5 1:1 4? . .. . 1:... ‘a..'..;. i”. ‘- . .fl v1. ‘ u 1...... v IE. Q» . a: Ir: . .1, )r l 13...; _ . .3 n32“ .. v, n». Dial .. 3:... v .4.» . 7'.) . I. ‘ {tel ~ . 1.? .34 Slyxtttt 1:1»... 2. 21.2121. ,y. . .. U‘.I , ., , .. a NeillVfi J: . ‘l .; aid...- club-:5 |:.l‘..l\. I: .n LIBRARIES MICHIGAN STATE UNIVERSITY EAST LANSING, MICH 48824-1048 This is to certify that the dissertation entitled THE RELATIONSHIP OF ATTACHMENT AND PATIENT- PROVIDER RELATIONSHIP QUALITY TO THREE EMPIRICALLY DERIVED CHRONIC-PAIN CLUSTERS presented by CAMILLE ANN HOOD has been accepted towards fulfillment of the requirements for the PhD. degree in Counseling Psychology / . fa"/I / I ' -’ major Professor’s Signature 4713/ a czflflj’ /’ ' ' / f Date ' x!“ MSU is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 2105 QEIRC/DatoDueJMd-pJS THE RELATIONSHIP OF ATTACHMENT AND PATIENT-PROVIDER RELATIONSHIP QUALITY TO THREE EMPIRICALLY DERIVED CHRONIC-PAIN CLUSTERS By Camille Ann Hood A DISSERTATION Submitted to ‘ Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Educational Psychology, and Special Education 2005 ABSTRACT THE RELATIONSHIP OF ATTACHMENT AND PATIENT-PROVIDER RELATIONSHIP QUALITY TO THREE EMPIRICALLY DERIVED CHRONIC-PAIN CLUSTERS By Camille Ann Hood Turk & Rudy’s (1988) empirically derived clusters of chronic-pain patients were originally created using psychological and social data from the Multidimensional Pain Inventory (MPI) (Kems, Turk, & Rudy, 1985). Labels for the clusters—dysfunctional, interpersonalIy distressed, and adaptive coper—were informed by salient distinguishing features. These chronic-pain subgroups have demonstrated robustness in populations with different medical diagnoses and using different instruments. Initial studies suggested this taxonomy have merit for tailoring chronic-pain assessment and treatment. Despite this, very little is known about the characteristics of individuals within each group. The relationship between attachment style and patient-provider relationship quality to each of three chronic-pain subgroups was examined. The sample population consisted of 141 chronic-pain patients collected at an outpatient neurology clinic located in a mid-sized city in the Midwest. Fearful and secure attachment styles were moderately related to the interpersonally distressed and adaptive coper subgroups. No relationships were demonstrated between dismissing or preoccupied attachment styles and the chronic-pain subgroups. The present study did not support a relationship between patient-provider relationship quality and any of the subgroups. However, a relationship was found between attachment and patient-provider relationship quality with individuals in the secure attachment group reporting better patient-provider relationship quality. The implications and limitations of this study’s findings are discussed, along with recommendations for future research. Thoughts regarding contribution counseling psychology can make to the study of attachment and chronic pain conclude this study. Capyfisht by CAMILLE ANN HOOD zoos To my husband, Dan, and our children, Katie and Madisen ACKNOWLEDGEMENTS One insidious aspect of writing a dissertation is that the commitment to the process all but precludes time to contemplate. However, having arrived at the end of this process, I relish reflecting on all of the individuals who have contributed to my success and the completion of this dissertation. I am deeply appreciative of Dr. Nancy Crewe’s willingness to mentor. It is through her direction and expertise that this dissertation was completed. I am especially appreciative of her support during the tough times and consider her both a role model and a fiend. I am indebted to the other members of my committee whose support has been unwavering. Dr. Gregory Spence-J ones was there for me at the inception of this project and contributed numerous hours helping me hone the research topic. He supported my desire to explore health psychology and modeled the role of a counseling psychologist in this field. I will always remember his statement, “1 good dissertation is a done dissertation.” This became my personal mantra, which I have passed on to others. I would also like to thank the other members of my committee, Drs. Kimberly Maier and Jean Baker, for their time and consultation. To all of my peers, colleagues, and friends, I extend my indebtedness for their unwavering support. It was through their commiseration, humor, and words of encouragement that I realized that “this too shall pass.” I extend my gratitude to the staff at Neurology of Battle Creek, who always asked if there was anything they could do to help me. Their support and professionalism vi permitted data collection to proceed fluidly. I also extend my appreciation to Dr. Mark Reckase for his humor and statistical knowledge. I mourned the passing of my parents while I was in graduate school. It was through the encouragement of my mother that I started this joumey--its end will be bittersweet. Mentioned almost last, but appreciated most, I honor my husband, Dan, with my deepest respect. His personal sacrifices are too numerous to mention and his willingness to be both mom and dad when I was in the depths of internship will always hold a warm spot in my heart. My children, Katie and Madisen’s maturity extends beyond their years as they had to learn to accommodate my mental and physical absence during parts of this process. Thank you for being such wonderful, strong women. I hope that, just as others have been role models for me, I can be a role model for you. Remember, you are never too old to learn—even when it feels like you are. vii TABLE OF CONTENTS LIST OF TABLES .............................................................................. xi CHAPTER ONE ....................................................................................................... 1 RESEARCH PROBLEM .......................................................................................... 1 Introduction ......................................................................................................... 1 Background ......................................................................................................... 2 Importance of the Study ....................................................................................... 4 Prevalence ........................................................................................................... 5 Complications to the Individual ........................................................................... 6 Cost to Society .................................................................................................... 9 Integrative Approach to Health Care and Its Management ................................. 10 Biopsychosocial Variables Related to Chronic Pain ........................................... 12 Empirically Derived Clusters of Chronic-Pain-Patients ...................................... 14 Attachment Style ............................................................................................... 18 Patient-Provider Relationship Quality ................................................................ 21 Summary ........................................................................................................... 22 Research Questions ................................................................................................ 23 Research Hypotheses ......................................................................................... 24 Assumptions ...................................................................................................... 26 Overview ........................................................................................................... 26 CHAPTER TWO .................................................................................................... 28 LITERATURE REVIEW ........................................................................................ 28 Overview ........................................................................................................... 28 The Status of Chronic Pain Treatment and Recommendations for Improvement ................................................................................................. 28 Empirically Derived Clusters of Chronic-Pain Patients as Taxonomy ................ 31 Three Subgroups of Chronic-Pain Patients ......................................................... 31 Research Utilizing Turk and Rudy’s Taxonomy ................................................. 33 Need for Additional Research on Variables Defining or Predicting Membership ................................................................................................... 35 Attachment ........................................................................................................ 36 Empirical Evidence Supporting Attachment Theory .......................................... 39 Research on Attachment Related to Medical Patients ......................................... 43 Physiological Explanations of Attachment ......................................................... 45 Psychological Explanations of Attachment ............................................. 47 The Relationship of Attachment to Patient-Provider Relationship Quality ......... 49 Patient-Provider Relationships ........................................................................... 53 Chronic-Pain Patient-Provider Relationships ..................................................... 55 Measurement of the Provider-Patient Relationship Construct ............................. S6 viii Perceived Quality of Information ....................................................................... 60 Affective Aspects .............................................................................................. 61 Patient Communication ......................................................................................... 63 Summary ............................................................................................................... 64 CHAPTER THREE ................................................................................................. 66 METHODS ............................................................................................................. 66 Introduction .................................................................................. 66 Research Hypotheses ......................................................................................... 66 Selection Criteria, Demographics, and Data Collection Overview of the Study Site ........................................................................... 68 Selection Process and Description of the Sample ............................................... 68 Instrumentation .................................................................................................. 72 Patient Information Questionnaire (PIQ) ............................................................ 72 The West Haven-Yale Multidimensional Pain Inventory (MPI) ......................... 72 The Patient Reactions Assessment (PRA) .......................................................... 73 Experiences in Close Relationships (ECR) ......................................................... 74 Procedures and Statistical Analyses ................................................................... 76 Procedures Used to Create Chronic-Pain Patient Subgroups ............................... 76 Statistical Procedures ......................................................................................... 78 Analyses of Classifiable Participants ................................................................. 78 Differences Among Subgroups .......................................................................... 78 Comparisons of the Independent Variables among Subgroups ........................... 80 Multinomial Logistic Regression ....................................................................... 81 Summary ........................................................................................................... 82 CHAPTER IV ......................................................................................................... 83 ANALYSIS OF RESULTS ..................................................................................... 83 Introduction ....................................................................................................... 83 Participants ........................................................................................................ 83 Classification of Participants into Chronic Pain Subgroups ................................ 85 Comparison of Chronic Pain Subgroups ............................................................ 87 Classification of Attachment Style ..................................................................... 92 Attachment Style and Chronic Pain Subgroup .................................................... 94 Patient Provider Relationship Quality and Chronic Pain Subgroup ..................... 97 Attachment Style and Patient Provider Relationship Quality .............................. 97 Attachment Style and Patient Provider Relationship Quality as Predictors of Chronic Pain Subgroup Membership .................................... 99 Summary ........................................................................................................... 99 CHAPTER FIVE ................................................................................................... 100 DISCUSSIONS AND CONCLUSIONS ................................................................ 100 Introduction ..................................................................................................... 100 ix Overview ......................................................................................................... 100 Attachment and Chronic Pain Subgroup .......................................................... 101 Characteristics of Attachment .......................................................................... 102 Characteristics of Chronic Pain Subgroups ...................................................... 106 Patient-Provider Relationship Quality .............................................................. 108 Attachment and Patient-Provider Relationship Quality .................................... 111 Limitations of the Study .................................................................................. 111 Recommendations for Future Research ............................................................ 114 Importance to Counseling Psychology ............................................................. 116 APPENDIX DOCUMENTATION AND LETTERS ................................................................... 119 REFERENCES .......................................................................................................... 142 LIST OF TABLES Table Page 4.1 Comparisons of Means, Standard Deviations, and Reliabilities from MPI ............................................................................................ 86 4.2 Gender Count and Percentages by Chronic-Pain Subgroup ................. 88 4.3 Marital Status by Chronic-Pain Subgroup ............................................ 89 4.4 Race/Ethnicity by Chronic Pain Subgroup ........................................... 90 4.5 Employment by Chronic-Pain Subgroup............ .................................. 91 4.6 Fisher's Linear Discriminant Functions Provided by Brennan et al. (1998) for Computation of Attachment-Style Categories .................. 93 4. 7 Attachment Style by Pain Subgroup .................................................... 96 4.8 Patient-Provider Relationship Quality by Attachment Style ................ 98 xi CHAPTER ONE RESEARCH PROBLEM Introduction Turk & Rudy’s (1988) empirically derived clusters of chronic-pain patients were originally created using psychological and social data fi'om the Multidimensional Pain Inventory (MPI) (Kems et al., 1985). Labels for the clusters— dysfunctional, interpersonally distressed, and adaptive coper—were informed by salient distinguishing features. These clusters have demonstrated robustness in populations with different medical diagnoses and using different instruments. Initial studies suggest this taxonomy have merit for tailoring chronic pain assessment and treatment. Despite this, very little is known about the characteristics of individuals within each group. This study examined the relationship between attachment style and the quality of the patient-provider relationship in each of the three clusters. Clusters of chronic-pain patients were examined fi'om a sample population collected at an outpatient neurology clinic located in a mid-sized city in the Midwest. Attachment style and quality of the patient-provider relationship were analyzed for potential relationships with the subgroups. The relationship between attachment style and the quality of patient-provider relationship was examined. Had significant relationships existed, attachment style and patient-provider relationship quality would have been examined to see if they improved the prediction of membership into a chronic pain cluster. Background Pain, and efforts to manage it, have long been a focus of human endeavor. Paradigms addressing pain have historically vacillated on a spectrum. One end views pain using a holistic model that emphasizes body-mind integration, while the other end uses a biomedical reductionism model in which the body and mind are seen as dualistic or separate. The paradigm from which pain is viewed is important because it serves as the template from which theories of etiology and treatment arise. The holistic model of pain is the result of a paradigm shifi occurring in the last century. The modem change in approach—from dualism to integration—started with the psychoanalytic work of Sigmund Freud. His work on hysteria and conversion reactions emphasized the perception of physiological maladies through a model focusing on the interactions of the mind and body (Gatchel, 1999). The integration approach was buttressed in 1955 by studies of the relationship between psychological factors and the health concerns of hypertension and coronary heart disease (National Research Counsel, 2001). Support for the utility of this model occurred when repeated attempts to understand the assignment of a cause of disease fi'om the sole contribution of medical and biological components proved inadequate (Turk & Flor, 1984). Further support for the integration approach grew as researchers demonstrated that biological and mechanistic explanations alone fail to account firlly for the inconsistency between level of pathology measured and disability (Magora & Schwartz, 1980; Turk, 1996). Research suggests some of that discrepancy is caused by psychological and social factors, supplying missing pieces of this puzzle (Bonica, 1953; Engel, 1968). A shift from psychodynamic to cognitive/behavioral models occurred in the areas of psychology, in general, and health psychology, specifically. It allowed the scientific method to be applied and heralded the transformation of previously ambiguous psychological and social constructs into measurable ones. Factors, such as social support and psychological stress, were defined and quantified. Pain, defined as an experience with objectionable sensory and emotional components related to tissue damage (Portenoy & Kanner, 1996) also became a measurable construct. Use of this model accelerated research into the interrelations of social factors and individual thought and behavior as they pertain to health issues, which resulted in more standardized nomenclature. Health psychology examines the realms of resilience, health, and health care as well as, disease, infirmity, and related interventions from a psychological perspective (Johnson, 1994). It has become a vital and diverse subject in clinical practice and research. Lamentably, much of the research in health psychology has been fragmented, failing to answer broad questions. This situation prompted the National Institutes of Health (NIH) to evaluate potential firture directions for the integration of psychology and