upon completion of questionnaires. 8. Voluntary Participation Your participation in this study is voluntary. You may stop participating in this study at any time. Your decision not to take part in this study or to stop your participation will not affect your medical care or any benefits to which you are entitled. If you decide to stop taking part in this study, you could do this without hesitation. 153 Your doctor, the Investigator and/or the Sponsor of this study may stop your participation in the study at any time if they decide that it is in your best interests. They may also do this if you do not follow instructions required to complete the study adequately. If you have other medical problems or side effects, the doctor and/or nurse will decide if you may continue in the research study. In the event of any new information becoming available that may be relevant to your willingness to continue in this study, you (or your legally acceptable representative, if relevant) will be informed in a timely manner by the Principal Investigator or his/her representative. 9. Compensation for Injury The questionnaires have been carefully chosen to be completely non-invasive – none of the questionnaires is likely to pose any risk of harm or injury to you. As such there are no compensation arrangements in place. 10. Confidentiality of Study and Medical Records Information collected for this study will be kept confidential. Your records, to the extent of the applicable laws and regulations, will not be made publicly available. However, Singapore Health Services (SHS) Centralised Institutional Review Board (CIRB) and Ministry of Health will be granted direct access to your original medical records to check study procedures and data, without making any of your information public. By signing the Informed Consent Form, you (or your legally acceptable representative, if relevant) are authorizing such access to your study and medical records. Data collected and entered into the Case Report Forms are the property of Singapore General Hospital and National University of Singapore. In the event of any publication regarding this study, your identity will remain confidential. 11. Who To Contact if You Have Questions If you have questions about this research study or there are any injuries sustained during the course of this study, you may contact, Dr. Marjorie Foo Wai Yin, Principal Investigator Department of Renal Medicine, Singapore General Hospital, Tel: (65) 63214436 Email: marjorie.foo.w.y@sgh.com.sg Dr Griva Konstadina, Co-Investigator Department of Psychology, National University of Singapore Tel: (65) 65163561 Email: psygk@nus.edu.sg The study has been reviewed by the SHS CIRB (the central ethics committee) for 154 ethics approval. If you want an independent opinion of your rights as a research subject you may contact the SingHealth CIRB Secretariat at 6323-7515. If you have any complaints about this research study, you may contact the Principal Investigator or the SingHealth CIRB Secretariat. 155 Appendix C Consent Form Appendix C: Consent Form Protocol Title: An evaluation of psychosocial and behavioural outcomes in patients on peritoneal dialysis regimes and their carers Principal Investigator & Contact Details: Dr. Marjorie Foo Wai Yin, Principal Investigator Department of Renal Medicine, Singapore General Hospital, Tel: (65) 63214436 Email: marjorie.foo.w.y@sgh.com.sg I voluntarily consent to take part in this research study. I have fully discussed and understood the purpose and procedures of this study. This study has been explained to me in a language that I understand. I have been given enough time to ask any questions that I have about the study, and all my questions have been answered to my satisfaction. _______________________ ____________________ Name of Participant Signature Date Witness Statement I, the undersigned, certify to the best of my knowledge that the participant signing this informed consent form had the study fully explained in a language understood by him / her and clearly understands the nature, risks and benefits of his / her participation in the study. _______________________ ____________________ Name of Witness Signature Date Investigator Statement I, the undersigned, certify that I explained the study to the participant and to the best of my knowledge the participant signing this informed consent form clearly understands the nature, risks and benefits of her participation in the study. _______________________ _______________________ Name of Investigator / Signature Person administering consent _ ____________________ Date 156 Appendix D Demographics Questionnaire Appendix D: Patient Demographic Information [1] What is your date of birth? (please write in MM/YYYY) __________________ [2] What is your first language? (please write in) _______ [3] How would you describe your ethnic background? 