Eric R. Neumaier, M.A. A dissertation submitted in partial fulfillment of the requirements for the degree of. Doctor of Philosophy

Transcription

1 Effects of an Integrated Behavioral Health Intervention on the Health Outcomes and Quality of Life of Patients with Kidney Stone Disease By Eric R. Neumaier, M.A. A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Counseling Psychology) at the UNIVERSITY OF WISCONSIN-MADISON 2012 Date of final oral examination: 08/16/12 The dissertation is approved by the following members of the Final Oral Committee: William T. Hoyt, Ph.D., Professor & Advisor, Counseling Psychology Kristina Penniston, Ph.D., RD, Associate Scientist, Urology, UW School of Medicine & Public Health Lori L. DuBenske, Ph.D., Psychologist, Psychiatry, UW School of Medicine & Public Health Takuya Minami, Ph.D., Associate Professor, Counseling Psychology Alberta M. Gloria, Ph.D., Professor, Counseling Psychology Daniel M. Bolt, Ph.D., Professor, Educational Psychology

2 i TABLE OF CONTENTS Acknowledgements v. Abstract vi. Chapter I. Introduction 1 Chapter II. Literature Review 9 Section I. Kidney Stone Disease (Urolithiasis) 9 A. Definition, Etiology, and Risk Factors of Stones 9-19 B. Prevalence Rates by Race, Sex, Age, & SES C. Symptoms 23 D. Quality of Life Effects 24 E. Psychological Effects 27 F. Diagnosis 32 G. Medical Treatment 33 H. Cost of Urolithiasis 37 I. Prevention 40 J. Chapter I Summary 44 Section II. The Primary Care Behavioral Health Model 45 A. History of Integrated Care 46 B. Definition and Description 48 C. Benefits of Integrated Health Care 52 D. Barriers to Integration 53 E. Common Behavioral Interventions 54 F. PCBH in a Secondary Care Clinic 58 Chapter III. Method 63 A. Participants 63 B. Design 63 C. Specific Aims & Hypotheses 63 D. Materials 66 E. Written Measures 68 F. Procedure 77 a) Sampling procedures 77 b) Random assignment 78 c) Control group 79 d) Treatment group 79 e) Integrated behavioral health invention 79 f) Consultation 81

3 ii g) Follow-up calls 85 h) Mental health crises 85 i) Documentation 86 j) BHC training 87 k) Supervision 88 l) Incentives 88 Chapter IV. Results 89 A. Patient Demographics 89 a) Intervention Goals 95 b) Follow-up Phone Calls 96 B. Provider Attitudes Towards Psychological Collaboration 97 a) Hypothesis I 97 C. Patient Attitudes Towards Psychological Intervention and Its Effects 99 on QOL a) Hypothesis II 99 b) Hypothesis III 105 c) Hypothesis IV 108 Chapter V. Discussion 116 A. Introduction 116 B. The Primary Care Behavioral Health Model 118 C. Generalizability of Results 119 D. Hypothesis I 122 E. Hypothesis II 124 F. Hypothesis III 126 G. Hypothesis IV 127 H. Limitations 132 I. Suggestions for Future Research 133 J. Conclusion 138 References 140 Appendix A: Provider Measures 149 Appendix B: Patient Measures 153 Appendix C: Sample Patient Handouts 171

4 iii TABLE OF TABLES Table 3.1 Procedure 66 Table 4.1 Patient Demographics (Categorical Variables), Chi- 92 Square, and Odds Ratio Results Table 4.2 Descriptive Statistics and Effect Size Results for Patient 94 Continuous Variables Table 4.3 Descriptive Statistics for Patient Intervention Goals 96 Table 4.4 Descriptive and Inferential Statistics for the DACC-MH 98 and Provider Attitudes Table 4.5 Pre- and Post-Intervention Comparisons for Combined 101 Continuous Variables for Completers Table 4.6 Summary of Intercorrelations for Continuous Variables for 102 Completers Table 4.7 Descriptive and Inferential Statistics for Biological 107 (Lithogenic) Risk Factors Table 4.8 Summary of Scale Intercorrelations, Descriptive Statistics, 109 and Test-Retest Reliability Results for Completers Table 4.9 Pre- and Post-Intervention Inferential and Descriptive 110 Statistics for QOL Scales Table 4.10 Summary of Intercorrelations for Residual Change (Gain) Scores for Completers 115

5 iv TABLE OF FIGURES Figure 2.1 The 5A s Model of Behavior Change in Primary Care 51 Figure 3.1 Procedure 81 Figure 3.2 Structure of the Initial Consultation Appointment Linked 83 with the 5A s Figure 4.1 CONSORT Diagram 91

6 v Acknowledgements I wish to send my sincerest thanks to Drs. Kris Penniston, Bill Hoyt, Lori DuBenske, and Steven Nakada for their tireless support and guidance on this project. I thank Drs. Kim Howard, Takuya Minami, Alberta Gloria, and Daniel Bolt for serving on the Committee and sharing their wisdom to improve this project. I thank Katrina Wojciechowski and Bridget Stroup for doing so much of the behind-the-scenes work. I also wish to thank Aaron Smith and Brian Padilla for selflessly agreeing to help me in my time of need, Dr. Glen Leverson for his initial statistical assistance, Dr. Heidi Beckman for opening me up to the field of health psychology, the UW Department of Urology for allowing a stranger into their Department, and my fellow stone patients who agreed to participate in this study. I hope this study is a small step in the direction of improving the lives of all those who suffer from this chronic condition. Finally, I wish to thank the UW Department of Counseling Psychology for the excellent training I have received and all of my family and friends who supported me. This project would not have been possible without you. Thank you again.

7 vi Abstract Urolithiasis (urinary tract or kidney stones) is a common problem with increasing prevalence. Due to high recurrence rates, it is considered a chronic disease and there is no cure. Stone patients report lower physical and psychological quality of life (QOL) than healthy adults. In addition, prevention strategies require patients to make significant behavioral and lifestyle changes. Due to stagnation in medical management, there is a need for new interventions, especially those designed to address QOL decrements. To our knowledge, the present study is the first to adapt aspects of the Primary Care Behavioral Health (PCBH) model of integrated medical-psychological care to a secondary care setting (Urology Stone Clinic) and to a sample of recurrent stone formers. Thus, the purpose of this study was to investigate the feasibility of implementing the PCBH model with patients with urolithiasis to improve patient health outcomes and QOL. One hundred and two (N = 102) recurrent stone formers were randomized into either a treatment-as-usual (TAU; e.g., labs, imaging data, prevention recommendations) control group or treatment group consisting of TAU plus the behavioral health intervention (one in-person meeting and two follow-up phone calls). It was hypothesized that the PCBH model would be successfully adapted to patients with urolithiasis and positively viewed, and that the intervention would decrease patients biological risk factors and improve patients health-related QOL. In addition, six urological providers (N = 6) participated and it was hypothesized that psychological collaboration would be positively viewed and increase psychological-mindedness. The results of all inferential hypothesis tests were non-significant. Various descriptive statistics and an interpretation of effect sizes provided some support for the feasibility of adapting the PCBH model to secondary care. The lack of significant findings was likely due in part to a small sample size and low power, although it may also have been a reflection of the decision to only sample recurrent stone formers. Directions for future research are discussed.

8 1 Chapter I. Introduction One in ten Americans experience limited mobility due to a chronic, disabling condition and seven in ten Americans die each year as a result of a chronic medical condition (Strosahl & Robinson, 2008). The most recognizable chronic health conditions in the U.S. include diabetes, hypertension, hyperlipidemia (high cholesterol), obesity, HIV/AIDS, chronic obstructive pulmonary disease (COPD), and depression. These conditions receive a great deal of media attention, public awareness, and research funding, as they are rightly recognized as important public health concerns. However, there are many other chronic health conditions that receive far less public attention and resources. One of these conditions, urolithiasis, or kidney stones, is overlooked. The average person likely has an understanding of what a kidney stone is, although they also likely view it as a male-dominated condition that, while often painful, is only a temporary and minor inconvenience. This stigma might be in part to blame for the few dedicated funding resources stone disease has received. It might also help explain the almost total lack of psychological literature on the effects of stone disease while there is a wealth of information on conditions such as breast cancer and HIV/AIDS. The lack of focus on stone disease becomes all the more surprising when one learns that stone disease is a chronic, not acute, condition that is increasing in prevalence, increasingly affecting women and younger adults, is very costly, and associated with significant mental health and quality of life (QOL) decrements. Kidney stones, simply put, are small, hard formations of minerals and acid salts (MayoClinic.com). There are five major types (Goldfarb, 2009; MayoClinic.com), although around 80% of all stones are comprised of calcium oxalate (Goldfarb). They form when the urine

9 2 become supersaturated with stone forming minerals and salts (MayoClinic.com). This forces the kidneys to filter an excess of these minerals where they bind to form stones (Goldfarb). Stones form due to a variety of genetic and environmental reasons. A little over half of the etiology is attributed to genetic factors while a sizeable portion is attributed to behavioral and lifestyle factors such as diet and exercise, water intake, stress, and a history of previous bariatric surgery (Goldfarb, 2009; MayoClinic.com). After a person forms a kidney stone, her/his likelihood of recurrence is 50-70% (Lotan, 2009). It is estimated that lifetime prevalence rates range from 10-13% for males to 5-7% for females (Goldfarb, 2009; Pearle, Calhoun, & Curhan, 2007). Stone disease typically affects people in the prime of their working years, significantly impacting both the patient and society (Saigal et al., 2005). Kidney stones are the second costliest urological disease, second only to urinary tract infections (Bergman, Gore, Singer, & Anger, 2010). The total annual expenditure surpasses prostate, bladder, and kidney cancers, urinary incontinence, and even erectile dysfunction (Bergman et al.). In 2000 alone, it is estimated that kidney stones cost as much as $5.3 billion to evaluate and treat in the U.S (Saigal et al., 2005). There are several reasons for this high, and increasing, cost. First, a key reason is that prevalence rates continue to rise. In the U.S., the adult prevalence rate has increased 37% in the past thirty years (Stamatelou, Francis, Jones, Nyberg, & Curhan, 2003). Caucasians are uniquely affected, having prevalence rates 2-3 times that of other racial groups, although the reasons for this are unclear (Stamatelou et al.). While this disease still primarily affects males (Scales et al., 2007; Stamatelou et al., 2003; Strope, Wolf, & Hollenbeck, 2010), women have become the fastest growing group diagnosed. For example, the inpatient hospital discharge rate for women increased 22% between the years while there was

10 3 no significant increase for men (Scales et al.). Second, the costs of treatments have continued to rise, despite the stagnation of medical care. Treatment has changed little in the last 25 years with no new technological breakthroughs expected (Lotan, 2009). Treatments, although partially effective in treating stones, do nothing to prevent future stones. Thus, there is no cure for stone disease. There is also a concern that the medical treatment of stone disease is actually regressing. Kerbl et al. (2002) voice concern that the most recent technologies are actually less effective than earlier technologies and that treatments have become more, instead of less, invasive. Third, there are significant indirect costs of stone disease. These costs can be broken down into two main categories, lost wages and productivity and psychosocial costs. One-third of employees with stone disease miss some work, missing an average of 19 hours per person (Saigal et al., 2005). Conservatively, this comes to 3.1 million lost workdays per year worth at least $775 million (in 2000 dollars) in lost productivity for insured employees (Saigal et al.). Although there is little research on the QOL and psychological decrements related to stone disease, depression is widely considered the most prevalent mental health issue faced by patients with chronic kidney disease, of which kidney stone disease is a part (Hoth, Christensen, Ehlers, Raichle, & Lawton, 2007). However, it is unclear to what extent this can be extrapolated to stone formers given the more medically serious nature of chronic kidney disease. Depression can greatly impact a patient s ability to adhere to treatment and prevention guidelines, increase suffering and the utilization of health care, decrease a return to full physical functioning and work, and may even increase the likelihood of stone recurrences (Diniz et al., 2007b). Therefore, there is little surprise that comorbid mental health concerns like depression can substantially increase both the direct and indirect costs of treatment (Greenberg et al., 2003). It is estimated that total annual expenditure

11 4 in treating depression was $70 billion in 2003 and an employee treated at least once for major depression cost employers 3.5 times that of the average employee (Greenberg et al.). Thus, QOL and psychological decrements are an important component of the indirect costs related to stone disease and are worthy of increased attention. As health care costs continue to skyrocket at a disproportionate rate and prevalence rates continue to rise, there is a great need to understand the basis of kidney stone costs and to explore cost-reduction options (Lotan). The QOL and psychological decrements related to stone disease have been found to consist of more than depression. Stone patients report lower physical and mental QOL than healthy adults (Penniston & Nakada, 2007). Women stone formers, in particular, appear to experience more decrements in QOL than their male counterparts. The number of comorbidities, number of painful episodes, and time since diagnosis and last stone event may also negatively impact QOL. Stone patients typically experience more anxiety and symptoms of depression than their healthy peers. Other psychosocial factors, such as stress, also have been shown to affect physical health and appear amenable to psychological interventions. Psychosocial decrements, in turn, can affect a patient s ability and motivation to adhere to complex treatment regimes (Diniz et al., 2007b). These findings suggest possible ways stone disease can affect patients mental health, the need to identify barriers to treating stone disease, and the need for more individualized and holistic treatments. These psychological barriers are prevalent in the primary care setting where the majority of Americans receive their health care. Every year, the rate of onset for mental and addictive disorders is believed to be around 27% (Strosahl & Robinson, 2008) and behavioral health conditions account for 15% of the overall disease burden in the U.S. (Murray & Lopez, 1996, as

12 5 cited in Hunter et al., 2009). It is estimated that 70% of all primary care visits are behavioralhealth related (Fries, Koop, & Beadle, 1993, as cited in Hunter et al.) and over 80% of all psychotropic medications are prescribed by nonpsychiatric providers (Beardsley, Gardocki, Larson, & Hidalgo, 1988, as cited in Hunter et al.). In an early investigation, Kroenke and Mangelsdorff (1989) followed 1,000 patients in a large primary care practice with the ten symptoms that comprise 40% of all of physician visits: chest, back, or abdominal pain, fatigue, dizziness, dyspnea, insomnia, headaches, edema, and numbness. After three years, only 16% of patients were found to have a biological cause for their symptoms and 10% of symptoms were suspected to have a psychological cause. The etiology of three-fourths of all symptoms remained unknown. Of those symptoms suspected to be psychological, depression, stress, anxiety, and grief were believed the most likely underlying issues. Two-thirds of patients underwent diagnostic testing or received referrals in an effort to find a biological cause for their symptoms, despite the aforementioned low success rate. The cost of diagnostic testing was high, especially for those reporting symptoms of headache and back pain. In the end, only 55% of patients with symptoms received any treatment, with pharmacological treatment outnumbering all others three to one. Perhaps unsurprisingly, symptoms frequently persisted and only half of symptoms saw any improvement. However, the authors found three variables that predicted symptom improvement: a biological etiology, symptom duration lasting less than four months, and having two or fewer symptoms. Thus, as patients report symptoms of a psychological nature, and it would be expected that a portion of the unknown etiologies also had a psychological nature, primary care physicians need training and assistance in treating these patients. However, medical school primarily trains physicians to diagnose and treat the 16% of patients that are found to

13 6 have a biological cause, not the majority that who present with symptoms of unknown etiology (Katon, 1995). Ignoring the potential impact of behavioral and mental health factors which may be underlining these unknown etiologies puts patients at risk of iatrogenic harm, or the risk of receiving unnecessary medical care that can be both costly and dangerous (Katon). It should be unsurprising then that primary care has been called the de facto mental health system in the U.S. (Blount & Miller, 2009; Katon, 1995). The utilization of primary care services has been found to improve health outcomes and lower health costs (Blount & Miller). The majority of mental health needs in this country are met solely in the primary care setting for numerous reasons (Blount & Miller; Strosahl & Robinson, 2008). These include the stigma of specialty mental health services, lack of access to mental health services, financial concerns, a belief that therapy is not or will not be effective, and long waits for mental health services (Blount & Miller; Strosahl & Robinson, 2008). To address these barriers, an integrated behavioral health model, the PCBH model, has been developed. The PCBH model of patient care combines physical and mental health needs into one model, facilitating direct collaboration between psychologists, called behavioral health consultants (BHCs), and medical professionals (Hunter et al., 2009). As five of the leading seven causes of death in the U.S. are partly related to unhealthy behaviors and stress (Blount & Miller, 2009), psychologists are uniquely positioned to improve America s healthcare. This model is considered the best opportunity to reach the greatest number and variety of patients with evidence-based behavioral health assessments and interventions (Hunter et al., p. 4). Psychologists utilize their expertise in mental health assessment, diagnosis, and treatment interventions to enact change consistent with medical recommendations. Targeting health-related

14 7 behaviors has also been suggested as one approach to addressing health disparities, as those of lower SES and racial/ethnic minorities are disproportionately affected by preventable risk factors (Whitlock, Orleans, Pender, & Allan, 2002). Previous research has demonstrated that the application of this model to primary care can lead to improved patient and provider satisfaction, improved health outcomes, improved patient quality of life, and significantly decreased health care costs. Despite the success of this model for primary care, there are few examples of its application to secondary care and it has never been applied to a population of kidney stone formers. Therefore, it was a goal of this study to investigate the feasibility of adapting the PCBH model to the needs of a secondary care clinic. In the proposed study, a total of one hundred and five patients with recurrent stone disease were randomized into two conditions, a treatment-asusual control group and a treatment group. The treatment group received a personalized behavioral health intervention guided by the consultation goals of the treating medical provider. For those in the treatment group, the behavioral health consultant (BHC) met with the patient s medical provider to establish consultation goals prior to the intervention. After forming consultation goals, the BHC met with the patient for between minutes to address these goals. The format of each behavioral health session was personalized but all sessions followed the five steps of the PCBH model: assess, advise, agree, assist, and arrange. These patients then received two follow-up phone calls, the first one-month post intervention and a second call at three-months post intervention. The purpose of the phone calls was to inquire into the patient s health, assess the patient s adherence to treatment/consultation recommendations, to provide relevant information or resources, and to discuss further recommendations. In order to assess

15 8 psychological and QOL outcomes, all patients completed a battery of paper-and-pencil measures pre- and post intervention. In order to assess biological risk factors as an outcome, the results of patients 24-hour urine analyses were compared on lithogenic risk factors pre- and post intervention. In addition, medical providers were surveyed pre- and post intervention regarding their attitudes towards consultation and managing mental health issues. This study had four main hypotheses with details presented in the Method section. First, psychological collaboration would be positively viewed by providers and increase their psychological-mindedness. Second, the PCBH model would be successfully adapted to patients with kidney stones and positively viewed. Third, the BHC intervention would improve patients biological risk factors for urolithiasis compared to treatment-as-usual. Fourth, the intervention would improve patients health-related QOL compared to treatment-as-usual. It was hypothesized that this intervention would lead to improved patient and provider satisfaction and improved health outcomes. It was hoped that by demonstrating the effectiveness of this model for kidney stone patients it may one day lead to this model s adaptation not only at the UW Hospital s Kidney Stone Clinic, but also at similar clinics across the country. With the increased incidence and prevalence of kidney stones, rising health care costs, and the stagnation of new medical management strategies, the need to develop improved interventions remains urgent.

16 9 Chapter II. Literature Review Section I. Kidney Stone Disease (Urolithiasis) Definition, etiology, and risk factors of stones. Definition. Urolithiasis (also called nephrolithiasis or kidney stones) is a condition in which small, hard formations of minerals and acid salts form (precipitate) in the urine (MayoClinic.com). The term urolithiasis applies to the formation of a stone anywhere in the urinary tract, such as the kidneys and bladder (Pearle et al., 2007). There are five main types of stones: calcium oxalate, calcium phosphate, uric acid, struvite, and cystine (Goldfarb, 2009; MayoClinic.com). Calcium oxalate stones are by far the most common, comprising around 80% of all stones (Goldfarb), although stones usually contain crystals of more than one type (MayoClinic.com). Uric acid stones account for 10% of stone case and are more common in patients with obesity and diabetes (Goldfarb). Cystine stones, originating from genetic metabolism abnormalities, comprise 1% of stone cases (Goldfarb). Struvite stones are most commonly seen in women (Goldfarb). Other, less common stones are composed of xanthine or 2,8-dihydroxyadenine (Pearle et al., 2007). Etiology. The pathophysiological basis (etiology) differs by stone type and location, with accompanying differences in risk factors (Pearle et al., 2007). West et al. (2008) argue that stone disease should be viewed as a systemic disease, as it represents the interaction of multiple risk factors (p. 746). Upper tract (kidney and ureter) stones are typically composed of calcium oxalate, calcium phosphate, uric acid, struvite, or cystine (Pearle et al.). Lower tract (bladder) stones are typically composed of uric acid or calcium phosphate (Pearle et al.). Less common stones may be medication-induced, such as indinavir [HIV medication] and ephedrine (Pearle et

17 10 al.). Scales et al. (2007), in their review of the Nationwide Inpatient Sample (NIS) data, found 69% of inpatients with stone disease were admitted for ureteral calculi (kidney stones of the ureter). Upper and lower tract stones have considerably different etiologies. Kidney stones form because of physiochemical or genetic abnormalities (e.g., metabolic disorders), causing the urine to become supersaturated with stone-forming salts, or in the case of struvite stones (MayoClinic.com), from a urinary tract infection with urease-producing bacteria (Goldfarb, 2009; Pearle et al.). Struvite stones in particular are known to be fast growing and large (MayoClinic.com). When the urine becomes supersaturated with these salts, the minerals begin to crystallize and form stones (MayoClinic.com). Similar to the etiology of many disorders, both physical and psychological, the causes can be viewed as the complex interaction of genetic and environmental factors. Genetic factors and blood disorders are one primary underlining cause (MayoClinic.com) and genetic factors are estimated to account for 56% of the stone phenotype (Goldfarb, 2009). For example, rare cystine stones are caused by a hereditary disorder leading to cystinuria, or excessive excretion of amino acids by the kidneys (MayoClinic.com). Other forms of stones are also viewed as having a hereditability component, as over 50% of stone patients have a first-degree relative with stone disease (Goldfarb). The most common phenotype among stone-forming relatives is hypercalciuria (Goldfarb), or an excess of calcium in the urine. As twin studies have yielded a great deal of insight into the genetic basis for many psychological disorders, they have also informed our knowledge of physical disease. In the case of kidney stones, twin studies have found that stones are twice as common in monozygotic twins than dizygotic twins (Goldfarb). However, science has yet to identify the specific genes responsible (Goldfarb).

18 11 The remaining 44% of the etiology of stone disease is related to environmental factors (Goldfarb, 2009). As the prevalence of stone disease continues to rise, this increase is likely related to environmental factors (Goldfarb) that are amenable to change. Individuals with calcium oxalate stones may have an excess of oxalate in their urine (hyperoxaluria) due to diet, as foods such as spinach, rhubarb, nuts, chocolate, and soy products are high in oxalate (Lieske, Kumar, Collazo-Clavell, 2008; MayoClinic.com; Sinha et al., 2007). Hyperoxaluria can also be caused by eating a diet low in calcium, as calcium prevents oxalate from being absorbed in the gastrointestinal tract (Goldfarb). There is evidence that eating a low-salt (sodium chloride) diet may reduce calcium excretion in calcium stone formers (Nouvenne et al., 2010), although it is unknown if this relates to the number of stones formed. Eating a diet high in animal flesh protein has also been implicated, as it may alter the chemical composition of the urine (Goldfarb). Certain surgical interventions, such as bowel resections and the Roux-en-Y (RYGB) gastric bypass, performed as a treatment for morbid obesity, are associated with increased risk for stones (Goldfarb; Kleinman, 2007; Lieske, 2009; Lieske et al., 2008; MayoClinic.com; Sinha et al., 2007). As will be discussed in greater detail later, several medical conditions with behavioral components are correlated with stone disease. The strongest evidence is for obesity and metabolic syndrome (Goldfarb; West et al., 2008). The relationship between diabetes and stone disease may be cyclical (West et al.). Although West et al. found hypertension to be the metabolic syndrome trait most associated with stones, they acknowledge, along with others (e.g., Goldfarb) that the link between hypertension and stones is not yet definitive. Kidney stones most often occur in otherwise healthy individuals (Pearle et al., 2007) and strike during an individual s prime working years, leading to substantial costs to society in terms

19 12 of lost productivity (Lotan, 2009; Saigal, Joyce, Timilsina, & Urologic Diseases in America Project, 2005). Bladder stones form primarily as the result of urinary stasis and/or recurrent bladder infection (Pearle et al.). Uric acid stones may form when an individual is dehydrated, eats a high-animal protein diet (West et al., 2008), has low urine ph caused by bowel fluid losses or by a diet with a high Potential Renal Acid Load (Goldfarb), has enzymatic defects in purine metabolism (West et al.), or has gout (MayoClinic.com). While kidney stones typically affect healthy individuals, bladder stones commonly occur in individuals with neurologic and/or anatomic abnormalities (Pearle et al.). Risk Factors. Risk factors are diverse and reflect both the genetic and environmental etiology of stone disease. Genetic/biological factors include increasing age, being Caucasian (Scales et al., 2007; Stamatelou et al., 2003), having a family history of kidney stone disease, and having a history of numerous medical conditions (e.g., inflammatory bowel disease, chronic diarrhea, renal tubular acidosis, cystinuria, hyperparathyroidism, obesity, and some urinary tract infections; MayoClinic.com). Despite correlational data of stone rates between the sexes, it is unclear if men and women actually have different biological risk factors and the historic gap in incidence rates between the sexes is shrinking (Scales et al., 2007). Environmental factors include frequent dehydration, low urine volume, working in certain professions such as teaching and truck driving (Goldfarb, 2009), possibly water hardness (Stamatelou et al., 2003), living in the South (Pearle, Calhoun, Curhan, & The Urologic Diseases of America Project, 2005; Stamatelou et al.), stress (Diniz, Schor, & Blay, 2006; Najem, Seebode, Samady, Feuerman, & Friedman, 1997; Walters, 1986), obesity, and eating a diet high in animal protein, oxalate, sugar, and/or salt (MayoClinic.com). A diet low in calcium, although

20 13 it may sound counterintuitive, is a risk factor (Goldfarb). This is because dietary calcium binds with oxalate in the intestine, preventing intestinal oxalate absorption, reducing the amount of oxalate in the urine (Goldfarb). Thus, in low calcium diets, an excess of oxalate is filtered by the kidneys, favoring the formation of calcium oxalate stones. Obesity and metabolic syndrome are perhaps the two most researched risk factors of stone disease. Metabolic syndrome, as defined by The American Heart Association and the National Heart, Lung, and Blood Institute, includes the presence of at least three of the following risk factors: hypertension, impaired glucose tolerance, abdominal obesity, gout, and hypertriglyceridemia (West et al., 2008). West et al., in their review of 18,000 adults from the Third National Health and Nutrition Examination Survey (NHANES III) data, found that hypertension was the trait associated with the greatest frequency of stones. After adjusting for age and other covariates, the presence of two or more metabolic syndrome traits significantly increased the likelihood of participants reporting having had a kidney stone. Increasing the number of traits to four or more was related to a two-fold increase in the odds of having stone disease. Despite the large body of research showing a relationship between stones and age, race, and sex, West et al. did not find an interaction between these variables and the relationship between metabolic syndrome traits and frequency of stones after adjusting for covariates. These covariates included medical history, medication use, age, race, history of gout, and SES. While the authors do not explicitly state why they believe no interactions were found, they do note the limitations of their sample. Most notably, that they utilized a cross-sectional design that prohibited temporal associations, lacked the ability to adequately adjust for multiple dietary

21 14 factors and medication use, and they lacked data on stone type. This evidence identifies obesity, metabolic syndrome, and increasing body weight as related risk factors for stones. Body weight is positively correlated with amount of oxalate excreted in the urine, and obesity is also associated with low urinary ph; both are risk factors for excessive uric acid in the urine (West et al., 2008). Taylor, Stampfer, and Curhan (2005), in their review of over 51,000 men, found that body mass might be directly related to stones. In their sample men weighing more than 220 lbs had a 44% increased risk of developing stones compared to men weighing less than 150 lbs. Consistent with previous research, weight and stones appear to be more positively correlated in women (Taylor et al.). It is believed that as women have a greater percentage of fatstoring (adipose) tissue for a given BMI than men, observed sex differences may be related to adiposity, not any underlining physiological differences (Taylor et al.). Similarly, as the relationship of body weight to diabetes is well documented, there should be little surprise that kidney stones and diabetes are also closely related. Kidney stones and diabetes have a cyclical relationship, with each being a risk factor for the other (West et al.). Insulin has been found to increase the excretion of calcium by the kidneys, to increase calcium absorption by the intestines, and to ultimately alter urine composition (Taylor et al., 2005). Diabetes has been found to be a specific risk factor for two stone types, calcium oxalate and uric acid (Goldfarb, 2009). As obesity has become one of the largest public health concerns with an estimated 30% of adults in the U.S. being obese (BMI > 30) (Matlaga et al., 2009), and considering the multifaceted impact of obesity on ones health, weight loss strategies are of vital importance. Due to the low long-term success rate of diet and lifestyle changes to treat medically complicated

22 15 obesity for many individuals, surgery is a popular option (Duffey et al., 2008; Lieske et al., 2008; Matlaga et al., 2009; Sinha et al., 2007). Various forms of bariatric surgery have been utilized for decades (Lieske et al.), although the Roux-en-Y gastric bypass (RYGB) is now the most commonly performed procedure in the U.S. (Sinha et al., 2007). The RYGB procedure has been shown effective in reducing weight and overall mortality (Lieske et al.), it also can lead to improvements in insulin resistance, sleep apnea, and hypertension, to name a few (Sinha et al.). However, it does come with known risks, including osteopenia, osteomalacia, and even rare neurological disorders (Sinha et al.). Despite these risks, the RYGB procedure is considered beneficial, safe, and effective (Sinha et al.). It is also a marked improvement over earlier, more invasive forms of bariatric surgery, such as jejunoileal (JI) bypass surgery (Lieske et al.). This particular procedure carried a high risk of developing calcium oxalate stones, amongst other complications such as liver disease and renal failure (Lieske et al.). The serious complications of JI bypass surgery led to its abandonment and the eventual rise of the RYGB procedure (Lieske et al.). However, as the popularity of this surgery exploded, physicians began to report increased incidences of stone disease in their post-surgery patients (Sinha et al.). These early antidotal reports sparked a great deal of concern, leading to several empirical investigations into the matter. Results of several studies have identified hyperoxaluria as the most common abnormality 6 months post-surgery, although calcium oxalate supersaturation, low urine volume, and low citrate concentrations (a protective factor) have also been noted (Duffey et al., 2008; Lieske et al., 2008; Sinha et al., 2007). Although these studies found a relationship between the RYGB procedure and lithogenic risk factors, Matlaga et al. (2009) were one of the first to show this does indeed lead to an increased likelihood of being diagnosed with a stone.