sociology into the study of health The Committee of Future Directions for Behavioral and Social Sciences Research was formed to address these matters (Singer & Ryff, 2001). This committee was assigned two tasks: evaluating the potential impact of behavioral and social science research on public health and disease prevention, and prioritizing areas of focus that would benefit from the integration of various disciplines. Addressing the second task, the committee responded with a list of ten thematic regions. Pertinent to this study are three regions: (a) increasing the understanding of behavioral and cognitive processes and social factors associated with the disease process, (b) improving the effectiveness of psychological and sociological interventions, and (c) identifying factors associated with resilience and well-being (Singer & Ryff, 2001). These recommendations encompass various aspects of health and disease and can be applied to diverse populations. To address effectively these thematic areas, research must systematically address current limitations existing in specific areas and populations. Keeping with the recommendations of the Committee of Future Directions for Behavioral and Social Sciences Research, this study incorporates the aforementioned regions, examining attachment style and patient-provider relationship quality in homogeneous clusters of chronic-pain patients. Importance of the Study Chronic pain is one of the most prevalent medical disorders in the United States (Harstall & Ospina, 2003; LaPlante & Carlson, 1996). On an individual level, pain decreases the quality of life by altering physical and social functioning. On a societal level, pain levies enormous costs via medical service usage and lost productivity. Its prevalence, associated complications to the individual, and resulting societal costs are just three detrimental aspects of chronic pain. To understand its impact, it is first necessary to understand how chronic pain differs from acute pain. Acute pain follows an expected course for the nature and extent of tissue damage. An individual’s response to acute pain is localized to the affected area and recovery proceeds within an expected time frame. Chronic pain, on the other hand, exists in intensity or duration above or beyond expected levels (Bonica, 1953). Furthermore, the effects of chronic pain are widespread. A better understanding of the systemic effects of pain can be garnered by reviewing the International Association for the Study of Pain (IASP’s) multiaxial approach in the assessment of chronic pain. This multiaxial approach creates specific diagnostic categories. Recommended areas for assessment include: (a) the region in which the pain exists, (b) the onset and intensity of the pain, (c) the systems of the body, (d) the etiology and duration, and (e) behavioral and psychosocial elements (IASP, 1986; Kems, Turk, & Rudy, 1985). This array of categories hints at the all-encompassing impact of chronic pain. Prevalence Despite the knowledge that pain is the number one reason individuals seek medical treatment (Stucky, Gold, & Zhang, 2001), a reliable comprehensive estimate of the prevalence of chronic pain is difficult to compute (V erhaak, 1998). In addition, research suggests that up to 75% of visits to primary care physicians for physical complaints may be because of psychosocial issues (Roberts, 1994). The following figures are used as indicators of the prevalence of chronic pain. Reviewing other studies, Clark (2002) reported chronic-pain prevalence ranged between 20% and 60% in general medical populations. He attributes the variance to differences in how chronic pain is calculated and the populations examined in these studies. For instance, examining a population of veterans using services at general medical clinics, Clark (2002) found the prevalence of chronic pain to be 48%. Between 75 and 85 million Americans affirm having chronic pain and an estimated 50 million are afflicted with disabilities (Berman & Sawyers, 1997). Of the elderly, 88% report chronic pain, while 68% of the middle-aged are estimated to have one or more chronic disorders, many associated with pain, which are not remediated by medical treatment (Hoffman, Rice, & Sung, 1996). Aronoff (2000) estimated that 80 to 85% of individuals over the age of 65 will have health issues predisposing them to pain. An annual estimate, based on the 1977-1978 calendar year, stated that more than 16 million people who went to physicians, it was for back pain alone (Cypress, 1983). Another estimate purports up to 85% of working adults will seek treatment or miss work because of musculoskeletal pain (Fordyce, 1995). Complications to the Individual The American Academy of Family Physicians (Physicians, 2005) listed the top 20 reasons for visits to primary care physicians, a list remarkable for the predominance of pain-specific reasons, including headaches, back pain, neck pain, disorders of the joints, and abdominal pain—which are among the most commonly noted reasons for missing work. Lost work time and the cost of medical visits and treatment clearly take chronic pain into the realm of economic impact. The Bureau of Labor Statistics (1998) estimated that in 1996, there were 425,000 episodes of upper extremity disorders resulting in periods of missed work Lower back pain ascribed to work-related injuries is so prevalent, it is described by Frymoyer (1991) as epidemic. Missed work because of chronic pain causes enormous financial strain to individuals and their families. Employees with acute pain also may miss work, but those absences tend to be understood better, have a predictable course, and an end point. In contrast, the nature of chronic pain is frequently misunderstood, its time course is unknown, and a definitive end point may not exist. The enduring nature of chronic pain frequently results in employment termination as individuals exhaust annual sick leave and holiday pay that is typically measured in days or weeks per year. Complicating this bleak outcome, a patient’s potential to earn money is stalled when financial resources are needed most. Buckle (1997) presented a timeline that delineated the development and presentation of chronic disorders and the concurrent change in labels associated with L the individual from worker to patient. He set time zero as exposure to work followed by onset of discomfort, onset of pain, medical consultation, referrals, and psychological evaluation. This timeline is often measured in years as the task of identifying physical and psychosocial causative agents is pursued. Turk and Nash (1996) estimate patients treated in pain clinics have endured pain for an average of seven years. In many cases, an accurate diagnosis (e. g., degenerative disk disease, multiple sclerosis, and rheumatoid arthritis) may lead to more effective treatments of the symptoms, but not a cure. One study estimated that 82% of chronic-pain patients who sought treatment at a pain clinic continued to deal with pain two years later (Croolc Weir, & Turk, 1989), and reports of individuals who have experienced chronic pain for up to 30 years are not uncommon (Turk & Nash, 1996). Unfortunately, in many cases, pain persists for the remainder of the patient’s life. Third-party payers, such as health maintenance organizations and insurance companies, also make a living with chronic pain difi'rcult. Ostensibly, the purpose of accurate assessment is to allow intervention based on correctly identified causative agents. Too often, however, accountability is the covert force driving assessment. In this realm, the political agenda of accountability is often at odds with optimal care. Buckle (197 7) astutely identified accountability as a clandestine reason for the search for causative agents and subsequent delays in treatment. In the United States, accountability is highly correlated with responsibility for bearing the financial burden. With money as a driving force, the motivation to discern malingerers from nonmalingers is high. Buckle suggested that the litigious nature of our society and the adversarial nature of the legal system pits employee against employer; insurance companies against patients; and employers, insurance companies, and patients against physicians. Chronic-pain patients may lose credibility under these conditions because their affliction is confirsed with malingering. Many patients report feeling as if others view them as “faking it” or exaggerating their symptoms for money or benefits. This cynicism may emanate from family members, as limited participation in family activities or household chores, strains close social relations. Increased levels of interpersonal distress (e. g., marital and familial discord) are reported at a time when support and comfort are essential (Schwartz, Slater, & Birchler, 1996). Intertwined with occupational, financial, and familial difficulties, psychological distress is also common in chronic-pain patients (Schwartz, Slater, & Birchler, 1996). Chronic pain is associated with higher rates of co-mor’oid depression, anxiety, and fear (Bennett et al., 1996; Lee, Cheung, Man, & Hsu, 1992). It is also associated with higher levels of guilt, pessimism, and self-criticism (Trief, Elliott, Stein, & Frederickson, 1987), and decreased social support (Rodin, Cravin, & Littlefiled, 1991). Morley, Williams, and Beck (2002) found relationships between chronic pain and negative self views, such as punishment and self-dislike. They also found that chronic pain was related to detrimental physical functioning, such as social withdrawal. Increased hopelessness, frustration, and demoralization occur as patients encounters repeat failures in their search for relief (Turk & Nash, 1996). Cost to Society The detrimental impact of chronic pain extends beyond the personal level to the societal level. As the prevalence of chronic pain is difficult to estimate, its cost to society is also difficult to measure. However, its enormous impact is shown in the following figures. Chronic low back pain alone is estimated to cause $8 billion annually in health-care costs (Bacon et al., 1994). Frymoyer (1993) estimated health- care costs to be $50 billion for chronic back pain when direct and indirect cost dollars are included. Gatchel and Epker (1999) stated that treatment for chronic-pain conditions accounts for most of the $70 billion in annual health care cost, despite evidence that it is under treated (Vidal, 2002) . Pain is a major contributor to rising health care expenditures, which comprise more than 14% of the gross domestic product despite the ongoing effort to contain costs (Mayer, Prescott, & Gatchel, 2000). Today the annual cost is estimated at $100 billion including hospital stays and decreased worker productivity (Kalb, 2003; Lipman, 2004). The massive impact of chronic pain on individuals and society suggests research related to chronic pain is warranted. It is estimated that as many as 7 5% of somatic complaints, such as pain and fatigue, seen in primary care settings are better accounted for by psychosocial factors than organic causes (Kroenke & Mangelsdorfi; 1989). Given the variability of individual response to chronic pain and its treatment, it is clear that chronic-pain research should address not only biomedical aspects, but also psychological and social factors as part of an integrative approach aimed at elucidating causes and effective treatments (Turk & Nash, 1996). Integrative Approach to Health Care and Its Management Complicating the treatment of patients with chronic pain is the fi'equency of inaccurate assessments of causative agents by patients and professionals. One back- pain study noted a weak association between the amount of pain experienced and the results ofMRI or X-ray imaging (Jarvik, Hollingworth, Heagerty, R., & Deyo, 2001). At times, chronic back-pain patients seeking to legitimize their pain tout abnormal imaging results as proof of their pain. However, many pain-fi'ee individuals show the same spinal abnormalities that are said to cause pain in symptomatic individuals (Turk & Gatchel, 2000). It is not uncommon for patients to experience the same or greater levels of pain, despite having surgery to correct the alleged causative agent. This paradoxical outcome is due, in part, to current limitations of knowledge and technology that hamper the medical professional’s ability to identify firlly the source(s) of chronic pain. Limited understanding of the impact of psychological and sociological factors also contributes to the inability to consistently predict which patients will experience pain relief from surgical intervention. These limitations lead to unnecessary and misguided surgery, which adds to the costs associated with chronic pain. Interventions to treat pain that are informed by the psychological and sociological influences on patients would reduce health-care costs (Gatchel et al., 2003). Compas, Haaga, Keefe, Leitenberg, and Williams’ (1998) review provided 10 evidence for the efficacy of health psychology interventions aimed at addressing psychosocial Concerns, though considerable work remains around the development, implementation, and assessment of intervention protocols. Thus, the practice of psychosocial intervention for improving health care is uneven. The public is familiar with some of the work conducted by health psychologists. Integrative techniques employed for cigarette smoking cessation, for example, are well known. Recommended techniques usually include a combination of support groups, patient education, and therapy. Compas and colleagues (1998) reviewed interventions following the criteria of Chambless and Hollen (1998) for determining empirically supported treatments. They reported that face-to-face psychosocial interventions, such as group behavior therapy programs (Hill, Rigdon, & Johnson, 1993) and intensive smoking cessation interventions, conducted in groups (Stevens & Hollis, 1989) are efficacious. The ubiquitous nature of these interventions—well known within the medical community—and their proven efiicacy helps make the use of psychosocial interventions to treat health concerns acceptable to the public. In contrast, integrative interventions for dealing with chronic pain are underutilized by the American medical establishment, despite the existence of empirically validated treatments including operant-behavioral therapy and cognitive- behavioral therapy (Compas, Haaga, Keefe, Leitenberg, & Williams, 1998). As in many fields, updates in the standard of care lags years behind theory and research. Nevertheless, interventions using biopsychosocial approaches for chronic pain are becoming used more widely as their efficacy becomes known. Contributing to the 11 improvement of these interventions is a constant refinement of integrative models of chronic pain. Biopsychosocial Variables Related to Chronic Pain Psychiatrist George Engel (1977) is frequently credited with initiating a comprehensive perspective to address health concerns. This perspective has been modified over the years and is currently referred to as the biopsychosocial model. In tandem with the development of the biopsychosocial model in medicine, an analog model was developed to understand chronic pain (Gatchel, 1999). This model provides a template for understanding the relationship among psychosocial factors and pain and has revolutionized treatment. One major way the biopsychosocial model has changed pain treatment is through its axiom that the experience of pain is similar regardless of the contributing factors. In other words, organic and psychogenic pains are equally debilitating. Prior to this model, organic and psychogenic pains were viewed as two mutually exclusive elements having different perceptual qualities. Organic pain refers to pain that originates in the body and that has an identifiable cause. Psychogenic pain is idiopathic; the level of pain experienced exceeds the quantity, quality, or duration expected. This second type of pain is referenced by some patients in statements such as: “My doctor believes the pain is all in my head,” and “I was referred to a psychologist because they don’t believe my pain is real.” These prototypical patient comments suggest that the conceptualization of real (i.e., organic) and imaginary (i.e., psychogenic) pain as separate entities continues to exist in the general population. The biopsychosocial model validates psychogenic pain and addresses it in treatment. 12 Rather than an either-or rubric, the biopsychosocial model supports an understanding of the interaction of both psychological and physiological factors in the interpretation of pain (Melzack & Wall, 1965). This comprehensive perspective serves as a framework for research to explain how an array of psychosocial variables, including beliefs, attitudes, emotions, perceptions, and social and environmental impact an individual’s pain experience. Incorporating the influence of these factors has changed the management of chronic pain. The sheer number of variables that influence the experience of chronic pain has led to a multimodal or integrated approach to pain treatment—one that involves the use of multiple disciplines to manage and reduce pain. Although the specific disciplines utilized vary, a common configuration is to use interventions fi'om medicine, physical therapy, and psychology. Numerous studies have shown improved outcomes with the inclusion of psychosocial interventions. The following three studies demonstrate prototypical findings on the effects of psychological interventions. Using progressive relaxation, along with covert imagery and cognitive refraining, Levendusky and Pankratz (1975) found a decrease in drug use and moderation of pain. In another study, a decrease in pain was reported after the use of cognitive reinterpretation (Rybstein-Blinchik, 1979). An improvement in depressive symptoms and in physical functioning was reported in patients completing a three-week inpatient program for managing pain (Jensen, Turner, & Romano, 1994). These studies, despite their positive outcomes, fall short in their ability to universally impact treatment of chronic-pain patients because while some patients show improvements, others do not. Another problem 13 with these programs is their inability to ameliorate permanently the entire spectrum of pain-related concerns. Improvements might be cited in one or more aspects of an individual’s life, (e. g., psychological, physical, familial, occupational, or social) but across-spectrum improvements are not reported. Programs applying treatment en masse without consideration of the unique needs of the participating individuals are partially responsible for these shortcomings. In an effort to address current limitations in treatment and to add to the theoretical foundation from which treatments can be customized, this study proposes to examine attachment style and patient perceptions of the patient-provider relationship—two factors important to understanding unique characteristics of chronic pain patient subgroups. Empirically Derived Clusters of Chronic-Pain-Patients Turk and Okifuji (2001) reported that research conducted on populations of chronic-pain patients has examined them as subgroups based upon biomedical classifications (e. g., back pain, temporomandibular joint pain, arthritis) or as subgroups based on psychopathological characteristics (e.g., depression, anxiety, personality disorders). This second scenario occurred because researchers employing cluster analysis utilized measures of psychopathology, such as the Minnesota Multiphasic Personality Inventory (MMPI) (Bradley, Prokop, Margolis, & Gentry, 1978; Armentrout, Moore, Parker, Hewitt, & F eltz, 1982; McCreary, 1985) to classify pain patients. Burns, Kubilus, Bruehl, and Harden (2001) noted that the problem with these analyses is the failure of the MMPI to measure comprehensively aspects of the 14 population necessary for meaningful grouping based upon integrated data This became obvious when comparing the scales of the MMPI to the International Association for the Study of Pain (IASP, 1986) taxonomy discussed above or that of the Multiaxial Assessment of Pain (MAP). The MAP examines variables reflecting patients’ physical and psychological status and behavioral