1 3 5 7 [4] Chinese 2 Indian 4 White 6 Do not wish to answer ____________ Malay Asian Other Other How would you describe your relationship status? 1 Married 2 Divorced 3 Widowed 4 Single 5 Living with partner 6 Other [5] What is your highest educational qualification? (please write in) ________________________ [6] Which of the following responses best characterises your current work activity or employment status? 1 2 3 4 employed full-time 5 employed part-time 6 self-employed 7 unemployed, laid off involuntarily 8 retired looking after home and family student other/none of the above [7] What approximately is the current estimated monthly income of your overall family? (Please remember your answers are confidential) 1 2 3 4. 5. 6. $ 0 -$ 2,000 $ 2,001 - $ 4,000 $ 4,001 - $ 6,000 $ 6,001 – above don’t know don’t wish to answer 157 [8] Which of the following best describes your housing: 1 2 3 4 5 6 [9] How long have you been on peritoneal dialysis? _______________________ 1 2 3 4   1-2 HDB flat 3-4 HDB flat HDB 5 room/executive/maisonette Condominium, including executive condominium and private apartment Terrace/ Semi – Detached / Bungalow other (please specify__________________________________) Less than 1 month 6 to 12 months 13 to 24 months More than 24 months (2 years)   158 Appendix E Medical Form Appendix E: Patient Medical Information I. Comorbidities (please tick any that apply and list any additional) Yes No Bone Disease Hypertension Diabetes (without end of organ damage) Diabetes (with end of organ damage) Myocardial infarction Congestive Heart Failure Peripheral Vascular disease Cerebrovascular disease Connective tissue disease Dementia Hemiplegia Leukemia Malignant lymphoma Malignant solid tumor in the last 5 years (metastatic) AIDS Peptic Ulcer Disease Liver disease (mild) Liver disease (moderate) Liver disease (severe) Renal disease (mild) Renal disease (moderate) Renal disease (severe) Please list any additional comorbid conditions_____________________________ ____________________________________________________________________ 159 II. Treatment Details Primary kidney disease diagnosis CAPD Type of current dialysis modality (Tick the appropriate column) APD III. Biochemical data (most recent 2 values recorded) Date1 ____________________ Time1 Date2 ____________________ Time2 Time1 Creatinine Potassium Phosphate Albumin Hemoglobin Kt/V Time2 IV. Medication details (oral and injectable) – please list prescribed medication (inc EPO) _____________________________________________________________________ _____________________________________________________________________ _____________________________________________________________________ _____________________________________________________________________ _______________________________________________________ 160 Appendix F Perrmission too use Table 2 Appeendix 6 161 162 Append dix G Perrmission to Use Table 3 Equaation 2 163 164 Appendix H Nonadherence Measures Appendix G: Nonadherence Measures QUESTIONS ABOUT PD EXCHANGES How often do you follow this regime? 1 Very often 2 3 Often Sometimes 4 5 Rarely Never 4 5 Rarely Never 4 5 Rarely Never Some people forget to perform PD exchanges/ session. Overall, how often does this happen to 1 2 3 Very often Often Sometimes you? Some say that they decide to skip one of their PD exchanges/session or adjust the length to suit their own needs. Overall, how often do 1 2 3 Very often Often Sometimes you do this? During the last 4 weeks how many time have skipped/missed one of your PD sessions/exchanges? _____________________________________________________ During the past 4 weeks how many times have you prolonged/shorten your PD session by 10 minutes ?____________________________________________________   QUESTIONS ABOUT TAKING YOUR MEDICINES How often do you follow this regime? Some people forget to take their medicines. Overall, how often does this happen to you? 1 2 3 Very often Often Sometimes 1 2 3 Very often Often Sometimes 4 5 Rarely Never 4 5 Rarely Never 4 5 Rarely Never Some people decide to miss out a dose of their medication or adjust it to suit their own needs. Overall, 1 2 3 Very often Often Sometimes how often do you do this? 165 QUESTIONS ABOUT DIET Have you been asked to follow a diet? How often do you follow these dietary recommendations? Some people forget to follow dietary recommendations. Overall, how often does this happen to you? 2 NO (ignore questions below) 1 YES 1 2 3 Very often Often Sometimes 1 2 3 Very often Often Sometimes Some people decide to adjust their 1 2 3 diet to suit their own needs. Overall, how often do you do Very often Often Sometimes this? 4 5 Rarely Never 4 5 Rarely Never 4 5 Rarely Never 166 Appen ndix I Reseearch Ethn nics Approvval Appendix H: Research Ethnics Approval A A 167 168 169 [...]