23 16 Matlaga et al. found that almost five years post-surgery, 7.65% of patients in the RYGB group experienced a stone, compared to only 4.63% in the matched control group, a statistically significant difference. It is believed that the increase in stone rates following the RYGB procedure is related to the malabsorption of calcium, leaving an excess of unbound oxalate to be filtered by the kidneys, causing hyperoxaluria and calcium oxalate stones (Semins et al., 2009). As it has become understood that the RYGB procedure can put patients at increased risk for stones, recent studies have begun investigating if the same holds true for another very popular bariatric surgery, gastric banding. This procedure places an inflatable tube around the stomach, just below the gastoesophageal junction, rather than surgical reconstruction of the gastrointestinal tract (Semins et al., 2009, p. 748). Compared to the RYGB procedure, it is seen as less invasive, less morbid, and is reversible (Semins et al., 2009), although it has a slower rate of weight loss and can lead to less improvement of diabetes (Penniston, Kaplon, Gould, & Nakada, 2009). The results of two recent studies (Penniston et al., 2009; Semins et al., 2009) suggest that gastric banding is not associated with an increased risk of kidney stones, likely related to the procedure not causing malabsorption. Therefore, in regards to postoperative kidney stone risk, gastric banding may be a safer procedure than RYGB (Penniston et al., 2009). The existing literature on bariatric surgery and urolithiasis risk, however informative, is not without its caveats. In the majority of studies, women participants outnumbered men, sometimes by as much as 5 to 1 (Duffey et al., 2008; Matlaga et al., 2009; Semins et al., 2009). This overrepresentation of women may complicate the interpretation of these results, as the majority of stone formers remain men despite the rapid increase in prevalence rates for women. Therefore, it is unclear if men have the same risk as women for developing stones post-bariatric

24 17 surgery. It is also unclear if there are different risk factors depending on type of RYGB procedure (proximal or distal) (Lieske et al., 2008; Sinha et al., 2007). Kleinman (2007) states that patients undergoing bariatric surgery are already at high-risk for stones considering the prevalence of preexisting conditions such as metabolic syndrome, obesity, and diabetes. He argues that instead of warning patients that stone disease is a risk factor of the RYGB procedure, providers should address the resulting hyperoxaluria postoperatively with appropriate preventative treatment. Finally, it has been suggested that the risks of RYGB may be cumulative, just as it was for the older JI bypass procedure, possibly predicting a sharp rise in stone rates for these patients in the coming years (Lieske et al.; Sinha et al.). Increasingly, attention is being paid to the impact of stressful life events on stone disease. Walters (1986) was amongst the first draw attention to this potential relationship. Najem et al. (1997) conducted an early study, interviewing 400 stone patients and healthy controls to assess any potential differences in experienced stressful life events. Regression analyses found three variables remained significantly different between patients and controls; these were patients who had lower annual family income, more events of stressful mortgage problems, and more emotional life events. However, the authors did not engage in any discussion of the believed mechanisms by which these variables impact stone disease. In addition, their study has several important limitations. Of largest concern is that 62% of the patients and controls were relatives due to their sampling procedure, although matching criteria were used so that there was no relationship between matched patients and controls. This lack of independence in their sample may have confounded results and according to the authors, likely underestimated the association between stressful life events and stone disease because the controls likely have been exposed to

25 18 some of the same stressful events as the patients (Najem et al.). Despite this, the authors believe it had a minimal effect on the generalizability their results. Also, with no known baseline to which to compare these results, it is difficult to say how these patients and controls may compare to a true random sample of their health and age matched peers (Najem et al.). Therefore, caution in interpretation of the results is warranted. More recently, Diniz et al. (2006) surveyed 194 Brazilian stone patients and controls, matched according to age and gender, to assess the number of stressful life events and their relative impact. Results confirmed that patients experienced a significantly greater number of stressful life events the year prior to their stone episode compared to the same year for controls. Patients also rated many events as having a greater impact on their lives, especially events related to work, finances, social relations, and physical health. While this was a well-conceived study, the authors did note some important limitations. The study participants were majority women (63%), despite the greater prevalence of men with stone disease, were majority Caucasian (70%), and approximately half were middle class. Diniz et al. state that women are more likely than men to report medical complaints, and the overrepresentation of women may have affected the results. Stress is believed to impact stone disease by decreasing preventative factors (urinary excretion of magnesium and citrate) and promoting risk factors (hyperoxaluria and low urine volume) (Najem et al.). Previous research has demonstrated that, in the 24 hours following a stressful event, numerous lithogenic supersaturation indices peak, including calcium oxalate and uric acid (Najem et al.; Walters). Lastly, the production of stress hormones during stressful events may lead to excessive calcium excretion (hypercalciuria), decreased urine volume (Najem et al.; Walters), and other lithogenic risk factors (Diniz et al.).

26 19 A final environmental cause, and potential risk factor, is the use of certain medications and over-the-counter supplements. A partial list of these includes high doses of calcium supplements, especially when not taken with meals; high-dose vitamin C; triamterene (diuretic); acetazolamide (glaucoma); and topiramate (epilepsy/migraines) (Goldfarb, 2009). These latter two medications may increase urine ph and decrease citrate excretion (Goldfarb), resulting in an increased risk for calcium phosphate stones. Prevalence rates by race, sex, age, & SES. Urolithiasis is a common problem with increasing prevalence in all developed countries (Stamatelou et al., 2003). As previously noted, this increase is likely related to societal changes that have encouraged unhealthy diets and sedentary lifestyles. Estimates of lifetime prevalence range from 10-13% for males to 5-7% for females (Goldfarb, 2009; Pearle et al., 2007), although it should be noted that Pearle et al. did not include stones formed in other parts of the urinary tract. Stamatelou et al., in their review of over 30,000 Americans surveyed between , estimated that 10-15% of the adult population would experience a kidney stone. They also found that prevalence rates continue to rise until the age of 70 for men and until the age of 60 for women. It has been estimated that every year 1.3 million adults ages in the labor force will require treatment for a stone (Saigal et al., 2005). Taylor et al. (2005), in their review of three different cohorts ranging in size from 45,000 to 101,000 adults and totaling over 280,000 combined person-years of follow-up calculated unadjusted incidence rates of stones between person-years. Although lifetime prevalence estimates vary, it is clear that the rate of individuals experiencing kidney stones continue to rise.

27 20 In the U.S., the adult prevalence rate has increased 37% in the past thirty years (Stamatelou et al., 2003). Recent data show a correlation between our country s obesity epidemic and stone disease (Strope et al., 2010; Scales et al., 2007; West et al., 2008). Once patients form a renal stone, their likelihood of recurrence is high. Five years after their first stone, 50% of patients will experience a second stone (Goldfarb, 2009). After 20 years the recurrence rate jumps to 80% (Goldfarb). However, recurrence rates are related to a myriad of factors, including stone type, age, race, and other health factors. Because of this high recurrence rate, kidney stone disease is considered a chronic medical condition (Penniston & Nakada, 2007; Saigal et al., 2005). Race. Caucasians have prevalence rates 2-3 times that of other racial groups, although the reasons for this are unclear (Stamatelou et al.). Scales et al. (2007), in their epidemiological review of over 1 million hospital discharges for stones between , found that 62% of patients were White, although the race for 1 in 4 patients was not given. However, Stamatelou et al. found that although prevalence rates increased for all racial groups, Caucasians saw the largest increases, with prevalence rates increasing 45% for women and 31% for men over the past thirty years (Stamatelou et al.). African Americans have the lowest levels of any racial group, regardless of age range, followed by the general population of Hispanics, and with Mexican Americans having the lowest numbers. It should be noted, however, that these are the only racial groups compared, although the reasons for this are not stated. Therefore, it is unknown how other racial groups, such as Asian, Pacific Islanders, Hmong, Native Americans, and bi- or multiracial individuals may compare.

28 21 Sex. Traditionally, the accepted ratio of stones in men to women was 3:1 (Scales et al., 2007). In their recent review, however, Scales et al. (2007) found that 59% of stone patients were men. While this disease still primarily affects males (Scales et al., 2007; Stamatelou et al., 2003; Strope et al., 2010), women have become the fastest growing group diagnosed. Epidemiological data show that the discharge rate for women increased 22% between the years while there was no significant increase for men (Scales et al.). In particular, renal calculi rates in women increased 21% while ureteral calculi rates increased 19% (Scales et al.). Men saw their renal calculi rates increase 12% while their ureteral calculi rates decreased by 10%. Despite these changes, in 2002 men and women had equal discharge rates of renal calculi while men still had twice the rate of ureteral calculi than women (Scales et al.). Stamatelou et al. (2003), in their review of data from the time periods and , found that prevalence rates increased 31% for Caucasian men and 45% for Caucasian women. Also during these time periods, prevalence rates were greater for men than women between the ages of Overall, the ratio of male to female inpatient visits changed from 1.7:1 to 1.3:1 between the years These ratios parallel other recent findings, with Strope et al. (2010) noting a change from 1.8 to 1.3. Strope et al., in their review of 107,000 discharges for stone events in the state of Florida, found utilization rates increased 22% for men and 52% for women between Although both sexes saw an increase, the growth rate was significantly greater for women. The rate of inpatient admissions increased 30% for women with no change observed in men. Women also outpaced men in outpatient utilization, with a 44% increase compared to men s 15%, a significant difference. Researchers have suggested that the increasing stone prevalence for women may be correlated with obesity rates that have increased more rapidly in women than

29 22 men (Scales et al.; Strope et al.). Obesity may also be a greater risk factor for stones disease in women than men, as obese women have almost twice the incident rates of obese men (Scales et al.). As previously mentioned, sex differences may relate to women s greater percentage of adipose tissue for a given BMI than men (Taylor et al., 2005). Other possible explanations include women s diets and working environments, which is related to fluid intake, becoming more similar to those of men (Strope et al.). Age. Kidney stones most commonly affect middle-age adults (MayoClinic.com; Stamatelou et al.; Scales et al., 2007), although they can develop at any point during the lifespan (MayoClinic.com). Scales et al. found a mean age was 48.1 (95% CI ) in their large sample of discharged patients, very similar to Strope et al. s (2010) mean age of 52. Men had a slightly higher mean age of 54 compared to women s mean age of 50, ages that remained consistent across the years reviewed (Strope et al., 2010). Prevalence rates begin to rise in men in their early 20 s, peaks between ages 40-59, and then decreases (Pearle et al., 2007). Prevalence rates appear to be more consistent across age groups for women (Pearle et al.). SES. While there is scant research on the topic, the few studies conducted provide equivocal findings. While some previous research has found a negative relationship between SES and prevalence of stone disease (Stamatelou et al., 2003), others have found no relationship between SES and the risk of stones (Najem et al., 1997). However, the results of Najem et al. should be viewed with some caution, as their study s numerous limitations cited above (e.g., mostly women, Caucasians) greatly reduce the generalizability of their findings. Any potential relationship between stones and SES may be explained by the impact SES status can have on multiple dimensions of health, including diet, working conditions, and access to health care.

30 23 Symptoms. Kidney stones are typically, though not unanimously, a very painful experience, causing significant physical and psychological distress. Individuals experiencing a kidney stone typically report flank (one-sided pain between the abdomen and back) pain in the abdomen, pelvis, and genitals. The pain typically progresses as the stone makes its way from the kidney down the ureter towards the bladder (Goldfarb, 2009). Thus, the location of the stone may be ascertained by the location of pain being experienced. Approximately 61% of individuals experience voiding urgency and suprapubic discomfort caused by a stone lodged at the ureterovesical junction, 24% experience flank pain with radiation to the genitals caused by a stone lodged between the ureteropelvic and the iliac vessels, and 11% experience flank pain caused by a stone lodged at the ureteropelvic junction (Goldfarb). The onset of pain may be sudden, followed by waxing and waning, and is often associated with nausea and vomiting without a fever (Goldfarb). Gross hematuria (visible blood in the urine) or microhematuria (blood in the urine only visible with a microscope) are also common symptoms. The pain experienced can be related to the stone size and type. Small stones (< 5 mm) may be only mildly discomforting for some, while others may experience extreme pain, nausea, and vomiting (Goldfarb, 2009). Approximately 98% of smaller stones pass spontaneously (Goldfarb). Larger stones may cause excruciating pain, are less likely to pass on their own, and typically require a longer course of treatment (Goldfarb). Struvite stones possess several unique qualities. As they typically remain in the kidney, they may not cause pain or blockage. Until the underlying urinary tract infection becomes symptomatic, they typically remain asymptomatic and undetected (Goldfarb). Thus, patients are typically unaware of them until they either become

31 24 symptomatic or are detected through imaging studies conducted during the course of unrelated treatment (Goldfarb). It is also possible that individuals with numerous other medical conditions may incorrectly attribute pain to stones (Pearle et al., 2007). However, many patients experiencing a recurrent stone typically make a correct diagnosis themselves (Goldfarb). Because patients with similar stone profiles may present with varying symptoms, it may be more beneficial for clinicians to address the impact of stone disease on patients quality of life (QOL). Quality of life effects. Given the high prevalence and recurrence rates of kidney stones, it is surprising that little attention has been paid in the literature to the psychological effects of stone disease. Only three previous studies have investigated the impact of stone disease on patients QOL (Bensalah et al., 2008; Diniz, Blay, & Schor, 2007a; Penniston & Nakada, 2007). Two more studies have investigated the stress of stone disease (Diniz et al., 2006; Najem et al., 1997), and investigations written from a psychological perspective are lacking. The little research that has been conducted has shown that stone patients report lower quality of life (QOL) than healthy U.S. adults (Bensalah et al., 2008; Diniz et al., 2007a; Penniston & Nakada, 2007). Penniston and Nakada (2007), in their survey of 189 stone patients, found that patients reported lower health related QOL for general health and bodily pain compared to healthy U.S. adults. Female patients reported more physical and psychological distress compared to their healthy gender norms relative to male patients and their norms. In addition, female patients reported lower QOL than male patients in numerous domains. Patients with other comorbidities such as diabetes mellitus type 2, hypertension, and musculoskeletal complaints reported lower QOL than stone patients

32 25 without these conditions. Surprisingly, neither the number of previous stones nor the number of previous shock wave lithotripsy (SWL) procedures affected QOL scores. Around the same time that Penniston and Nakada (2007) were conducting the first QOL study in the U.S., Diniz et al. (2007a) were conducting a very similar study in Brazil. They too utilized the SF-36 to measure QOL, although Diniz et al. focused their study only on a sample of recurrent stone formers (history of 2 or more stones). Despite this different sample of stone formers, the results of Diniz et al. were strikingly similar to those found by Penniston and Nakada. They found that stone patients reported lower physical and mental health than matched healthy controls. Interestingly, they found three variables to be predictors of QOL; lower SES, time since diagnosis, and number of painful stone episodes. Patients with lower SES status presented with lower mental and physical health, perhaps because SES is related to many determinants of health. They also found a significant correlation between time since diagnosis and number of recurrent painful stone episodes, with the number of episodes increasing with time. These patients with more painful episodes were then more likely to present with QOL decrements related to physical capacity and vitality. While Penniston and Nakada did not find a similar QOL impact for the number of stones, Diniz et al. s findings may be influenced by their sample of only recurrent stone formers and their focus on painful episodes, not just number of stones. In explaining the impact of painful episodes, the authors note the large body of health literature that has found patients suffering from conditions associated with recurrent pain are more likely to report poorer physical and mental health. Also similar to previous QOL studies, Bensalah et al. (2008) interviewed and surveyed 155 stone patients, utilizing the same survey (SF-36) as Diniz et al. (2007a) and Penniston and

33 26 Nakada (2007). Perhaps then it is not surprising that Bensalah et al. found similar results to these studies, with a few notable additions. Stone patients reported lower physical and mental QOL than healthy U.S. adults. Similar to Diniz et al. s study of time since initial diagnosis, the authors controlled for patients time interval from last stone event. In doing so, they found that patients who experienced a stone event in the last month reported even lower physical and social QOL than the uncontrolled patient group. Those patients whose stone event was more than one month ago had a comparable profile with healthy U.S. norms, except patients still reported lower physical QOL. Also similar to Penniston and Nakada, patients with other comorbidities such as obesity and a recent stone event reported lower QOL. Female stone patients reported lower physical QOL than male patients. Again, surprisingly the number of previous stone events did not affect QOL scores. While there continued to be no effect on QOL for having undergone the SWL procedure, having undergone percutaneous nephrostolithotomy (PCNL) or ureteroscopy (URS) did negatively impact QOL. The number of previous ureteral stents was associated with lower physical, and especially mental, QOL. This may be related to the significant level of discomfort that stents can produce (Bensalah et al., 2008). The authors conclude that increasing BMI, age, and number of surgical procedures are the greatest predictors of QOL for stone patients. Despite the few studies of stone patients QOL, a few general conclusions can be drawn. Of most importance, all studies have found that stone patients report lower physical and mental QOL than healthy adults. Women stone formers, in particular, appear to experience more decrements in QOL than their male counterparts. The number of comorbidities, number of painful episodes, and time since diagnosis and last stone event may also negatively impact QOL.

34 27 It is unclear why the number of previous stones has not been found to impact QOL. The authors offer few hypotheses, although the findings regarding painful stone episodes and treatment types may offer some insight. It is possible that the presence or absence of pain may more directly impact physical and metal QOL than simply the presence of a stone, as stones may be asymptomatic or painless. It is also noteworthy that although no study has found the SWL procedure (a minimally invasive procedure) to impact QOL, PCNL, URS, and stent placements (all of which are highly invasive procedures) have been shown to negatively impact QOL. Together, these results highlight the negative impact stone disease can have on patients lives and the need for treatment to go beyond stone removal. Psychological effects. Aside from the QOL studies, there is a paucity of research into the psychological effects of kidney stone disease. The reasons for this lack of attention to stone disease in the psychological literature are unclear. It may be a result of psychology s focus on other, more well-known and funded diseases, including breast cancer and HIV. It is also possible that stone disease has erroneously been presumed to not affect mental health, thus not requiring psychological investigation. The little research that does exist into the psychological effects of stone disease typically comes from existing medical studies. Diniz, Blay, and Schor (2007b) conducted one of the only known studies investigating the mental health of recurrent stone patients in Brazil. They surveyed 64 recurrent stone patients and controls, matched for age and sex, utilizing the State-Trait Anxiety Inventory (STAI) and the Beck Depression Inventory (BDI). All patients were recurrent stone formers with a history of at least two stones, although at least two months had passed since their last stone event. For symptoms of anxiety, stone patients

35 28 were less likely to fall into the low anxiety state group and more likely to fall into the high state anxiety group. For symptoms of depression, patients were more likely to fall into the moderate and severe depression categories than controls, with 59% of patients reporting symptoms of depression (compared to 28% for controls). Patients were also found to have a 3.74 times higher probability of having symptoms of depression than controls. Significant positive correlations were found between anxiety-state, anxiety-trait, and symptoms of depression and number of painful stone episodes. Age and sex difference did not influence scores of anxiety or depression. Although this study was not conducted in the U.S. and focused on recurrent stone formers, it still highlights the impact of stone disease on the mental health of stone formers. Not only does stone disease appear to negatively impact the lives of patients, it also negatively impacts society. Depression, especially when associated with a painful chronic medical condition, is known to substantially increase both the direct and indirect costs of treatment (Greenberg, Leong, Birnbaum, & Robinson, 2003). Therefore, treatment approaches need to take into account both the physical and psychological aspects of stone disease (Diniz et al., 2007b). Depression is widely considered the most prevalent mental health issue faced by patients with chronic kidney disease (Hoth et al., 2007), although this term does include many medical conditions, not just kidney stone disease. In patients with a severe form of chronic kidney disease, end-stage renal disease (ESRD) with hemodialysis, the prevalence rates of depression range from 20-30% (Hoth et al.). As previously mentioned, Diniz et al. (2007b) found 59% of their sample of 32 recurrent kidney stone formers reporting symptoms of depression, with 34.4% reporting symptoms of moderate depression and 6.2% reporting symptoms of severe depression. Penniston and Nakada (2007), in their study of 189 patients QOL, found rates of depression to

36 29 be 19% in women and 5% in men. It has been estimated that the rates of depression in individuals with chronic and serious medical conditions to be 20-25% during the course of their medical condition (American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders (DSM; 4 th ed.,text Revision; 2000). Together, this body of evidence suggests that patients with varying severity of kidney disease are disproportionately affected by symptoms of depression with rates greater than the general population and that depressive symptoms likely affect patients level of functioning (Hoth et al.). In order to identify patients with chronic kidney disease, of which stone disease is but one part, at risk for poor adjustment to managing a chronic illness, Hoth et al. (2007) investigated the relationship between social support, the Big Five factor of Agreeableness, and depressive symptoms over time. They surveyed 59 patients with moderately impaired renal function at baseline and approximately18 months later. The authors utilized the Social Provisions Scale to assess social support, the Beck Depression Inventory to assess depression, and the NEO-Five Factor Inventory at baseline-only to assess Agreeableness. The average patient scored in the mild to moderate depression range at baseline, although the authors note a large standard deviation in scores. As predicted, the authors found that patients high in Agreeableness with greater social support saw a decrease in depressive symptoms over the 18 months. Patients high in Agreeableness but with lower social support actually saw an increase in depressive symptoms over time. Also as predicted, for patients low in Agreeableness, the degree of social support over time did not significantly moderate depressive symptoms. Although traits such as Agreeableness are not typically considered malleable, the authors defend this study of personality variables as an opportunity to consider individual differences. Since it is unlikely that those patients low in

37 30 Agreeableness can increase their degree of Agreeableness, providers can use this information to tailor treatment programs. This study offers important insight into the nature of social support and suggests that increasing social support may not be beneficial for every patient. Instead, patients known to be low in Agreeableness might find individualized treatments more beneficial (Hoth et al.). While one typically thinks of mental disorders when discussing psychological functioning, one interesting study has investigated the impact of clinical decision making on anxiety in stone patients. Margalith and Shapiro (1997) surveyed 96 Israeli patients with a ureteral calculus and found that, in general, patients anxiety did not decline if they were offered the opportunity to choose their stone treatment. Instead, they found that doctor-patient interaction (provision of information) without the patient choosing the treatment procedure actually reduced the patient s anxiety. However, highly educated patients and those with a passive coping style saw their anxiety reduced if perceived they had participated in the decision-making process, even if they actually had not participated. The authors conclude that patients are likely to see their anxiety reduced if they believe they have received relevant information from the treating urologist, regardless of if they participate in the clinical decision-making process. As previously noted, highly educated patients and patients with passive coping styles appear to be the exceptions. Margalith and Shapiro believe, similar to Hoth et al. (2007), that treatment should be tailored to individual patient characteristics and in the least, include the provision of information to patients. In doing so, each patient s mental health needs can be included as integral part of the treatment they receive, likely improving outcomes.