and social functioning. Rather than incorporating factors to portray a complete picture of potential individual variance—such as social disruption, interference with life, physical activities, and pain severity—the MMPI measures facets of psychopathology. Subgroups emerging from cluster analysis on instruments measuring psychopathology alone have limited heuristic value because of the limited scope fi'om which the clusters were derived. To remedy this concern, Turk & Rudy (1988) used cluster analysis on data {Tom the Multidimensional Pain Inventory (MPI; Kerns, Turk, & Rudy 1985). This instrument assesses a wide spectrum of variables, and is grounded in a cognitive behavior model of pain. Arrays of symptoms are assessed using nine empirically derived scales. These scales are grouped into three sections. The first section contains five subscales—Pain Severity (PS), Interference (I), Support (S), Life Control (LC), and Affective Distress (AD). The combination of these scales measures a chronic pain patient’s perceptions of cognitive, affective, and social aspects of pain. The second section has three subscales, which measure the behavioral responses of others to pain behavior—Puni shing Responses (PR), Solicitous Responses (SR) and Distracting Responses (DR). The last section is a checklist of activities and has one scale—General Activity (GA). Using cluster analysis, Turk and Rudy (1988) discovered three distinctive subgroups which they labeled using monikers identifying 15 primary characteristics of each group—dysfirnctional, interpersonally distressed, and minimizers/adaptive copers. These empirically derived clusters were replicated in the same study. Other researchers (Rudy, Turk, Zaki, & Curtin, 1989; Turk & Rudy, 1990) have replicated the findings of Turk and Rudy (1988) suggesting the robustness of these clusters. The Dysfirnctional cluster profile is comprised of individuals who indicated high levels of pain severity (PS), psychological distress (PD), and interference with their life (I) while also reporting lower levels of perceived control (LC) and general activity (GA). The cluster profile labeled Interpersonally Distressed reflects scores indicating problems with social support (8) and an increased perception of punishing responses from others (PR), along with lower scores on the Solicitous Responses (SR) and Distracting Responses (DR) scales. The third cluster profile, Minimizes/Adapters, indicated lower pain severity (PS), life interference (I), and affective distress (AD) scores with higher scores for perceived life control (LC) and general activity (GA). It appears that individuals in this group are adapting to, or coping better with their pain than individuals in the other two clusters. As a caveat, the authors note the possibility that these individuals were minimizing or responding in a socially desirable manner. Therefore, they labeled this third group minimizers/adaptive copers. Subsequent research has supported the idea that this group is more accurately distinguished as adaptive copers rather than nrinimizers. Burns, Kubilus, Bruchl, and Harden (2001) extracted a fourth empirically derived cluster of chronic-pain patients building on the work of Turk and Rudy (1988) and Weinberger, Schwartz ,and Davidson’s (1979) work on repression using 16 measures of defensiveness. Guided by theory suggesting the existence of patients who use repression as a defense mechanism, they used the Multidimensional Pain Inventory (MP1) and Balanced Inventory of Desirable Responding (BIDR; Paulhus, 1984) as a means of detecting this group. A subset of the dysfunctional group, this cluster was labeled Repressors because of the juxtaposition of higher scores on pain severity and interference with life with lower scores on cognitive and affective symptoms of anxiety and depression. Consistent with theory, individuals using repression as a defense mechanism report higher levels of physical distress with lower levels of psychological distress. Since no other studies to date have validated the findings of Burns, Kubilus, Bruehl, and Harden (2001) and the fourth group is a subset of the dysfirnctional group, this study will focus on the three clusters originally derived by Turk and Rudy (1988). The robustness of Turk & Rudy’s empirically derived clusters of chronic-pain patients has been replicated in validation studies (Turk & Rudy, 1988), studies using different types of pain (Rudy, Turk, Zaki, & Curtin, 1989; Turk & Rudy, 1990), and studies using different constructs (Jamison, Rudy, Penzien, & Mosley Jr., 1994; Strong, Ashton, & Stewart, 1994). The ability to recreate these clusters under different conditions suggests their merit as a unique classification taxonomy. The comprehensive nature of the criteria establishing these clusters hints at their heuristic value in comparison to groups based on pathology alone. Drawing on factors portraying a complete picture of potential individual variance—including pain severity, social disruption, interference with life functioning, and quality of physical activities—is consistent with a Multiaxial Assessment of Pain (MAP) based 17 taxonomy which views the impact of chronic pain as affecting the individual as a whole. Researchers studying empirically derived clusters of chronic-pain patients hypothesize that increased understanding of this taxonomy may serve to improve basic and applied research (Main, Wood, Hollis, Spanswick, & Waddell, 1992; Talo, Rytokoski, & Puukka, 1992). Of particular interest is the potential to tailor treatments to match patient characteristics (Turk, 1990). Analogous to the ability to transfirse blood, based on understanding individual blood types, pain treatment matches based on an individual’ 3 taxonomy type offers the potential for more effective treatment, and an improved outcome and quality of life. The next logical step is to understand better the relationships between each of the clusters and other pertinent variables. Initial forays have been undertaken examining the relationship between these groups and psychological factors, such as pain-related anxiety, acceptance of pain, and perceptions of the future (HellstrOm, J ansson, & Carlsson, 2000; McCracken, Spertus, J aneck, Sinclair, & Wetzel, 1999). There exists, however, a paucity of studies examining the relationships between empirically derived chronic pain groups and other variables. Finding variables contributing to or predictive of group membership would advance understanding of group characteristics. Two unique but related variables are attachment style and patient perceptions of the quality of the patient- provider relationship. Attachment Style Bowlby’s seminal work on the relationships between a child and caregiver explains: (a) the development of attachment (e.g., emotional bonds) between the 18 infant and caregiver, (b) the development of internal working models, self and other, and (c) the patterns of subsequent attachments that form based on these internal schemas (Bowlby 1969/1982). Contemporary attachment theory, an extension of Bowlby’s worlc explains factors that promote or obstruct individual development, based on internalized schemas developed during infant-caregiver interactions. It also explains how identity formation and psychological functioning are related, and it has been used to link past bonding styles with current and firture attachments. More recently, attachment theory has been examined as a predictor of health and treatment outcomes (Ciechanowski, Katon, Russo, & Walker, 2001; Ciechanowski, Walker, Katon, & Russo, 2002; Dozier, 1990; Kotler, Buzwell, Romeo, & Bowland, 1994). Interactions between an infant and parent instill a mental template or internal working model that becomes increasingly stable as interaction patterns are repeated. This template is comprised of two components reflecting views of the self and the other (Bartholomew & Horowitz, 1991; Bowlby, 1969). The view of self determines whether sufficient self-worth exists to be cared for by others. The view of another, in essence, determines the trustworthiness of others to provide care. These two components combine to determine subsequent attachment styles throughout one’s life. Attachment styles are obtained by assigning positive and negative weights to both components. These four combinations result: 1) positive self-view, positive view of others; 2) positive self-view, negative view of others; 3) negative self-view, positive view of others, and; 4) negative self-view, negative view of others. The following four attachment types are respectively obtained: secure, dismissing, preoccupied, and fearful. Bowlby (1973) proclaimed the existence of a causal l9 relationship between experiences individuals have with their parents and their subsequent capacity to establish affectionate bonds with others. In other words, the templates of perception around self and others created during initial bonding directly influence the types of subsequent relationships formed throughout life. Chronic-pain patients are fi'equently involved in relationships with providers that permit the expression of attachment style. Activation of internal working models occurs especially during periods of stress (e.g., the stress experienced with chronic pain) (Bowlby, 1988). Several studies have demonstrated links between attachment style and health care utilization and the patient-provider relationship (F eeney, 2000; Gerlsma & Luteijn, 2000; Noyes Jr. et al., 2003). Maunder & Hunter (2001) reported that insecurely attached individuals are associated with increased disease risk. They also found support, albeit less robust, for an association between insecure attachment style and treatment nonadherence. Attachment style has been studied in psychology for some time (e. g., Pistole, 1999; Hazan & Shaver, 1994a, 1994b) however, the study of attachment in the arena of psychosomatic medicine is more recent. There exists a paucity of research in the area of attachment, as it relates to psychosomatic illness, and few articles exist relating attachment styles to chronic pain. The importance of attachment as a predictor of psychological status (e. g., Arrnsden & Greenberg, 1987; Bartholomew & Horowitz, 1991; Kobak & Hazan 1991; Simpson, 1990) and initial findings suggesting a relationship between attachment style and health outcomes (Dorfrnan-Botens, 1994; F eeney, 2000) point toward attachment’s merit to help understand chronic-pain patients. The long-term nature of chronic pain 20 means frequent interactions with health care providers. Understanding how such patients perceive these relationships is important. Patient-Provider Relationship Quality The importance of the quality of the patient-provider relationship is amplified in chronic-pain patients for several reasons (Kaplin, Greenfield, & Ware, 1989). First, the nature of chronic pain often means that rather than effecting a cure, the goal of medical intervention is to manage the symptoms. For effective management, a solid patient-provider relationship must exist in order to facilitate the communication of several types of information. It is not enough to communicate effectively the treatment regimen alone because much of the responsibility for implementation falls on the patient. Motivating the patient to adhere to treatment comes into play. In some cases this may be as simple as explaining how a procedure works or why it is necessary. Patients with fibromyalgia, for example, often are recommended to exercise as part of their treatment. The provider’s perspective on long-term pain management warrants regular exercise to decrease pain and prevent secondary deconditioning. A patient’s perspective, however, often mired in short-term sensations, focuses on feelings of general malaise and temporary discomfort while exercising, impeding adherence to this part of treatment. By conveying information that allows the patient to understand better the reason for exercise—and thus overcome their obj actions—the provider can improve the likelihood of the treatment program being followed. A high-quality patient-provider relationship allows for both the transmission and reception of important communication. Supporting self-care in tandem with 21 assurances of proper medical treatment is referred to as collaborative management (Von Kortf, Gruman, Schaefer, Curry, & Wagner, 1997). It is the establishment of a quality working relationship that permits collaboration so that optimal health care, such as goal setting, management of self-care, and follow-up, occur. The long-term nature of relationships with chronic-pain patients and their association with treatment outcomes highlight the importance of understanding patient-provider relationship quality in this population. Summary Turk and Okifuji (2002) stated that the emergence of consistent cluster patterns reproduced in numerous pain disorders (e.g., F ibromyalgia Syndrome, Temporomandibular Disease, and Chronic Low Back Pain) suggested the existence of psychosocial diagnoses separate from biomedical diagnoses. There is currently a limited understanding of variables that define or predict membership into these clusters. Awareness of how empirically derived clusters, used as the basis of psychosocial diagnoses in a population of chronic-pain patients (Turk & Rudy, 1988), relate to variables associated with treatment success and failure would address existing limitations in research. Research in this area is valuable for two reasons. First, previous research has found that patients grouped on the basis of similar psychosocial factors varied in their outcomes to identical treatment (Epker & Gatchel, 2000; Rudy, Turk, Kubinski, & Zaki, 1995). Research delineating differences among the subgroups would further current understanding of unique characteristics of each cluster and theoretically explain variations in treatment outcomes. Secondly, the 22 nature of each cluster—dysfunctional, interpersonally distressed, and adaptive coper—suggests differences in levels of communication style, and ability to establish working relationships. These factors are associated with attachment style as exemplified by the subscales—Trust, Communication and Alienation fi'om one measure of attachment (Armsden & Greenberg, 1987). Understanding the relationship between chronic pain patient subgroups and attachment may increase the understanding of core differences between types of pain patients. Attachment theory purports that qualities of subsequent relationships are impacted by attachments formed early in life. As applied to this study, attachment style should be predictive of perceptions of the patient-provider relationship quality. Perceptions of the patient- provider relationship quality and attachment style are both associated with symptom perceptions, health care utilization, and ultimately treatment outcome—areas crucially important to successful treatment of chronic-pain patients (Ciechanowski, Walker et al., 2002; McCarberg, B. & Wolfe, J ., 1999; Peters, Simon, Folen, Umphress, & Lagana, 2000). A better understanding of the relationship of each cluster to attachment and patient-provider relationship quality would ultimately contribute to customizing treatment to fit the psychosocial characteristics of individuals in the subgroups. Research Questions The primary purpose of this research is to establish whether relationships exist between attachment and patient-provider relationship quality, and each of three empirically derived chronic pain clusters. If significant relationships existed in both of the independent variables and the chronic pain clusters, then additional analyses 23 were planned to see if attachment and patient-provider relationship quality predicts membership into one of the chronic pain clusters. Keeping with the NIH’s Committee of Future Directions for Behavioral and Social Sciences Research (Singer & Ryff, 2001) recommendations and the aforementioned purposes of this study, the following research questions are proposed: Regarding attachment and patient-provider relationship quality, this study first asks: I. In a chronic pain population, is the prevalence and nature of attachment comparable in the three chronic pain subgroups? 2. In a chronic pain population, is the nature of patient-provider relationship quality comparable in the three chronic pain subgroups? And if significant differences are found: 1. Do the variables attachment style and patient-provider relationship quality improve the predicted membership in a chronic pain subgroup? Research Hypotheses Three subgroups of chronic-pain patients—dysfirnctional, interpersonally distressed, and adaptive coper—were created as the nominative criterion variable. Four types of attachment were identified: secure, preoccupied, dismissing, and fearfirl. The attachment variable is a nominative predictor. Preoccupied, dismissing, and fearfirl attachment types are referred to as insecure for the purposes of prediction, as the literature does not allow for confident predictions to be made about individual types of insecurely attached individuals and the criterion variable. Hypotheses 24 suggest that the absence of secure attachment will be related to the criterion variable as indicated below. Preoccupied individuals score low on avoidance and high on anxiety; dismissing individuals score low on anxiety and high on avoidance, and fearfirl individuals score high on both avoidance and anxiety (Brennan, Clark, & Shaver, 1998). Given these differences, each of the three insecure types will be examined separately if a significant relationship is found between insecure attachment and either the dysfunctional or interpersonally distressed subgroups. Quality of the patient-provider relationship is a continuous independent variable. Analysis of data will be guided by the following hypotheses. The research question asking, “In a chronic pain population, is the prevalence and nature of attachment comparable in the chronic pain subgroups?” is addressed with the following hypotheses. W: Chronic pain patients who are classified as dysfirnctional or interpersonally distressed are more likely to have an insecure attachment style such as preoccupied, dismissing or fearfirl. We]: Chronic-pain patients who are classified as adaptive copers are more likely to have a secure attachment style. The research question asking, “Within a chronic pain population, is the nature of patient-provider relationship quality comparable in each of the chronic pain subgroups?” is addressed with the following hypotheses. Hmthesis 3: Chronic pain patients who are classified in the dysfunctional or interpersonally distressed subgroups are more likely to report poor patient-provider relationship quality. 25 Hymthesis 4: Chronic-pain patients who are classified in the adaptive copers group are more likely to report better patient-provider relationship quality. If both attachment style and patient provider relationship were significant, the additional research question research question asking, “Are differences in attachment and patient provider relationship quality predictive of membership in a chronic pain subgroup?” would have been addressed with the following hypothesis. Hypothesis 5: Attachment style and patient provider relationship quality will improve the prediction of members who are classified in the chronic pain subgroups. Assumptions 1. It will be assumed that the participants responded to the self-report measures accurately. 