... and intervening accordingly may be more cost-effective Two broad categories have been proposed to classify causes of nonadherence: intentional and unintentional (Clifford, Barber, & Horne, 2008) This intentional- unintentional typology is predominantly used in investigating medication-taking behaviors (Daleboudt, Broadbent, McQueen, & Kaptein, 2010; Unni & Farris, 2011) Both forms of nonadherence. .. distinguished intentional and unintentional nonadherence Nonadherence on APD vs CAPD Only one known study compared nonadherence difference between APD and CAPD patients (Bernardini et al., 2000) Home visit supply inventories were used to evaluate nonadherence to dialysis exchanges in this study and PD modality was identified as an independent predictor of nonadherence, with more nonadherence reported in. .. roles in determining intentional and unintentional nonadherence Concern beliefs have been reported to affect both intentional and unintentional nonadherence (Daleboudt et al., 2010; Unni & Farris, 2011), whereas necessity beliefs were mainly associated with intentional nonadherence (Schuz et al., 2011; Unni & Farris, 2011) There is ample evidence suggesting that self-efficacy is correlated with nonadherence. .. unintentional nonadherence despite their important implications for intervention Determinants of Nonadherence Previous studies on determinants of nonadherence tend to focus on demographical and clinical variables, overlooking the effects of psychosocial variables (Karamanidou, Clatworthy, et al., 2008; Russell, Knowles, & Peace, 2007) However, identifying demographical and clinical associates of nonadherence. .. users is rising steadily in recent years in many parts of the world, like Canada, US, Singapore and Switzerland (Blake, 1999; Dell'Aquila, Berlingo, Pellanda, & Contestabile, 2009; Mehrotra, 2009) In Singapore, incident rate of PD patients choosing APD had increased from 3% in 1999 to 50.5% in 2008 (SRR, 2010) Patients are motivated to choose APD mainly due to the autonomy it provides instead of medical... 2006) Nonadherence with salt and fluid was a critical reason for PD drop-out (Kawaguchi et al., 2003) Intentional and Unintentional Nonadherence Increasing knowledge is a standard way employed to reduce nonadherence in intervention programmes However, nonadherence is problematic even among those with good knowledge (Lee & Molassiotis, 2002; Nerbass et al., 2010) Clarifying causes of nonadherence and intervening... dialysis, probably in the same order of importance as medical indicators (Bander & Walters, 1998) Dialysis regimen is extremely complicated and time-consuming, involving regular clinical visits, attending dialysis sessions, taking a variety of medications, limiting water intake and paying great attention to food choices As treatment complexity has been cited as the most important reason affecting patients ... study in another type of chronic illness identified forgetfulness, accidentally overdose and the unavailability of medication as reasons for unintentional nonadherence and intentional nonadherence was mainly caused by side effects, social activities, eating out, drinking alcohol or traveling (Eliasson, Clifford, Barber, & Marin, 2011) To the best of our knowledge, no studies concerning ESRD patients. .. are ambulatory during this period) absorbing wastes, toxins and excess water from the blood and then is drained out of the body together with the wastes Then the infusion process begins again and the procedure repeats The draining and infusion process is called an exchange, taking 30 to 60 minutes depending on the patient’s health status These repeated exchanges can be performed either manually by patients. .. patients nonadherence (Donovan, 1995), it is not unexpected that it is easy for dialysis patients to be nonadherent Nonadherence studies in dialysis patients are greatly hindered by a lack of consistent standards for measuring nonadherence Common measures of nonadherence include: (a) report from patients or medical staff, (b) biological and biochemical markers, (c) electronic monitoring and (d) checking ... dietary guidelines the most difficult to adhere to Intentional nonadherence occurred more frequently than unintentional nonadherence for dialysis and diet; intentional and unintentional nonadherence. .. quality of life was the strongest predictor of unintentional nonadherence to dialysis Self-efficacy was the strongest predictor of intentional and unintentional nonadherence to both medication and. .. Clarifying causes of nonadherence and intervening accordingly may be more cost-effective Two broad categories have been proposed to classify causes of nonadherence: intentional and unintentional