38 31 Previously discussed, the impact of stress on the development and management of stone disease remains an important psychological variable. The exact mechanisms by which stress influences stone disease is not fully understood, but it is likely a cyclical relationship. Highly stressed individuals may increase their lithogenic risk through both biological mechanisms (Najem et al., 1997; Walters, 1986) and psychosocial factors, such as engaging in unhealthy or risky behaviors. Likewise, previous research has demonstrated that stone patients experience more physical and mental health decrements than their healthy counterparts. Living with a chronic medical condition, especially one that can be unpredictable and painful, likely also increases patients stress level. Although research is lacking on the impact of living with stone disease on life behaviors, some insight can be gleaned from the broader health psychology literature. In one often-cited study (Andersen et al., 2008), women diagnosed with breast cancer were randomized to either a control or psychological intervention condition. The intervention was a psychologist-led small group designed to reduce stress, teach healthy behaviors, and improve treatment adherence. The results of this intervention were striking. After a median follow-up period of 11 years, women in the intervention were 45% less likely to have had their breast cancer return and were 56% less likely to have died from cancer. While breast cancer is considerably different than stone disease, the authors predict similar mechanisms for the impact of stress on health as is discussed in the stone literature. In a second example, researchers investigated the impact of aerobic exercise, stress management training, and usual care on general distress and depression with patients with ischemic heart disease (IHD), the leading cause of death in the U.S. (Blumenthal et al., 2005). The authors found that patients in the aerobic exercise and stress management training conditions had improved psychosocial

39 32 functioning, less distress, fewer symptoms of depression, and improved markers of cardiovascular risk than patients receiving usual care. They concluded that these interventions might eventually lead to improvements in clinical outcomes by improving functioning and altering risk factors. In summary, stone patients typically experience more anxiety and symptoms of depression than their healthy peers. Several personality characteristics have been shown to influence patients mental health, such as degree of Agreeableness and having a passive coping style. Other psychosocial factors, such as stress, also have been shown to affect physical health and appear amenable to psychological interventions. Psychosocial decrements, in turn, can affect a patient s ability and motivation to adhere to complex treatment regimes (Diniz et al., 2007b; Hunter et al., 2009). These findings highlight the many ways stone disease affects patients mental health, the need to identify barriers to treating stone disease, and the need for more individualized and holistic treatments. Diagnosis. As previously discussed, most kidney stones are diagnosed when patients experience extreme pain and seek medical attention. However, asymptomatic stones are typically discovered during imaging studies for unrelated reasons (Goldfarb, 2009). As kidney stones may not have a specific presentation during an exam, a physical examination is necessary to rule out other conditions (Goldfarb; Pearle et al., 2007). Imaging technology has become the standard method to diagnosis a stone and is required to make a definitive diagnosis, unless a stone has already passed or been surgically removed (Pearle et al.). A computed tomography (CT) scan is considered the diagnostic gold standard (Goldfarb, p. 4). This is because it has superior

40 33 sensitivity and specificity, is useful in making differential diagnoses if a ureteral stone is not found, and is fast and relatively safe (Goldfarb). Its main drawback is its cost, as it is more expensive than other methods, although this is primarily an artifact of insurance billing practices (Goldfarb). Another drawback is the radiation exposure from the scan itself, and there is concern about exposing patients to dangerous dosages of radiation over time. Other imaging technologies, such as ultrasonography and intravenous pyelography (IVP), have some benefits, but are, unfortunately, overall less effective as a diagnostic tool for stone disease. In addition to imaging techniques, a complete blood count, urinalysis, and specific serum chemistries are also recommended (Goldfarb). These tests assist in taking a complete health assessment and understanding a patient s specific stone etiology (Goldfarb). In order to further understand a patient s stone etiology and to assist in prevention planning, several techniques exist to determine stone mineral composition. As previously mentioned, about 80% of stones contain calcium and calcium oxalate (Goldfarb). Medical treatment. Historically, the end goal of treatment of urolithiasis has been a stone-free patient (Penniston & Nakada, 2007). This goal has typically been achieved through a range of medical interventions that have changed in popularity over time. Surgical or other urological treatment is necessary in approximately 20-25% of patients with a diagnosis of nephrolithiasis (Goldfarb, 2009; Saigal et al., 2005). For the remainder of patients, surgical interventions are not always necessary to treat the presence of a stone, as approximately 98% of smaller stones pass spontaneously (Goldfarb, 2009). Larger stones, however, are less likely to pass on their own and typically require a longer course of treatment (Goldfarb). Treatment is warranted if the stone is

41 34 associated with pain, obstruction, infection, growth despite medical treatment, and being of a large size (Pearle et al., 2007). There are numerous treatment options, each depending on patient need and preference. Extra-corporeal shock wave lithotripsy (ESWL) is the most common treatment in the U.S, comprising between 49-54% of all procedures (Pearle et al., 2005). It is considered the first-line treatment for stones up to 1 cm in diameter located in the proximal ureter (Goldfarb). This procedure uses shock waves to break up the stone into smaller fragments that are then passed in the urine (MayoClinic.com). While this procedure is considered the only completely non-invasive treatment option (Pearle et al., 2007, p. 285), it is not without side effects. It may cause moderate pain, hematuria, bruising on the back or abdomen, bleeding around the kidney and other adjacent organs, and discomfort as the stone fragments pass through the urinary tract (MayoClinic.com). Previous research has found no QOL decrements for patients undergoing the ESWL procedure (Bensalah et al., 2008; Penniston & Nakada, 2007). The second most-common treatment, ureteroscopy, comprises around 40-42% of all procedures (Pearle et al., 2005). It is usually used to treat ureteral stones (Pearle et al., 2007) and may be more successful in removing stone fragments, reducing the likelihood of recurrence (Goldfarb, 2009). In this procedure, a ureteroscope (a thin tube equipped with a light and camera) is passed through the urethra and bladder into the ureter (MayoClinic.com). The stone is then either snared and removed or is fragmented with a small laser into smaller fragments that will spontaneously pass (MayoClinic.com). This procedure can be associated with considerable discomfort and pain. Typically following this procedure, a stent running from the kidney to the bladder is placed inside the patient for around a week (Damiano et al., 2005). It is believed the stent aids in passage of urine and stone fragments post-surgery and prevents scar tissue from

42 35 narrowing the ureter (stricture) (Damiano et al.; Nabi, Cook, N'Dow, & McClinton, 2007). Stent placement, however, is also associated with considerable patient discomfort (Bensalah et al., 2008; Damiano et al.). Damiano et al. found that patients having had a stent reported significantly higher pain, discomfort, and urinary symptoms, more symptoms of depression and anxiety, and reduced QOL compared to controls. These results are similar to the findings of Bensalah et al. who found that the number of previous ureteral stents was associated with lower physical, and especially mental, QOL. Perhaps not surprisingly then, Nabi et al. found that twothirds of patients who had previously received a stent said they would prefer not to have a stent in the future. In addition to the numerous physical and psychological decrements reported after stent placement, a recent meta-analysis found almost no benefits of stenting on clinical outcomes (Nabi et al.). The authors could find little difference between those that had received a stent and those that had not, perhaps highlighting a lack of clinical guidelines organizing their usage. As if the QOL decrements and lack of clinical benefits or guidelines were not enough data to question the use of stents, stenting also increases treatment costs (Nabi et al.). However, the authors provide no hypotheses as to why stenting remains so popular despite this evidence. Instead, they caution that recommendations to discontinue the practice of stenting are premature given the poor quality of extant research. A third treatment option, although infrequently used, is percutaneous nephrostolithotomy (PCNL), comprising 5-6% of procedures (Pearle et al., 2005). Lack of surgical expertise in this procedure among non-academic urologists is one reason it may be used less frequently. In this invasive procedure, a small incision in the back is made, allowing the surgical removal of the stone (MayoClinic.com). PCNL is typically performed to remove large stones or stones

43 36 associated with anatomic abnormalities (Pearle et al., 2007). It is recommended as the first-line treatment for patients with staghorn calculi for which ESWL is ill advised (Goldfarb, 2009). A fourth treatment, open surgery, continues to see a decline in its use. It is estimated to compromise only 1-2% of all stone treatments today (Pearle et al., 2005; Pearle et al., 2007). In recent years, treatment rates have risen, coinciding with the rise in prevalence rates of stone disease. Between 1994 and 2000, hospital outpatient visits increased by 40% and between 1992 and 2000 physician office visits increased by 43% (Pearle et al., 2007). However, as outpatient procedures have become the norm, inpatient hospitalization rates decreased 15% during the same period. Despite the list of surgical options available, there are numerous reasons to be concerned about the current treatments of urolithiasis. First, although the minimally invasive ESWL procedure remains the most widely used treatment for stones in the U.S., some have raised concerns about the rise in popularity of ureteroscopy (Kerbl et al., 2002). Analyzing Medicare data between the years 1988 and 2000, Kerbl et al. noted that ureteroscopy saw a 13% increase while ESWL saw a 3% decrease in percent of all stone procedures. The authors state that newer, second- and third- generation shockwave lithotripters are less powerful, and consequently less effective, than the first-generation machines. This reduced effectiveness, coupled with advancements in ureteroscopy equipment and a lower cost than ESWL, may have contributed to the increased use of ureteroscopy. It is also worth restating that as previously reported, the ESWL procedure does not appear to negatively impact patient QOL while having undergone PCNL or ureteroscopy was found to negatively impact QOL (Bensalah et al., 2008). Second, despite the abilities of surgeries to treat an existing stone, it does nothing to prevent future stone

44 37 occurrences. Thus, there is no cure for stone disease. Third, there have been very minimal advances in surgical procedures over the past thirty years (Kerbl et al.; Pearle et al., 2007) with few future prospects (Lotan, 2009), suggesting the stagnation or even regress of medical treatments. Finally, the field is beginning to recognize that a stone-free patient may not be a sufficient end-goal and has begun considering the impact of stone disease on QOL issues (Bensalah et al., 2008; Penniston & Nakada, 2007). Bensalah et al. go as far as saying that, QOL should represent a primary end point in the care of stone forming patients (p. 2242). This recent focus on patient QOL requires that new, more encompassing interventions be created (Penniston & Nakada). Cost of urolithiasis. Kidney stones are the second costliest urological disease, second only to urinary tract infections (Bergman et al., 2010). The total annual expenditure surpasses conditions that receive far more publicity than kidney stones, including prostate, bladder, and kidney cancers, urinary incontinence, and even erectile dysfunction (Bergman et al.). In the year 2000 it is estimated that urolithiasis treatment cost between $ billion in the U.S., the last year figures are available (Lotan, 2009; Saigal et al., 2005). This calculates to an estimated $3,494 per patient in 2000, comparable with other chronic conditions such as diabetes and asthma and exceeding some conditions such as irritable bowel syndrome (Saigal et al.). The price of urolithiasis is impacted by its high, and rapidly increasing, prevalence rates and high recurrence rates (Lotan). Costs are also driven by the almost quarter of patients with urolithiasis that require costly treatments (Goldfarb, 2009). Urolithiasis is somewhat unique compared to other chronic conditions in that it affects adults in the prime of their working years. As such, the direct and indirect costs of the

45 38 disease are borne primarily by employers and employer-provided health insurance, unlike other conditions that affect older adults and Medicare (Saigal et al.). The average stone patient s annual health care expenditure was found to be more than twice that of an employee without this condition, even after adjusting for factors such as other comorbid conditions (Saigal et al.). As health care costs continue to skyrocket at a disproportionate rate and prevalence rates continue to rise, there is a great need to understand the basis of urolithiasis costs and to explore costreduction options (Lotan). The direct medical costs of stone disease can be broken down into physician office visits, surgeries, and medications (Lotan, 2009). Surgery compromises the largest single cost (Lotan), with the average ureteroscopy costing $1,425 and the average ESWL costing $2,295 in 2000 dollars (Saigal et al., 2005). Rates are even higher for PCNL and open surgery. For patients with an insurance claim of nephrolithiasis, 18% of their total expenditure went towards medication expenses (Saigal et al.). Even though medical treatment of urolithiasis has shifted to primarily outpatient surgeries, reducing hospital stays, costs have continued to rise (Lotan). As previously mentioned, causes of these include the increasing number of procedures due to the increasing prevalence rates and the popularity of stenting following ureteroscopy, despite its questionable benefit. The indirect costs of urolithiasis can be broken down into two main categories, lost wages and productivity and psychosocial costs. Lost wages and productivity has received the majority of the attention in the stone literature. Saigal et al. (2005), in their review of employer claim data from 25 large U.S. employers, found that 1% of employees were treated for urolithiasis, with one-third of these employees missing work. For those that missed work an

46 39 average of 19 hours per person was missed. The amount of work loss varied as a function of the care required, as patients requiring inpatient hospitalization lost an average of 47.9 hours of work while patients treated at an ambulatory care facility lost an average of 5.1 hours. Conservatively, this comes to 3.1 million lost workdays per year worth at least $775 million (in 2000 dollars) in lost productivity for insured employees (Saigal et al.). Although the urologic literature has largely ignored the cost of QOL and psychological decrements related to stone disease, estimates of this can be drawn from the broader health psychology literature. As previously discussed, depression is widely considered the most prevalent mental health issue faced by patients with chronic kidney disease, of which kidney stone disease is a part (Hoth et al., 2007). Depression can greatly impact a patient s ability to adhere to treatment and prevention guidelines, increase suffering and the utilization of health care, decrease a return to full physical functioning and work, and may even increase the likelihood of stone recurrences (Diniz et al., 2007b). Therefore, there is little surprise that comorbid mental health concerns like depression can substantially increase both the direct and indirect costs of treatment (Greenberg et al., 2003). It is estimated that total annual expenditure in treating depression was $70 billion in 2003 and an employee treated at least once for major depression cost employers 3.5 times that of the average employee (Greenberg et al.). Thus, QOL and psychological decrements are an important component of the indirect costs related to stone disease and are worthy of increased attention in the literature. In an effort to investigate avenues for cost containment, Lotan (2009) identified two potential areas, lower health care-related costs and a decrease in prevalence rates. In recent years attempts have been made to reduce health care costs by reducing physician reimbursement rates, although this has not significantly impacted health care costs (Lotan). Instead, the author states

47 40 that the majority of treatment costs can be attributed to the high cost of equipment. To control costs he suggests improved decision-making process regarding inpatient admissions, avoiding time delays in performing required surgeries, and encouraging more spontaneous passage of stones. In order to decrease prevalence rates, Lotan highlights the need for primary prevention, an idea supported by other authors (e.g., Saigal et al., 2005). This includes assisting patients to make a range of behavioral and lifestyle changes, although he notes that research in this is lacking. Therefore, increased attention to prevention efforts has the potential to not only improve patient care, but reduce costs as well. Prevention. As much of this review has addressed the high prevalence and recurrence rates of stone disease and its related costs, the importance of effective primary and secondary prevention interventions cannot be overstated. Most of these interventions take the form of either drug therapies or a range of behavioral and lifestyle modifications. Of the latter, increasing fluid intake is the most-recommended prevention strategy in the literature. It is recommended that patients increase their fluid intake to a minimum of L/d (68-85 oz/d) in order to achieve a urine output of at least 2 L/d (Goldfarb, 2009). The reason for this is simple: increasing fluid intake makes urine less supersaturated with stone-forming minerals and salts such as calcium oxalate (Goldfarb). Although it is logical to be concerned that diluting the urine would also dilute the concentration of beneficial compounds like citrate, this has not been supported by research (Meschi et al., 2004). Increasing fluid intake is also highly recommended because it is cheap and safe. However, the literature fails to mention that frequent urination can lead to decrements in QOL, such as on long car rides, flights, during movies, meetings, classes, while sleeping, and

48 41 other events where the need for frequent bathroom breaks is embarrassing, inconvenient, or impractical. QOL decrements aside, there are a number of studies on its effectiveness for secondary prevention. Borghi et al. (1996) conducted a well-known longitudinal study investigating the impact of high water intake on stone recurrence rates. They randomized stone formers into two treatment groups. One group was encouraged to have a daily urine output of at least 2 liters and a second group was told it was not necessary to make any special changes. These two groups there then compared with a group of healthy controls. Interestingly, stone formers in both groups had significantly lower urine output than controls at baseline, suggesting a possible cause of initial urolithiasis. After five years, the patients in the high water intake group had significantly fewer stone recurrences (12%) than patients who did not alter their water intake (27%). In order to further understand what lithogenic risk factors might differentiate patients who had a reoccurrence from those who did not, urine analyses found hypercalciuria to be the only factor. Borghi et al. conclude that increasing water intake is an effective prevention strategy, although for patients with hypercalciuria dietary and pharmacological interventions are also warranted. In a similar study, Carvalho, Ferrari, Renner, Vieira, and Riella (2004) investigated the stone clinic effect, or the common recommendations encouraging stone formers to increase their fluid intake and limit their intake of salt and protein. They instructed a group of 25 stone formers to follow the stone clinic effect for one year. At the end of this time, 80% of patients reduced their number of stones and calcium oxalate excretion from baseline. As the body of this study is only available in Portuguese, it prevents this author from analyzing the merit of their findings. Since the average interval between stone occurrences reported in the literature is anywhere between 2-5 years (Borghi et al., 1996; Goldfarb, 2009), it might be

49 42 unlikely to expect the majority of patients to have had a recurrence so soon. Therefore, the lack of stones seen at the end of the year may not be attributable to the stone clinic effect. The study was also based on a small number of participants (N = 25), potentially impacting power and generalizability. Despite the popularity of increased fluid intake for secondary prevention, there are no known studies of its effectiveness for primary prevention (Goldfarb). In fact, others have gone as far as saying that, primary prevention is not practical at this time (Pearle et al., 2007, p. 318). This likely is related to the difficulty of designing such a study with adequate control. Dietary changes are also a popular secondary prevention recommendation (Goldfarb, 2009; Meschi et al., 2004), although some also consider it a key primary prevention strategy (Meschi et al.). As previously discussed, dietary risk factors include a diet low in calcium and citrate and high in oxalate, animal protein, and salt (sodium chloride). Dietary changes have been shown to be particularly important in patients with idiopathic calcium oxalate stones (Nouvenne et al., 2010; Penniston & Nakada, 2007) and with obese patients (Meschi et al.; Penniston & Nakada), although some debate remains as to the specific nature of dietary recommendations required (Penniston & Nakada). It is generally advised to tailor nutrition recommendations to each patient s individual risk factors, including her/his 24-h urine composition, stone type, comorbidities, and body habitus. Thus, there may be no one size fits all diet for all stone formers. Meschi et al. provide ten general dietary recommendations: Keep weight under control, limit animal proteins, limit sugar intake, limit fat intake, maintain normal calcium intake, limit oxalate intake, reduce sodium chloride intake, increase alkaline potassium intake, increase vegetable fiber intake, and increase water intake. Previous research has shown that many patients are able to adhere to some dietary changes and that these changes correspond to a reduction in

50 43 urinary risk factors (Nouvenne et al.; Penniston & Nakada). However, not all patients are able to make clinically relevant dietary changes and it has been difficult for some studies to measure patient compliance (Penniston & Nakada). It is also unknown if patients are able to adhere to dietary changes long term, a significant issue given the chronic nature of stone disease (Meschi et al.; Penniston & Nakada). Thus, although dietary changes appear effective as a measure of secondary prevention in reducing stone reoccurrence rates, variable adherence to diet could limit its clinical effectiveness (Penniston & Nakada). In addition to dietary changes, a range of other lifestyle changes is often recommended. These include getting regular exercise, losing weight (Meschi et al.), maintaining fluid intake, taking medications, and adhering to dietary recommendations. Maintaining a healthy weight is particularly effective, as even small reductions in weight have been shown to reduce lithogenic risk (Meschi et al.). As demonstrated from the broader health psychology literature, patients often struggle to enact, and adhere to, complex lifestyle changes long-term. Drug therapy is also a category of secondary prevention strategies (Goldfarb, 2009). For patients with calcium oxalate stones, thiazides are the most commonly recommended, as they reduce urinary calcium excretion by stimulating renal calcium reabsorption (Goldfarb), thus reducing the amount of calcium excreted in urine. Common thiazides include chlorthalidone, hydrochlorothiazide, and indapamide. Although thiazides have been shown to reduce recurrence by 30-35% over three years compared to placebo, 15-30% of patients taking thiazides still experience stone recurrence (Goldfarb). Research suggests that thiazides use requires at least two years to show any benefit (Goldfarb). In addition, thiazides may cause hypokalemia,

51 44 necessitating that many patients on them also take potassium citrate or potassium-sparing agents (Goldfarb). While numerous prevention strategies, mostly for secondary prevention, have been shown effective, none are entirely effective. Dietary and lifestyle strategies are often unsuccessful due to limited patient understanding, time, and self-efficacy in making changes. Thus, the clinical effectiveness of extant prevention strategies is limited, requiring an investigation of adherence barriers and the formation of new prevention strategies. Section I summary. Urolithiasis is a chronic medical condition, and one of increasing prevalence. Women are the fastest-growing group affected and also appear to experience the greatest QOL decrements. For patients who do experience a kidney stone, there are limited treatment options available. In the past 30 years treatments have become more invasive and, in the case of ESWL, less effective. As prevalence rates have increased and the costs of treatments have risen, there is little surprise that the overall cost of treating urolithiasis continues to grow. Not only have the direct costs of treatment grown, but so have the indirect costs in the form of loss productivity. Additionally, patients have been shown to experience numerous QOL and psychological decrements related to stone disease, including depression, anxiety, and physical pain. In some cases, the invasive treatment options available have been shown to increase both mental and psychological suffering. Surgical treatments do nothing to prevent future stone occurrences and development of new treatments has stagnated. Prevention strategies are only partially effective, and require patients to make complex behavioral and lifestyle changes. Many patients are either unable or unwilling to adhere to these complex changes and adherence barriers are only partially

52 45 understood. Therefore, there is a strong need for improved treatment and prevention strategies, especially ones that address the QOL and psychological decrements that have traditionally received little attention. Section II. The Primary Care Behavioral Health Model As the previous section discussed, urolithiasis is a serious chronic medical condition associated with both physical and mental health concerns. Medical treatments are limited in their effectiveness and do nothing in the way of preventing reoccurrences. Prevention strategies require patients to make significant behavioral and lifestyle changes. For even the most mentallyfit patient, making significant behavioral and lifestyle changes can be a daunting challenge. Behavioral interventions have been shown to be a very effective method for addressing common unhealthy behaviors and change behavioral risk factors on a variety of medical conditions (Whitlock et al., 2002). The use of behavioral interventions by mental health professionals is not a new concept, as it dates back several decades. Mental health professionals first employed their expertise in behavioral change in the primary care setting, a setting where the majority of Americans receive the bulk of their health care, both physical and mental (Robinson & Strosahl, 2009). Their work lead to the creation of a new model of integrated health care, the primary care behavioral health (PCBH) model (Robinson & Strosahl). The PCBH model combines patients physical and mental health needs and facilitates collaboration of medical providers with psychologists. It is considered the best opportunity to reach the greatest number and variety of patients with evidence-based behavioral health assessments and interventions (Hunter et al., 2009, p. 4). Psychologists utilize their expertise in mental health assessment, diagnosis, and treatment to enact change consistent with medical recommendations. Targeting health-related

53 46 behaviors has also been suggested as one approach to addressing health disparities, as those of lower SES and racial/ethnic minorities are disproportionately affected by preventable risk factors (Whitlock et al.). Previous research has demonstrated that the application of this model to primary care can lead to improved patient and provider satisfaction, improved health outcomes, improved patient quality of life, and significantly decreased health care costs (Katon, 1995; Robinson & Strosahl). Despite the success of this model for primary care, there are few examples of its application to secondary care. In addition, there is no previous study of the application of this model to a population of kidney stone formers. If the PCBH model can be successfully adapted to a secondary care clinic, it could translate well to other settings and may greatly improve patient care while reducing costs. History of integrated care. While it is widely accepted today that the body and mind are intertwined, this is a relatively young concept in the scientific literature. The beginnings of integrated practice are somewhat mysterious, with only speculation on the origins of this model (Robinson & Strosahl, 2009). It likely began sometime in the 1960s and gained more popularity in the 1970 s due to the requirements of family practice residencies (Robinson & Strosahl). The first cited study on the effects of mental health treatment on physical health dates back only to 1965 (Mumford, Schlesinger, Glass, Patrick, & Cuerdon, 1984). Similar studies proliferated until 1984 when Mumford et al. published their seminal meta-analysis of the effects of mental health treatment on medical utilization rates. They found that patients who received mental health treatment were more likely than controls to utilize less inpatient medical care, thus leading to a reduction in health care costs. They did note, however, that mental health treatment lead to an increase in the

54 47 utilization of outpatient care consisting of physician office visits, lab test, and prescription drugs. This increase, though, was mitigated by the substantially higher costs of inpatient treatment. The cost offset of mental health treatment was the greatest for those patients 55 years and older, leading the authors to conclude that, underutilization of mental health services by the elderly may result in needless suffering among the elderly and needless cost to society (p. 1156). While Mumford et al. s (1984) seminal meta-analysis provided much-needed empirical support for the benefits of mental health treatment on utilization of medical treatment, many of these early studies were of questionable quality. Katon (1995) lists several critiques of this early literature including poor matching of treatment and control patients, no measurement of the effects of the psychotherapy, a lack of manualized psychotherapies, and a lack of investigation of the mechanisms by which medical costs were decreased. Relevant to the current project, it should also be noted that these early studies did not investigate an integrated or collaborative treatment model (Katon). The mental health care that the patients received was all located in traditional mental health settings, not integrated in the primary or secondary care systems. Integrated primary care, or the collaboration of medical and mental health professions to create one treatment plan, gained a more permanent foothold in the 1980 s in large, progressive HMO s like Seattle s Group Health Cooperative (Robinson & Strosahl, 2009). While challenging to initiate, initial applications of this model saw increased patient and provider satisfaction and increased patient adherence. Although integrated care added to treatment costs, this added cost was lower than the overall costs of a usual care patient (Robinson & Strosahl). The successful development of this model highlighted several early pitfalls that were later corrected. These included trying to retain the structure of traditional mental health, focusing on

55 48 diagnosis rather and functional and QOL improvements, and failing to conduct comparative research (Robinson & Strosahl, 2009). From these mistakes the model has been adapted to incorporate a more short-term consultation focus, to engage primary care providers, and to set minimal, realistic goals with patients (Robinson & Strosahl). The proliferation of this model is a testament to its multiple benefits and long-term sustainability. Definition and description. Robinson and Strosahl (2009) named this new model the Primary Care Behavioral Health Model (PCBH) because it conveys the essence of a new approach to behavioral health service delivery - behavioral health services as primary rather than specialty services (p. 63). This model requires a shift from the traditional way mental health has been delivered in the U.S. and instead employs a consultation approach (Strosahl & Robinson, 2008). The PCBH model conceptualizes health according to the biopsychosocial model. It defines good health as consisting of effective prevention and immediate access to care. The mission of the PCBH model is to improve the overall health of the population (Strosahl & Robinson). The authors see the behavioral health consultant, or BHC, achieving this goal through two mechanisms: (1) by augmenting the usual preventive and direct care for behaviorally based problems; (2) through educational interventions and changes to the system of care that improve the primary care system s ability to provide such care (p. 88). The BHC is co-located in the practice, is viewed as a core member of the medical team, and whose services are considered routine care (Strosahl & Robinson). The role of the BHC is to act as a consultant to the referring medical provider, usually a primary care physician (PCP), not as a therapist to the patient. As part of the shift from traditional mental health care, the PCP is the BHC s primary customer (Strosahl & Robinson).

56 49 Functioning as a consultant also means that PCPs retain responsibility for a patient and s/he remains on a PCP s caseload. Similarly, the PCBH model de-emphasizes making mental health diagnoses and instead places a focus on setting realistic and concrete goals to improve patients functioning and QOL (Strosahl & Robinson). The primary outcome of the BHC s work, broadly defined, is to enhance the PCP s ability to assist the patient. The goal is not to cure the patient of all distress but rather to improve the patient s functioning and provide support to the physicianpatient relationship (Strosahl & Robinson). Strengthening this relationship is key, as individual, brief, behavioral interventions have been shown to have only a modest impact on behavioral change (Whitlock et al., 2002). Previous research has found that between 5-15% of patients in an intervention will make clinically significant behavioral changes. On a population level, overall risk factors post-intervention have been shown to change 1-20% (Whitlock et al.). While the individual effects of behavioral interventions may be small, when provided in the context of primary care, they have the ability to reach the majority of the population. Thus, the overall impact of behavioral strategies becomes far greater when applied on this broader scale and has the potential to significantly improve public health (Whitlock et al.). Behavioral interventions are typically brief due to the time-limited nature of primary care, as consultation appointments usually last minutes (Strosahl & Robinson). The number of sessions is also typically low, with an average of 1-3 sessions per patient. With such limited time devoted to each individual patient, a typical BHC sees 12 or more patients on an average day (Strosahl & Robinson). Such frequent contact with PCPs has an additional benefit in that it allows BHCs to teach PCP s core behavioral interventions, increasing the likelihood that PCP s will utilize these skills with all their patients (Strosahl & Robinson).