2. Scoring of the instruments will be done accurately and in accordance with standardized procedures. Overview This chapter provided a purpose for this study, based upon the gaps and limitations in the understanding of characteristics defining chronic-pain patients. Brief background information was provided, along with data augmenting the need for research related to chronic pain. The integrative approach to chronic-pain treatment was discussed as a precursor to understanding the importance of biopsychosocial variables related to chronic pain. Empirically derived clusters of chronic-pain patients were discussed as a way to create homogeneous groups, based upon psychosocial 26 variables. An argument was advanced for the need to understand more about characteristics defining individuals within each cluster. Two variables, attachment and perception of patient-provider relationship quality, were presented as psychosocial variables with the potential to define and predict membership into a chronic pain cluster. Research questions, hypotheses, and assumptions followed. Chapter H examined literature applicable to this research study and the research questions specifically 27 CHAPTER TWO LITERATURE REVIEW Overview This chapter reviewed relevant literature and research related to this study. It first discussed shortcomings in the current treatment of chronic pain, followed by limitations in the understanding of unique psychosocial characteristics of homogeneous chronic pain clusters. Attachment and patient-provider relationship quality were discussed as two variables worthy of firrther study because of their potential to define and predict membership in a cluster, with relevant theories and research for each of the two variables. It is argued that a comprehensive understanding of unique cluster characteristics is a necessary precursor for tailoring treatment to individual chronic-pain patient needs. The Status of Chronic Pain Treatment and Recommendations for Improvement Although the biopsychosocial model of pain provides a basic understanding of chronic pain, detailed information necessary for comprehensive and effective treatment has proven elusive. Clinicians, and researchers alike, have commented on the inadequacies of chronic pain treatment. Long and Webb (1980) pointed to the ongoing parade of health care and allied health professionals sought out by chronic- pain patients as evidence of the shortcomings of their treatment. Bendelow and Williams (1996) concluded that the majority of chronic-pain patients are dissatisfied or fi'ustrated with the results of their treatment. In tandem with these feelings, patients 28 feel a pervasive sense that medicine, in spite of its technological advancements, has failed them. Chronic-pain programs (Jamison, 1996; Johansson, Dahl, Jannert, Melin, & Andersson, 1998; Seville, 1996) utilized diverse treatment modalities (Bendelow & Williams, 1996) to address the psychosocial aspects of chronic pain. The efficacy of chronic-pain treatment was reported as limited, regardless of the program evaluated (Turk & Rudy, 1988). A study conducted by the Mayo clinic examined responses from the Health Status Questionnaire (Wetzler & Radosevich, 1992), collected 13 years after patients attended Mayo’s Pain Management Center (Maruta, Malinchoc, Offord, & Colligan, 1998). The authors reported that 68% of patients felt they had poor physical health-related quality of life as measured by pain levels, physical health, and physical and social firnctioning. These findings are in agreement with several previous studies (Swanson, Floreen, & Swenson, 1976a; Swanson, Swenson, Maruta, & McPhee, 1976b). On the positive side, patients in the Mayo study reported better emotional and mental health than the normative group and higher levels of employment compared to pretreatment. These findings suggest that, although there are lasting contributions garnered through participation in pain-treatment programs, comprehensive improvement in perception of pain is not common for most patients. Pertinent to this study, the majority of patients continue to have pain and health- related quality of life concerns despite treatment. Research-design limitations are partially responsible for meager outcomes in chronic-pain research. One acknowledged confound is the assumption that the sampled population is homogeneous. Turk and Rudy (1988) suggested that the actual 29 heterogeneity of the population obscures the efficacy of the examined treatment program because homogeneous subgroup outcomes are not separately examined. Furthermore, research identifying unique characteristics of subgroups has been limited and, at times, misdirected. This situation has impeded the development and implementation of customized treatment because psychosocial characteristics, which could inform treatment methods, have not been delineated adequately. Examining chronic-pain patients as a homogeneous sample and failing to understand the subgroup characteristics perpetuates the status quo—the folly of which has been recognized by researchers. The call for tailored treatment has been repeatedly articulated. Examining older chronic-pain sufferers, Helme (2001) recommended the uniqueness of patients be ascertained. Lister (1996), addressing physicians, recommended comprehensive treatment be tailored to meet the needs of the individual. Evers, Kraaimaat, Van Riel and De Jong’s (2002) study on the effectiveness of cognitive-behavioral therapy in chronic-pain patients concludes that treatments centered on patient characteristics holds promise for improving outcomes. The discovery of empirically derived clusters of chronic-pain patients appears to be a step toward addressing recommendations for tailored assessment and treatment. These clusters are formed based on psychological and social characteristics of individuals with chronic pain, incorporating the contributions of diverse factors recognized in the International Association for the Study ofPain’s (IASP) definition of pain (Portenoy & Kanner, 1996). 30 Empirically Derived Clusters of Chronic-Pain Patients as Taxonomy Psychologists have attempted repeatedly to identify meaningful psychological and social characteristics of chronic-pain patients. Initially, researchers operationalized separate psychological and social variables in an effort to detect differences among patients. Advancing this line of research, Turk and Rudy (1988) were among the first to develop a taxonomy of chronic-pain patients, based on a comprehensive assessment that included both psychological and social factors. Using cluster analysis on Multidimensional Pain Inventory (MPI) data (e. g., pain-related interference, perceived life control, affective distress, social support, general activity, and responses by significant others), they discovered three distinctive clusters or subgroups. These groups were given labels based on identifying characteristics-— dysfirnctional, interpersonally distressed, and adaptive coper. Turk and Rudy’s seminal article conveyed two goals related to the development of this taxonomy. First, they hoped to provide a classification system that would facilitate the assessment of chronic-pain patients in a meaningful way. Next, they envisioned that this taxonomy would serve as the foundation fiom which effective treatments could be created. Three Subgroups of Chronic-Pain Patients The three subgroups created using cluster analysis on data from the MPI are referred to as dysfunctional, interpersonally distressed, and adaptive coper (Kems et al., 1985). Items fi'om the MPI are quoted as a way to understand characteristics of individuals who are categorized into each of the subgroups. 31 The dysfirnctional profile is comprised of individuals reporting high levels of pain severity combined with a perceived low level of control. Pain severity was measured by questions including, “Rate the level of your pain at the present moment.” The individuals also reported a high level of pain interference in their life, including decreases in their social, household, and outdoor activities, as well as a high level of psychological distress. Individuals in the interpersonally distressed cluster report problems with social support from their significant other, perceiving them to be less attentive and less helpful individuals than the other two subgroups, as indicated by low scores on questions such as, “How supportive or helpfirl is your significant other to you in relation to your pain?” They have higher scores on their perception of punishing responses from significant others, including displays of anger, frustration, or irritation. Interpersonally distressed individuals also reported that their significant other does not help them by using distractions to take their nrind off the pain, such as reading to them, involving them in activities, or talking. The third classification is labeled adaptive coper. Compared to the other two groups, individuals in this group appear to be firnctioning better overall. Individuals reported lower levels of pain severity and of pain interfering with their life activities. They also reported less affective distress. In response to the question, “How much has your pain changed the amount of satisfaction or enjoyment you get from participating in social and recreational activities?” individuals in this group tended to indicate that there has been little or no change. Compared to individuals in the other clusters, these individuals reported that pain impacts their life to a lesser degree. 32 Research Utilizing Turk and Rudy’s Taxonomy Several studies have been conducted identifying differences among the clusters. For example, clusters have been shown to differ based on the amount of analgesic medications used, with patients classified as dysfunctional reporting greater usage (Lousberg, Greonman, & Schmidt, 1996). This same study compared cluster membership with MMPI results, finding that interpersonally distressed subjects have an elevated neurotic triad and more passive-aggressive characteristics. Two behavioral dimensions of adjustment have been studied: pain-related anxiety and acceptance (McCracken et al., 1999). McCracken’s group found greater pain-related anxiety and less pain acceptance to be more characteristic of the dysfirnctional group than the other groups. Pain-related anxiety and acceptance were suggested to be key factors on which to focus in order to move an individual from a dysfunctional to adaptive classification. This study is important because it predicts the possibility of change through remediation of specific characteristics. Another study examined the relationship of numerous outcome variables to each of the three clusters in an 18-month, follow-up study (Bergstrom, Jensen, Bodin, Linton, & Nygren, 2001 ). The dysfirnctional group was found to have a higher rate of work absence and to utilize health—care resources more fi'equently than adaptive copers. Both dysfirnctional and interpersonally distressed groups indicated poorer health than adaptive copers. Interestingly, this study also found insignificant differences in improvement among the groups over the 18 months. Other studies have, however, shown differences in improvements. For example, examining psychological variables in temporomandibular joint dysfunction patients, more 33 interpersonally distressed and dysfunctional patients reported positive changes than adaptive copers (Rudy, Turk, Kubinski, & Zaki, 1995). Past research has demonstrated unique characteristics among the chronic-pain subgroups, and it has also demonstrated that the clusters can be created under a variety of circumstances. Clusters similar to those created by Turk and Rudy (1988) have been created using different instruments. In a creative study Bergstroem, Bodin, Jensen, Linton, and Nygren (2001) collected data using several instruments, including the Swedish version of the MPI (MPI-S), to generate clusters. When Bergstroem’s group examined data from the noncluster generating instruments, they found a patient profile comprised of lower pain severity levels, less pain interference in daily activities, higher levels of control and lower levels of affective distress when compared to the other two profiles. The characteristics of this group paralleled those of the MPI-S created adaptive coper group. Similarities were also found between profiles created using the MPI-S and noncluster creating data for dysfirnctional and interpersonally distressed groups. In another study Jamison, Rock, and Parris (1988) created a three-cluster solution using the Symptom Checklist-90 (SCL-90) on a population of heterogeneous pain patients. Subsequent analysis found no difference in demographics or physical pathology among the groups. Recreating clusters using other instruments suggested the clusters are most likely a reflection of real world phenomena, rather than an idiosyncrasy of the MPI. Turk and Rudy (1990) examined the robustness of the clusters by attempting to recreate the taxonomy using a variety of physical conditions, including low-back pain, headache, and temporomandibular disorders. Again, three 34 clusters emerged regardless of the malady exarrrined, providing evidence that this taxonomy can be generalized. Variances in the percentage of patients belonging to each cluster, based on the physical condition, were observed. Some groups, such as the low-back pain group, had a higher percentage of individuals categorized as dysfunctional because of higher levels of disability and functional compromise associated with this condition. This was interpreted as support for this taxonomy’s discriminant validity. Although this threecluster solution has been recreated using different instruments and with different populations, there are limitations. In a study using the Brief Symptom Inventory (Derogatis & Spencer, 1982), a modified version of the SCL-90-R, only two distinct clusters were found (Geisser, Perna, Kirsch, & Bachman, 1998). Need for Additional Research on Variables Defining or Predicting Membership The three-cluster profile for chronic-pain patients showed promise as a template for tailoring treatment. Despite a growing body of research in this area, very little is actually known about characteristics that define or predict membership in a group (McCracken et al., 1999). Clarification of factors identifying characteristics associated with, or predictive of, patient subgroup membership is important for more accurate assessment and diagnosis, and improving treatment outcomes. Two variables, attachment-style and patient-provider relationship quality, may clarify important differences among the subgroups. 35 Attachment Attachment theory is proposed as a potential conceptual framework from which to understand an individual’s perceptions and behaviors associated with health. Additionally, attachment theory aids in understanding varying individual patterns of health-care utilization (Ciechanowski, Katon, Russo, Dwight-Johnson, 2002). Prior to delving into the research addressing attachment style as it is applied to medical studies and patient-provider relationships, it is important to understand the theoretical and empirical support for attachment theory. Bowlby and Ainsworth (Bowlby, 1969/1982, 1973), examining care provider- infant dyads, suggested that different attachment types developed in infancy based on the nature of the dyad’s interactions. These interactions (or lack thereof) are the basis for the creation of internalized working models of self and other. The model of self is dichotomous. Positive reflections of the self manifest as acceptance of intimacy, a sense of autonomy, and feelings of worth. Negative self-reflections manifest as uncertainties regarding intimacy and independence, and doubts about self worth. The model of other is also dichotomous, with others viewed as trustworthy and available or not (Bartholomew & Horowitz, 1991). The concepts of selfand other influence the nature of subsequent close relationships and are associated with various mental and physical health conditions. A satisfying, consistent, and continuous relationship with a care provider facilitates secure attachment. Individuals with a secure attachment type reap benefits lasting throughout their life, such as a sense of well-being and resiliency (Sroufe, 1983; Weinfield, Srouge, Egeland, & Carlson, 1999). Subsequent attachments are 36 initiated and maintained with positive expectations regarding the self and others. These expectations translate into viewing one’s self as accepting and worthy of comfort while simultaneously viewing the other individual as a potential source of comfort and assistance. Contrasting this, an individual who is insecurely attached enters relationships with one of three negative permutations of the self or other. Bartholomew and Horowitz (1991) separated insecure attachment types into three subtypes, referred to as preoccupied, fearful, and dismissing. These subtypes are based on the internalized models of self (positive/negative) and other (positive/negative). A positive model of other with a negative model of self is referred to as preoccupied; a negative model of other with a negative model of self, fearfirl; and a negative model of other with a positive model of self is termed dismissing. An internalized negative model of self incorporates social discomfort, fears of intimacy, and feelings of unworthiness; while the internalized negative model of other includes concerns about trustworthiness and rejection. Individuals with insecure attachment styles have difficulties initiating and maintaining subsequent relationships because of their trepidation. Bowlby (1973, 1980) suggests that an insecure attachment type predisposes or increases one’s vulnerability to psychopathology. Research supports this assertion. Insecure attachments are associated with numerous Axis I and II diagnoses, such as anxiety and substance abuse disorders (Fonagy, Leigh, Steele, Steele, Kennedy, Mattoon, Target, & Gerber, 1996), eating disorders (Cole-Detke & Kobak, 1996), borderline personality disorder (Fonagy et al., 1996; Patrick et al., 1994), and antisocial personality disorder (Rosenstein & Horowitz, 1996; Fonagy et al., 1996). 37 Examining the relationship between attachment and mental health, West, Rose, Verhoef, Spreng, and Bobey (1998) found women who were classified as anxiously attached were more likely to report symptoms of depression. Two fundamental tenets of Bowlby’ 3 work have been incorporated into contemporary attachment theory: (a) the relative stability of attachment style over the lifespan, demonstrated through the recapitulation of initial attachment style in subsequent adult-pair relationships, and (b) the existence of associations between attachment and well-being and mental health. Contemporary attachment theory extends Bowlby’s work, and supports the use of attachment as a framework for addressing process and outcome concerns in populations of medical patients. It draws parallels between the concepts put forth by Bowlby and Ainsworth in infant-care provider dyads and patient-care provider relationships. Explaining the differences between attachment relationships and other social relationships, Ainsworth (1985) put forth four characteristics unique to attachment relationships: (a) secure base behavior, (b) proximity seeking, (c) separation protest, and (d) safe-haven behavior. All four characteristics can be demonstrated in the patient-provider relationship. Secure base behavior refers to viewing the care-provider as a unique individual offering comfort and security not found in others. Having a secure base permits exploration because there is a place of comfort available should a situation become uncomfortable. In contrast to the behavior of the child, in which exploration refers to examining phenomena within a certain range of the attachment figure, secure 38 base behavior in a patient includes exploration of intellectual and emotional facets that typically would not be discussed in a more casual relationship. Proximity seeking behavior is demonstrated when a patient becomes concerned about his/her symptoms or discomfort, and makes calls and appointments in an effort to seek comfort and support. The presence of one’s provider and his or her assurances and explanations serves to decrease the patient’s anxiety. Separation protest is seen when separation occurs involuntarily. When a patient cannot see his/her provider within an acceptable timeframe, or when insurance does not permit the patient to see a desired provider, expressions of frustration may occur. The fourth characteristic, safe-haven behavior, is frequently observed in crisis or emergency situations. Safe-haven behavior refers to seeking out the provider when a health concern or condition is perceived as threatening to one’s well-being or life. Key themes in patient-provider attachment are accessibility to the provider and feelings of comfort and reassurance when the provider is attentive (F eeney & Noller, 1996). Research on child, adolescent, adult, and patient-provider relationships supports the use of contemporary attachment theory as an aid in understanding medical patients. Empirical Evidence Supporting Attachment Theory The following research studies are presented to demonstrate the continuing effects of attachment over an individual’s life. This array of studies demonstrates the credibility of generalizing attachment theory across different populations. Predictability over a five-year period, based on the attachment styles of infants, was shown by Strage and Main (1985). Their study involved observing and 39 categorizing infant attachment style based upon nonverbal behavior, and doing the same five years later using verbal interactions. This study is remarkable because predictability occurred despite the change in communication modalities used to assess attachment. Augmenting the findings of Strage and Main; Main, Kaplan, and Cassidy (1985) reported the stability of attachment style over the first six years of life. Studying the effects of early attachment in later life stages, researchers have investigated both adolescent and adult populations. Goldberg (2000) stated, “Transference is the aspect of an interpersonal relationship in which a person’s behavior toward another is influenced by relationships to important figures from childhood and infancy, usually parents and caretakers” (p. 1165). Examining adolescent attachment, Allen and Land (1999), suggested that attachments are not relinquished fiom parents during this period, but rather transferred to peers. During this developmental period, peers have substantial influence on each other. In an effort to separate from parents, adolescents depend inordinately on peers and fitting-in as talismans. Concurrent with the change from parents to peers as attachment objects is the change from hierarchical to bilateral attachments. In the former, the individual is the recipient of care, while in the latter a more evenly divided ratio of care receiving and giving is experienced. Studies on adolescent attachment are important for understanding the use of attachment theory as it applies to medical patients for three reasons. First, it supports contemporary attachment theory, which predicts correlation between attachment type and later attachments. Researchers studying adolescent attachment view parental attachment as the driving force for the nature of later peer attachments (Hartup, 1992; 40 Weiss, 1982; Tesclr, 1983). Supporting and extending this supposition, a 20-year longitudinal study initiated when the participants were 12 months old found 72% received the same attachment classification as young adults (Waters, Merrick, Treboux, Crowell, & Albersheim, 2000). Second, adolescent research confirms attachment theory’s prediction of an association between attachment and subsequent self—concept. For example, adolescent research on attachment found the quality of parental relationships more highly related to self-esteem than those of peers (Greenberg, Siegal; Leitch, 1984; O’Donnell, 1976). Finally, adolescent research suggested that attachment relationships develop under variable conditions of perceived power hierarchies. This finding supports the application of attachment theory in situations in which the power hierarchy differs from that found in the infant- parent dyad. In patient-provider relationships a range of relationship styles exist from paternalistic or authoritarian (similar to the infant-parent hierarchy) to patient- centered or collaborative (similar to the peer hierarchy). Research supporting the utility of attachment theory in different types of relationships lends credence to its use in patient-provider relationships. Hazen and Shaver (1987; Shaver & Hazerr, 1988) have extended the application of attachment theory to adult relationships including adult intimate and family relationships. Their research supports Ainsworth (1979) and Weiss’s (1991) contention that the nature of some adult relationships meets the criteria for attachment. Research suggests that committed relationships, such as those found in marriage and some nonmarried relationships (e.g., cohabiting, homosexual) meet the 41 criteria for attachment. Attachment also exists in professional relationships, such as the professor-student (Lopez, 1997) and the therapist-client (Pistole, 1994, 1999). Despite the relatively stable attachment classification predicted by attachment theory, the correlation is not perfect. Bowlby’ s theory suggests that attachment classification is subject to environmental impact. Waters, et al. (2000) stated that negative life events, such as parental divorce or loss, physical assault or molestation by a member of the family, and life-threatening illness of either the parent or child, can be instrumental in changing one’s attachment classification. The authors reported 44% of individuals experiencing a negative life event changed classification between infancy and early adulthood. In comparison, only 22% of individuals whose mother reported an absence of negative life events changed attachment classification. Research demonstrates that while some events impact change in attachment classification, others, such as lifestyle, do not (Hamilton, 2000). Hamilton’s study reported the stability of attachment as 77%. Attachment as a variable is attractive both because of its relative consistency over time and its potential for change. Its consistency over the life span contributes to its predictive value. On the other hand, attachment type is not unyielding; attachment theory contends that the detrimental impact of early unhealthy relationships can be modified through experiencing healthy relationships. Supporting this tenet, studies have suggested that interventions may serve to change individuals with an insecure attachment history (Egeland, Jacobvitz, & Sroufe, 1988; M Main, Kaplan, & Cassidy, 1985). Individuals who experience this transformation have been labeled as “earned secure” adults (Pearson, Cohen, Cowan, & Cowan, 1994). This line of 42 thought applied to chronic-pain patients suggested the possibility of successful intervention in chronic-pain patients who are insecurely attached. Bowlby (1969/1982, 1973, 1980) argued that the purpose of attachment is to ensure species survival. The nature of illness, and its potential consequence, death— the antithesis of survival—may serve as a stimulus activating attachment in the patient-provider relationship. Given attachment theory’s heuristic value and its applicability in other professional relationships, researchers have utilized attachment theory to study the patient-provider relationship (Ciechanowski et al., 2001). Research on Attachment Related to Medical Patients The last twenty years has witnessed a proliferation of research about the incidence, evaluation, process, and outcome of biopsychosocial influences on medical symptoms (Katon, Sullivan, & Walker, 2001), including attachment, which has been examined in the context of various patient populations and health outcomes. It has been argued that a better understanding of attachment may contribute to a better understanding of many of the other psychosocial variables because they are interpersonal or relational in nature (Basler, Grzesiak, & Dworkin, 2002). The following studies reflect some of the research on attachment as it relates to health. A study examining childhood factors found that childhood psychological trauma including physical, sexual, and emotional abuse perpetrated by a primary caregiver was related to postsurgical outcome. In this study, 95% of patients who had successful spinal surgery reported no childhood abuse by a primary caregiver, while only 15% of patients who reported three or more of these types of trauma had 43 successful spinal surgery (Schofferman, Anderson, Hines, Smith, & White, 1992). In a related study, Schofferman’s group (1993) found a relationship between childhood psychological trauma and chronic pain. A study on attachment and health outcomes found that avoidant attachment style may increase distress because of the tendency to avoid the support of others during times of need. The author suggested use of this coping pattern may increase one’s propensity toward illness (Kotler et al., 1994). Studying the relationship between attachment and symptom reporting, Feeney and Ryan (1994) found that ambivalent attachment was the highest predictor of illness symptom reporting, while avoidant attachment was inversely related to visits to health professionals. Interestingly, no association was found for ambivalent attachment. Feeney and Ryan, pointing to the differences between avoidant and ambivalent attachment, suggested that grouping the two into a general category (i.e., insecure attachment) would overlook important health implications. Another study found that attachment style affected the impact of chronic low- back pain on mental health, with individuals in the avoidant and anxious-ambivalent categories reporting higher levels of affective distress than those in the securely attached category. Furthermore, this study demonstrated that attachment style was related to both the assessment of pain and the type of strategy employed to cope with it. Individuals who were labeled securely attached viewed their pain in a less ominous manner and relied on more active coping strategies than insecurely labeled individuals (Mikulincer & Florian, 1995). 44 Physiological models and psychological explanations exist to describe associations between attachment and health. Two physiological explanations for the impact of attachment type on health are discussed, followed by psychological explanations. Physiological Explanations of Attachment Attachment has been examined in both animal and human studies. The early work of researchers utilizing a physiological perspective affirmed a link between attachment and health. A series of studies on rats provided a physiological explanation for the long-term health effects of early attachment bonds (Adler & Grota, 1969; Hoffer, 1994; Levine, 1957, 1962; Levine, Haltmeyer, Kaas, & Penenberg, 1967). In an experimental design study, rat pups were removed from their mother for a short period of time (i.e., a few minutes). This caused an increase in vocalization in the pups, which caused the mothers to lavish more attention on the pups once returned than if they had not been temporarily removed. Over the rat pups’ lives, repeated physiological measures of their response to stress were evaluated. One important finding was that the amount of glucocorticoid released during stressful conditions was less than in rat pups that did not receive the additional attention.1 The results of this study indicated that long-term benefits exist as a result of increased maternal comfort. ' The hormone glucocorticoid is related to stress levels. In situations ofacute stress its benefits include improving reaction times. However, in situations of chronic stress, prolonged exposure to glucocorticoid 1m been associated with temporary and permanent memory impairment (McEwen & Sapolsky. 1995; Sapolsky. 1996). 45 In a modified condition, rat pups were again temporarily removed; however, the duration was increased to 3 to 6 hours per day. In this scenario, the rat mothers ignored the pups upon their return. These pups showed increases in their glucocorticoid levels during period of stress and, as in the first scenario, the impact was lifelong. This series of studies supports the lifelong impact of early maternal care and ofi‘ers a biological explanation for differences in behavior and health related to attachment. A physiological explanation of differences based on attachment type in humans was proposed by Nachmias, Gunner, Mangeldorf, Panitz, and Buss (1996). Their research suggested mothers of insecurely attached children activate the hypothalamic-pituitary-adrenocortical (HPA) system via intrusive maternal behaviors. Activation of the HPA system resulted in an increase of cortisol levels when the child perceived a novel situation as potentially threatening. Increased levels of cortisol were also present when the children viewed themselves as possessing inadequate cOping resources to address novel situations. High cortisol levels were associated with emotional distress, such as anxiety, fear, feelings of loss, and a decreased sense of internal locus of control, and with behavioral inhibition in insecurely attached children. In contrast, securely attached children did not experience increased cortisol levels in the presence of their mothers and experienced emotional syntony. Kennedy, Kiecolt-Glaser, and Glaser’s study (1988) argued that attachment has a physiological impact on individuals. Their study substantiated attachment’s role in mediating the relationship between stressors and immune functioning. This study 46 and others suggested that attachments developed through social interactions influence one’s biology. These studies supported the existence of one or more physiological mechanisms facilitating attachment’s influence on health. Our current understanding explains secure attachment’s provision of protection from anxiety in stressful situations and reduction of vulnerability to stress related diseases such as asthma, dermatitis, and irritable bowel syndrome. Conversely, insecure attachment through the same mechanism increases vulnerability to stress related diseases and pain. Details of the physiological mechanisms underlying the associations between attachment and health are not firlly understood. The fields of biomedicine and psychology are researching the associations between attachment and health in an efl‘ort to advance current understanding. Having briefly examined physiological explanations for the relationship between attachment and health, psychological explanations follow. Psychological Explanations of Attachment Stuart and Noyes’ (1999) study of somatization cleverly illustrated the connection between attachment and physical complaints and elucidated the importance of understanding a patient’s attachment typez. The authors surmised that having attachment needs addressed via health care provision in childhood predisposes the individual to the development of later chronic illness or pain as a way of expressing their emotional needs. Unfortunately, this ineffective communication 2 Somafimtion’shallmarksarephysieal symptomsinexoessofwhatisexpectedfmmthegiven medical condition, and co-morbid psychological and functioml distress. A partial list of diagnoses under this broad category includes somatimtion disorder, conversion disorder, hypochondriasis, and 47 method increases already heightened levels of distress in two primary ways. First, the core issue of having one’s emotional needs met is circumvented because of the overt fours on medical treatment. Second, regardless of the quality or quantity of care provided, insecurely attached individuals typically interpret care as insufficient because of their persistent view of others as being unable to meet their needs (Bartholomew, 1997). Chronic pain may exacerbate the patient’s fear of separation, which catalyzes their attempts to alleviate it by seeking support from medical providers. This pattern is inadvertently perpetuated as the need for an emotional connection remains unmet, resulting in increased levels of distress that are subsequently expressed through further somatic communication (Bretherton, 1992). Behavioral models have been proposed to explain the relationship between attachment and health. West, Livesley, Reiffer, and Sheldon (1986) suggested that stressfirl environmental events impact individuals differently based on their attachment type. Attachment type affects both one’s desire to seek social support, and one’s repertoire of coping behaviors. Securely attached individuals tend to seek social support from others and use effective coping mechanisms. Conversely, insecurely attached individuals are unable to appropriately utilize social support, acting excessively dependent on or avoidant of social support. Additionally, their coping styles tend to be less effective partly because they expect failed social support and they perceive self-inadequacy. Research examining the childhood antecedents of hypochondriasis found that separation anxiety in childhood was correlated with levels of hypochondriacal symptoms in adulthood (Noyes Jr. et al., 2002). These findings support the 48 interpersonal model of hypochondriasis, which suggested that adversity in early childhood, which results in insecure attachment, increases one’s vulnerability to psychopathology and may lead to care-seeking behavior later in life (Rutter, 1997). The unverifiable complaints of hypochondriacs, similar to the complaints of chronic- pain patients, may result in the health care provider withholding treatment. As in other models, the unmet needs of the patient may result in escalation of symptoms or doctor-shopping as the patient seeks to ameliorate their discomfort. The Relationship of Attachment to Patient-Provider Relationship Quality Describing the relationship between attachment and pain complaints, Kolb (1982) stated that not only is attachment a factor, but understanding it greatly improves pain management. According to Kolb, patient complaints regarding pain can be viewed as approach behaviors, or methods of entering the patient-provider relationship. Provider responses deemed appropriate by the patient result in a reduction of subjective pain and increased functioning. However, provider responses deemed inappropriate (e.g., requesting an increase in daily activities despite reports of pain) result in either increased dependency or withdrawal. These behaviors are explained by attachment theory, and relate to the patient’s perception of a lack of understanding or support. Within the patient-provider relationship, the overt communication between patient and provider is about pain. Covertly, however, much of the communication is about the relationship. Ifthe provider’s response does not meet the needs of the patient—on both levels—the patient may seek new caregivers to attempt to get his/her needs meet. Statements such 49 as, “My doctor doesn’t understand me,” and “I have been to several doctors, but none of them understand how much I suffer” are indicative of patients’ believing their medical needs have gone unmet. In a slightly different scenario, pain patients may become threatened when they believe treatment is being withheld, such as when pain medication is reduced, invasive procedures are denied, and referrals to different providers are made. Kolb suggested that, despite verbiage stating dissatisfaction in medical treatment, manifested patient anger originates fiom perceived unmet relationship needs. The patient’s reaction in these cases is often defamation of the medical staff or threats of legal action. These reactions arise unconsciously because of separation anxiety, but are not recognized as such. From the patient’s perspective, their responses are justified, given their perception of the medical treatment they have received. Rarely is either the patient or the provider aware of the covert reasons for the behaviors that occur within the relationship. In a third scenario, Kolb suggested that pain patients present in a manner that increases the probability of care-taking behaviors by their providers. Initially, the chance to respond in a helpful manner is welcomed by the provider. Over time, however, constant requests for help may backfire, as the provider perceives that treatment is inefl‘ective. In these situations, the patient may be maligned with labels such as malingerer, drug seeker, or hypochondriac. Ultimately, many of these patients are referred elsewhere. This rebuff is perceived as confirmation of unworthiness in individuals whose internal model of self incorporates such feelings. This, in turn, as suggested by Stuart and Noyes (1999) above, reinforces the individual’s negative 50 internalized model of self that, in turn, increases emotional distress and attempts to communicate via pain symptoms. Kolb (1982) argued that interpreting the actions of pain patients using attachment theory would increase our understanding of chronic pain and potentially improve patient-provider relationships and treatment interventions. Figures provided for successful pain management