57 50 In a typical intake appointment, the BHC first consults with the PCP regarding consultation goals (Hunter et al., 2009). Consultation goals can cover a wide range of issues, from comorbid mood disorders to poor treatment adherence to difficulties coping with chronic pain. Optimally, the PCP will do a warm hand-off where s/he introduces the BHC as a part of routine care, thus reducing any stigma associated with receiving mental health services (Strosahl & Robinson, 2008, p. 88). With consultation goals in mind, the BHC meets with the patient for minutes gathering information, possibly conducting focused assessments, and codeveloping a treatment plan with the patient (Hunter et al.). After the appointment the BHC provides written and/or verbal feedback to the PCP regarding the patient s functioning, assessment of consultation goals, and co-developed treatment plan. As necessary, the BHC may be involved in follow-up efforts with the patient, including phone calls and scheduling future sessions. For patients with common chronic medical conditions such as diabetes, hypertension, obesity, and chronic pain, the BHC may be involved in the patient s care for months or years (Hunter et al.). The BHC does not just inform the PCP what s/he will be doing with the patient, but also makes recommendations to the PCP on what s/he could be doing to further support the patient and the co-developed treatment goals. The structure of a behavioral health session is guided by the 5 A s model. This model, originally consisting of only four A s, was developed by the National Cancer Institute to guide smoking cessation (Whitlock et al., 2002). Later, a fifth A (arrange ) was added by the Canadian Task Force on Preventative Health Care. The 5 A s, assess, advise, agree, assist, and arrange have been applied to clinical trials and brief primary care interventions covering numerous behaviors (Hunter et al., 2009; Whitlock et al.). This model is considered flexible, patient-

58 51 centered, and applicable to every patient and problem (Hunter et al.). Taken from Hunter et al. and Whitlock et al., each phase entails: Figure 2.1. From Integrated Behavioral Health in Primary Care: Step-by-step Guidance for Assessment and Intervention, by C. L. Hunter et al., 2009, p. 6. Copyright 2009 by the American Psychological Association. Reproduced with permission. The use of APA information does not imply endorsement by APA. No further reproduction or distribution is permitted without written permission from the American Psychological Association. Assess: Gathering information regarding behavioral health risks, physical symptoms, cognitions, and relevant environmental variables (e.g., family, friends, work environment). The goal for the BHC is to understand what variables may be impacting the patient s health and functioning and identify which variables the patient is willing or able to change in order to decrease symptoms and improve functioning.

59 52 Advise: The BHC informs the patient of intervention options based on the earlier assessment. Patients are given specific, personalized, and clear advice about their health risks and benefits of change. The goal of this phase is for the BHC to describe the proposed intervention and expected outcomes. Agree: In this stage the BHC and patient collaboratively agree on treatment goals and an action plan. The goals and plan may be based on the BHC s recommendations or created by the patient. Whatever is selected, it is based on the patient s willingness to change. Assist: This phase involves the actual provision of the intervention strategies designed to help the patient meet their treatment goals. Strategies include providing information, teaching new skills and problem-solving techniques, increasing social support and self-efficacy, and even recommending assessment for psychopharmacological treatment. Arrange: At the end of the session, the BHC informs the patient if follow-up is necessary and if so, with whom and when. Follow-up may be with the BHC, the PCP, or with a referral, such as to specialty mental health treatment. During follow-ups the BHC provides further support and assistance. The treatment plan may also be modified. Benefits of integrated health care. As previously mentioned, there are two obvious benefits of an integrated primary care model, in that it allows mental health services to reach a broader range of the population and that it affords PCPs additional training in behavioral interventions (Robinson & Strosahl, 2009; Whitlock et al., 2002). However, there are many more benefits of integrated health care that may not seem as obvious. Integrated care allows for patients to receive more services and reduces the number of referrals medical providers need to make to specialty mental health referrals (Hunter

60 53 et al., 2009). This, in turn, affords improved collaboration between providers, as integrated treatment plans are unlikely between medical providers and specialty mental health providers. As previously discussed, Mumford et al. (1984) found that patients who received mental health treatment were more likely than controls to utilize less inpatient medical care. This lead to a reduction in health care costs, one of the largest selling points of integrated health care. Therefore, it should not be surprising that previous research has found that not effectively addressing behavioral and mental health needs leads to higher medical costs, as well as poorer medical, functional, and behavioral outcomes (Strosahl & Robinson, 2008, p. 85). The PCBH model, in particular, has also been shown to be highly adaptable. For example, the U.S. military and large HMO s such as Kaiser Permanente has employed it. It has been shown to be effective in addressing the needs of both common primary care conditions and in managing chronic illnesses (Blount & Miller, 2009; Strosahl & Robinson). It has even been shown effective for common mental health concerns like depression. Integrated care has been found to improve adherence, treatment satisfaction, and depressive outcomes for patients with major depression (Katon, 1995). For patients with minor depression, there was a lack of primary gains but secondary gains were found, as patients reported increased adherence and treatment satisfaction (Katon). Barriers to integration. Much of the literature discusses barriers to integrated care and these are important to mention, as the present study would undoubtedly confront some of these. Barriers include physicians lack of time due to high caseload demands, physicians limited training in behavioral interventions, a lack of mental health professionals trained in behavioral medicine, patients

61 54 stigma of mental health services, a lack of knowledge on how to locate services, and concerns of financial cost (Blount & Miller, 2009; Strosahl & Robinson, 2008; Wong & Marshall, 2010). The PCBH model is one obvious way to overcome many of these barriers. Common behavioral interventions. Hunter et al. (2009) outline nine empirically-supported interventions they believe form the core of a BHC s practice. These interventions are designed to be utilized by the BHC in the brief timeframe of a behavioral health appointment with the goal that patients can practice these skills in their daily lives. Patients that are unable or unwilling to practice these interventions outside the appointment are referred to specialty mental health services. Each of these nine interventions have been shown to be effective for addressing the symptoms and functional impairments commonly seen in primary care, although it is reasonable to expect them to also be applicable to the same concerns seen in secondary care. The nine interventions are: (a) relaxation training, (b) goal setting, (c) cognitive disputation, (d) motivational interviewing (MI) strategies, (e) problem solving, (f) self-monitoring, (g) behavioral self-analysis, (h) stimulus control, and (i) assertive communication. Interventions (d) through (g) are considered especially useful in enhancing patient motivation and treatment adherence. Each of these interventions will be briefly summarized from Hunter et al. (a) Relaxation Training: This core behavioral strategy is considered very versatile, can be easily taught and learned, and can show an immediate effect on physiological symptoms. Relaxation exercises are broken down into three categories, breathing, muscle relaxation, and visual imagery exercises. The

62 55 authors provide written scripts, patient handouts, and sample exercises for each category. (b) Goal Setting: Patients typically come to a medical appointment with physical complaints in mind, although as previously discussed, it is not always possible to find physical causes for these complaints. Goal setting allows patients to set clear, realistic, and concrete goals to address their complaints. After the BHC assists a patient in setting goals, it is recommended that these goals be written down in behavioral terms, along with the strategies that will be utilized and the target dates for completion. It is important to note that goals should always be co-created, as a patient is unlikely to achieve a goal s/he feels is unimportant or unobtainable. (c) Cognitive Disputation: While cognitive behavioral therapy (CBT) is unquestionably the predominant theoretical orientation in medical and behavioral settings, not all aspects of CBT are amenable to BHC use. However, many CBT techniques are both applicable and effective in brief interventions. The BHC will want to explore the thoughts and beliefs the patient holds that may be contributing to the patient s distress. This assessment phase can also be broken down into three cognitive areas to explore: predictions, expectations, and evaluations. The goal is to encourage patients to think critically about how their thoughts and behaviors may be playing into their complaints. One common strategy is to provide patients with a thought-behavior-consequence log to complete. The BHC needs to carefully evaluate a patient s ability to be insightful and think critically about the

63 56 influence of their thoughts and behaviors. For patients that are not able to demonstrate this, a referral to specialty mental health may be warranted. (d) Motivational Interviewing: In recent years, much has been written about the benefits and application of MI strategies in the provision of both medical and mental health services. Patient nonadherence has been found to be a complex and multifactor concept. Patients may struggle with treatment adherence due to complex treatment plans, lack of knowledge, understanding, or self-efficacy, poor patient-provider communication, treatment side effects, and psychosocial barriers. After the BHC identifies the patient s likely barriers to adherence, it is recommended the BHC emphasize with the difficulty of the situation and offer to work on overcoming these barriers together. MI strategies are designed to assist patients in recognizing problems, address the nature of ambivalence, eliciting selfmotivation for change. Three common MI strategies are examining readiness to change, the importance of change, and self-efficacy for change. The authors provide written scripts, patient handouts, and sample exercises for these strategies. (e) Problem Solving: This intervention involves teaching patients a valuable skill that they can apply to future situations. Hunter et al. suggest a seven-step model that contains the following steps: define the problem, brainstorm possible solutions, critically evaluate each possible solution, select the best option, implement the chosen solution, assess the outcome, and fine-tune the solution as necessary. The authors provide a patient worksheet that outlines this model.

64 57 (f) Self-monitoring: Teaching patients this intervention offers them a method for tracking their progress towards a goal, staying focused on change, and rewarding progress. This intervention may employ several methods, such as reporting behaviors on a calendar, keeping tallies of behaviors, and charting behaviors on a graph. (g) Behavioral Self-analysis: This intervention employs the ABC model, where A stands for antecedents, B stands for the behavior itself, and C stands for consequences. A BHC can guide a patient through the ABC model, asking the patient to give examples of each step. As this model helps identify barriers to treatment, together the patient and BHC can decide on methods for addressing these barriers. (h) Stimulus Control: This is very similar to the ABC model above. It asks patients to identify a stimulus that occurs before the target behavior, thus allowing steps to be taken to alter this behavior into a desired behavior. An example of this is asking a patient wishing to change the behavior of eating junk food what triggers the behavior to start eating junk food. The patient would then list these factors, categorizing them into behaviors, emotions, thoughts, people, and environment in a manner similar to the CBT log. The authors provide an example stimulus control log to help patients track these variables. Finally, the BHC assists the patient in deciding on methods to overcome the identified triggers. (i) Assertive Communication: Communication difficulties are rarely a patient s presenting concern, although issues related to this can negatively impact a

65 58 patient s health. It can hamper the physician-patient relationship, exacerbate symptoms, reduce functioning, and impede goal attainment. In order to help patients identify maladaptive communication patterns, the authors recommend the use of two models, the HARD model and the XYZ formula. The HARD model stands for honest, appropriate, respectful, and direct. Using this acronym, the model shows the healthy and unhealthy aspects of nonassertive, assertive, and aggressive communication styles. The XYZ formula helps patients connect how they feel with how they behave. Patients answer the questions When I feel X, When you do Y, In situation Z, I would like. By connecting feelings and behaviors, patients can identify behaviors they wish to change. PCBH in a secondary care clinic. Since the field s seminal article was published in 1965, the literature on integrated health care, the PCBH model, and the mind-body connection has blossomed. The PCBH model has expanded to the U.S. Air Force and Navy, the VA system, and large HMOs such as Kaiser Permanente (Strosahl & Robinson, 2008). Despite the proliferation of this field, scant literature exists on integrated health care in a secondary care clinic setting. The reasons for this are unclear. With no literature discussing this phenomenon to draw hypotheses from, any ideas are pure speculation. But anecdotal evidence might suggest that the nature of secondary care makes it an unlikely place for the birth of integrated health care. In many secondary clinics, especially those surgical in nature, patients may only be seen once or twice by the surgeons, limiting any potential benefits that arise from long-term relationships. Due to their limited interpersonal interaction with patients, some of who may not even be conscious when they meet due to

66 59 emergencies, may encourage surgeons to value technical skills over interpersonal skills. Therefore, it may possible that surgeons are the least psychologically-minded of all medical providers. There is some research support for this, as Thombs, Adeponle, Kirmayer, and Morgan (2010) found that compared to other physicians, surgeons had less-positive attitudes towards managing mental health problems and engaging in psychiatric consultation. Whatever the reasons, secondary care remains a forgotten but important direction for integrated care. A large body of research supports the existence of mental health and QOL decrements for the conditions treated in secondary care, making it all the more surprising this area of medical care has been so overlooked. However, some research does exist on the application of an integrated care model to secondary care. The most relevant example to the present study takes place in another surgical area, oral and maxillofacial surgery. This is similar in that it involves a surgical setting, although these patients are presenting primarily with traumatic injury, as opposed to a chronic medical condition. Wong and colleagues (Wong, Marshall, & Shetty, as cited in Wong & Marshall, 2010 [original source not published]) surveyed medical providers and orofacial trauma patients at a large, urban level-1 trauma center, the Los Angeles County and University of Southern California Medical Center. Goals of the study included both provider and patient (a) interest in establishing a collaborative care model in a surgery setting (b) perceived barriers to its establishment. Patients were screened for three of the most common issues this population faces, PTSD, major depression, and alcohol use disorder (AUD). In their sample, they found that 35% met the criteria for major depression, 34% met the criteria for PTSD, and 31% met the criteria for AUD. Perhaps unsurprisingly, comorbidity rates were high among those meeting the criteria for at least one disorder, with 80% meeting the criteria for at least 2 disorders and 50% meeting

67 60 the criteria for all three disorders. However, despite the overwhelming need for mental health services, only 8% reported receiving any ongoing treatment, and for all of these patients treatment had begun before their injury. With this need in mind, the authors interviewed patients meeting the criteria for at least one disorder regarding their interest in a psychologically-focused aftercare program. Patients expressed a great deal of interest, with 48% stating they were very interested and with only 16% stating they were not interested. When asked about perceived barriers to using this aftercare program, the two most common reported were lack of knowledge about where to find services (81%) and concerns about financial costs (71%). Surgeons were similarly surveyed (Zazzali et al., 2007, as cited in Wong & Marshall) and they strongly agreed there was a need for an aftercare program that addressed patients psychosocial needs. Surgeons also agreed that their department could be doing more to address patients needs and that they recognized the benefits of such a program. However, when they were asked how receptive they would be to such an aftercare program placed in their own department, surgeons were surprisingly less enthusiastic, rating themselves as only moderately receptive. Colleagues and administrative staff were even less enthusiastic about the placement of the aftercare program in their department. Finally, when asked where the best location was for this program, surgeons overwhelmingly said specialty mental health and community-based settings rather than something in the surgical setting. While these results are hard to explain, they may be related to surgeons citing structural and attitudinal barriers to the implementation of this program as more salient than patients (Chandra et al., 2008 [original source not availale], as cited in Wong & Marshall). Surgeons also tended to underestimate their ability to influence patients behaviors, with more patients stating they would be influenced by their doctor than the doctors believed.

68 61 Thus, Wong and Marshall conclude that educating surgeons on the considerable influence they have on patients to change their health behaviors could lead to more patients receiving psychosocial services. While Wong and Marshall (2010) describe the most relevant application of the integrated care literature to the present study, it falls short of providing any empirical data on the actual implementation of such a service. Because of this lack of relevant literature, one similar study is worth mentioning. Ranatunga et al. (2010) describe an integrated care implementation in a rheumatology clinic, although it differs in scope from the present study. As opposed to utilizing the PCBH model, this study followed the chronic care model. Briefly, this model has six components, self-management support, delivery system design, decision support, clinical information systems, and interaction with the healthcare system and the community. This model differs from the PCBH model in many ways. First, it is guided by theories of patient self-efficacy and places an emphasis on patient self-management. Second, it does not involve mental health professionals. Instead, it employs physicians to do the bulk of intervention by utilizing techniques such as motivational interviewing. Nurses receive the primary responsibly of conducting follow-up phone calls, assessing the success of the intervention, and modifying treatment goals. Third, sessions tend to be much longer, with initial appointments ranging from minutes and follow-up appointments lasting minutes. This study s results are difficult to interpret, as the authors state that patients baseline scores already exceeded their intervention goals and that patients had already undergone numerous previous therapeutic interventions. This led the authors to suggest that perhaps this intervention would be best for newly diagnosed patients. While they did find some benefits, such as some improved measures

69 62 of osteoarthritis, amount of time exercising, and increase self-efficacy, other important goals were not met. However, changes in measures of osteoarthritis and exercise did not significantly improve and only scores of self-efficacy saw statistically significant improvement. In addition, post-intervention the clinic found it difficult to maintain aspects of the intervention, particularly the follow-up phone calls, bringing the long-term feasibility of this model into question. As this brief review of the literature of integrated care in secondary care clinics demonstrates, there is much work yet to be done. Wong and Marshall s (2010) report of several related studies suggests that patients receiving medical treatment in secondary care do present with mental health and QOL decrements, and that they are receptive to receiving mental health services. However, patients and surgeons readily acknowledge the numerous barriers to the broader provision of mental health services in secondary care. They also report on the surprising finding that, at least in their surgical clinic, surgeons and other employees are only marginally accepting of co-locating mental health services in house. It may be that mental health professionals need to do further education with their medical colleagues to demonstrate the benefits of co-located and integrated services. Finally, while the glaring theoretical and methodological limitations of Ranatunga et al. (2010) require caution in their interpretation, it does provide evidence of a movement towards providing mental health services in secondary care. Providing further support, Strosahl and Robinson (2008) highlight the PCBH model s ability to improve the lives of those living with chronic diseases, suggesting that the model is applicable to patients with urolithiasis.

70 63 Chapter III. Method Participants One hundred and five (N = 105) adult patients from the UW Health Urology Clinic who are recurrent stone formers (patients that have >1 kidney stone event) or are at high-risk for recurrence (e.g. abnormal 24-hr urine analysis) consented to participate in the following study. Three (n = 3) requested to be removed from the study without explanation, for a total number of 102 participants (N = 102). Fifty-two participants (n = 52) were randomized to the control and fifty (n = 50) were randomized to the intervention group. Design This study employed a pretest-posttest control group design wherein participants were randomly assigned to either a control (treatment-as-usual) or treatment group. Both prior to and after the intervention all participants completed the paper-and-pencil measures. For those patients that had stone-related biomarker and imaging data already contained in their charts, this information was reviewed pre- and post-intervention as well. Specific Aims and Hypotheses As this study was largely exploratory in nature due to no previous applications of the PCBH model to a kidney stone population, a greater emphasis was placed on achieving specific aims than a priori hypotheses. However, this study did hope to achieve two specific aims along with three a priori hypotheses. The study had two specific aims: (a) determine the feasibility of adapting the PCBH model to patients receiving care for kidney stones in a secondary care clinic, and (b) investigate the effects of the intervention on stone formers (SF) health-related QOL and lithogenic outcomes. Details are presented in Table 3.1. All power calculations are based on a

71 64 standard estimated effect size of.50 (moderate) for a psychological intervention.the study s four hypotheses for providers and patients were as follows: Hypothesis 1: Psychological collaboration would be positively viewed by providers and increase their psychological-mindedness. Utilizing a modified version of the Doctors Attitudes Toward Collaborative Care for Mental Health (DACC-MH; Thombs et al., 2010) Scale, this hypothesis was tested by measuring urological providers pre- and post-intervention attitudes towards psychological consultation and management of mental health problems. It was hypothesized that providers attitudes towards both would improve from baseline to post-intervention. (b.) Utilizing an author-designed survey, provider satisfaction with the urologic care they provide was measured at baseline and postintervention. It was hypothesized that providers would be more satisfied with the level of care they provide post-intervention after having a BHC involved. (c.) The same survey was also used to measure provider attitudes towards working with a BHC, with it being hypothesized that attitudes would become more favorable post-intervention. Hypothesis 2: The PCBH would be successfully adapted to patients with kidney stones and positively viewed. Utilizing an author-designed survey, the following were also measured: (a.) patient satisfaction with urologic care, (b.) patient knowledge about stone disease and prevention, (c.) patients self-efficacy in managing their stone disease, and (d.) success of the BHC in working with patients. Surveys were administered at baseline (pre-intervention) and at 3 months (end-ofintervention). (e.) The number of patients the BHC was able to reach during the intervention and attrition rates were also used as a further measure of feasibility. It was expected that the BHC

72 65 would reach the minimum number of patients needed for appropriate statistical power (N = 100; see Table 3.1) and that attrition would be lowest for those in the intervention condition. Hypothesis 3: The BHC intervention would improve patients biological risk factors for urolithiasis compared to treatment-as-usual. Biological risk factors were analyzed by conducting a chart review of 24-hour urine analysis results, a test that measures thirteen lithogenic risk factors. While the majority of patients would have this lab data, it was understood that this data might not be available for all patients. Hypothesis 4: The intervention would improve patients health-related QOL compared to treatment-as-usual. Health-related QOL was assessed on six dimensions using validated gold standard paper-and-pencil instruments: (a.) depression (Brief Patient Health Questionnaire; Brief PHQ; Spitzer, Kroenke, & Williams, 1999), (b.) panic syndrome (Brief PHQ), (c.) pain (Short-form McGill Pain Questionnaire; SF-MPQ-2; Dworkin et al., 2009), (d.) stress (Perceived Stress Scale; PSS-10; Cohen & Williamson, 1988), (e.) life satisfaction (Satisfaction With Life Scale; SWLS; Diener, Emmons, Larsen, & Griffin, 1985), (f.) stone-former QOL (WI Stone-QOL; Penniston, 2010). These measures were given at baseline (pre-intervention) and after 3 months (post-intervention).

73 Table 3.1 Summary of Hypotheses and Outcome Measures Hypothesis Outcome Measure Time N Effect Size (Cohen s d) 1a. Urological providers attitudes towards collaboration & MH management 1b. Provider satisfaction with urologic care 1c. Provider attitudes towards working with a BHC 2a. Patient satisfaction with urologic care 2b. Patient knowledge about stone disease & prevention 2c. Patients self-efficacy in managing their stone disease Doctors Attitudes Toward Collaborative Care for Mental Health (DACC- MH) Scale Author-designed survey Author-designed survey Author-designed survey Author-designed survey Author-designed survey A & B intervals* A & B intervals* A & B intervals* A & B intervals* A & B intervals* A & B intervals* 2d. Success of the BHC Author-designed survey B interval* 2e. Feasibility of adaptation: intervention condition would have a lower attrition rate Review of attrition rates B interval* 3. Biological risk factors 24-hour urine analyses A & B intervals* 4a. Depression Brief PHQ A & B intervals* 4b. Panic syndrome Brief PHQ A & B intervals* 4c. Pain SF-MPQ-2 A & B intervals* 4d. Stress PSS A & B intervals* 4e. Life satisfaction SWLS A & B intervals* 4f. Stone-former QOL WI Stone-QOL A & B intervals* Note: *A = baseline B = end-of-intervention (3 mos.) d = Cohen s d effect size Materials Est. Power At baseline and post-intervention, some participants had a biomarker of stone parameters (lithogenic risk) in their medical record as a part of their standard urologic care. Stone 66

74 67 parameters were measured using a common reliable and valid measure, 24-hour urine analysis results. This test required patients to collect their urine for 24-hours, after which a small sample was then sent to the lab. The test typically measures the following thirteen lithogenic risk factors: calcium (mg./day), oxalate (mg./day), uric acid (mg./day), citrate (mg./day), ph, total volume (L/day), sodium (meq./day), potassium (meq./day), sulfate (mmol./day), phosphorus (mg./day), magnesium (mg./day), ammonium (meq./day), and creatinine (mg./day) (Pak, Peterson, & Poindexter, 2001). In a sample of 225 recurrent kidney stone formers on their usual diets, Pak et al. analyzed the results of two random 24-hour urine analyses to measure reliability. Unfortunately, the authors do not list the average time between the two random tests, although the time interval was one week in other subsequent tests. They found that for eleven of the thirteen risk factors, there was no significant difference between the two tests, with an r >.68. For each of the thirteen factors, Pearson correlations between the two tests ranged from.56 to.94 and all were without statistically significant differences, except for magnesium and ammonium. For patients with abnormal or normal findings at the first testing, these findings were reproduced in 81% of patients at the second testing. The authors conclude that 24-hour urine analyses over a short time show few differences in mean values, are highly positively correlated, and are capable of detecting abnormal or normal results. It was expected that almost all patients would have 24-hour urine results available in their chart at baseline, as this is a part of standard care. Baseline 24-hour urine results were compared to the patients test results post-intervention. It was expected that many patients would already have the results of a second 24-hour urine in their charts by post-intervention. Research suggests that multiple 24-hour urine analyses can improve diagnostic accuracy (Pak et al., 2001).

75 68 Written Measures A copy of all written measures can be found in Appendix A for providers and Appendix B for participants. The author has obtained permission to utilize each of these measures by their respective copyright owners for those measures that are not in the public domain. Doctors Attitudes Toward Collaborative Care for Mental Health Scale (DACC-MH; Thombs et al., 2010) is a brief, 8-item scale designed to assess doctors attitudes towards two core components of collaborative care, Attitudes Towards Psychiatry Consultation and Attitudes Toward Management of Mental Health Problems. However, with the authors permission, the first component has been changed in the present study to attitudes towards psychological consultation to reflect the shift in providers from psychiatry to psychology. Each component subscale consists of four items. A sample item from the consultation component is, I would welcome more contact with psychologists. A sample item from the management component is, Management of emotional problems is an important part of my care of chronic outpatients. Each item is rated on a dichotomous scale consisting of agree or disagree. Raw subscale scores are obtained by simply counting the number of agrees and two items are reversed scored (disagree is counted), items six and eight. As this scale is very new, there is only one study on its reliability and validity to date, the authors original validation study. In this study the authors administered the DACC-MH to 225 medical doctors in a London hospital, of which 138 where physicians and 87 were surgeons. Confirmatory factor analysis resulted in the removal of several items and results show a model with good fit. Correlation between the two subscales was.49. The Cronbach s alpha for the total scale was.70,.67 for the Consultation subscale, and.65 for the Management subscale. Internal consistency reliability at pre-intervention (n = 6) for the

76 69 present sample was α =.44 for the total scale,.47 for the Consultation subscale, and.54 for the Management subscale (Table 4.4). Internal consistency reliability at post-intervention (n = 6) for the present sample was α =.48 for the total scale,.88 for the Consultation subscale, and.82 for the Management subscale (Table 4.4). In the present study, test-retest reliability was r = -.41, p =.42 for the Consultation subscale and r =.08, p =.86 for Management subscale. The authors also conducted a known-groups validity test, and as expected from previous research, physicians scored more positively on both subscales than did surgeons. On the Consultation subscale, physician mean was 3.1 (SD = 1.1) while the surgeon mean was 2.7 (SD = 1.4). On the Management subscale, the physician mean was 3.5 (SD =.9) while the surgeon mean was 3.1 (SD = 1.2). Thombs et al. list several limitations of the scale, include the dichotomous rating scale, no linking of doctor ratings to their actual behaviors, and the lack of data on its use in various settings. To address the first limitation, the present study created a 7- point Likert scale, with 1 = strongly disagree and 7 = strongly agree. Items six and eight (management subscale) remain reverse-coded. A mean score was then calculated for each subscale and the total scale, with higher scores representing greater interest in consultation and managing mental health concerns, respectively. The authors conclude that it is a good measure of the two core components of collaborative care and that the scale may be useful to evaluate medical doctors readiness to engage in collaborative care. Short-Form McGill Pain Questionnaire-2 (SF-MPQ-2; Dworkin et al., 2009) is a 22-item measure of major symptoms of both neuropathic and non-neuropathic pain consisting of four dimensions: continuous pain, intermittent pain, neuropathic pain, and affective descriptors. Each of these dimensions consists of six items except the last, which consists of four. There are no

77 70 reverse coded items. To score each subscale, a mean of the item ratings of the scale is calculated. A total score can be calculated by taking a mean of all item ratings. The lower the score, the less pain a patient is experiencing. An example of a continuous subscale item is throbbing pain, an example of an intermittent subscale item is shooting pain, an example of a neuropathic pain subscale item is hot-burning pain, and an example of an affective descriptors subscale item is tiring-exhausting. The SF-MPQ-2 is based on both the original MPQ, widely considered the gold-standard in pain scales, and the first version of the Short Form-MPQ (SF-MPQ). There is a considerable body of literature documenting the use, reliability, and validity of these measures. This revised, second version of the SF-MPQ was created to capture neuropathic pain, a dimension previously lacking. The authors also revised the rating scale, choosing a 0-10 format, in order to make this scale more relevant for longitudinal studies and clinical trials. SF-MPQ-2 was validated on a sample of 882 adult patients with a wide range of pain disorders and on 226 patients with painful diabetic peripheral neuropathy. The non-diabetic neuropathy participants were recruited using a web-based survey posted on the American Chronic Pain Association (ACPA) website while the patients with painful diabetic neuropathy were recruited through a randomized clinical trial. A confirmatory factor analysis demonstrated a good fit of the four subscales. The Cronbach s alpha for the total scale ranged from.91 to.95, alpha ranged from.73 to.87 for the continuous pain subscale,.85 to.87 for the intermittent pain subscale,.78 to.83 for the neuropathic pain subscale, and.77 to.86 for the affective descriptors subscale. Internal consistency reliability for the present sample was α =.91 at pre-intervention and α =.94 at post-intervention. The α coefficients for the completers are presented in Tables 4.8. This scale has also demonstrated

78 71 discriminant and construct validity by comparing the relevant subscales with other appropriate measures. Test-retest data was not calculated in this study, but it has been previously calculated for the SF-MPQ. Grafton, Foster, and Wright (2005) administered it five days apart with patients with osteoarthritis awaiting surgery and calculated high intra-class correlations for the subscales ranging from However, it should be noted that these results are not directly comparable with the SF-MPQ-2 as the two versions utilize slightly revised subscales. In the present study, test-retest reliability was r =.53, p =.00. Brief Patient Health Questionnaire (Brief PHQ; Spitzer et al., 1999) is a shortened 8-item (with numerous sub-items) version of the longer PHQ. It is primarily designed to detect depression and panic disorder, as it consists of the PHQ-9 s depression screen, a panic measure, and items on stressors and women s health. Participants rate their responses on a variety of forced-choice responses and complete one open-ended question, What is the most stressful thing in your life right now? For the depression screen, it asks how bothered an individual has been in the past two weeks by nine items. These items compromise the diagnostic criteria for a depressive episode and include items such as, little interest or pleasure in doing things. Participants rate these nine items on four-choice Likert scale ranging from not at all to nearly every day. It then asks one item regarding panic, In the last 4 weeks, have you had an anxiety attack suddenly feeling fear or panic?, rated yes or no. If the answer is yes, four follow-up items are also asked to be completed. The next item asks for a global sense of how difficult these problems have been for a person. The following question asks participants to rate in the last four weeks if they have been bothered by a list of 10 stressors. An example is, worrying about your health, rated by three Likert choices ranging from not bothered to bothered a lot. The

79 72 remaining questions ask about possible abuse, global stress, and women s health. This variety in response styles is a useful aspect of the Brief PHQ, as it allows for both categorical and dimensional analyses (Hahn, Reuter, & Härter, 2006). Scores for the nine-item depression screen are scored on a scale from 0-4, with possible total scores ranging from 0 to 27. A score up to 5 is considered mild depression, a score of 10 represents moderate, a score of 15 represents moderately severe, and a score of 20 or more represents severe depression (Spitzer et al., 1999). For scores considered moderate or above, treatment is recommended. The four items that measure panic are only considered to represent a positive case of panic disorder if yes is rated for all four items. The questions regarding stressors and women s health are not considered diagnostic and are not scored. However, they remain clinically useful. Research has found the Brief PHQ to have excellent reliability and validity (Spitzer et al., 1999). As one of the most common screeners, there is a vast body of research supporting its clinical usefulness, adaptability to various conditions, and strong reliability and validity. In its initial reliability and validity study (Spitzer et al.) it was administered to 3,000 patients across the U.S. Of relevance to the current study, the PHQ was found to not under- or overdiagnose psychiatric disorder and the correlation between a PHQ diagnosis and the independent diagnosis of a mental health professional was found to be r =.84. Thus, the authors conclude the PHQ has valuable diagnostic validity. However, Spitzer et al. do not provide alpha coefficients for unknown reasons. Internal consistency reliability for the present sample was α =.75 at preintervention and α =.80 at post-intervention. The α coefficients for the completers are presented in Tables 4.8. In the present study, test-retest reliability was r =.24, p =.01.