treatment underscore the need for improvement with outcomes for complete treatment failure approximately equal to those for complete success—20.3% vs. 19.8%, respectively (Maruta, Swanson, & Swenson, 1979). Descriptions of the groups within Turk and Rudy’s (1988) psychosocial pain taxonomy are suggestive of both different attachment styles and different health care provider relationship quality. The most obvious examme is the interpersonally distressed group—whose very label asserts relationship problems. Individuals in this group report problems with social support fi'om at least one significant other, such as a spouse or family member. Attachment theory suggests that these individuals are more likely to have an insecure attachment style. Furthermore, because of their insecure attachment, they are also more likely to experience problems with their health care providers. The individuals in the subgroup labeled adaptive coper are characterized by their perception of lower levels of pain, higher levels of activity, and minimal interference by pain in their daily activities. Adaptive copers appear to be better adjusted overall in comparison to the other two groups. Empirical examination of securely attached individuals repeatedly shows similar associations between secure attachment and measures of psychological resilience (F eeney & Noller, 1990) and 51 physical well-being (Feeney & Ryan, 1994). For example, intimacy, passion, and commitment are associated with secure attachment (Levy & Davis, 1988) and secure attachment is associated with higher levels of marital adjustment (Kobak & Hazen, 1991). As predicted by attachment theory, securely attached individuals employ constructive techniques of handling negative feelings, such as distress and hostility in social situations (Kobak & Sceery, 1988). Similar to individuals who are securely attached, adaptive copers report lower levels of psychological distress as measured by lower scores on the Affective Distress Scale (Turk & Rudy, 1988). The similarities in these two populations suggest that individuals who fit the adaptive profile are likely to be securely attached. Increased levels of psychological distress and pain characterize individuals in the dysfunctional group. Similarities are seen between chronic-pain patients who are labeled dysfirnctional (Turk & Rudy, 1988) and individuals who are insecurely attached, with the latter experiencing increased levels of psychopathology and medically unexplained symptoms (Garralda, 1992; Hotopf, Mayou, & Wadsworth, 1999). Despite theory and existing research on attachment and other medical conditions, no other research was found examining potential relationships between attachment types and chronic pain subgroups by this author. Attachment’s value in understanding chronic-pain patients is twofold. First, attachment theory has proven usefirl as a predictor of both mental and physical health outcomes. Second, attachment theory offers a model for understanding patient- provider interactions. Having discussed the importance of attachment, the merits of understanding patient-provider relationship quality are presented next. 52 Patient-Provider Relationships Medical care today is provided by various caregiver roles—medical doctors, physician assistants, nurse practitioners, etc. In a review of the literature on factors influencing the relationship between a patient and provider, Sihvonen (1989) suggested that central elements impacting relationships are the same, regardless of the type of provider caring for the patient. The patient-provider relationship is frequently studied in health psychology. Numerous studies support the idea that the patient-provider relationship is important to improving compliance (Buller & Buller, 1987; Heszen—Klemens & Lapinska, 1984; Willson & McNamara, 1982). Noncompliance issues include failure to follow dietary restrictions, perform routine testing, and take medication as directed (Hausman, 2001). For example, the type of relationship a patient has with his/her health-care provider was associated with his/her intention take medications as directed (Goldring, Taylor, Kemeny, & Anton, 2002). Patient-noncompliance is estimated to be a contributing factor in 22% of cases seen in the emergence department and a significant contributor to health care costs (Murphy & Coster, 1997; Olshaker et al., 1999). Other research suggesting that it occurs in a high percent of the patient population, with rates up to 80% (Golden & Johnston, 1970; Grandinetti, 1998; Health, 1997; Stimson, 1974). Other studies have linked the patient-provider relationship to treatment and health outcomes, including one reporting that the absence of a quality patient- provider relationship was a barrier to obtaining therapeutic goals (Becker & Maiman, 1980). For example, a recent study concluded that the continuity of a patient-provider 53 relationship increases the number of preventive care visits (Ettner, 1999). Another demonstrated a relationship between satisfaction with the provider and utilization of health-care services (Zastowny, Roghrmnn, & Cafferata, 1989). Not only is the patient-provider relationship related to frequency of visits, it is also related to patient satisfaction, which is related to communications patterns within the patient-provider dyad (Sullivan & Beeman, 1982). Patients evaluate their medical care more on interpersonal and informational skills of their provider than his or her treatment abilities (Deale & Wessely, 2001). Patients reporting that they liked their provider were more likely to report better physical and mental health after a visit (I. A Hall, Horgan, Stein, & Roter, 2002). Conversely, poor patient-provider relationships are associated with poor health-care provision and patient dissatisfaction (Schwenk, Marquez, Lefever, & Cohen, 1989). Furthermore, when patients were regarded by the health-care provider as difficult, they were more likely to be viewed as having psychosomatic symptoms (Hahn, Thompson, Wills, Stern, & Budner, 1994). These studies demonstrate an array of important outcomes with which patient-provider relationship quality is linked. Qualities of the patient-provider relationship were also associated with specific treatment improvements. For example, interpersonal relationship ratings of one’s health care provider were associated with quality of care for depression (Meredith, Orlando, Humphrey, Camp, & Sherboume, 2001), with an open dialogue style reported as improving the treatment of depression (Anonymous, 2000). Kaplan, Greenfield, and Ware (1989) measured health physiologically, behaviorally, and subjectively to examine the effects of the patient-provider relationship on the 54 outcomes of chronic disease. They concluded that perceptions of the patient-provider relationship were an important influence on health-care outcomes, regardless of how health was measured. Understanding of the importance of the patient-provider relationship has grown as health care moves toward placing a greater emphasis on treating the patient as a whole (Bell et al., 2002). This importance is duly noted in chronic-pain patients. Chronic-Pain Patient-Provider Relationships The importance of the quality of the patient-provider relationship is amplified in chronic-pain patients (Kaplin et al., 1989), with a necessary precursor to effective pain management being a sound patient-provider relationship. When working with chronic-pain patients, it is not enough simply to communicate the treatment regimen because much of the responsibility for implementation falls on the patient. This fact necessitates using motivation in tandem with education to increase treatment adherence. In some cases this means explaining both what is expected and how or why it is necessary. The treatment recommendation of exercise for fibromyalgia patients is illustrative. Providers tend to focus on reducing long-term pain, often prescribing regular exercise to decrease overall pain and prevent secondary deconditioning. The provider assumes that the implied benefits are understood by the patient, and that they provide necessary motivation for the patient to comply—to exercise. A consideration frequently not communicated or understood by the patient is the potential for an increase in discomfort while exercising. Feelings of general malaise and discomfort while exercising impede adherence. This common communication breakdown occurs 55 because trust and communication—two facets of a quality patient-provider relationship—have not been developed. Quality patient-provider relationships are paramount because they foster the transmission and reception of important information, which optimizes participation in self-care. Supporting self-care along with assurances that proper medical interventions will occur is referred to as collaborative management. Establishing a quality relationship will facilitate collaborative management and improve patient adherence and outcome (Von Korff et al., 1997). Measurement of the Provider-Patient Relationship Construct Despite a burgeoning interest in patient-provider relationships, a single agreed-upon definition of the patient-provider relationship has proven elusive. Chaitckil, Kreitler, Skaked, Schwartz, and Rosen (1992) state that the patient- provider relationship is complicated to study—whi ch impedes consensus—because of aspects, such as its hierarchical nature, emotionally laden focus, and requirement of cooperation prior to treatment. Difficulties in understanding the patient-provider relationship can be grouped as follows: (a) the unique multidimensional nature of the construct, (b) its dyadic nature, and (c) political forces. This list, while not exhaustive, attempts to organize concerns related to the patient-provider relationship found in the literature. The multidimensional nature of relationship quality is partly responsible for assessment difficulties. Rather than a direct measure, patient-provider relationship quality is a derived construct. This means it is a composite of several other measures 56 that conceptually capture its nature. The next logical question asks, “What elements best measure patient-provider relationship quality?” A lack of consensus regarding an answer is evident in the number of different conceptual constructs used to measure the patient-provider relationship, e. g., (Galassi, Schanber, & Ware, 1992; Ware, Snyder, & Wright, 1976; Wolf, Putnam, James, & Stiles, 1978). Patient satisfaction is one of the most fiequently used measures of the patient-provider relationship (Kaplin et al., 1989) along with patient-provider communication (Kraytman, 1979). Trust, too, has been used to assess the nature of the patient-provider relationship (M. A. Hall, Dugarr, Zhang, & Mishra, 2001; Mechanic & Meyer, 2000). Examination of measures of the patient-provider relationship suggested that in addition to the use of different subscales, the relative weight given to a subscale also varies (Galassi et al., 1992; Ross, Steward, & Sinacore, 1995; Stump, Dexter, Tierney, & Wolinsky, 1995; Wolf et al., 197 8). Another factor confounding this construct is its bilateral or dyadic nature (Jaffe, 1958). Research suggested that each individual in the relationship has difl‘erent criteria for measuring its qualities (DiMatteo, 1979; DiMatteo & DiNicola, 1978; Freeman, Goldman, & Cecil, 1980). One focal difference between patients and doctors is that patients perceive psychosocial factors as more meaningful, while physicians perceive technical and diagnostic factors as being more important (Davis, 1968; Korrnan, Stubblefield, & Martin, 1964). Engel (1988) and others (Mollernan, Krabbendarrr, & Annyas, 1984) stated that, in addition to each of the parties having difi‘erent values, each party has more than one need that should be measured, further complicating the measurement of patient-provider relationship quality. For example, 57 the relational needs of the patient have been subdivided into two types: a need for information related to the nature of the illness and a need to feel understood. Further underscoring the different perspectives of patients and physicians, articles discuss the necessity for physicians to respond to patients’ relationship needs (Robins & Wolf, 1988) by increasing bi-directional information exchanges (Hausman, 2001) and improving communication as a method for decreasing malpractice risk (Levinson, Roter, Mullooly, Dull, & Frankel, 1997). A review of the literature clearly demonstrated differences in perspectives between patients and providers impede consensus regarding how to measure patient-provider relationship quality. A third aspect influencing what factors encapsulate the construct of patient- provider relationship quality is political in nature. Cline (2003) referred to this political aspect as the influence of macro-level systems (e. g., technology, health-care organizations, media, and social influences) on micro-level communication patterns (e. g., verbal messages, dyadic interactions). The influence of managed care over the last 50 years on the role of psychiatrists and psychologists provides an example. Discussing the role of the provider in psychological treatment in the 19503, Rogers (1951) stated that the provider’s role was to establish rapport and provide a secure environment so that the patient could undertake the journey necessary for insight and change. Today, the provider’s role has evolved in response to competition, psychopharmacology, and managed care, such that it is focused on fact-gathering first and the patient’s comfort second. Whereas the psyclriatrist’s role used to be more psychoanalytic in nature, today it is more focused on medication management. 58 Psychiatric visits today typically include an initial session lasting between 50 and 90 minutes in which historical and physical information is gathered followed by additional sessions focused mainly on checking medications, which take about 15 minutes. Cruz and Pincus (2002) stated that few medical specialties have changed as much as psychiatry with respect to clinical roles over the last 30 years. Attempts to establish rapport and trust today are balanced with or secondary to data collection (Regestine, 2000). However, it is not possible to address the patient’s emotional complexities without listening to their concerns. This transformation changes the nature of patient-provider interactions and therefore the qualities that are important for assessing relationship quality. Conceptual difl‘erences in the construct of patient-provider relationship quality necessitate close attention to instrument selection. Three elements of patient-provider relationship quality important to understanding chronic-pain patients—quality of medical information, affective aspects, and the patient’s ability to communicate with their health care provider—are discussed next. 59 Perceived Quality of Information Understanding the chronic pain patient’s perception of the quality of his/her provider’s information is important because he/she tend to desire more medical information and use it differently than patients with acute afflictions (Cassileth, Zupkis, & Sutton-Smith, 1980). For individuals suffering with chronic pain, information fortifies their sense of control, an important aspect to chronic-pain management. For example, an individual with chronic pain may choose to attend their child’s wedding, despite receiving information from their caregiver that attending will exacerbate the pain they experience the following day. Increased awareness allows the individual to adjust his/her schedule to manage their pain. In this case, the individual may schedule activities, such as letter writing or watching videos, the day after the wedding as a way to manage their pain. Awareness of the consequences and the ability to adjust to them increases the person’s sense of control and decreases their perception of pain. For individuals with chronic pain, managing their pain is tantamount to managing their life. Another function of information for chronic-pain patients is validation. Chronic pain is a disability in which individuals often feel isolated and misunderstood. An exchange of information with a caregiver can assuage these feelings by decreasing self-doubt and guilt and increasing one’s sense of control. Many individuals believe modern medicine, with its many technological advances, should be able to cure or remedy all maladies. Examples of medical success are foremost in the minds of most individuals facing personal strife, in spite of conditions such as cancer, AIDS, or birth defects that are not fully remediable. Given this, the goal to minimize discomfort rather than eliminate it becomes unacceptable to many who suffer chronic pain. When a physician’s best efforts fail to eliminate all their pain, patients may conclude that the provider is incompetent, that they need additional treatment, or that they have not adequately communicated their situation. The belief that something else can be done to improve their condition is common and given the subjective nature of pain, it is virtually impossible for the patient to know if they have maximized their pain treatment. Effective communication is vital to discern concerns that could lead to beneficial changes in treatment versus those reflecting denial of the situation. Many factors influence the number of visits required for chronic pain treatment, such as fluctuations in the intensity of pain because of changes in the individual and environment; the side effects of medication, and the fact that not all individuals respond similarly to medications. Most individuals with chronic pain require several visits to their care provider before reaching optimal levels of comfort and firnctioning. Over time, situations, such as firrther physical deterioration or Opioid tolerance, decrease the efficacy of once successfirl treatment, which causes the patient to return to their provider. The nature of chronic pain treatment—including frequent interactions with providers, and the patient’s need for information—makes communication an important measure of the patient-provider relationship. Affective Aspects Patients who perceive their provider as displaying concern and taking time to communicate explanations are more likely to comply with the therapeutic regimen (Falvo, Woehlke, & Deichmann, 1980). Factors contributing to a physician’s 61 approachability—such as inquiring into patient concerns, worries, and expectations; taking time to answer patient questions in layman’s terms; and being cordial—— produce increased patient satisfaction and compliance (Peck & King, 1982). These studies suggest that affective aspects of the patient-provider relationship are important to patient care. The importance of relationship quality and strength of connection to outcome is underscored in Safran, Montgomery, Chang, Murphy, and Rogers’ (2001) longitudinal observational study. The authors stated that measures of the patient- provider relationship should include aspects of patient trust, quality of communication, perception of being known by the provider, and interpersonal treatment. These elements share qualities that denote affective exchanges in the relationship and were predictive of a patient’s tendency to remain in the care of his/her provider. The conclusion suggests “that medical practices and health plans cannot afford to ignore (that) the essence of medical care delivery involves the interaction of one human being with another” (Safran et al., 2001) p.142. A measure of the quality of the relationship built in psychotherapy is referred to as the working alliance or therapeutic alliance (Horvath & Symonds, 1991). It is suggested that elements contributing to this alliance—such as the provider’s ability to initiate and maintain an interpersonal dialogue, demonstrate empathy and understanding, and forego judgment—contribute significantly to the success of treatment (Cruz & Pincus, 2002). Many of these same qualities are measured in the Roter Interactional Analysis System, an instrument designed to analyze verbal exchanges in the physician—patient dialogue (Roter, 2000). This instrument codes 62 verbal utterances into two main categories—task oriented or technical elements and afi‘ective or socioemotional elements. Data coded directly from audiotapes permitted emotional elements conveyed via inflection and tone to be included. The inclusion of emotional elements in this instrument underscores the researchers’ understanding of the importance of affect as a contributing factor to health communication and treatment outcomes. The importance of affect in the study of patient-provider F relationships is also supported by attachment theory. m—Au.:;_v.