80 73 Hahn et al. (2006), in a study of convergent and discriminant validity for the Brief PHQ, found it to discriminate well and even surpassed the specificity of other convergent measures. Recent investigations have found it particularly useful, especially when compared to other frequently used screeners (Rief, Nanke, Klaiberg, & Braehler, 2004) for the detection of mental illness in patients with chronic illnesses (Hahn et al., 2006). The Satisfaction With Life Scale (SWLS; Diener et al., 1985) is a brief 5-item self-report scale that measures a general sense of life satisfaction. Life satisfaction is defined as a judgmental process in which individuals assess the quality of their lives on the basis of their own unique set of criteria (Shin & Johnson, 1978, as cited in (Pavot & Diener, 2009, p. 102). The SWLS does not assess life satisfaction in specific domains but rather allows respondents to subjectively weight their own relevant life domains (Pavot & Diener). An example item is, In most ways my life is close to my ideal. Participants rate the degree to which they agree or disagree with the 5 items using a 7-point Likert response scale where 1 = strongly disagree and 7 = strongly agree. Scores can be calculated to yield both absolute and relative measures of life satisfaction. A score between 5 to 9 is indicative of being extremely dissatisfied with life, 15 to 19 slightly dissatisfied, 20 is considered neutral, 21 to 25 slightly satisfied, and 26 to 30 satisfied. Most individuals fall in the slightly satisfied to satisfied range (23 to 28). Over twenty years of research have found this scale to be applicable to a wide range of populations, including both inpatients and outpatients (Pavot & Diener, 2009). For three medical norms relevant to the present study, disabled students have had means between 20.8 to 24.3 (SD 8.4 to 7.4), inpatient veterans a mean of 11.8 (SD 5.6), and outpatient means between 19.3 to 25.2 (SD 6.7 to 7.0). It has been found to have good convergent validity with other assessments

81 74 of subjective-well being. In addition, it has demonstrated discriminant validity with measures of emotional well-being. Factor analysis results found a single factor that accounted for 66% of the variance. The Cronbach s alphas in numerous studies have ranged from.79 to.89. Internal consistency reliability for the present sample was α =.88 at pre-intervention and α =.89 at postintervention. The α coefficients for the completers are presented in Tables 4.8. The SWLS shows sufficient temporal stability. In various studies it has ranged from.83 to a two-week interval to.54 over a 4-year period. However, it also demonstrates sufficient sensitivity to be clinically useful during the course of an intervention (Pavot & Diener). In the present study, test-retest reliability was r =.71, p =.00. Clinical populations have been shown to score lower in life satisfaction, although this increases during the course of treatment. It is recommended that the SWLS be included in a battery of diagnostic or emotional well-being measures, such as in the present study, because it assesses the conscious and subjective evaluation of a respondent s life (Pavot & Diener). Perceived Stress Scale (PSS-10; Cohen & Williamson, 1988) is a short, 10-item measure of how much unpredictable, uncontrollable, and overloaded respondents perceive their lives. The PSS-10 is not situation specific, but instead is more interested in general life circumstances. An example questions is, In the last month, how often have you been upset because of something that happened unexpectedly? The strengths of this scale are its short time to administer and ease of scoring. Participants rate each item on how they have felt in the past month using a five-point Likert rating scale ranging from 0 never to 4 very often. After reverse scoring four items (4, 5, 7, 8), a total score is calculated by adding up all responses, with scores ranging from 0 to 40. The higher the score, the higher the perceived stress. Cohen and Williamson, in their large

82 75 reliability and validity study, interviewed 2,387 adults. They found a mean score of (SD = 6.35) and significant demographic differences. Women reported more stress than men, stress decreased as age increased, stress declined with income, but only to a moderate point, increased education was associated with less stress, and identifying as a racial or ethnic minority was associated with increased stress. Relevant to the current study, the poorer the health of the respondent, the greater stress they reported. The PSS-10 demonstrated construct validity with a number of scales, demonstrating a correlation of.21 (p <.0001) with a measure of health care utilization. The Cronbach s alpha was.78. Internal consistency reliability for the present sample was α =.91 at pre-intervention and α =.90 at post-intervention. Recently, the test-retest reliability was measured in a sample of 793 Brazilian university teachers and scores were found to range from.68 to.86 (Reis, Hino, & Añez, 2010). In the present study, test-retest reliability was r =.56, p =.00. The α coefficients for the completers are presented in Tables 4.8. The PSS-10 is the most widely used psychological instrument for measuring the perception of stress (Sims et al., 2008). It has been utilized with a wide-range of clinic issues from eating behaviors (Sims et al.) to surviving SARS (Lee et al., 2007). As this scale underwent construct validity with measures of health and health care utilization, it is well suited to the present study. Also, as research suggests a possible connection between stress and kidney stones (Diniz et al., 2006; Najem et al., 1997; Walters, 1986), this is an important variable to measure. Wisconsin Stone-QOL (Penniston & Nakada, 2010) is a short, 28-item scale designed to assess the unique health-related QOL of stone forming patients. This scale is the first QOL measure for stone formers, as no previous measure existed (Bergman et al., 2010). Based on extensive focus groups of stone formers, QOL for this population has been conceptualized as

83 76 consisting of the following seven domains: (a.) fatigue/activity level, (b.) sleep, (c.) social life, work, (d.) treatment concerns, general, (e.) pain, physical symptoms, (f.) social and sexual function, (g.) general emotions. Under each of the seven domains, there are between 3-8 statements the participant rates, for a total of 28 ratings. Participants are asked to write how true each item is for themselves in the past four weeks. Each item is rated on a 5-point Likert scale where 1 represents very true and 5 represents not true at all. Although the wording of the scale does change based on the item, 1 always represents the most-inclusive extreme and 5 always represents the least-inclusive extreme (e.g., not at all, never ). A sample statement is, I don t feel the usual freedom to travel or to attend or participate in social events. A total mean score is calculated, with a higher mean score representing greater health-related QOL. Individual domain means can also be calculated if desired. In a recent validation study at the University of Wisconsin Urology Clinic, 99 adult stone formers at or around a stone event had a total mean score of 3.27 (SD not listed). The total scale s Cronbach s alpha was.97. The Cronbach s alpha for each domain was.91. No test-retest reliability figures are available yet, although it is expected these will be available in the near future. However, 59 of the participants were surveyed again one month later, although this survey had a slightly different direction, asking them about their stone experiences in general, not in the past four weeks. The total scale s Cronbach s alpha was.96 and the domain-specific Cronbach s alpha was.90. Internal consistency reliability for the present sample was α =.93 at both pre-and post-intervention intervals. In the present study, test-retest reliability was r =.47, p =.00. The α coefficients for the completers are presented in Tables 4.8. Future factor analysis and tests of convergent and discriminant validity are also planned.

84 77 Author-designed questionnaire. Through an author-designed questionnaire participants were asked a number of demographic questions as well as questions designed to measure some of the above-stated hypotheses. Participants were asked to supply their age, sex, race, sexual orientation, relationship status, level of educational attainment, employment status, household income bracket, number of kidney stones occurrences, stone treatment type(s), number of ER visits, medication usage, perception of disclosure of information by the medical providers, and perception of participation in decision-making. For those in the intervention group, a separate post-intervention survey was created which asked them their satisfaction with the study, how useful and beneficial they found the BHC, and for their recommendations regarding the future use of a BHC in the urology clinic. Although this author-designed questionnaire was based on the research team s prior experience and based in the literature, it was not pilot tested prior to being utilized in this study. Procedure Sampling procedures. It was initially planned to recruit patients from both the UW Hospital and Meriter Hospital, but as the study s implementation progressed, it was learned that data collection would not be possible for logistical reasons at Meriter. Patients already receiving urologic services for their kidney stone(s) at the UW Hospital that met the study s selection criteria were mailed a consent form prior to their urologic appointment. However, patients were instructed to not complete the consent form until they arrived at their appointment and had the opportunity to discuss the study with a member of the research team. Once these participants finished meeting with their urologic provider, and their provider felt they could possibly benefit from the BHC s services, the provider asked the BHC to meet with the patient. Before the BHC

85 78 met with the patient, he asked the provider for a consultation goal(s), such as to assist the patient in making dietary or lifestyle changes recommended by the provider. Once the BHC met with the patient, he introduced the study, answered any questions, and obtained written consent if s/he wished to participate. In order to avoid creating differences between the control and intervention groups through the use of separate consent forms, all participants received the same consent form. However, there was a concern that once patients in the control group realized they were not receiving the intervention, it might lead to differential attrition rates between the two groups. Despite this limitation, it was necessary in order to create equivalent treatment groups. A total of six urologic practitioners (urologists, nurse practitioners, and physician assistants) participated in this study. Figure 4.1 presents a CONSORT diagram regarding patient recruitment and retention. Random assignment. Those patients that consented were randomly assigned to one of two conditions, treatment or control (treatment-as-usual), utilizing the technique of minimization. This technique is similar to stratified randomization but allows for improved balance of participant characteristics. Patients were randomized according to four variables: number of stone occurrences (< 2 or > 2), sex (M or F), clinic site (UW or Meriter Hospitals), and treating urologic provider. Therefore, all participating practitioners had patients that were in both conditions. While this design might have been expected to weaken the effect size for the treatment condition, it was necessary in order to provide a sufficient unit of analysis for power analyses. Thus, in order to obtain a power of.70 (for patient measures; used for practical measures) with an estimated effect size of.50, a minimum N of 100 was required, with half being randomly assigned to each condition.

86 79 Control group. Participants randomly assigned to the control group received treatmentas-usual (TAU) as identified by their medical needs. The TAU for kidney stone patients consists of an appointment with a urologist or other urologic practitioner (e.g., Nurse Practitioner, Physician Assistant, nephrologist), completing an initial 24-hour urine analysis, and at least one follow-up appointment, typically scheduled six-months to a year from the initial appointment. Depending on a patient's medical needs, TAU may also include informational handouts, an evaluation and a meeting with a dietitian, and/or medication. The patients received no further interventions. Patients completed the paper-and-pencil measures during this initial appointment. After the completion of the intervention period (approximately three months), control group participants were mailed the paper-and-pencil measures to complete for a second time as a posttest measure. For those patients that had a second 24-hour urine analysis results in their charts at this time, a pre- post analyses was conducted. Treatment group. Participants randomly assigned to the treatment group received an integrated behavioral health invention in addition to TAU. They also received two follow-up phone calls at 1-month and 3-months post-intervention. The initial in-person intervention was provided by three psychologists-in-training under the supervision of a licensed psychologist experienced in behavioral health interventions. The primary BHC conducted all follow-up phone calls. Integrated behavioral health intervention. As the PCBH model is a collaborative model, it was key that a patient s medical provider and the BHC first establish consultation goals. These goals could vary widely, from comorbid mood disorders to poor treatment adherence to difficulties coping with chronic pain. After consultation goals were established, the

87 80 BHC met with the patient for a brief minute appointment, guided by the Five A s model, during which the behavioral intervention was delivered and consultation goals addressed. See Figure 3.1. After the intervention, two follow-up phone calls were conducted (see Follow-ups section below).

88 81 Figure 3.1: Procedure Invitation and recruitment Eligible patients with an upcoming urologic visit were mailed a study description and consent form prior to their appointment Urology provider identified eligible patient who might benefit from BHC at his/her usual clinic visit Patient invited to participate during visit and provided informed consent Baseline activities Each subject: a.) Completed medical history and demographics questionnaire b.) Completed psychometric assessments c.) 24-hour urine analysis results were reviewed if available Randomization Based on demographic and other data obtained, participants were randomized to 1 of 2 groups: TREATMENT-AS-USUAL (control) INTERVENTION BHC consulted with urologic provider re: medical goals & pt s concerns BHC reviewed pt s history, met with pt to discuss goal(s), completed standardized protocol (5 A s) Chart reviewed (when available) & 1- month follow-up phone call (BHC) Chart reviewed (when available) & 3- month follow-up phone call (BHC) 3-month psychometric measures mailed & second 24-hour urine analysis resulted reviewed if available Consultation. The first step of the intervention was consultation. After each patient in the treatment group met with her/his primary urologic practitioner, the practitioner consulted with the BHC regarding the patient s treatment. Each consultation was in-person and focused on any

89 82 particular concerns the practitioner felt might be impacting the patient s health. In order to identify what would be the most likely issues providers would request consultation with the BHC on, the authors conducted a survey (April 2010) of UW Health urology providers. Results showed providers would request BHC assistance for issues related to chronic pain, noncompliance with recommended diet or behavioral changes, overweight/obesity, coping with a chronic/terminal illness, and non-compliance with medication. In order to address these consultation issues, the BHC met with the patient to jointly form goals, provide information or resources, implement behavioral interventions, conduct risk assessments, and assist with followup or relapse prevention strategies. The BHC did not provide psychotherapy, case management, or social work services nor conduct any psychological assessments beyond the predetermined battery. In consultation with the treating urologic practitioner, the BHC created a list of 1-3 primary consultation goals to be achieved during the intervention. After the intervention, the BHC documented the nature and outcome of the intervention in the patient s electronic medical chart (see Documentation for further details). Although patients might have also been receiving services from an allied professional, such as a dietitian, as part of their overall treatment, the BHC did not directly consult with them. However, the BHC did conduct a chart review of each patient as an aspect of the intervention (both at initial consultation and at follow-ups, when available), thus any patient concerns expressed by an allied professional may have been addressed in the intervention. Prior to starting the behavioral health intervention with the BHC, patients completed the paper-and-pencil measures. The BHC then met with each patient for approximately minutes in order to obtain relevant psychosocial history and to achieve the co-created

90 83 consultation goals. As previously discussed, the 5 As model of behavioral health consultation were followed with each patient to create a standardized method. A sample behavioral health intervention for a patient with a consultation goal of addressing chronic pain issues is presented below. Figure 3.2 Figure 3.2. From Integrated Behavioral Health in Primary Care: Step-by-step Guidance for Assessment and Intervention, by C. L. Hunter et al., 2009, p. 22. Copyright 2009 by the American Psychological Association. Reproduced with permission. The use of APA information does not imply endorsement by APA. No further reproduction or distribution is permitted without written permission from the American Psychological Association. Assess: The BHC greeted the patient, introduced himself, and the reason for their meeting (consultation goal of addressing chronic pain). The BHC then asked the patient a series of questions to gather information regarding experiences with chronic pain, physical symptoms, and any behavioral health risks. The BHC assessed functioning and identified which variables the patient was willing or able to change in order to decrease symptoms and improve functioning. Advise: The BHC summarized that he understands the patient recently underwent ureteroscopy for stone removal and the patient s indwelling stent is causing intolerable pain

91 84 despite opioid medications. He also summarized the patient s frustration with the long wait prior to having a surgery date scheduled and the negative impact this wait has had on the patient s job and finances. To address nonpharmacological options for addressing the chronic pain issue, the BHC suggested two possible strategies. First, he described to the patient the Gate Control model of pain (Appendix C) to help the patient understand what he might be doing to contribute to the experiences of pain. The patient was also given the handout Understanding Chronic Pain. The BHC introduced the proposed intervention with The 5 Steps for Managing Intense Pain Episodes and the Monitoring Pain worksheet. The BHC stated that this worksheet was a simple, effective, and broad tool for helping the patient manage the pain and that other patients have found it very useful. Thus, the BHC conveyed it would very likely be useful to his chronic pain issues. Agree: Together, the BHC and patient collaboratively agreed to follow The 5 Steps for Managing Intense Pain Episodes. They also agreed that the patient would begin completing the Monitoring Pain worksheet. The BHC reinforced the patient s willingness to change and to try some new strategies. Assist: The BHC provided the patient with copies of all the discussed worksheets. Together, they reviewed each of the five steps of managing pain and the BHC asked the patient to give an example of each step in order to gauge patient understanding. The BHC next explained how to accurately fill out the Monitoring Pain worksheet. Arrange: At the end of the session, the BHC informed the patient he would follow-up with the patient in one month via phone. The BHC informed the patient that during the follow-up he and the patient would check-in regarding current functioning and pain, review the completed

92 85 Monitoring Pain worksheet, and assess the application of the five management steps. The BHC also provided the patient with his contact information in case the patient has any questions about the agreed upon treatment plan. Finally, the BHC reminded the patient that there would be a second follow-up phone call in three months. Follow-up calls. Patients in the treatment condition received two follow-up phone calls from the BHC as previously mentioned. The first call took place one month after the in-person intervention. The purpose of this follow-up was to inquire into the patient s health, assessment of patient s adherence to treatment/consultation recommendations, assessment of patient strengths and current needs, assessment of patients stage of change, provision of relevant information or resources, discussion of further recommendations, and a preview of the final follow-up in two months. The purpose of the second and final follow-up call was the same as the first, with the addition of a brief discussion highlighting the end of the intervention, a reminder that the patient may contact the supervising psychologist if s/he has any questions or concerns throughout the remainder of the study, and a brief thank you for her/his participation. Also, the BHC conducted a chart review, when the chart was available, of each patient prior to the follow-up phone calls. This was done in order to gain a more complete picture of a patient s current health and needs, as well as to address any patient concerns expressed by an allied professional. Mental health crises. At any point during the intervention, if the BHC identified the patient is currently experiencing a mental health crisis, the BHC followed applicable legal and ethical standards. A mental health crisis was defined as the patient being in imminent danger of harming her/himself or some one else or the patient s admittance of perpetrating on-going abuse of a child or vulnerable adult. In such incidences, the BHC would have immediately paged the

93 86 supervising psychologist for consultation. Additional steps, as appropriate, might have included admittance to the ER and/or involvement of the police or Child Protective Services. During the course of the study, only one patient expressed a serious mental health concern (passive suicidal ideation) requiring the BHC to consult immediately with the supervising psychologist. In this case, the patient was determined to not be at imminent risk of self-harm and was provided with a list of psychotherapy resources. The BHC also provided additional phone follow-up calls with this patient to ensure the patient was receiving the level of care that he needed. Finally, the patient s referring provider was informed of the patient s status in case the provider was contacted by the patient. Documentation. As was required by hospital policy as well as the standard of care, the BHC documented all patient interactions in the patient s electronic medical chart. In each note the BHC summarized the results of the intervention, including relevant history and functioning, assessment results, clinical impressions, treatment recommendations, and follow-up plans. The note also explicitly addressed the status of each of the consultation goals. Hospital policy determined that follow-up phone calls should not be documented in the patient s electronic medical chart, as these calls were conducted solely by the BHC and were primarily a part of a research protocol. However, the BHC did document in a secure database all aspects of the follow-up calls as outlined in the Follow-up calls subsection. Any additional consultation goals set by the treating urologist were also explicitly documented. As it is the standard for mental health related documentation in a medical setting, all documentation was written in a manner sensitive to a patient s privacy. For example, the BHC might have documented the stress created by a patient s intention to divorce her/his spouse as

94 87 stress related to martial discord that is negatively impacting her/his health. However, all legal and ethical documentation standards were followed, such as in instances of mandated reporting (e.g., suicidal or homicidal ideation). BHC training. The provider of all interventions was one of three BHCs, e.g. an advanced PhD-level student in counseling psychology at UW-Madison. Each student held two degrees in psychology and had successfully completed multiple years of supervised clinical training. The first BHC, one of the study s principle investigators, was an advanced doctoral student who had completed his coursework, passed his preliminary examinations, and had several years of clinical training in a hospital setting. This BHC recruited 70 (n = 70; 68%) of the study s 102 participants and conducted all follow-up phone calls. The study s second and third BHCs were approached to aid in data collection prior to the primary BHC s move from Madison to complete his predoctoral internship. The two additional BHCs were also advanced doctoral students in counseling psychology at UW-Madison. Both students had completed several years of coursework and clinical training. They each had clinical training placements at the hospital where data was collected, authorizing them to be involved in the provision of clinical services. The primary BHC provided training for these additional BHCs that included an extensive discussion of the study and its design, provision of all study materials, and provision of a binder with standardized patient handouts to be utilized during the intervention. The primary BHC or a fellow research team member also personally introduced each BHC to the clinic and the urologic providers. Throughout the course of the study, the primary BHC was available by phone and to support the secondary BHCs with any issues that arose. The second BHC recruited 12 (n = 12; 12%) and the third BHC recruited 20 (n = 20; 20%) of the study s 102 participants.

95 88 Supervision. As the BHC was not a licensed psychologist, the provision of services was supervised by a licensed psychologist according to all applicable legal and professional standards. This included the BHC meeting with the supervising psychologist weekly or biweekly for one hour of individual supervision where patient care and needs were discussed. The supervising psychologist also served as a contact for any patient inquires or concerns, a consultant in handling any patient mental health crises that might have arisen, and the official contact for any patient mental health concerns during the entire course of the study. Finally, the supervising psychologist reviewed and signed all documentation written by the BHC before it was placed in the medical chart. The supervising psychologist was a licensed clinical health psychologist in the State of Wisconsin, affiliated with the UW Hospital and Clinics, and had extensive experience in the provision of behavioral health interventions. Incentives. As this study was not grant funded, it was not possible to offer patients financial incentive. However, participants were reminded during the study that their participation would add to our knowledge of how to best address stone patient s needs. Patients were not billed in any way (i.e. no third-party billing) for their participation, other than the charges incurred as part of their regular medical treatment of their kidney stone(s).

96 89 Chapter IV. Results Patient Demographics One hundred and five (N = 105) recurrent stone formers were recruited into this study. Three (n = 3) requested to be removed from the study, for a total number of 102 participants (N = 102). Fifty-two participants (n = 52) were randomized to the control and fifty (n = 50) were randomized to the intervention group. Of the 102 participants, 56 returned the post-test survey packet for a 55% response rate. Figure 4.1 presents a CONSORT diagram of the number of patients that were eligible to participate, contacted to participate, who consented to participate, and those who completed the study. Categorical demographic variables, as well as chi-square and odds ratio results, are presented in Table 4.1 for those that completed the study ( completers ) and those that did not return the final survey ( attritors ). All polychotomous variables (e.g., number of previous stones) were dichotomized in the most clinically meaningful manner. Table 4.2 lists the descriptive statistics and Cohen s d results for the continuous variables for the completers and attritors. In most respects, the completers and attritors were demographically similar and had comparable scores on the continuous variables. Demographically, there were two exceptions that negatively impact generalizability. Retirees were less likely to be an attritor and those with four or more past stone-related ER visits were less likely to be a completer (see Table 4.1). All independent-samples t-tests between the completers and attritors were non-significant. Cohen s d results were in direction of completers reporting greater satisfaction with their urologic care, fewer symptoms of depression, and greater stone-related QOL at baseline than did the eventual attritors, although the differences were not significant (see Table 4.2). Also, there was no

97 90 difference between groups in terms of knowledge about stone disease, χ 2 = 0.51, p =.48, OR = 3.27 (see Table 4.1). Together, these results provide some support for the generalizability of the study s findings and provide insight into factors that may have led participants to not complete the study.

98 Analysis Follow up Allocation Enrollment 91 Figure 4.1. CONSORT Diagram Assessed for eligibility (n = 327) Excluded (n = 222) Not meeting inclusion criteria (n = 0) Refused to participate (n = 35 1 ) No BHC available to recruit at time of appointment (n = 187) Randomized (n = 105) Allocated to control (n = 53) Received allocated intervention (n = 53) Allocated to intervention (n = 52) Received allocated intervention (n = 52) Lost to follow up (n = 0) (no mid-study f/u required for controls) Withdrew from study (n = 1) (Pt did not give a reason) Lost to follow up (n = 8) (unable to reach for phone f/u) Withdrew from study (n = 2) (Pts did not give a reasons) Analyzed (n = 29) Did not return final survey (n = 23) (for unknown reasons) Analyzed (n = 27) Did not return final survey (n = 23) (for unknown reasons) Note. 1 This number is a best estimate, as the number of patients that declined to participate was not recorded.