‘.g~‘ . Attachment theory views affect as the vehicle that facilitates bonding. Given the importance of affect in both attachment theory and the patient- provider relationship, carefirl consideration was given to its measure in this study. Patient Communication Patient-provider communication is a cornerstone to the practice of medicine (Kaplin et al., 1989). In a comprehensive report on four clinical trials, Kaplin (1989) demonstrated that specific aspects of patient-provider communication were related to better health outcomes, regardless of whether health was measured physiologically, behaviorally, or subjectively. The authors concluded that communication, as a measure of the patient-provider relationship, must be considered in light of patient health outcomes. The subjective nature of pain increases the importance of verbal communication. Chronic-pain patients need to feel comfortable discussing their pain and its impact on their life. Assessing the patient-provider relationship quality by measuring the patient’s level of comfort with initiating communication is therefore important to this study. 63 This study prOposed to examine patient-attachment style and patient-provider relationship quality. The unique characteristics and needs of chronic-pain patients support the inclusion of the patient’s perspective of relationship quality. F urtherrnore, attachment theory’s tenet of affect serving as the mechanism in which initial relationship quality influences subsequent relationships suggests the inclusion of a measurement of this element. Lastly, given the importance of communication in the treatment of chronic-pain patients, its assessment in this study is also essential. These elements combined influenced the selection of the measure of patient-provider relationship quality. Perceived quality of information provided by the provider, patient communication, and affective components of the patient-provider relationship are important to the measure of the quality of the patient-provider relationship in this study. Summary This chapter systematically reviewed the literature supporting the need for this study. Limitations in chronic pain treatment were explored providing justification for improving current understandings of factors that may lead to improved treatment. The use of empirically derived clusters of chronic-pain patients was presented as a viable taxonomy for understanding important psychosocial characteristics of homogeneous subgroups of chronic-pain patients. Next, psychosocial characteristics of these homogeneous groups were presented, along with limitations in the corpus of research. Arguments were presented in support of the need to study attachment and patient- provider relationship quality to understand further the characteristics associated with, and predictive of, homogeneous groups of chronic-pain patients. Attachment theory and related research were discussed with specific attention given to findings related to medical patients. Attachment’s associations with patient-provider relationship quality were presented. Details about research related to patient-provider relationship quality were discussed, including problems defining the construct. Recommendations were made about elements to be included in the measure of patient-provider relationship quality in a population of chronic-pain patients. The next chapter discusses the research design, procedures, and statistics used to analyze the data collected in this study. 65 betv qua] begi inch instr subg adat Stat follc ratio this: (level. relatic Chronil CHAPTER THREE METHODS Introduction This study examined the existence and nature of potential relationships between two predictor variables—attachment style and patient—provider relationship quality and each of three empirically derived chronic-pain subgroups. This chapter begins by restating the hypotheses. It then describes the methodology of this study including selection criteria, data-collection procedures, demographics, and instrumentation. Procedures including techniques used to create three clusters (i.e., subgroups) of chronic pain patients—dysfirnctional, interpersonally distressed, and adaptive coper—and procedures for determining attachment style are presented. Statistical analyses of classifiable participants and differences among subgroups follow, along with comparisons of the independent variables among subgroups. The rationale for the use of logistic regression is discussed next. A summary concludes this chapter. Research Hypotheses Four primary hypotheses were tested in this study. A fifth hypothesis was developed and would have been tested had both attachment and patient-provider relationship quality demonstrated significance. The research question asking, “Is there a relationship between attachment and chronic pain subgroup?” was addressed with the following hypotheses. 66 W: Chronic-pain patients who are classified as dysfunctional or interpersonally distressed are more likely to have one of these insecure attachment styles: preoccupied, dismissing or fearful. Hypothesis 2: Chronic pain patients who are classified as adaptive copers are more likely to have a secure attachment style. The research question asking, “Is there a relationship between patient-provider relationship quality and chroniccpain subgroup?” was addressed with the following hypotheses. Hypothgis 3: Chronic pain patients who are classified in the dysfunctional or interpersonally distressed subgroups are more likely to report poor patient—provider relationship quality. Hypothesjsfi: Chronic pain patients who are classified in the adaptive copers group are more likely to report better patient-provider relationship quality. If both attachment style and patient-provider relationship quality were significant, the additional research question research question asking, “Are differences in attachment and patient provider relationship quality predictive of membership in a chronic-pain subgroup?” would have been addressed with the following hypothesis. W: Attachment style and patient provider relationship quality will improve the prediction of members who are classified in the chronic pain subgroups. 67 The last hypothesis was not examined because only one of the independent variables demonstrated significance. Selection Criteria, Demographics, and Data Collection Overview of the Study Site Initial recruitment efforts were to take place at two sites, but one of the clinics elected not to participate in data collection. Therefore, participants in this study were recruited solely from Neurology of Battle Creek in Battle Creek, Michigan. University Committee for Research Involving Human Subjects (UCRIHS) and appropriate Internal Review Board (IRB) approvals were obtained prior to starting data collection. Selection Process and Description of the Sample A Provider Assistance Request form that outlined the purpose of the study and the need for volunteer participants was given to the staff of Neurology of Battle Creek. A copy of this form as well as all other materials used in this study are included in the Appendix. Volunteer participants were recruited using three methods designed to protect patient confidentiality, which all called for the patient to contact the researcher and not vice versa. First, patients were contacted by the medical facility through a letter on office stationary signed by the office manager. It explained the nature of the research and included a stamped postcard addressed to the researcher’s post office box. Secondly, patients who met the inclusion criteria were informed of the research by a medical staff member during their visit, and given the opportunity to obtain a survey packet during their office visit. The third method of recruitment involved the 68 patient’s contacting the researcher to request a survey packet after reading a recruitment advertisement posted at the facility. Survey packets included a participant recruitment form, two copies of the informed consent form, a compensation form, a list of medical and mental health referrals, demographic questionnaire, and instruments of attachment, care provider relationship quality, and psychosocial factors related to pain. A stamped, pre- addressed envelope was also enclosed. The participant recruitment form informed prospects regarding the purpose of the study and basic inclusion criteria. It directed them to the informed consent form for additional information on potential benefits and risks associated with participation and precautions taken to protect their confidentiality. The informed consent form also provided contact information. Participants were instructed to return a signed copy, and to retain a copy for their files in case they had additional questions. Participants were informed that no treatment differences or preferences would be given to those participating, and that individual responses would not be shared with their health-care providers. Instruments were presented in a counterbalanced order. Completion of the questionnaire was expected to take 30 minutes based on sample trials and compensation was given in the form of $10 gift certificates to local establishments. To ensure confidentially, only the informed consent form had name and address (so that compensation could be sent). Along with the consent form, all the other documents were coded with the same identification number. No identifying information such as patient or care provider’s names appeared on the survey packets. ’69 The informed consent forms and the survey packets were securely stored separately. A master list was created so return rates could be tracked. It was stored with the completed informed consent forms. Debriefing procedures consisted of enclosing a letter with their compensation, thanking them for their participation and repeating the contact information provided in the informed consent form. A list of health resources was provided in case the participant wanted to discuss issues that came to their awareness as a result of participating in the research. The initial number of participants was expected to be 120. Exclusion criteria was modeled after the procedures used by Burns, Kubilus, Bruchl and Harden (2001). Participants were excluded if they had a history of or current: (a)) schizophrenia or bipolar disorder; (b) substance dependence, including alcohol; and/or (c) inability to complete the forms independently because of intellectual difiiculties or insufficient education. Participants under the age of 18 were excluded. The definition of chronic pain requires that the pain exists for three or more months. No participants were eliminated fi’om analysis for this reason. Bipolar patients were omitted, based on the results of a previous study which showed differences in health-related quality of life factors compared to patients without this co-morbid condition (Arnold, Witzeman, Swank, McElroy, & Keck Jr., 2000). Participants who did not complete the forms fully were omitted from analysis. Answers on the demographic questionnaire were examined to ensure that participants had at least an eighth-grade education. They were also qualitatively examined to determine if participants were able to read and write sufiiciently well to understand the items in the instruments. Using the F lesch- 70 Kincaid Grade Level readability score, grade reading levels of 6.7, 6.0, 7.5, were determined for the Experiences in Close Relationships Questionnaire, Multidimensional Pain Inventory, and Patients Reactions Assessment, respectively. Given the assumptions of logistic regression, 50 subjects per independent variable (Aldrich & Nelson, 1984; Grimm & Yarnold, 1995) or 100 classifiable subjects were required for this study, given the two predictor variables—attachment and patient-provider relationship quality. This study sought initially to gather 120 participants to compensate for cases omitted for the reasons stated above, and for cases that were not classifiable into one of the three pain categories. This study’s criterion variable was the chronic-pain subgroup—dysfimctional, interpersonally distressed and adaptive coper. These categories are mutually exclusive but not exhaustive (Grimm & Yarnold, 1995). In other words, though a participant can only be in one category, not every participant belongs to one of the categories. In order to meet the statistical requirement of exhaustion, analyses conducted after the creation of the subgroups included only those participants who were classifiable. Previous studies reported that between 67% (McCracken et al., 1999) and 95% (Turk & Rudy, 1988, 1990) of participants are classifiable. The highest percent classified in previous studies would require this study to gather a minimum of 105 subjects, while the lower figure, 67%, suggests that 149 participants were needed. The initial plan was to analyze the data after 120 surveys had been collected, and to continue collection as needed. This incremental technique was not implemented because of the rapid return of surveys. A total of 170 surveys were collected. 71 Instrumentation Patient Information Questionnaire (PIQ) A demographic and chronic pain questionnaire was developed to provide information necessary to describe the sample population. Information gathered from this questionnaire was used to qualify participants, based on the above-mentiOned criteria and to see if differences existed among the chronic-pain subgroups. The West Haven-Yale Multidimensional Pain Inventory (MPI) The outcome variable, chronic pain subgroup, was operationalized using a cluster analysis of the data from the MPI (Kems et al., 1985). (The procedures used to create the clusters are discussed later in this chapter.) This measure provides a comprehensive understanding of the pain patient, providing information on the severity of pain, the compromises in life quality, and levels of physical and social activities. It has demonstrated good psychometric validity and reliability and was normed on a population of chronic-pain patients with a mean age of 50.5 years (SD = 14.2) and primarily male (88%). The average duration of chronic pain was 10.2 years (SD = 10.5) and the pain etiology was heterogeneous (Kems et al., 1985). Means and standard deviations for each of the subscales of the MPI will be reported in a table and compared to the normative data reported by the authors. Sample items were provided in the previous chapter. Instructions for scoring the subscales state that items specific to each subscale should be added and the total divided by the number of items. In the case of missing data, the sums were divided by the number of answered items. A scale with more than 72 25% of the questions unanswered would have eliminated the case from analysis. There were no cases eliminated for this reason. Other studies utilizing this instrument with pain patients have supported its validity and reliability when used on patients with different sources of pain (Bernstein, Jaremko, & Hinkley, 1995; Rudy et al., 1995; Turk & Rudy, 1988). In an analysis of nine self-reported measures used in the assessment of chronic-pain patients, the MPI was reported to be among the best to capture various facets of pain including perception, affective distress, and firnctionality (Mikail, DuBreuil, & D'Eon, 1993). The Patient Reactions Assessment (PRA) The predictor variable, patient-provider relationship quality, was measured using the Patient Reactions Assessment (PRA) (Galassi et al., 1992). This instrument uses three indices to capture the multidimensional qualities of patient-provider relationships. Each index is composed of five questions. The patient’s perception of the quality of information communicated by the provider and their subsequent understanding is addressed in the Patient Information Index (PII). A sample item reads, “I understand the medical plan for me.” The second subscale, Patient A flecrive Index (PAI), measures patient perceptions regarding the provider’s levels of respect and caring for them. “My provider is warm and caring toward me,” exemplifies items in this subscale. The last subscale, Patient Communication Index (PCI) examines the patient’s perceived ability to communicate information to their provider about their concerns and illness. A sample item reads, “It is hard for me to ask how treatment is going.” A seven-point Likert scale with anchors, Very Strongly Agree and Very 73 Strongly Disagree, is utilized. Seven of the questions are reverse coded. The authors report high internal consistency. The PRA coefficient alpha reported by the authors is .91 overall. Total PRA mean is 88.16 (SD = 12.85) with subscale means for the P11, PAI, and PCI of 29.24 (SD = 4.54); 30.49 (SD = 4.95); and 28.34 (SD = 6.32), respectively. It has demonstrated concurrent validity. Confirmatory factor analysis supports a relationship between the PRA and the conceptual model grounding the use of these subscales. This instrument was chosen over others such as the Medical Interview Satisfaction Scale (Wolf et al., 1978) and Patient Satisfaction Questionnaire (Ware et al., 1976) because of the unique perspective it offers on the patient-provider relationship. Other scales capture the provider’s contributions to the patient-provider relationship without considering the patient’s perceptions of their contributions to the process. Given the study’s focus on patient characteristics, it was important to assess patient beliefs regarding perceived contributions to the relationship process. Secondly, this instrument was designed to measure patient-provider relationships, rather than strictly patient-physician relationships, asking the patient to answer questions as they related to their most recent encounter with the medical professional with whom they spent the most time. This more accurately reflected the staffing found in chronic-pain treatment settings. Experiences in Close Relationships (ECR) The predictor variable attachment was measured using the Experiences in Close Relationships instrument which consists of 36 questions (Brennan et al., 1998). Individuals were asked to indicate on a seven-point Likert scale how much they 74 agreed or disagreed with each statement. This measure was created as an all-purpose reply to questions the authors received regarding which attachment instrument of the many available should be used by researchers. They conducted a principal component analysis of 60 subscales fi'om 14 available attachment measures (Armsden & Greenberg, 1987; Brennan & Shaver, 1995; Camelley, Pietromonaco, & Jafi’e, 1994; Carver, 1994; Collins & Read, 1990; Feeney, Noller, & Hanrahan, 1994; Griffin & Bartholomew, 1994; Hindy, Schwartz, & Brodsky, 1989; Onishi & Gjerde, 1994; Rothbard, Roberts, Leonard, & Eiden, 1993; Shaver, 1995; Simpson, 1990; Wagner & Vaux, 1994; West & Sheldon-Keller, 1994) to understand the underlying structure. Two major dimensions or factors emerged, (a) avoidance and anxiety, which are very highly correlated with avoidance and discomfort with closeness measures (r = .95), and (b) anxiety and fear of rejection measures (r = .95), respectively. The two factors are almost uncorrelated, r = .11 and conceptually equivalent to Bartholomew’s four- category model of attachment. Secure individuals score low on both avoidance and anxiety; dismissing individuals score low on anxiety and high on avoidance; preoccupied individuals score low on avoidance and high on anxiety, and fearful individuals score high on both avoidance and anxiety. Internal consistency reported by the authors for the avoidance and anxiety scales are .94 and .91, respectively. The present study found excellent internal consistencies of .91 and .94, respectively. Sample items included: “I want to get close to my partner, but I keep pulling back.” and “I worry a lot about my relationships.” Two versions of this measure are available (romantic and close relationship) and are highly associated, xz = 1049.07, p = .0001. This study used the close 75 relationship version, which uses the word ‘close’ instead of ‘romantic,’ so that individuals who are not in romantic relationships can be included. When analyzed using item-response theory, the ECR scales demonstrated higher test information firnctions compared to other measures of attachment (F raley, Waller, & Brennan, 2000). Discriminant validity, along with construct and predictive validity, are good (Shaver & Mikulincer, 2002). Detailed instructions for scoring the instrument and computing the attachment-style categories provided by the authors were used in this study (Brennan et al., 1998). Procedures and Statistical Analyses All data gathered and entered into the computer was identified by an identification number to protect confidentiality. Data entry was done solely by the author. Procedures Used to Creme Chronic-Pain Patient Subgroups Some studies reported that 92 to 95% of chronic-pain patients could be classified as dysfunctional, interpersonally distressed, or adaptive (Turk & Rudy, 1988, 1990). McCracken et al. (1999) reported a lower figure, stating that 67.4% of patients in their study were classified using this taxonomy. 