99 92 Table 4.1 Patient Demographics (Categorical Variables), Chi-Square, and Odds Ratio Results Characteristic Completers Attritors n % n % χ 2 p OR 95% CI Race [0.15, 93.65] Caucasian 56 (100%) 45 (98%) Multiracial 0 1 (2%) Sex [0.52, 2.49] Male 31 (55%) 24 (52%) Female 25 (45%) 22 (48%) Group [0.49, 2.34] Control 29 (52%) 23 (50%) Intervention 27 (48%) 23 (50%) Sexual orientation [0.77, 13.04] Heterosexual 53 (95%) 39 (85%) Non-heterosexual 3 (5%) 7 (15%) Relationship status [0.62, 3.08] Married 36 (64%) 26 (57%) Unmarried 20 (36%) 20 (43%) Educational attainment [0.65, 3.15] B.S. degree or more 33 (59%) 23 (50%) Some college or fewer 23 (41%) 23 (50%) Employment status [0.051, 0.71] Employed/Other 41 (73%) 43 (93%) Retired 15 (27%) 3 (7%) Household income [0.35, 1.98] <$49, (27%) 14 (30%) >$50, (73%) 32 (70%) Number of previous stones [0.73, 3.54] < 3 31 (55%) 20 (43%) > 4 25 (45%) 26 (57%)

100 Table 4.1 Continued Characteristic Completers Attritors n % n % χ 2 p OR 95% CI Types of stones [0.63, 3.00] Calcium oxalate 30 (54%) 21 (46%) All other/unsure 26 (46%) 25 (54%) ER visits [1.73, 13.78] < 3 50 (89%) 29 (63%) > 4 6 (11%) 17 (37%) Stone treatments [0.87, 4.26] No treatment 31 (52%) 18 (39%) Treatment/Unsure 25 (48%) 28 (61%) Medication use [0.51, 2.82] Yes 18 (32%) 13 (28%) No 38 (68%) 33 (72%) Knowledge about Stones [0.52, 20.69] Knowledgeable 49 (91%) 6 (75%) Unknowledgeable 5 (9%) 2 (25%) 93

101 Table 4.2 Descriptive Statistics and Effect Size Results for Patient Continuous Variables Variable Completers Attritors M SD M SD Cohen s d Satisfaction Total Management Self-efficacy Success of BHC SF-MPQ-2 (pain) Brief-PHQ(depression) SWLS 2 (life satisfaction) PSS (stress) WI Stone-QOL Note. 1 Post-intervention measure only; all other comparisons are at pre-intervention. 2 Higher scores represent greater life satisfaction and stone-related QOL; lower scores represent higher functioning for all other scales. All independent samples t-tests between groups were non-significant at the p <.05 level. 94

102 95 Overall, participants that completed the study were a relatively homogenous group. All identified as White (N = 56) and most identified as heterosexual (95%) and married (64%). Participants were also very educated, with 59% having a college degree, including 32% that had an advanced degree. Most were employed full-time (37%), with a sizeable group being retired (27%). Socioeconomically, 73% had a household income of at least $50,000. In regards to stone disease, participants who completed the study were almost evenly split in the number of previous stones, with 55% having fewer than 4 previous stones. Calcium oxalate, the most-common stone type overall (Goldfarb, 2009), was also the most-common stone type in this sample, with 54% reporting at least one. However, a noticeably large group, 29%, were unsure of their past stone type. Two reasons for this include not being told by their providers or a patient having too small of stone fragments to collect. Almost half (52%) have not required any stone treatments to date and a majority of participants (68%) were not taking a medication to prevent future stones. Intervention Goals In order to better describe the intervention, Table 4.3 lists the co-created patient goals set during the initial in-person meeting. As most patients had multiple goals, the number of goals is greater than the number of patients in the intervention group (n = 50). The most common were related to increasing fluid intake, making dietary changes, and medication adherence. A noteworthy number of patients shared psychological and psychosocial concerns that were beyond the scope of the intervention and required a referral to specialty mental health services.

103 96 Table 4.3 Descriptive Statistics for Patient Intervention Goals Goals n % Increase fluid intake 26 52% Make global dietary changes 21 42% Decrease salt intake 6 12% Decrease acid load in diet 1 2% Increase dietary calcium 7 14% Medication adherence 6 12% Improve overall treatment adherence 3 6% Decrease anxiety/stress/deep breathing 6 12% Practice sleep hygiene strategies 2 4% Increase coping regarding surgery 1 2% Practice CBT pain management strategies 4 8% Increase motivation for change 10 20% Practice mindfulness strategies 2 4% Increase exercise 3 6% Improve disease-related coping/qol/adjustment 7 14% Increase social support 3 6% Provided psychotherapy referral resources 3 6% Improve communication with providers 4 8% Improve organizational skills 2 4% Insurance issues 1 2% Note. Goals were created by patients in the intervention group only. The number of goals exceeds the number of patients (n = 50), as most patients had multiple goals. Follow-up Phone Calls Follow-up phone calls reviewed progress towards initial goals and occasionally required modification. For example, during a follow-up call a young patient reported he had not filled a prescription designed to aid in stone prevention because he had recently lost his insurance and could not afford the monthly cost. The BHC, with the patient s permission, asked the treating provider if there was a cheaper medication that could be prescribed instead. The provider was able to change the prescription to a medication on the national $4/month plan (offered by pharmacies such as Walgreen s and Wal-Mart) and once the provider explained this change to the patient, he agreed to fill this prescription. In another example, the patient informed the BHC that she found the initial meeting to be very inspiring, as she went from feeling hopeless about

104 97 being able to manage her stone disease to feeling empowered that she could make lifestyle changes. In the majority of follow-up phone calls the patient reported that she or he was making satisfactory progress towards their goals, although the study did not employ any objective measures of progress for confirmation. In a third example, a patient stated he had only made soso progress towards his dietary goals, although he was proud that he had stopped eating fast food and was dining out less. In this incidence, the BHC probed what the patient meant by soso and worked with the patient to address the barriers he saw to making further dietary changes (e.g., lack of time or knowledge to cook at home). The patient and BHC then agreed on a modified plan to best address these barriers and to continue making dietary changes. The BHC frequently offered positive reinforcement and encouragement for the progress patients had made towards their goals or assistance in modifying goals. Provider Attitudes Towards Psychological Collaboration Six (N = 6) urologic providers (e.g., urologist, nephrologist, clinical dietician, PA, and NP) participated in this study, with all (N = 6) completing the post-intervention survey. Hypothesis I: Psychological collaboration would be positively viewed by providers and increase their psychological-mindedness. Utilizing a modified version of the Doctors Attitudes Toward Collaborative Care for Mental Health (DACC-MH; Thombs et al., 2010) Scale, independent-samples t-tests were conducted (see Table 4.4) to compare urological providers pre- and post-intervention attitudes towards psychological consultation and management of mental health problems. Paired-sample t- tests were not conducted, as surveys were administered anonymously to protect confidentiality. Contrary to hypothesis, providers attitudes towards psychological consultation did not significantly improve from pre-intervention (M = 5.92, SD = 0.38) to post-intervention (M =

105 , SD = 0.83), t(10) = -1.12, p =.29. Although providers attitudes towards psychological consultation had a notable effect size, d = 1.08, in the predicted direction of providers attitudes becoming more positive over time, it was not significant. Contrary to hypothesis, providers attitudes towards their responsibility and capability to manage mental health problems did not significantly change from pre-intervention (M = 5.58, SD = 0.75) to post-intervention (M = 5.29, SD = 1.23), t(10) = 0.50, p =.63, d = The negative direction of the effect size, indicating providers felt even less responsible and capable for managing mental health problems at postintervention, was not significant. Thus, this first aspect of Hypothesis I was not supported. Table 4.4 Descriptive and Inferential Statistics for the DACC-MH and Provider Attitudes Variable n M SD t p d α (a.). Pre-intervention Total Consultation Subscale Management Subscale (a.) Post-intervention Total Consultation Subscale Management Subscale (b.) Satisfaction Pre Satisfaction Post (c.) BHC full-time Pre BHC full-time Post Success in collaborating Post Helpfulness of feedback Post Providers were also asked pre- and post-intervention to rate their satisfaction with the urologic care they provided. This was assessed by one question asking, How satisfied are you that your patients currently have access to all the tools that they need with respect to their stone disease? An independent-samples t-test was conducted to measure changes in provider satisfaction with urologic care. Again, a paired-sample t-test was not conducted, as data were collected anonymously. Contrary to hypothesis, providers satisfaction with access to all the tools they need in respect to stone disease did not significantly change from pre-intervention (M

106 99 = 4.83, SD = 1.17) to post-intervention (M = 4.67, SD = 1.37), t(10) = 0.23, p =.83, d = This second aspect of Hypothesis I was not supported. Contrary to hypothesis, providers comfort level of having a BHC as a full-time member of her/his treatment team did not significantly change from pre-intervention (M = 6.50, SD = 0.55) to post-intervention (M = 6.00, SD = 0.71), t(9) = 0.42, p =.22, d = Although the observed effect size was not in the predicted direction, indicating that providers became less comfortable with having a BHC as a full-time team member over time, results were not significant. Post-intervention study feedback provided some insight into the negative effect size and is discussed in the following chapter. Also, the increased standard deviation from pre- to post should be noted, as it is possible that providers attitudes became more polarized over time. At post-intervention and as hypothesized, providers rated the BHC as very successful in being able to collaborate in order to enhance the overall treatment provided to their patients (M = 6.00, SD = 0.71). Providers also rated the BHC s feedback as very helpful at post-intervention, (M = 6.50, SD = 0.84). T-test results do not support this third aspect of Hypothesis I, although the descriptive results do provide some support. Patient Attitudes Towards Psychological Intervention and Its Effects on QOL Hypothesis II: The PCBH model would be successfully adapted to patients with kidney stones and positively viewed. One-way analyses of covariance (ANCOVA) were conducted to control for pre-test scores. For each ANCOVA, the group (control or intervention) was the independent variable, the post-score the dependent variable, and the pre-score was the covariate (Table 4.4). Cohen s d s were also calculated as a measure of effect size utilizing pre- post difference scores to control for any differences at pre-intervention (Table 4.4).

107 100 Utilizing an author-designed survey to assess patient satisfaction with their urologic care, patients were asked six questions with an accompanying seven-point Likert rating scale, with a higher score signifying greater satisfaction. As these six questions were highly correlated at preintervention for all 102 participants (r s =.40 to.90, all significant at p <.01) and positively if less uniformly well-correlated for the completers (r s =.13 to 78) at post-intervention, they have been combined into a single measure of patient satisfaction. Correlations for the completers are presented in Table 4.5, with the items 1-6 comprising the combined satisfaction measure. A p Contrary to hypothesis, at post-intervention there was no difference in satisfaction scores between the control group (M = 5.69, SD = 1.02) and the intervention group (M = 5.54, SD = 1.21), both groups saw a reduction in their satisfaction scores from pre- to post-intervention, intervention group saw a trend towards a larger decrease in satisfaction scores, d = -.28, although this difference was not significant. Thus, this first aspect of Hypothesis II was not supported. Patient knowledge about stone disease and prevention methods was assessed with four open-ended questions. Answers were then coded as either knowledgeable or unknowledgeable. Of the completers, 91% were coded as knowledgeable compared to 75% of the attritors. Patient knowledge remained equivalent between the control and intervention groups at post-intervention, χ 2 = 0.79, p =.38, OR = 3.27, not supporting this second aspect of Hypothesis II.

108 101 Table 4.5 Pre- and Post-Intervention Comparisons for Combined Continuous Variables for Completers Control Intervention Pre Post Pre Post Cohen s Variable M SD M SD M SD M SD F p d Satisfaction Total * ** Difference Management Selfefficacy Difference Success of BHC Note. *With-in group t-test significant at the p <.05 level. **p <.001.

109 102 Table 4.6 Summary of Intercorrelations for Continuous Variables for Completers Item Satisfied with care -.70**.86**.36**.33*.37**.31.28*.29*.28* 2. Treated whole me.66** -.77**.52** **.28.50**.42**.41** 3. Recommend UW.59**.78** -.35**.27*.33*.38*.33*.33*.33* 4. Information tailored to you 5. Told treatment options 6. Asked treatment preference 7. Confidence in adherence 8. Confidence in prevention 9. Control over stones 10. Given prevention information 11. Set treatment goals.31*.46*.51** -.33*.33*.40*.53**.45**.48** **.57** -.44** **.38**.36**.45**.57** -.39* ** *.45** ** * ** ** **.32*.18.45**.69** -.55**.12.29*.27*.63**.74**.39**.37*.46**.49** Reach treatment goals 13. Helpful techniques ** **.92** -

110 Note. 1 Full text of each item appears in Appendix B, Author Designed Questionnaire. 2 Postintervention data only. Pre-intervention correlations are shown above the diagonal; postintervention correlations are shown below the diagonal. *p <.05; **p <

111 104 Utilizing an author-designed survey to measure patients self-efficacy in managing their stone disease, patients were asked four questions with an accompanying seven-point Likert rating scale. As these four questions were significantly correlated at pre-intervention for all 102 participants (r s =.23 to.74, all significant at p <.05) and moderately correlated for the completers (r s =.37 to.69, all significant at p <.05) at post-intervention, they have been combined into a single measure of self-efficacy. Correlations for the completers are presented in Table 4.6, with the items 7-10 comprising the combined self-efficacy in stone management measure. A p Contrary to hypothesis, at post-intervention there was no difference in self-efficacy scores between the control group (M = 4.65, SD = 1.25) and the intervention group (M = 4.86, SD = 1.27), d = 0.27, indicating greater self-efficacy for the intervention group), this difference was not significant. Therefore, this third aspect of Hypothesis II was not supported. At post-intervention, patients in the intervention group were asked to rate the BHC s success in meeting their needs. Utilizing an author-designed survey, patients were asked three questions with an accompanying seven-point Likert rating scale. Patients were also asked two questions with forced choice responses. The first of these questions, Please rate the degree of progress you made toward reaching your treatment goals, offered three choices: achieved none/some/all of my goals. The second question, which asked participants their recommendation regarding the future work of a BHC, offered four detailed choices that are presented in Appendix B. The three Likert-response questions were highly correlated with each other (r s =.87 to.92,

112 105 all significant at p <.01). Correlations are presented in Table 4.5, with the items comprising the success measure. As hypothesized, patients rated the BHC as helpful in setting goals (M = 5.48, SD = 1.69) and reaching the set goals (M = 5.62, SD = 1.83). They also rated the techniques used by the BHC as helpful (M = 5.86, SD = 1.68), as hypothesized. For the first of the forced-choice measures of success, the majority of participants (71%) stated they achieved some of their goals compared to all/none of their goals, χ 2 = 14.86, p =.001. For the second measure, patient recommendations regarding the future work of a behavioral health consultant in providing services to patients with kidney stone disease, participants were almost split in their recommendations. Eight (38%) recommended that the BHC stay involved and to the same degree as now while 11 (52%) recommended that the BHC become more involved in working with patients, χ 2 = 14.62, p =.002. Thus, these results support the fourth aspect of Hypothesis II. Contrary to hypothesis, attrition rates did not differ between the intervention and control groups, 44% vs. 46%, OR = 1.07, p =.86. Hypothesis III: The BHC intervention would improve patients biological risk factors for urolithiasis compared to treatment-as-usual. One-way analyses of covariance (ANCOVA) were conducted to control for pre-test scores. For each ANCOVA, the group (control or intervention) was the independent variable, the post-score the dependent variable, and the pre-score was the covariate (Table 4.7). Cohen s d s were also calculated as a measure of effect size utilizing pre- post difference scores to control for any differences at pre-intervention (Table 4.7). Biological risk factors were analyzed by conducting a chart review of 24-hour urine analysis results, a test that measures thirteen lithogenic risk factors. A total of 24 patients (N =

113 106 24) had this lab data for both the pre- and post-interventions periods, with 12 (n = 12) being in the control group and 12 (n = 12) being in the intervention group. The total number of risk factors was counted for each patient at pre- and post-interventions periods. A p Contrary to hypothesis, at post-intervention there was no difference in the number of lithogenic risk factors between the control group (M = 3.25, SD = 1.71) and the intervention group (M = 2.92, SD = 2.19), d = -0.28, indicating greater reduction in lithogenic risk factors for the treatment group), this difference was not significant. Thus, Hypothesis III was not supported.

114 107 Table 4.7 Descriptive and Inferential Statistics for Biological (Lithogenic) Risk Factors Control Intervention Pre Post Pre Post Variable M SD M SD M SD M SD F p d Risk Factors ** ** Difference Note. n = 12. **With-in group t-test significant at the p <.001.

115 108 Hypothesis IV: The intervention would improve patients health-related QOL compared to treatment-as-usual. Health-related QOL was assessed on six dimensions, five of which utilized validated gold standard paper-and-pencil instruments. One-way analyses of covariance (ANCOVA) were conducted to control for pre-test scores. For each ANCOVA, the group (control or intervention) was the independent variable, the post-score the dependent variable, and the prescore was the covariate (Table 4.9). Cohen s d s were also calculated as a measure of effect size (Table 4.9). Table 4.6 lists the intercorrelations of the five QOL measures for the control group, along with post-intervention scale means and standard deviations, pre- and post-intervention alpha scores, and test-retest correlation coefficients. Table 4.8 lists the same results for the intervention group.

116 109 Table 4.8 Summary of Scale Intercorrelations, Descriptive Statistics, and Test-Retest Reliability Results for Completers Measure SF-MPQ * * (pain) 2. Brief-PHQ.42** ** ** (depression) 3. SWLS ** ** (life satisfaction) 4. PSS (stress) 5. WI Stone-QOL (stone-related QOL) ** Pre-intervention M (SD) 1.13 (1.31) 3.66 (3.34) (6.72) (3.02) 4.07 (0.66) Post-intervention M (SD) 1.02 (1.41) 2.24 (3.44) (6.64) (2.95) 4.24 (0.092) Pre-intervention α Post-intervention α Test-retest r.53**.24*.71**.56**.47** Note. Pre-intervention correlations are shown above the diagonal; post-intervention correlations are shown below the diagonal. *p <.05; **p <.01.

117 110 Table 4.9 Pre- and Post-Intervention Inferential and Descriptive Statistics for QOL Scales Control Intervention Pre Post Pre Post Cohen s Variable M SD M SD M SD M SD F p d 2 SF-MPQ Brief-PHQ (Depression) SWLS a PSS WI Stone a QOL Combined QOL 1 Note. 1 This combined measure was created by converting each scale s total score to a z- score, then taking the mean of all five scales z-scores. All scales except the SWLS and the WI Stone-QOL were reversed, thus, a higher score reflects greater QOL. a Predicted direction is positive, as higher scores represent greater life satisfaction and stonerelated QOL. 2 All d s were calculated using mean difference (post pre) scores.

118 111 Pain was measured utilizing the Short-form McGill Pain Questionnaire (SF-MPQ-2; Dworkin et al., 2009). A total pain score was calculated for each patient at pre- and postintervention intervals, with a higher score signifying greater pain. In addition, a score can be calculated for each of the four dimensions: continuous pain, intermittent pain, neuropathic pain, and affective descriptors. A p. Thus, an ANCOVA was not conducted. ANCOVA s were also not run for each of the four pain dimensions for the same reason. A possible explanation for this might be intervention group s extremely low test-retest reliability coefficient (r =.061), while the control group saw a moderate correlation, r =.67, p <.01. There was a trend for both groups to have similar pain scores at post-intervention, d = (see Table 4.9). Depression was measured utilizing the Brief Patient Health Questionnaire (Brief PHQ; Spitzer, Kroenke, & Williams, 1999). A total depression score was calculated for each patient at pre- and post-intervention intervals, with a higher score signifying greater depression. A p Contrary to hypothesis, at post-intervention there was no difference in depression scores between the control group (M = 3.15, SD = 2.89) and the intervention group (M = 5, SD = 4.51), d = 0.092, Brief PHQ results did not support Hypothesis IV.

119 112 Panic syndrome was measured utilizing the Brief Patient Health Questionnaire (Brief PHQ; Spitzer, Kroenke, & Williams, 1999). The four items that measure panic are only considered to represent a positive case of panic disorder if yes is rated for all four items. Thus, no scale score is calculated, as this measure simply creates a diagnostic dichotomy (i.e., either positive for panic disorder or not). Only two patients in the study scored positive for panic disorder (one from each group and one at pre- and one at post-intervention). Due to this small n, no inferential statistics were run. Life satisfaction was measured utilizing the Satisfaction With Life Scale (SWLS; Diener, Emmons, Larsen, & Griffin, 1985). A total life satisfaction score was calculated for each patient at pre- and post-intervention intervals, with a higher score signifying greater life satisfaction. A p Contrary to hypothesis, at postintervention there was no difference in life satisfaction scores between the control group (M = 25.79, SD = 5.49) and the intervention group (M = 22.85, SD = 7.16),, indicating greater reduction in for the treatment group), although this difference was not significant. Thus, SWLS results did not support Hypothesis IV. Stress was measured utilizing the Perceived Stress Scale (PSS-10; Cohen & Williamson, 1988). A total stress score was calculated for each patient at pre- and post-intervention intervals, with a higher score signifying greater perceived stress. A p

120 113 Contrary to hypothesis, at post-intervention there was no difference in perceived stress scores between the control group (M = 20.48, SD = 2.15) and the intervention group (M = 19.85, SD = 3.71),, indicating greater reduction in for the treatment group), this difference was not significant. Thus, PSS results did not support Hypothesis IV. Stone-former QOL was measured utilizing the WI Stone-QOL (Penniston, 2010). A total mean score was calculated for each patient at pre- and post-intervention intervals, with a higher mean score representing greater health-related QOL. A p Contrary to hypothesis, at post-intervention there was no difference in QOL scores between the control group (M = 4.36, SD = 0.93) and the intervention group (M = 4.09, SD = 0.62), did not support Hypothesis IV. health-related QOL post-intervention, d = Thus, WI Stone-QOL results

121 114

122 115 Table 4.10 Summary of Intercorrelations for Residual Change (Gain) Scores for Completers Item Biological Risk - 2. SF-MPQ Brief-PHQ (Depression) PSS * - 5. Panic Syndrome ** SWLS * WI Stone-QOL ** - Note. *p <.05; **p <.01.

123 116 Chapter V. Discussion The present study is the first known application of the Primary Care Behavioral Health (PCBH) model with a group of kidney stone formers. Its primary purpose was to investigate the feasibility of adapting the PCBH model to the management of patients with urolithiasis in order to improve patient health outcomes and quality of life. Although this study found few statistically significant results related to intervention outcomes, most likely due to its very modest scale for an intervention study, my findings supported the intervention s feasibility and its acceptability by providers and patients. As hypothesized, providers rated the BHC as very successful in being able to collaborate with them, and they rated the BHC s feedback as very helpful. It should be noted, however, that these were post-intervention only descriptive statistics and were not capable of being subjected to inferential testing. In regards to inferential results, measures of feasibility and acceptability failed to provide support for the intervention. Providers attitudes towards psychological consultation were in the predicted direction of becoming more positive over time (d = 1.08), although contrary to hypothesis, this change was not statistically significant. Also contrary to hypothesis, providers comfort level of having a BHC as a full-time member of her/his treatment team did not significantly change. Although the negative effect size (d = -0.91) indicates that providers became less comfortable with having a BHC as a full-time team member over time, t-test results were not significant. For patients, they rated the BHC as helpful in setting and reaching treatment goals, as hypothesized. They also rated the techniques used by the BHC as helpful. Again it should be noted that these were post-intervention only descriptive statistics and were not capable of being subjected to inferential testing. However, chisquare results did support that a significant majority of participants (71%) believed they achieved at least some of their goals (compared to none or all of their goals) and a significant majority

124 117 (52%) recommended that the BHC become more involved in working with patients (compared to becoming less involved or staying at the same level of involvement) Although few stated they achieved all of their goals, it might be unrealistic to expect a majority of participants would have achieved all of their behavioral health goals in the three-month span of the intervention, as some goals were more complex or longitudinal in nature. This latter point is discussed in further detail as an area for future studies to address. Inferential results failed to provide further support regarding the feasibility and acceptability of the intervention for patients. Despite a negative effect size indicating (d = -.28) lower patient satisfaction with urologic care for the intervention group, this difference was not significant. Also contrary to hypothesis, there was no difference in attrition rates between groups. In an effort to go beyond simply measuring the feasibility and acceptability of the intervention, this study tested three hypotheses regarding the intervention s ability to improve providers psychological-mindedness and patients health-related QOL and biological outcomes. Contrary to predictions, there was no significant change in providers psychological-mindedness and all patient health-related QOL and biological outcomes were not significant. As previously mentioned, this may be a direct result of the study s small sample size. Therefore, it seemed advisable to report and interpret effect sizes in addition to significance tests in order to describe their direction and magnitude. For patient outcomes, effect sizes were in the predicted direction for the intervention group for the variables self-efficacy in managing stones d = 0.27), and lithogenic risk factors d = -0.28). However, several variables did not have effect sizes in the predicted direction and results numerically favored the control group. For providers, this primarily was a decreasing trend in their psychological-mindedness (i.e., attitudes towards their responsibility and capability to manage mental health problems, d = -0.39). For patients,

125 118 variables with effect sizes not in the predicted direction included satisfaction with urologic care (d = -.28) and life satisfaction. Given that all intervention outcomes related to providers psychological-mindedness and patient health-related QOL and biological outcomes were not significant, it is possible that any observed effect sizes could be explained as the result of sampling error. While the quantifiable benefits of adapting the PCBH model to Urology and patients with urolithiasis are equivocal, it is clear that this model, whose efficacy is well known in primary care, can be adapted to secondary care. The broader question that remains is if kidney stone patients, or at least recurrent stone formers (as opposed to the newly diagnosed, for example) are a population best served by integrated care given the unique challenges of this secondary-care setting. The Primary Care Behavioral Health Model The concept of medical and mental health providers collaborating together on patient care is decades old (Robinson & Strosahl, 2009). Although it is a not a new concept, it remains an evolving one. It has evolved from traditional mental health settings to the first beginnings of colocated care and finally in the 1980 s, to integrated primary care (Robinson & Strosahl, 2009). The PCBH model differs from traditional mental health not only in the setting in which it is delivered, but in who it identifies as the patient and in its goals as well (Strosahl & Robinson, 2008). Employing a consultation approach and conceptualizing patients from the biopsychosocial model, the BHC acts as a consultant to the referring medical provider. When meeting with a patient, the BHC assists her or him in setting realistic and concrete goals to improve functioning and QOL (Strosahl & Robinson, 2008). Increased patient and provider satisfaction with care and increased patient adherence are among the most robust benefits for this

126 119 model (Robinson & Strosahl, 2009). Although the brief interventions employed by the BHC have only been shown to have a small effect (it is estimated that 5 15% of intervention patients make clinically significant change), integrated care allows mental health services to reach many more individuals than the traditional mental health system (Whitlock et al., 2002). Few scholars debate the aforementioned benefits of the PCBH model, and perhaps because of this, it has continued to proliferate. It is now commonplace in the U.S. military, the VA system, large HMO s (Strosahl & Robinson, 2008), and large, inner-city safety-net medical systems. This model is widely considered to be the future of health care and continues to receive a tremendous amount of attention, both from scholars and the community (Clay, 2012). However, the attention remains focused on primary care, possibly ignoring the potential benefits of adopting this model for the unique challenges of secondary care. In addition, there is no known study of the application of this model to a population of kidney stone formers when similar interventions been utilized with other chronic diseases (Ouwens, Wollersheim, Hermens, Hulscher, & Grol, 2005). Thus, it was reasonable to presume that the PCBH model could have been successfully adapted to patients with kidney stone disease and that these patients would see the same benefits of increased satisfaction and greater QOL. Generalizability of the Results This study involved a sample of patients with recurrent kidney stones with the intention of generalizing to all patients suffering from this disease. Completers of the study were largely a homogeneous group, with all identifying as White (N = 56), and most identifying as heterosexual (95%) and married (64%). Although previous epidemiological studies on stone formers did not collect sexual orientation or martial status, Caucasians have been found to have prevalence rates 2-3 times that of other racial groups (Stamatelou et al., 2003). In one recent epidemiological

127 120 review, 62% of stone formers were White (Scales et al., 2007). Compared to the current sample s 100% White-identified sample, our lack of racial diversity remains a limitation. Given that a large body of literature exists describing the barriers racial and ethnic minorities (REMs) face to receiving equal healthcare, such as poverty, unemployment, and being disproportionately affected by the criminal justice system (Jones, Crump, & Lloyd, 2012), the additional support provided by this intervention could be of especial benefit to this population. In regards to sex, completers were 55% male, very similar to Scales et al. s epidemiological finding of 59%. It is of particular importance that our sample accurately reflects the gender breakdown of stone disease, as previous research has found that female patients report more physical and psychological distress and lower QOL than male patients (Penniston & Nakada, 2007). As stones typically occur during the prime working years (Goldfarb, 2009; Lotan, 2009), it was fitting that 37% of the completers were employed full-time and 73% were employed at least part-time. Also, a sizeable percent (27%) of completers were retired, consistent with the roughly 33% of stone formers Pearle et al. (2007) found to be retirement-aged. Although retirees were less likely to be an attritor (see Table 4.1), our percentage of retirees appears consistent with national trends. It is possible that retirees are more likely to complete research due to having increased free time or different values regarding research compared to working and younger generations. Moreover, our sample was uncharacteristically highly educated compared to the U.S. population. A total of 59% had a college degree, including 32% that had an advanced degree. These rates are 2-3 times the U.S. average of 30% of the population with a bachelor s degree and 11% having an advanced degree (U.S. Census Bureau, 2011). Previous epidemiological studies on stone formers did not report data on education rates. This unusually high rate of educational attainment limits the generalizability of the results to all US stone formers and may be a partial

128 121 explanation for some of the non-significant results found. It is possible that highly educated patients can benefit the least from short, behavioral and psychoeducational interventions, such as the ones provided in this study. Thus, the educational attainment of participants is a variable worth additional consideration in future studies. Socioeconomically, 73% of completers had a household income of at least $50,000. According to the U.S. Census Bureau (2010), the median household income in 2010 for Whites was $52,480, with 78% of the U.S. population with this median income being White. There were no significant differences between completers and attritors socioeconomically, with previous research finding equivocal results regarding the relationship between SES and the risk of stones (Najem et al., 1997; Stamatelou et al., 2003). Interestingly, those patients with four or more past stone-related ER visits were less likely to be a completer. While other stone-related factors, such as number of previous stones, did not significantly impact completion rates, it is unclear why the number of ER visits decreased the likelihood to complete. One possible explanation may be that those who frequent ER s more often to receive their stone care may have a more severe or unpredictable form of stone disease, poorer coping skills, or lack additional health care resources such as a primary care provider or insurance. This, in turn, may have decreased their energy or willingness to complete another stone-related activity such as this study. Or it may also be possible that those with more ER visits have received a greater amount of stone-related health care over the years, thus reducing their interest in completing this study. In addition, Cohen s d results were in the direction of completers reporting greater satisfaction with their urologic care, fewer symptoms of depression, and greater stone-related QOL at pre-intervention than did the eventual attritors, although t-tests were not significant (see

129 122 Table 4.2). As this may suggest that the completers were experiencing less distress at preintervention, it is possible this motivated them to complete the study as well as provided them with the energy to do so. However, this can also be seen as a limitation, as those patients that were in the greatest distress were less likely to complete the study, thus we lack the ability investigate the effects of the intervention on them. It is possible that these more distressed patients did benefit the most from the intervention, but by not returning the final survey this improvement could not be captured. Hypothesis I Contrary to hypothesis, providers attitudes towards psychological consultation and their attitudes towards their responsibility and capability to manage mental health problems did not significantly change post-intervention. These results have several possible explanations. First, providers attitudes towards psychological consultation were in the direction (d = 1.08) of becoming more positive over time. This may suggest that if a larger sample of providers was used, thus increasing power, t-tests might have been significant. Also, independent samples t- tests, instead of paired-sample t-tests, were required since providers scores were collected anonymously to protect confidentiality. Future studies may wish to code providers in a similar manner as patients in order to measure individual changes over time, as it is likely that the provider group will always be smaller, and thus have lower power, than the patient group. Also contrary to hypothesis, providers attitudes towards their responsibility and capability to manage mental health problems did not significantly change post-intervention (d = ). The negative effect size indicates a decreasing trend in providers sense of responsibility and capability for managing mental health problems. This may suggest, for whatever reason(s), that consulting with BHCs made providers feel more inadequate to manage the behavioral and

130 123 mental health issues of their patients. However, it is possible that this decreasing trend suggests a more positive phenomenon was occurring. One possible explanation is that as the urologic providers became increasingly comfortable with consulting with a mental health provider and having one co-located in their clinic, their views of their own responsibility to manage mental health issues decreased. However, the intervention did not change provider satisfaction with access to all the tools that they need with respect to their stone disease, d = Again, by not having a BHC as a permanent member of their team, providers might be expected to be less satisfied post-intervention instead of more satisfied, as they were aware the BHC was only temporary. It is also possible that providers may have interpreted this question differently than intended, as they may be more concerned with medical tools of treatment, such as new surgical techniques, than they are with psychological tools. Although providers rated the BHC as very successful in being able to collaborate in order to enhance the overall treatment provided to their patients and the BHC s feedback as very helpful, providers comfort level of having a BHC as a full-time member of her/his treatment team did not significantly change (d = -0.91). The negative direction of the effect size is puzzling, as it is in the direction of providers becoming less comfortable with a full-time BHC. Taken at face value, it suggests that the providers are strongly not interested in having a BHC as a full-time team member, presumably because the BHCs were unsatisfactory. One provider did note on the post-intervention survey that she or he felt that feedback was not always given in a timely manner, reducing its usefulness. Indeed, navigating a complex electronic records system along with IRB and ethical requirements was a significant challenge for the part-time BHCs. Additional logistical limitations, such as very limited clinic and administrative space for both providers and the BHCs, could have also added to dissatisfaction with the BHCs. However,

131 124 given that providers post-intervention mean was 6.0 (SD = 0.71) out of a maximum of 7.0, still within the strongly agree category, and the t-test was non-significant, this is perhaps an oversimplification. As previously mentioned, the slightly increased standard deviation from pre- (SD = 0.55) to post (SD = 0.71) should be noted, as it is possible that providers attitudes became more polarized over time. Thus, it may be true that for some providers, they became less interested over time with the idea of having a BHC as a full-time team member while others became more strongly interested. In this instance, having matched providers scores pre-post would have allowed further inferential statistics to be run. Also, providers were not asked if they were interested in having a BHC as a part-time team member. It is possible that the majority of providers would welcome a BHC as a team member, but perhaps feel there is not the need at this time for the BHC to be full-time. Finally, it is possible that it will simply take time for most providers to become comfortable with collaborating with a mental health professional, to allow another professional to be directly involved in the patient care they provide, and to learn what a BHC does and how she or he could even improve patient care. Given the part-time presence of the BHCs in the Urology clinic, the three-month span of the intervention might have been too short for most providers to buy-in to the PCBH model as of yet. Also related to the multiple BHCs part-time presence, it is possible that it providers never felt the BHCs were integrated into the clinic, possibly reducing engagement with this model. Hypothesis II The study s second hypothesis, that the PCBH would be successfully adapted to patients with kidney stones and positively viewed, received mixed support. The 5 A s framework of conducting an integrated behavioral health visit was shown to work as well in a busy urology clinic as it does in the often-hectic primary care setting. The BHCs all found this to be a useful

132 125 format for guiding each individual intervention and it helped ensure that many important aspects of patient care, such as setting collaborative goals, were followed. But beyond subjective measures of feasibility, one must consider the objective results. Unexpectedly, both patient groups reported a decrease in their satisfaction with their urologic care at post-intervention, with intervention group seeing a larger decrease in satisfaction scores (d = -.28), although this difference was not significant. The trend towards a greater decrease for the intervention group is counter to the findings of previous research, as one of the PCBH s greatest strengths is improving patient satisfaction (Robinson & Strosahl, 2009). There are several possible explanations, although gathering additional qualitative patient feedback would likely have been beneficial in providing direction. First, although attrition rates were similar between the two groups, patients in the intervention group could have felt more burdened by the added participation required, thus decreasing their satisfaction. Second, in the post-intervention survey some intervention group patients reported feeling the BHC repeated material they already knew (perhaps why no change was seen in knowledge scores as well), that the intervention was not individually tailored to them, that the surveys took longer than advertised, and they were confused by what participation even entailed along with the study s timeline. Additionally, some participants reported being confused as to why a mental health professional was meeting them during their urology appointment and sometimes presumed the BHC was a physician or allied health professional. Third, the course of the study required the use of three different BHCs and this lack of continuity in patient care could have had a negative impact. Given that this study was the first of its kind in a Urology Department and with a stone former population, some areas for improvement are expected. However, the intervention group s trend towards decreasing satisfaction suggests that any future studies will want to give this

133 126 variable increased attention, possibly through additional patient feedback and those ideas outlined in the Future Research section below. One trend that was in the predicted direction, although not statistically significant, was the intervention group s trend towards improving self-efficacy scores d = 0.27). As the intervention was intended to co-create goals with patients as well as assist them in making these behavioral changes, it would be expected that patients self-efficacy in managing their stone disease would improve. Another benefit that was expected was patients rating the BHC as helpful in setting and reaching goals as well as finding the BHC s techniques helpful. In further support of this, the majority of participants (71%) stated they achieved some of their goals. In terms of the future of this model, a majority (52%) recommended that the BHC become more involved in working with patients while a sizeable number (38%) recommended that the BHC stay involved and to the same degree as now. These data illuminate difficult-to-explain findings such as intervention patients reporting lower satisfaction. Patients may have felt the BHC was doing enough in light of their numerous critiques listed above as well as felt dissatisfied because the BHC was not doing more. For example, one patient provided the feedback that she or he would have liked the BHC to have shown her or him how to cook meals in line with stone former dietary recommendations. Hypothesis III One of the unique aspects of this study afforded by its interdisciplinary approach is that it focused not only on psychological and QOL variables, but also on biological risk factors. Unfortunately, without extramural funding to order biological tests the study was limited to those biological data ordered as a part of treatment-as-usual. Thus, only a small number of participants (n = 24) had these data at pre- and post-intervention for comparison. The small sample for these

134 127 tests weakened power and was a contributing factor in the non-significant ANCOVAs. Therefore, a d = -0.28, indicating greater reduction in lithogenic risk factors for the intervention group), this difference was not significant. One likely explanation for this was the study s short three-month time frame, as this was likely too short of a time frame to see measurable changes in a patient s lithogenic profile. A more appropriate time frame would likely be three-years (Kambadakone, Eisner, Catalan, & Sahani, 2010), allowing for lithogenic factors to be measured longitudinally. If replicated in a larger sample and over a longer period of time, this decreasing trend would be a valuable finding, as many funding agencies, as well as providers, likely emphasize quantifiable biological improvements over improvements in psychological or QOL variables. Hypothesis IV Health-related QOL has a variety of potential definitions. In the present study, it was assessed on six dimensions that were believed to be the most relevant for stone formers: pain, depression, panic syndrome, life satisfaction, stress, and stone-related QOL. It was hypothesized that the intervention would improve each of these variables compared to treatment-as-usual. Results, however, did not support this hypothesis. Pain, an important variable to include given the high levels of pain associated with some stone events, saw ANCOVA was not conducted. Panic syndrome, in hindsight, was a poor variable to include. Its relevance for stone formers is dubious; assessing anxiety is a more relevant choice for a future study. Although all hypothesis-testing ANCOVAs were nonsignificant, several noteworthy effect sizes are discussed, as they describe the direction and magnitude of between-group comparisons.

135 128 The present study included a gold standard measure of pain, the Short-form McGill Pain Questionnaire (SF-MPQ-2; Dworkin et al., 2009). As previously noted, the ANCOVA was not conducted. Cohen s d results suggest no difference between the groups, d = This was a disappointing finding, as pain was a key QOL variable of this investigation. One possible cause for this is the intervention group s extremely low test-retest reliability coefficient (r =.061), a finding that is difficult to explain. The design of the SF-MPQ-2 may also be partially responsible, as it only asks participants to rate their pain in the past week, likely making it sensitive enough to only capture the pain of an ongoing, acute stone event. [Recall that stone disease is typically only painful when someone has an acute stone event or status post a stone event (Goldfarb, 2009).] Thus, this measure likely lacks the sensitivity to capture those stone formers who typically experience a high level of pain when they do experience a stone, presuming they were not experiencing a stone event at the time of the surveys. Perhaps a better measure would be to ask patients their pain level during their last stone event or their average pain during a stone event. Conversely, it might be unrealistic to expect an intervention of this type to improve pain proactively if a person is not experiencing an acute stone event at the time of the intervention. Panic syndrome, as measured by the Brief PHQ (Spitzer, Kroenke, & Williams, 1999), was an extraneous variable that should be removed from future studies. It was initially included as a proxy variable for anxiety, as panic and anxiety are related, although they also are quite dissimilar. The very small number of patients who scored positive for the condition, perhaps suggesting that panic disorder is uncommon in stone formers, also limited its relevance in the current study. To more accurately capture anxiety, a short measure such as the GAD-7 (Spitzer, Kroenke, Williams, & Lowe, 2006) should be utilized in future studies.

136 129 Contrary to hypothesis, at post-intervention there was no difference in perceived stress scores between groups as measured by the PSS-10 (Cohen & Williamson, 1988). Although the difference between groups was not statistically significant,, indicating greater reduction in for the treatment group. Stress was included as a QOL variable in this study due to its possible correlational and/or causal relationship with stone disease (Diniz et al., 2006; Najem et al., 1997). If it is included as a variable in future studies, an effort should be made to investigate by what mechanism(s) the intervention could impact stress. This would be a significant contribution to the literature, as previous research is equivocal whether the relationship is correlational and/or causal (Diniz et al., 2006; Najem et al., 1997). One theory is that by helping stone formers feel there is something they can do to prevent future stones and that they have someone (the BHC) to assist them in making changes, the stress they experience regarding stone disease would decrease. As previous research has found that stone formers have higher stress levels than controls (Diniz et al.), future studies may wish to make this comparison by the addition of a stone-free control group. Contrary to hypothesis, at post-intervention there was no difference in stone-related QOL scores, as measured by the WI Stone-QOL (Penniston, 2010), between groups., d = 0.011, although this was not statistically significant. Equivalent change for both groups was not hypothesized. If QOL is included in a larger replication study, it is possible that this is a variable that naturally improves overtime, or at least over the short-term period of the study (i.e., following an acute stone-formation event). In support of this, Rabah, Alomar, Binsalah, and Arafa (2011) found that stone patients 4-10 months post-lithotripsy, matched with health controls, had equivalent health-related QOL, except for the domains of pain and social functioning. However, it should be mentioned that lithotripsy

137 130 is arguably the least invasive of all stone treatments and these results may not be replicated with more invasive treatments, such as ureteroscopy. It is also important to note that the WI Stone- QOL asks patients to rate their QOL in the past month, perhaps not capturing any QOL changes that occurred in the first two months of the intervention s three-month span. Regardless, as the first-known measure of QOL with stone patients, it assesses the dimensions of QOL predicted to be relevant to stone formers and is an important measure for future studies to retain. epression is widely considered the most prevalent mental health issue faced by patients with chronic kidney disease (Hoth et al., 2007) and likely this is true for kidney stone disease as well.

138 131 Contrary to hypothesis, at post-intervention there was no difference in life satisfaction scores, as measured by the SWLS (Diener, Emmons, Larsen, & Griffin, 1985), between groups., indicating greater reduction in for the intervention group, although this difference was not significant. chronic, life-long disease that often requires significant lifestyle changes and has a high likelihood of reoccurring. It would be interesting to measure changes in this variable over time, as it is possible that life satisfaction scores could improve as patients gain greater control over their disease. Also, future studies should focus on research design improvements and patient feedback post-intervention to understand the mechanisms by which the intervention could increase distress, at least in the short-term. Given the non-significant findings above and the overall low power of this study, it was decided to create a combined measure of QOL by converting each scale s total score to a z-score, then taking the mean of all five scales z-scores. All scales except the SWLS and the WI Stone- QOL were reversed, thus, a higher score reflected greater QOL. Counter to expectations, at postintervention the QOL scores between the control group (M = , SD = 0.60) and the intervention group (M = , SD = 0.66) did not differ, F(1, 46) = 0.14, p =.71, d = (see Table 4.9).

139 132

140 133 uthordesigned questionnaire, although based on previous experience and grounded in the literature, was not pilot tested prior to this study and may have weaker psychometric properties than the gold-standard scales utilized. Suggestions for Future Research Although suggestions for future research have been spread throughout this manuscript to address specific topics as they were presented, several recommendations can be made for providers and patients. The PCBH model differs from traditional mental health in numerous ways, one of those being that providers are the BHC s primary customer (Strosahl & Robinson, 2008). Therefore, any attempt to adapt the PCBH model to secondary care needs to consider the needs and feedback of providers. In the present study, there were no statistically significant provider variables, making it advisable, with appropriate reservation, to interpret effect sizes and descriptive statistics. Providers appeared to approve of some aspects of the BHC s work, rating the BHC as successful in being able to collaborate and to provide helpful feedback. There was also a non-significant trend of providers attitudes towards psychological consultation becoming more positive over time, perhaps related to their non-significant trend in

141 134 feeling less responsible and capable for managing mental health problems. However, there was a non-significant trend of providers becoming less comfortable with having a BHC as a full-time team member over time. Interpreting these trends, it could be suggested that providers are open to collaborating with a mental health professional, but are less sure if they would want to integrate a BHC into their practice full-time. Future research should place a greater emphasis on understanding providers views and how a BHC could better fit into their practice. This could be accomplished in numerous ways, such through meetings and shadowing. If there was an overall interest in a BHC in their practice, all parties could agree on the role of the BHC. The BHC should, however, be able to dedicate more time in the clinic and be able to address previous concerns, such as untimely patient feedback. Through improving the BHC s understanding of providers needs and feedback, the feasibility of future applications of this model could be improved. Equally as important as fully understanding the needs of providers is the importance of understanding the training needs of counseling psychologists to serve as BHCs. As has been discussed throughout this paper, integrated care differs from traditional mental health in numerous ways, including the setting, who the patient is, how goals are created, and in the length of appointments (Strosahl & Robinson, 2008). Unfortunately, most psychology training programs do not train students to meet the unique needs of integrated care (Bluestein & Cubic, 2009). In order to lay the groundwork for student training in this burgeoning area, the American Psychological Association (APA) created the Interdivisional Task Force for a Primary Care Psychology Curriculum in 1999 (McDaniel, Schroeder, Belar, & Hargrove, 2002). This Task Force published a list of twelve recommendations to guide educational programs and individual practitioners interested in designing educational and training experiences. The twelve core areas

142 135 of knowledge and skills for primary care psychology are: biological, cognitive, behavioral and developmental, and sociocultural components of health and illness; health policy and health care systems; common primary care problems; clinical assessment of common primary care conditions; clinical interventions in primary care; interprofessional collaboration in primary care; and ethical, legal, and professional issues in primary care (McDaniel et al, 2002). In their conclusion, the APA Task Force called for more faculty role models, broad interdisciplinary experiences, organizational collaboration, increased funding of this model, and the need for psychology to effectively communicate its own unique role in the primary care setting (McDaniel et al, 2002). Related to this latter point, those in the field have felt an increasing pressure for psychologists to define their roles or risk being replaced by less-trained, but lower cost, Master-level mental health clinicians (L. DuBenske, personal communication, August 16 th, 2012). Some of needs outlined by the APA Task Force are currently being met by innovative programs such as the Eastern Virginia Medical School s Clinical Psychology Internship Program ((Bluestein & Cubic, 2009) and counseling psychology doctoral programs at Ball State University, Virginia Commonwealth University, and University of Denver, to name a few. In addition, educational training programs with a focus on health psychology are becoming more commonplace, such as the University of Colorado s Ph.D. Program in Clinical Health Psychology, although this program s website states they are currently seeking APA-accreditation as a clinical psychology program. Along with the feasibility of adapting this model to providers and the training of BHC s, the feasibility of adapting this model to kidney stone patients was also a central focus. Unfortunately, all inferential statistics testing the feasibility of this model were non-significant. Again this makes it advisable, with appropriate reservation, to interpret effect sizes and

143 136 descriptive statistics. One key measure of feasibility was patient satisfaction and this saw a larger decrease for the intervention group, although this difference was not significant. While this nonsignificant trend towards decreasing satisfaction could be related to the additional requirements of participating in the study, it may also represent a trend towards general dissatisfaction with working with a BHC. Although descriptive statistics suggest that patients positively viewed the BHC s assistance, it is important that any future study better understand the connection between patient satisfaction and the intervention. Future studies should attempt to ascertain what aspects of the intervention patients find beneficial or unhelpful, and which patients fall into these views. A second topic of feasibility involves the targeted patient population. While this study focused on recurrent stone formers, with the assumption that these patients were likely experiencing a more severe form of the disease and thus, in greater need of assistance, results call this assumption into question. Results could be interpreted to suggest that recurrent stone formers are less likely to benefit from this type of intervention, perhaps because they are already quite knowledgeable on a whole about stone disease and may benefit less from psychoeducational or behavioral interventions. One idea for future research might be to not only compare a treatment and control group, but to add a third group of newly diagnosed stone formers. Utilizing the intervention with the newly diagnosed might prove to be a fruitful direction for this model. In addition, the intervention and model employed in this study, which could perhaps be called the Secondary Care Behavioral Health model (SCBH), might be useful if applied to other chronic conditions. Indeed, in Ouwens et al. s (2005) review of similar integrated care interventions, they found them targeted at patients with chronic diseases as diverse as stroke, diabetes, and chronic heart failure.

144 137 A third topic of feasibility involves the co-created patient goals of the intervention. When co-creating goals, there was not a particular focus by the BHC to create only short- or long-term goals, but goals that were mutually agreed upon. Future studies may wish to have patients only create short-term and concrete goals. For example, if the goal is to increase daily fluid intake, participants could be asked to estimate what their current fluid intake is in liters and then to set a small goal, such as drink one additional glass of water a week. As a measure of validity, participants reported fluid intake could be compared to an objective measure, their estimated fluid intake reported in their 24-hour urine analysis. This might be a more concrete goal than simply increase fluid intake. Additional directions for future research are also related to the overall feasibility and benefits of the intervention. One of the most probable reasons for the non-significant findings was the small sample size and accompanying low power. Future studies need to utilize a much larger sample and one calculated on a more modest effect size. Also, as previous research has found that women report more physical and psychological distress and lower QOL than men (Penniston & Nakada, 2007), this larger sample should also investigate potential gender effects. Second, there is a need to streamline the research procedures. Many participants stated they found the timeline of the study or specific participation requirements to be confusing. Future studies could be clearer with patients regarding what their participation will entail and when, perhaps by giving them a diagram and handout of the study, similar to those presented in the Method and Appendix subsections. Third, many participants felt the study s measures were too lengthy and this may have negatively impacted post-intervention completion rates. Future studies should utilize the results of this study to refine the design of the required larger study, such as to remove panic syndrome as a variable. Fourth, future studies need to employ a longitudinal design

145 138 over a span of at least three years (Kambadakone et al., 2010), as it would be unlikely for lithogenic risk factors to change in a few months. Fifth, securing funding for a larger study would allow the researchers to offer participant incentives for participation, likely increasing participation and follow-up rates. Sixth, future studies should address how to provide more direct and efficient feedback to providers, addressing one of their primary complaints. This point also highlights the value of obtaining qualitative feedback from providers and patients, as it can capture the insight of participants. Seventh, as some participants felt they had to meet with a BHC even if they felt they did not need it, future studies should attempt to better screen and meet with only those patients that agree they would be willing to meet with a BHC. Interventions could also be designed to make patients feel they have been tailored to them. Conclusion This study was the first known application of the PCBH model to patients with kidney stone disease and one of the first studies to focus exclusively on the mental health and QOL needs of stone formers. Although limited by a small and homogeneous sample and a lack of funding, it has provided valuable data on the feasibility and benefits of this model. Although all inferential comparisons of benefits of this model were non-significant, effect sizes and descriptive statistics provide some insight into the direction and magnitude of these comparisons. Despite this study s small N and consequent limited statistical power, descriptive results suggest that the PCBH model can be adapted to secondary care and that providers and patients can positively view aspects of it. It is hoped that this study, although modest in scale, has laid the groundwork for future studies on addressing the needs of kidney stone patients..

146 139 References Andersen, B. L., Yang, H. C., Farrar, W. B., Golden-Kreutz, D. M., Emery, C. F., Thornton, L. M., et al. (2008). Psychologic intervention improves survival for breast cancer patients: A randomized clinical trial. Cancer, 113, Bensalah, K., Tuncel, A., Gupta, A., Raman, J. D., Pearle, M. S., & Lotan, Y. (2008). Determinants of quality of life for patients with kidney stones. J Urol., 179, Bergman, J., Gore, J. L., Singer, J. S., & Anger, J. T. (2010). Readability of health related quality of life instruments in urology. J Urol., 183, Blount, F. A., & Miller, B. F. (2009). Addressing the workforce crisis in integrated primary care. J Clin Psychol Med Settings., 16, Bluestein, D., & Cubic, B. A. (2009). Psychologists and primary care physicians: A training model for creating collaborative relationships. Journal of Clinical Psychology in Medical Settings, 16, Blumenthal, J. A., Sherwood, A,, Babyak, M. A., Watkins, L. L., Waugh, R., Georgiades, A., et al. (2005). Effects of exercise and stress management training on markers of cardiovascular risk in patients with ischemic heart disease: A randomized controlled trial. JAMA, 293, Borghi, L., Meschi, T., Amato, F., Briganti, A., Novarini, A., & Giannini, A. (1996). Urinary volume, water and recurrences in idiopathic calcium nephrolithiasis: A 5-year randomized prospective study. J Urol., 155, Carvalho, M., Ferrari, A. C., Renner, L. O., Vieira, M. A., & Riella, M. C. (2004). Quantification of the stone clinic effect in patients with nephrolithiasis. Rev Assoc Med Bras., 50(1),

147 140 Clay, R. A. (2012, May). Integrated care is the future. Monitor on Psychology, 43(5), Cohen, S., & Williamson, G. (1988). Perceived stress in a probability sample of the United States. In S. Spacapam & S. Oskamp (Eds.), The social psychology of health: Claremont Symposium on applied social psychology. Newbury Park, CA: Sage. Damiano, R., Autorino, R., De Sio, M., Cantiello, F., Quarto, G., Perdonà, S., et al. (2005). Does the size of ureteral stent impact urinary symptoms and quality of life? A prospective randomized study. Eur Urol., 48, Diener, E., Emmons, R., Larsen, R., & Griffin, S. (1985). The Satisfaction With Life Scale. Journal of Personality Assessment, 49, Diniz, D. H., Blay, S. L., & Schor, N. (2007a). Quality of life of patients with nephrolithiasis and recurrent painful renal colic. Nephron Clin Pract., 106, Diniz, D. H., Blay, S. L., & Schor, N. (2007b). Anxiety and depression symptoms in recurrent painful renal lithiasis colic. Braz J Med Biol Res., 40, Diniz, D. H,, Schor, N., & Blay, S. L. (2006). Stressful life events and painful recurrent colic of renal lithiasis. J Urol., 176, Duffey, B. G, Pedro, R. N., Makhlouf, A., Kriedberg, C., Stessman, M., Hinck, B., et al. (2008). Roux-en-Y gastric bypass is associated with early increased risk factors for development of calcium oxalate nephrolithiasis. J Am Coll Surg., 206, Dworkin, R. H., Turk, D. C., Revicki, D. A., Harding, G., Coyne, K. S., Peirce-Sandner, S., et al. (2009). Development and initial validation of an expanded and revised version of the Short-form McGill Pain Questionnaire (SF-MPQ-2). Pain, 144(1-2), Goldfarb, D. S. (2009). In the clinic: Nephrolithiasis. Annals of Internal Medicine, 151, ITC2-2- ITC2-16.

148 141 Grafton, K. V., Foster, N. E., & Wright, C. C. (2005). Test-retest reliability of the Short-form McGill Pain Questionnaire: Assessment of intraclass correlation coefficients and limits of agreement in patients with osteoarthritis. Clin J Pain, 21, Greenberg, P. E., Leong, S. A., Birnbaum, H. G., & Robinson, R. L. (2003). The economic burden of depression with painful symptoms. J Clin Psychiatry, 64, Suppl 7, Hahn, D., Reuter, K., & Härter, M. (2006). Screening for affective and anxiety disorders in medical patients: Comparison of HADs, GHQ-12 and brief-phq. GMS Psycho-Social- Medicine, Hoth, K. F., Christensen, A. J., Ehlers, S. L., Raichle, K. A., & Lawton, W. J. (2007). A longitudinal examination of social support, agreeableness and depressive symptoms in chronic kidney disease. J Behav Med., 30, Hunter, C. L., Goodie, J. L., Oordt, M. S., & Dobmeyer, A. C. (2009). Integrated behavioral health in primary care: Step-by-step guidance for assessment and intervention. Washington, DC: American Psychological Association. Jones, D. J., Crump, A. D., & Lloyd, J. J. (2012). Health disparities in boys and men of color. American Journal Of Public Health, 102(Suppl 2), S170-S172. Kambadakone, A. R., Eisner, B. H., Catalano, O. A., & Sahani, D. V. (2010). New and evolving concepts in the imaging and management of urolithiasis: Urologists' perspective. Radiographics, 30(3), Katon, W. (1995). Collaborative care: Patient satisfaction, outcomes, and medical cost-offset. Family Systems Medicine, 13, Kerbl, K., Rehman, J., Landman, J., Lee, D., Sundaram, C., & Clayman, R. V. (2002). Current management of urolithiasis: Progress or regress? J Endourol., 16,

149 142 Kleinman, J. G. (2007). Bariatric surgery, hyperoxaluria, and nephrolithiasis: A plea for close postoperative management of risk factors. Kidney Int., 72, Kroenke, K., & Mangelsdorff, A. D. (1989). Common symptoms in ambulatory care: Incidence, evaluation, therapy, and outcome. Am J Med., 86, Lee, A. M., Wong, J. G., McAlonan, G. M., Cheung, V., Cheung, C., Sham, P. C., et al. (2007). Stress and psychological distress among SARS survivors 1 year after the outbreak. Can J Psychiatry, 52, Lieske, J. C. (2009). Gastric bypass procedures and renal calculi--how should we counsel patients and bariatric surgeons? J Urol., 182, Lieske J. C., Kumar, R., & Collazo-Clavell, M. L. (2008). Nephrolithiasis after bariatric surgery for obesity. Semin Nephrol., 28, Lotan, Y. (2009). Economics and cost of care of stone disease. Advances in Chronic Kidney Disease, 16, Margalith, I., & Shapiro, A. (1997). Anxiety and patient participation in clinical decisionmaking: The case of patients with ureteral calculi. Soc Sci Med., 45, Matlaga, B. R., Shore, A. D., Magnuson, T., Clark, J. M., Johns, R., & Makary, M. A. (2009). Effect of gastric bypass surgery on kidney stone disease. J Urol., 181, MayoClinic.com (n.d.). Kidney stones. Retrieved from McDaniel, S. H., Schroeder, C., Belar, C. D., & Hargrove, D. (2002). A training curriculum for professional psychologists in primary care. Professional Psychology: Research & Practice, 33(1), 65.

150 143 Meschi, T., Schianchi, T., Ridolo, E., Adorni, G., Allegri, F., Guerra, A., et al. (2004). Body weight, diet and water intake in preventing stone disease. Urol Int., 72(Suppl 1), Mumford, E., Schlesinger, H. J., Glass, G. V., Patrick, C., & Cuerdon, T. (1984). A new look at evidence about reduced cost of medical utilization following mental health treatment. Am J Psychiatry, 141, Nabi, G., Cook, J., N'Dow, J., & McClinton, S. (2007). Outcomes of stenting after uncomplicated ureteroscopy: Systematic review and meta-analysis. BMJ, 334(7593), 1-7. Najem, G. R., Seebode, J. J., Samady, A. J., Feuerman, M., & Friedman, L. (1997). Stressful life events and risk of symptomatic kidney stones. Int J Epidemiol., 26, Nouvenne, A., Meschi, T., Prati, B., Guerra, A., Allegri, F., Vezzoli, G., et al. (2010). Effects of a low-salt diet on idiopathic hypercalciuria in calcium-oxalate stone formers: A 3-mo randomized controlled trial. American Journal of Clinical Nutrition, 91, Ouwens, M., Wollersheim, H., Hermens, R., Hulscher, M., & Grol, R. (2005). Integrated care programmes for chronically ill patients: A review of systematic reviews. International Journal for Quality in Health Care, 17(2), Pak, C. Y., Peterson, R., & Poindexter, J. R. (2001). Adequacy of a single stone risk analysis in the medical evaluation of urolithiasis. J Urol.,165, Pavot, W., & Diener, E. (2009). Review of the Satisfaction With Life Scale. Assessing wellbeing: The collected works of Ed Diener (pp ). New York, NY US: Springer Science + Business Media. Pearle, M. S., Calhoun, E. A., Curhan, G. C., & The Urologic Diseases of America Project. (2005). Urologic diseases in America project: Urolithiasis. J Urol., 173,

151 144 Pearle, M. S., Calhoun, E., & Curhan, G. C. (2007). Urolithiasis. In: Litwin MS, Saigal CS, (Eds). Urologic Diseases in America. US Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Washington, DC: US Government Printing Office,; NIH Publication No [pp ] Penniston, K. L., Kaplon, D. M., Gould, J. C., & Nakada, S. Y. (2009). Gastric band placement for obesity is not associated with increased urinary risk of urolithiasis compared to bypass. J Urol., 182, Penniston, K. L., & Nakada, S. Y. (2010, October). A QOL instrument for patients with urolithiasis: Conceptual foundation and item selection. Poster presented at the 84 th annual meeting of the American Urological Association, San Francisco, CA. Rabah, D. M., Alomar, M., Binsaleh, S., & Arafa, M. A. (2011). Health related quality of life in ureteral stone patients: Post-ureterolithiasis. Urol Res., 39(5), Ranatunga, S., Myers, S., Redding, S., Scaife, S. L., Francis, M. D., & Francis, M. L. (2010). Introduction of the chronic care model into an academic rheumatology clinic. Qual Saf Health Care, [Epub ahead of print]. Reis, R.S., Hino, A. A., & Añez, C. R. (2010). Perceived stress scale: Reliability and validity study in Brazil. J Health Psychol., 15, Rief, W., Nanke, A., Klaiberg, A., & Braehler, E. (2004). Base rates for panic and depression according to the Brief Patient Health Questionnaire: A population-based study. Journal of Affective Disorders, 82,

152 145 Robinson, P., & Strosahl, K. (2009). Behavioral health consultation and primary care: Lessons learned. Journal of Clinical Psychology in Medical Settings, 16, Saigal, C. S., Joyce, G., Timilsina, A. R., & Urologic Diseases in America Project. (2005). Direct and indirect costs of nephrolithiasis in an employed population: Opportunity for disease management? Kidney Int., 68, Scales, C. D., Jr., Curtis, L. H., Norris, R. D., Springhart, W. P., Sur, R. L., Schulman, K. A., et al. (2007). Changing gender prevalence of stone disease. The Journal of Urology, 177, Semins, M. J., Matlaga, B. R., Shore, A. D., Steele, K., Magnuson, T., Johns, R., et al. (2009). The effect of gastric banding on kidney stone disease. Urology, 74, Sims, R., Gordon, S., Garcia, W., Clark, E., Monye, D., Callender, C., et al. (2008). Perceived stress and eating behaviors in a community-based sample of African Americans. Eat Behav., 9, Sinha, M. K., Collazo-Clavell, M. L., Rule, A., Milliner, D. S., Nelson, W., Sarr, M. G., et al. (2007). Hyperoxaluric nephrolithiasis is a complication of Roux-en-Y gastric bypass surgery. Kidney Int., 72, Spitzer, R., Kroenke, K., & Williams, J. (1999). Validation and utility of a self-report version of PRIME-MD: The PHQ Primary Care Study. JAMA: Journal of the American Medical Association, 282, Spitzer, R. L., Kroenke, K., Williams, J. B. W., & Lowe, B. (2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Arch Internal Med., 166,

153 146 Stamatelou, K. K., Francis, M. E., Jones, C. A., Nyberg, L. M., Jr., & Curhan, G. C. (2003). Time trends in reported prevalence of kidney stones in the United States: Kidney International, 63, Strope, S. A., Wolf, J. S., Jr., & Hollenbeck, B. K. (2010). Changes in gender distribution of urinary stone disease. Urology, 75, Strosahl, K., & Robinson, P. (2008). The primary care behavioral health model: Applications to prevention, acute care and chronic condition management. Collaborative medicine case studies: Evidence in practice (pp ). New York: Springer Science + Business Media. doi: / _8. Taylor, E. N., Stampfer, M. J., & Curhan, G. C. (2005). Obesity, weight gain, and the risk of kidney stones. JAMA, 293, Thombs, B. D., Adeponle, A. B., Kirmayer, L. J., & Morgan, J. F. (2010). A brief scale to assess hospital doctors' attitudes toward collaborative care for mental health. Can J Psychiatry, 55, U.S. Census Bureau. (2010). Median Income in the Past 12 Months (In 2010 Inflation-Adjusted Dollars). Retreived from /productview.xhtml?pid=acs_10_1yr_s1903&prodtype=table U.S. Census Bureau. (2011). Educational Attainment in the United States: Retreived from Walters, D. C. (1986). Stress as a principal cause of calcium oxalate urolithiasis. Int Urol Nephrol., 18, West, B., Luke, A., Durazo-Arvizu, R. A., Cao, G., Shoham, D., & Kramer, H. (2008). Metabolic syndrome and self-reported history of kidney stones: The National Health and

154 147 Nutrition Examination Survey (NHANES III) American Journal of Kidney Diseases, 51, Whitlock, E. P., Orleans, C. T., Pender, N., & Allan, J. (2002). Evaluating primary care behavioral counseling interventions: An evidence-based approach. Am J Prev Med., 22, Wong, E. C., & Marshall, G. N. (2010). Barriers to the collaborative care of patients with orofacial injury. Oral Maxillofac Surg Clin North Am., 22,

155 148 Appendix A: Provider Measures Urologic Provider Survey Pre-Intervention 1. I would welcome more contact with psychologists Strongly disagree Strongly agree 2. I would like more help in providing psychological and social care Strongly disagree Strongly agree 3. I would like to know more about what psychologists have to offer in the management of medical or surgical patients Strongly disagree Strongly agree 4. I would like more contact with the health psychology service Strongly disagree Strongly agree 5. Management of emotional problems is an important part of my care of chronic outpatients Strongly disagree Strongly agree 6. When psychological factors appear to be an important cause of the presenting problem, I confine myself to physical assessment Strongly disagree Strongly agree 7. Hospital doctors should be able to use psychological methods like discussion of anxiety or problems Strongly disagree Strongly agree

156 Hospital doctors are not responsible for emotional care of patients Strongly disagree Strongly agree 9. How satisfied are you that your patients currently have access to all the tools that they need with respect to their stone disease? Very unsatisfied Very satisfied 10. How comfortable would you feel having a behavioral health consultant as a full-time member of your treatment team? Very uncomfortable Very comfortable * Feel free to write additional comments in the space below: END OF SURVEY Thank you!

157 150 Urologic Provider Survey Post-Intervention 1. I would welcome more contact with psychologists Strongly disagree Strongly agree 2. I would like more help in providing psychological and social care Strongly disagree Strongly agree 3. I would like to know more about what psychologists have to offer in the management of medical or surgical patients Strongly disagree Strongly agree 4. I would like more contact with the health psychology service Strongly disagree Strongly agree 5. Management of emotional problems is an important part of my care of chronic outpatients Strongly disagree Strongly agree 6. When psychological factors appear to be an important cause of the presenting problem, I confine myself to physical assessment Strongly disagree Strongly agree 7. Hospital doctors should be able to use psychological methods like discussion of anxiety or problems Strongly disagree Strongly agree

158 Hospital doctors are not responsible for emotional care of patients Strongly disagree Strongly agree 9. How satisfied are you that your patients currently have access to all the tools that they need with respect to their stone disease? Very unsatisfied Very satisfied 10. How successful would you rate the behavioral health consultant in being able to collaborate with you in enhancing the overall treatment provided to your patients? Very unsuccessful Very successful 11. How helpful would you rate the behavioral health consultant s feedback? Not at all helpful Very helpful 12. How comfortable would you feel having a behavioral health consultant as a full-time member of your treatment team? Very uncomfortable Very comfortable * Feel free to write additional comments in the space below: END OF SURVEY Thank you!

159 152 Appendix B: Patient Measures Patient Survey Pre-Intervention SF-MPQ-2 R. Melzack and the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT), All Rights Reserved. SF-MPQ-2 contact information and permission to use: MAPI Research Trust, Lyon, France. Internet:

160 153

161 154

162 155 Satisfaction with Life Scale (Diener et al., 1985) Below are five statements with which you may agree or disagree. Using the 1-7 scale below, indicate your agreement with each item by placing the appropriate number on the line preceding that item. Please be open and honest in your responding. The 7-point scale is as follows: 1 = strongly disagree 2 = disagree 3 = slightly disagree 4 = neither agree nor disagree 5 = slightly agree 6 = agree 7 = strongly agree 1. In most ways my life is close to my ideal. 2. The conditions of my life are excellent. 3. I am satisfied with my life. 4. So far I have gotten the important things I want in life. 5. If I could live my life over, I would change almost nothing.

163 156 Perceived Stress Scale- 10 Item (Cohen & Williamson, 1988) The questions in this scale ask you about your feelings and thoughts during the last month. In each case, please indicate with a check how often you felt or thought a certain way. 1. In the last month, how often have you been upset because of something that happened unexpectedly? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often 2. In the last month, how often have you felt that you were unable to control the important things in your life? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often 3. In the last month, how often have you felt nervous and "stressed"? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often 4. In the last month, how often have you felt confident about your ability to handle your personal problems? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often 5. In the last month, how often have you felt that things were going your way? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often 6. In the last month, how often have you found that you could not cope with all the things that you had to do? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often 7. In the last month, how often have you been able to control irritations in your life? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often 8. In the last month, how often have you felt that you were on top of things? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often 9. In the last month, how often have you been angered because of things that were outside of your control? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often 10. In the last month, how often have you felt difficulties were piling up so high that you could not overcome them? 0=never 1=almost never 2=sometimes 3=fairly often 4=very often

164 157 Wisconsin Stone-QOL THE WISCONSIN LIVING WITH KIDNEY STONES QUESTIONNAIRE This questionnaire is designed to understand quality of life of patients with urinary tract (kidney) stones. The questions below ask about how your problems with kidney stones have affected you during the past month. Some questions may look very similar or have similar wording, but each one is different. Please answer the questions as honestly as possible. THE QUESTIONNAIRE IS 2-SIDED. Although you may have a number of physical or medical problems, please do your best to think only about your problems related to kidney stones. All information is confidential. Thank you for your input! 1. In the last 4 weeks, how true for you are the following statements that people with urinary tract or kidney stones sometimes make? Very true Mostly true Somewhat Not at A little true true all true A.) My energy level during the day is less than usual B.) I feel very tired or fatigued C.) My activity is limited Below are problems people with kidney stones sometimes report about their sleep patterns. In the last 4 weeks, how true are any of these problems for you? Very true Mostly true Somewhat Not at A little true true all true A.) Trouble getting to sleep or with waking up while trying to sleep B.) Needing to get up frequently while sleeping to urinate C.) Poor quality sleep or not feeling rested after sleeping D.) Difficulty returning to sleep Because of kidney stones, how true for you over the last 4 weeks are the following? Somewhat Very true Mostly true A little true true A.) I don t feel the usual freedom to travel or to attend or participate in social events B.) I force myself to go to work or school, to exercise, or to fulfill other responsibilities C.) I have missed work or family time, or lost leisure or recreation time D.) I make frequent adjustments or changes to my daily schedule E.) I have less ability than usual to focus on my work, family, or other Not at all true

165 158 commitments or interests 4. How often have you felt the following in the last 4 weeks because of kidney stones? Always Somewhat or almost Very often Hardly at all often always A.) Problems or difficulties sticking to the diet recommendations advised B.) Problems tolerating or taking prescription medications as directed Not at all, never C.) Concern about my weight D.) Unhappy with my appearance E.) Concern about my health generally Below are some physical symptoms that people with urinary tract or kidney stones commonly report. In the last 4 weeks, how often have you felt these symptoms? Always Somewhat or almost Very often Hardly at all often always A.) Nausea, stomach upset or cramps B.) Physical pain C.) Urinary frequency (feeling like you have to go more than usual) D.) Urinary urgency (sudden or unstoppable urge to urinate) Not at all, never Because of kidney stones, in the last 4 weeks, how true are the following for you? Very true Mostly true Somewhat Not at A little true true all true A.) I have less interest in sex or less sexual contact than usual B.) I need to make special arrangements when traveling C.) I have less interest than usual in socializing or being around others In the last 4 weeks, because of your kidney stones, how much have you felt the following? Very much Quite a lot Somewhat A little bit A.) Frustrated with my situation B.) Worried about what is wrong now D.) Anxious or nervous about what might go wrong in the future E.) Annoyed at the nuisances and inconveniences of my situation F.) Reduced ability, compared to usual, to cope with everyday issues or responsibilities Not at all, never F.) More irritable than usual

166 159 A few questions about you PLEASE CIRCLE YOUR RESPONSE -- WITHIN THE LAST 4 WEEKS 1.) Did or do you currently have stones in your urinary system? Not Yes No sure 2.) Did you have any pain or symptoms related to kidney stones? Not Yes No sure 3.) Did you go to the Emergency Room or urgent care because of kidney Not Yes No stones? sure 4.) Did you have a traumatic or very upsetting life event in the last 4 weeks? Not Yes No sure 5.) Were you hospitalized or otherwise seriously affected by some other health problem not related to kidney stones? Yes No Not sure Copyright 2010 by Penniston & Nakada. Reproduced with permission. No further reproduction or distribution is permitted without written permission from the authors.

167 160 Author-designed Questionnaires Questionnaire #1: All Participants at Pre-Intervention Demographic Questionnaire Directions: Please answer these questions. Complete the blank or check the box that best describes you. 1. Today s date: / / mm dd yyyy 2. What is your racial or ethnic background? (Please check all that apply) Caucasian/White African American/Black Native American Pacific Islander Middle Eastern 3. How would you describe your sexual orientation? Heterosexual Queer Bisexual Lesbian/Gay Questioning 4. What is your current relationship status? Single Married (Incl. civil unions) Divorced Coupled/partnered Cohabitating Latino/a Multiracial Asian Other Man who has sex with men Woman who has sex with women Unsure Other Engaged Widowed Separated Other 5. What is the highest grade in school that you completed? 8 th grade or less Technical degree/aa Some high school College degree (BA/BS) High school grad/ged Advanced degree (MA, PhD, MD) Some college 6. What is your current employment status? Homemaker Unemployed Retired On disability On leave of absence 7. What is your family household income (from all sources)? Less than $20,000 Between $20,000 and $49,999 Full-time employed Part-time employed Full-time student only Other Between $50,000 and $99,999 $100,000 or more

168 Number of previous stones: None One Two Three Four or more 9. What type of stones have you had (check all that apply)? Calcium oxalate Uric acid Calcium phosphate Struvite Cystine Not sure Other (identify if possible): 10. How many trips to the Emergency Room have you had related to kidney stones? None Three One Four or more Two 11. What treatment(s), if any, have you undergone to treat/remove a stone(s) (check all that apply)? No treatment/passed spontaneously Extra-corporeal shock wave lithotripsy (ESWL) Ureteroscopy Percutaneous nephrostolithotomy (PCNL) Open surgery Not sure Other 12. Do you currently take any medications for the treatment or prevention of kidney stones? Yes No 13. How satisfied are you with the urological care you have received to date from UW Health Urology? Very unsatisfied Very satisfied 14. The care that I received from UW Health Urology treated the whole me, not just my stone(s) Strongly disagree Strongly agree 15. Based on my stone experiences at UW Health Urology, I would recommend it to others with stone disease Strongly disagree Strongly agree 16. What causes kidney stones? (Please be as detailed as possible.)

169 What treatment options are available to you? Please list all that apply. If you are not in need of treatment at this time, please skip this question and proceed to # What are the risks and benefits of the treatment options available to you? 19. How can you prevent future stones? 20. How confident are you that you can effectively adhere to your recommended prevention plan? Not at all confident Very confident 21. How confident are you can prevent developing future stones? Not at all confident Very confident 22. Please rate how true the following is for you: I have total control over my stone disease Strongly disagree Strongly agree

170 How confident are you that the medical information you received from your providers was tailored to you? Not at all confident Very confident 24. I was given all the information I need to prevent future stones Strongly disagree Strongly agree 25. I was told by my urologic provider(s) what treatment options were available to me Strongly disagree Strongly agree 26. To what extent do you think that the doctor asked for your preference of the treatment you will receive for the stone? I am sure that the doctor did not ask my opinion I am sure that the doctor asked my opinion * Feel free to write additional comments in the space below: END OF SURVEY Thank you!

171 164 Questionnaire #2: Control Group Only at Post-Intervention Participant Survey Packet Dear Participant, Thank you very much for your participation to date in our research study on the effects of kidney stone disease on patient quality of life. Your participation will greatly add to our knowledge of how stone formers experience this condition and what care might best serve you. Attached is the final survey packet consisting of 10 pages. It is very important that you complete all pages and return this so that we have complete data for you. Many pages are very short and it is estimated that it will only take minutes for you to complete. Enclosed is a postage paid envelope for you to mail the survey back to me. For those that I have not received a packet back by in two weeks, I will send you a friendly reminder. Once I have received your packet, I will be conducting short one-on-one interviews. The purpose of these interviews is to gain an in-depth understanding of patients personal experiences living with and managing kidney stones in their own words. Participation in interviews is strictly voluntary. Please check below regarding your interest. Interested in being interviewed Not sure yet Not interested in being interviewed If you have any questions, please do not hesitate to contact me at (608) Thank you again and I look forward to receiving your packet soon! Kristina Penniston, Ph.D., R.D. Name Birth Date Today s Date (Packet included SF-MPQ-2, Brief PHQ, SWLS, PSS, WI Stone-QOL, and the same Author- Designed Questionnaire as given at Pre-Intervention)

172 165 Questionnaire #3: Intervention Group Only at Post-Intervention Participant Survey Packet Dear Participant, Thank you very much for your participation to date in our research study on the effects of kidney stone disease on patient quality of life. Your participation will greatly add to our knowledge of how stone formers experience this condition and what care might best serve you. Attached is the final survey packet consisting of 12 pages. It is very important that you complete all pages and return this so that we have complete data for you. Many pages are very short and it is estimated that it will only take minutes for you to complete. Enclosed is a postage paid envelope for you to mail the survey back to me. For those that I have not received a packet back by in two weeks, I will send you a friendly reminder. Once I have received your packet, I will be conducting short one-on-one interviews. The purpose of these interviews is to gain an in-depth understanding of patients personal experiences living with and managing kidney stones in their own words. Participation in interviews is strictly voluntary. Please check below regarding your interest. Interested in being interviewed Not sure yet Not interested in being interviewed If you have any questions, please do not hesitate to contact me at (608) Thank you again and I look forward to receiving your packet soon! Kristina Penniston, Ph.D., R.D. Name Birth Date Today s Date (Packet included SF-MPQ-2, Brief PHQ, SWLS, PSS, WI Stone-QOL, and the following Author-Designed Questionnaire with additional questions regarding the intervention)

173 166 Demographic Questionnaire Directions: Please answer these questions. Complete the blank or check the box that best describes you. 1. Number of previous stones: None One Two Three Four or more 2. What type of stones have you had (check all that apply)? Calcium oxalate Uric acid Calcium phosphate Struvite Cystine Not sure Other (identify if possible): 3. How many trips to the Emergency Room have you had related to kidney stones? None Three One Four or more Two 4. What treatment(s), if any, have you undergone to treat/remove a stone(s) (check all that apply)? No treatment/passed spontaneously Extra-corporeal shock wave lithotripsy (ESWL) Ureteroscopy Percutaneous nephrostolithotomy (PCNL) Open surgery Not sure Other 5. Do you currently take any medications for the treatment or prevention of kidney stones? Yes No 6. How satisfied are you with the urological care you have received to date from UW Health Urology? Very unsatisfied Very satisfied 7. The care that I received from UW Health Urology treated the whole me, not just my stone(s) Strongly disagree Strongly agree 8. Based on my stone experiences at UW Health Urology, I would recommend it to others with stone disease Strongly disagree Strongly agree

174 What causes kidney stones? (Please be as detailed as possible.) 10. What treatment options are available to you? Please list all that apply. If you are not in need of treatment at this time, please skip this question and proceed to # What are the risks and benefits of the treatment options available to you? 12. How can you prevent future stones? 13. How confident are you that you can effectively adhere to your recommended prevention plan? Not at all confident Very confident 14. How confident are you can prevent developing future stones? Not at all confident Very confident

175 Please rate how true the following is for you: I have total control over my stone disease Strongly disagree Strongly agree 16. How confident are you that the medical information you received from your providers was tailored to you? Not at all confident Very confident 17. I was given all the information I need to prevent future stones Strongly disagree Strongly agree 18. I was told by my urologic provider(s) what treatment options were available to me Strongly disagree Strongly agree 19. To what extent do you think that the doctor asked for your preference of the treatment you will receive for the stone? I am sure that the doctor did not ask my opinion I am sure that the doctor asked my opinion 20. How helpful would you rate the behavioral health consultant in helping you set realistic treatment goals? Not at all helpful Very helpful 21. What were the treatment goal(s) you set with the behavioral health consultant? 22. How helpful would you rate the behavioral health consultant in helping you reach your treatment goals? Not at all helpful Very helpful

176 How helpful would you rate the techniques used by the behavioral health consultant? Not at all helpful Very helpful 24. Please rate the degree of progress you made toward reaching your treatment goals. I achieved none of my goals I achieved some of my goals I achieved all of my goals 25. Please rate your recommendations regarding the future work of a behavioral health consultant in providing services to patients with kidney stone disease. I would not recommend that a behavioral health consultant work with patients for kidney stone disease. I would recommend the behavioral health consultant stay involved, but to a more limited degree. I would recommend the behavioral health consultant stay involved, and to the same degree as now. I would recommend the behavioral health consultant become more involved in working with patients. 26. Optional: How could the behavioral health consultant improve? * Feel free to write additional comments in the space below: END OF SURVEY Thank you!

177 170 Appendix C: Sample Patient Handouts From Integrated Behavioral Health in Primary Care: Step-by-step Guidance for Assessment and Intervention, by C. L. Hunter et al., Copyright 2009 by the American Psychological Association. Reproduced with permission. The use of APA information does not imply endorsement by APA. No further reproduction or distribution is permitted without written permission from the American Psychological Association.

178 From Integrated Behavioral Health in Primary Care: Step-by-step Guidance for Assessment and Intervention, by C. L. Hunter et al., Copyright 2009 by the American Psychological Association. Reproduced with permission. The use of APA information does not imply endorsement by APA. No further reproduction or distribution is permitted without written permission from the American Psychological Association. 171

179 From Integrated Behavioral Health in Primary Care: Step-by-step Guidance for Assessment and Intervention, by C. L. Hunter et al., Copyright 2009 by the American Psychological Association. Reproduced with permission. The use of APA information does not imply endorsement by APA. No further reproduction or distribution is permitted without written permission from the American Psychological Association. 172