0f the patients in the McCracken et al. study, 14% were not considered because of missing data, while 10.4% of patients were hybrids: individuals with characteristics common to more than one subtype; and 11% were anomalous: individuals whose scores did not conform to expectations. The breakdown of the classifiable patients was 32% dysfunctional, 22% interpersonally distressed, and 49% adaptive (McCracken et al., 1999). These ers utilized a computer program, MPI Microcomputer Program, Version 2, 76 (Rudy, 1989b) to score and classify responses fiom the Multidimensional Pain Inventory (MPI). This program reads the scores fiom the MP1 and reports raw scale and standardized scores for each of the 13 subscales. MPI subscale statistics, such as means, standard deviations, and correlations, are also produced. Using Turk and Rudy’s ( 1986; 1988) empirically-derived taxonomy system, the program computes multivariate classification statistics (i.e., subgroup profiles) for each case and classifies each case into one of three subgroups—dysfirnctional, interpersonally distressed, and adaptive coper. Cases presenting with undeterminable profiles are labeled unclassifiable. In more than 800 cases tested, an unclassifiable label occurred 5% of the time (Rudy, 1989a). This research classified patients using version three of the same program (Rudy, 2004). Profiles on the MMPI were significantly different in predicted directions for the subgroups created via the MPI. For example, the dysfunctional and interpersonally distressed group means were significantly elevated in comparison with the adaptive coper group mean on scales 4, 6, 7, and 8 (Psychopathic Deviate (Pd), Paranoia (Pa), Psychasthenia (Pt), and Schizophrenia (Sc), respectively) (Etscheidt, Steger, & Braverman, 1995). The distribution of chronic-pain patients into subgroups is uneven. In one study of classified pain patients, approximately 32% were dysfunctional, 22% were interpersonally distressed, and 46 % were adaptive copers (McCracken et al., 1999). In Bergstrom, Jensen, Bodin, Linton and Nygren’s study (2001) the distribution was 35, 23, and 42 percent, respectively. 77 Statistical Procedures Analysis first examined the percent of patients that were classified. The percent of nonclassifiable participants is reported in Chapter Four, along with a comparison of this study’s classifiable percentages and similar studies. All remaining analyses were conducted on participants who were classified into one of three chronic pain subgroups. Analyses of Classifiable Participants Three main analyses were conducted on the classifiable data with descriptive statistics conducted first, followed by inferential statistics. First, the three subgroups were examined for differences based on demographic factors. Second, univariate group comparisons were conducted analyzing the independent variables. A Chi- square test of association was used to determine if attachment style differed among the chronic-pain subgroups, and an ANOVA was used to determine if the mean of patient-provider relationship quality differed in the chronic pain subgroups. This study was designed to examine whether attachment and patient-provider relationship quality improved predicted membership into one of the subgroups if both variables demonstrated significance. Unfortunately, only one of the variables was significant, so modeling using logistic regression was not conducted. Differences Among Subgroups The three subgroups were analyzed to see if differences existed, based on demographic factors. Gender, marital status, race, and work status are categorical variables. Chi-squared tests of association were conducted to see if differences 78 existed between observed and expected fi'equencies. Contingency tables are provided in the next chapter for each categorical demographic factor. Nondirectional tests were used with the following null and alternative hypotheses. Ho: Observed fiequency distributions are equal to the expected distributions in each category. H1: Observed frequency distributions are not equal to the expected distributions in each category. The assumptions of Chi-square were checked. The Chi-square test assumes that each observation is unique to a single category, and that the observations are independent with respect to other observations. The samming procedures used prevented the occurrence of dependent observations. Continuous demographic characteristics; age, duration of pain, pain level, and education were analyzed using ANOVA. The following null hypothesis and alternative hypothesis were tested. Ho: There are no differences in the population means among the three chronic pain subgroups: P1 = u; = r13 H1: Differences exist in at least one pair of population means: pi - up qt 0 for some i Three assumptions of AN OVA are independence, normality and homogeneity of variance. Sampling procedures addressed the first two assumptions. The assumption of homogeneity of variances (i.e., oz; = 622 = 023) was statistically tested. 79 A previous study showed that the subgroups did not differ on age, education or duration of pain using analysis of variance (ANOVA) (McCracken et al., 1999). No differences in that study were found among the subgroups on gender, marital status, or race using Chi-squared analyses. However, in this same study, the subgroups differed on work status with more adaptive copers currently working when compared to the other two subgroups (McCracken et al., 1999). The finding of pain level (intensity) between subgroups was significantly different in other studies (McCracken et al., 1999; Turk & Rudy, 1988). It was expected that this study would demonstrate similar findings. However, while no differences were found on work status, differences were found between the groups for the variables age and level of pain. Correlations were checked between these variables and patient-provider relationship quality with no significant relationships noted. Also no significant relationships were found between these variables and attachment. Comparisons of the Independent Variables among Subgroups Univariate comparisons were conducted for each of the three chronic-pain subgroups and the independent variables—attachment and patient-provider relationship quality. Chi-square test of association was used to analyze the relationships among chronic-pain subgroups and four categories of attachment— secure, fearful, preoccupied, and avoidant. Results are reported through a contingency table in the next chapter. Patient-provider relationship quality was explored using ANOVAs. Since none of the relationships were significant Tukey post-hoc tests (a = 0.05) were not conducted. 80 An ANOVA was conducted between the two independent variables, attachment and patient-provider relationship quality. This F test was used to test whether the means of relationship quality differed across the attachment categories. It does not however, provide information about the strength of the association. To determine how much variance in patient-provider relationship quality can be accounted for by attachment type, an omega-square (62 statistic was used. Multinomial Logistic Regression This study proposed to further explore the unique contributions of attachment and patient provider relationship quality in the prediction of chronic pain subgroup membership. A review of multivariate statistics determined that logistic regression was the most appropriate statistical method for a number of reasons (Tabachnick & Fidell, 2001). Logistic regression addresses the research question of whether group membership can be predicted by a number of variables. It also provides information on the probability of membership into a particular cluster given any possible configuration of the independent variables. This analysis however, was not conducted because only one of the variables was significant. 81 Summary The primary purpose of this study was to examine the existence and nature of relationships between attachment and patient-provider relationship quality and each of three empirically derived chronic-pain groups. Research hypotheses were stated followed by the methodology. Selection criteria, data-collection and measures were discussed. Procedures and statistical analysis were presented with an emphasis on F creation of clusters and the classification of attachment style. Chapter Four discusses the results of the analyses. 82 CHAPTER IV ANALYSIS OF RESULTS Introduction This chapter presents the findings of this study using the order of the hypotheses for organization. First, characteristics of the sample population and the classification of participants into chronic-pain subgroups are discussed. Discrete demographic variables (i.e., gender, race, marital status, and employment status) and continuous variables (i.e., age, education, duration of pain, and pain level) were examined to see if differences existed between the subgroups based on these factors. An explanation of how attachment style was classified is provided. Next, attachment style and patient-provider relationship quality were examined for significant relationships with the chronic-pain subgroups. Findings of a relationship between attachment style and patient-provider relationship quality are presented. Examination of the ability of attachment style and patient provider relationship quality to predict membership into a chronic-pain subgroup was not conducted because only one of the variables demonstrated significance. Participants A total of 226 surveys were distributed between January 1 and March 15, 2005. The vast majority were mailed after the patient had returned a postcard. Fewer than 20 surveys were given out by health—care providers. From this, 170 were returned for a return rate of 69%. Data were entered solely by the author and inspected to ensure accuracy. 83 Surveys were excluded from analysis in two steps. First, cases were excluded if they did not meet the inclusion criteria. The cases excluded, and for which criterion, are as follows: (a) incomplete data, four; (b) minor, one; (c) schizophrenic, one; (d) chemical or alcohol dependency, two; and (e) bipolar disorder, five. No cases were eliminated because their duration of pain was less than the three-month minimum or because their level of education was considered too low to comprehend the survey questions. After this step, there were 157 participants remaining. In a second exclusion step, cases that were not classifiable into one of three chronic-pain subgroups were eliminated. The Multidimensional Pain Inventory (MPI) Microcomputer Program Version 3 (Rudy, 2004) recognized two types of patient profiles as atypical. The hybrid label was given to profiles that included features of more than one cluster. In this sample there were 10 hybrid profiles. A second atypical profile, labeled anomalous, represented participant responses that are inconsistent with expected MPI scores. In this sample, six cases received the anomalous label. None of the cases were unclassifiable because of missing data. After the second exclusion step, 141 cases remained. The following information reflects only these cases. The age range was from 18 to 84 with a mean of 52.46 (SD = 14.95). The education range was from 8th grade to PhD. with a mean of 13.55 years of education (SD = 2.13). In the sample, 77% were female. Relationship status was 15.6% single, 63.1% married, 14.2% divorced, and 7.1% widowed. Racial/ethnic status: 92.3% White, 3.5% Afiican American, and 4.2% of the data missing. The employment breakdown was 46.1% not working or retired, 14.9% working 1 to 31 hours per week, 84 35.5% working 32 or more hours per week and 2.8% of the data missing. The mean time since the onset of pain was 11.19 years (SD = 11.37). A histogram of this variable demonstrated a positive skew. The mean pain intensity (pain level) was 5.63 (SD = 2.31) on a scale of O to 10 with 10 being the worst possible pain. The following body regions and related pain diagnoses/etiologies were reported: 29 head and neck (migraine, motor vehicle accident [MVA], cervical disc problems); 11 face (trigeminal neuralgia, injury, unknown); 44 extremities, including feet, legs, arms, and hands (neuropathies, multiple sclerosis, arthritis); 37 back (degenerative, ruptured and herniated discs, MV A); 28 body or multiple sites (fibromyalgia, arthritis, lupus, multiple sclerosis, unknown); and 13 joints (arthritis, lupus, injury). Some participants reported more than one area of pain so that the total is greater than the 141 cases. Classification of Participants into Chronic Pain Subgroups The authors of the Multidimensional Pain Inventory (Kems et al., 1985) reported internal consistencies ranging fiom .70 to .90. The present study’s reliability was .57 to 93. Reliabilities, means, and standard deviations fiom the author’s study and the present study for each of the 12 scales are reported in Table 4.1. Overall, 89.8% of cases were classifiable into one of the chronic-pain subgroups using the MPI Microcomputer Program (Rudy, 2004). This program used Turk and Rudy’s (1986; 1988) empirically derived taxonomy system to compute a multivariate classification statistic for each case. It then classified each case into one of three subgroups—adaptive coper, dysfirnctional, or interpersonally distressed. This study’s classification percentage is roughly in the middle of classification percentages 85 Table 4.1 Comparisons of Means, Standard Deviations, and Reliabilities from MPI Beak M. .32 Reliabilityb 1. Interference 9 3.84 (3.74)‘: 1.18 (1.26) 0.93 (0.90) 2. Support 3 4.18 (4.31) 1.22 (1.47) 0.79 (0.83) 3. Pain severity 3 3.61 (3.55) 1.18 (1.11) 0.85 (0.72) 4. Self-control 2 4.12 (3.63) 1.15 (1.57) 0.79 (0.79) 5. Negative mood 3 3.55 (3.23) 0.84 (1 .32) 0.75 (0.73) 6. Punishing responses 4 2.29 (0.97) . 1.34 (0.94) 08940.84) 7. Solicitous responses 6 3.73 (2.57) 1.28 (1.15) 0.84 (0.78) 8. Distracting responses 4 2.69 (1.72) 1.15 (1.11) 0.75 (0.74) 9. Household chores 5 4.21 (2.71) 1.37 (1.30) 0.88 (0.86) 10. Outdoor work 5 2.24 (1.19) 1.16 (1.04) 0.80 (0.77) 11. Activities away 4 3.23 (1.79) 0.89 (0.83) 0.57 (0.70) 12. Social activities 4 2.89 (1.94) 0.90 (0.95) 0.59 (0.74) a = Total items in scale b = Cronbach’s alpha c = Numbers in parentheses are from Kems, Turk and Rudy’s study (1985). The distribution of chronic pain patients into subgroups was expectedly uneven with 73 cases (52%) classified as adaptive copers; 33 (23%) dysfunctional; and 35 (25%) interpersonally distressed. In a classification study of pain patients treated at a university pain management center, approximately 46% were adaptive copers, 32% were dysfirnctional, and 22% were interpersonally distressed (McCracken et al., 1999). In Bergstrom, Jensen, Bodin, Linton and Nygren’s study (2001) conducted on Swedish individuals referred to a vocational rehabilitation program, the distribution was 42, 35, and 23 percent, respectively. Compared to other studies the percent of members in the dysfunctional subgroup was less, while the percent in the interpersonally distressed subgroup was more. 86 Comparison of Chronic Pain Subgroups An underlying goal of sample selection was to obtain a sample that was relatively homogeneous in nature. The ex-post facto design of this research prevented random assignment of subjects, increasing the potential for confounding variables to influence the findings. Therefore, the three chronic pain subgroups were compared on variables thought to have a potential impact on the results of statistical testing of the r independent variables—attachment and patient-provider relationship quality. A Chi- squared test of association was used for the discrete variables with no significant associations: Gender x2 (2, N = 141) = .480 (p < .787); marital status 7(2 (6, N = 141) b = 3.862 (p < .695); race x2 (4, N = 136) = 5.058 (p < .281); and work status fl (4, N = 136) = 6.720 (p < 1.51). Detailed information on the count and percentage is presented in Tables 4.2, 4.3, 4.4, and 4.5, respectively. Continuous demographic characteristics were analyzed using ANOVA with age and pain level demonstrating differences between the subgroups; F (2, 138) = 3.946, p < .022 and F (2, 136) = 9.872, p < .000, respectively. No differences were found between the chronic-pain subgroup and either education or duration of pain; F (2, 138) = 2.674, p < .073; F (2, 135) = .870, p < .421, respectively. The finding of differences in ages between the subgroups is unexpected, as other studies have shown no differences (McCracken et al., 1999; Turk & Rudy, 1988). The finding of differences in pain level (intensity) between subgroups is consistent with the findings of other studies (McCracken et al., 1999; Turk & Rudy, 1988). McCracken et al. 87 Table 4.2 Gender Count and Percentages by Chronic Pain Subgroup Gender Total Male Female Chronic- Pain Adaptive Count 22 5 1 73 Subgroup o . . . ”W‘tmnpm“ 30.1% 69.9% 100.0% Subgroup Dysfirnctional Count 1 l 22 33 o . . . ”m‘mnp‘m‘ 33.3% 66.7% 100.0% Subgroup Interpersonally distressed Count 9 26 35 o . . . ”Mth‘npa‘“ 25.7% 74.3% 100.0% Subgroup Total Count 42 99 141 o . . . ”mm“ Pam 29.8% 70.2% 100.0% Subgroup 88 Table 4.3 Marital Status by Chronic Pain Subgroup Marriage Total Single Married Divorced Widowed Chronic- Pain Subgroup Adaptive Count 10 50 7 6 73 %WithinPain 13.7% 68.5% 9.6% 8.2% 100% Group Dysfimctional Count 6 18 7 2 33 %WithinPain 18.2% 54.5% 21.2% 6.1% 100% Group Interpersonally Distressed Count 6 21 6 2 35 %withinPain Subgroup 17.1% 60.0% 17.1% 5.7% 100% Total Count 22 89 20 10 41 %withinPain Subgroup 15.6% 63.1% 14.2% 7.1% 100% 89 Table 4.4 Race/Ethnicity by Chronic Pain Subgroup Caucasian African Missing Total American Chronic- Pain Subgroup Adaptive Count 67 2 l 70 % within Pain Group 95.7% 1 .4% 2.9% 100% Dysfunctional Count 30 3 0 33 % with Pain Subgroup 90.9% 9. 1% 0% 100% Interpersonally Count 33 O O 33 Distressed % with Pain Subgroup 100% .O% .0% 100% Total Count 130 5 1 136 % within Pain Subgroup 95.6% 3.7% .7% 100% 9O Table 4.5 Employment by Chronic-Pain Subgroup Employment Total None 1-31 32 or Hours more hours per per week week Chronic- Pain Subgroup Adaptive Count 3 1 15 23 69 % within Pain Group 44.9% 21.7% 33.3% 100% Dysfunctional Count 19 l 12 3 2 ' % with Pain Subgroup 59.4% 3.1% 37.5% 100% Interpersonally Count 15 5 15 3S Distressed % with Pain Subgroup 43.9% 14.3% 42.9% 100% Total Count 62 21 50 136 % within Pain Subgroup 45.6% 15.4% 36.8% 100% 91 (1999) showed that more adaptive copers were currently working compared to the dysfunctional and interpersonally distressed subgroups, while this study did not show any differences. Correlations were checked between the potentially confounding factors (age and pain level) and patient provider relationship quality with no significant relationships noted; r = .018, p < .829; r = -.029, p < .731. This suggested that these potentially confounding variables do not have a relationship with patient provider relationship quality and therefore would not offer a reasonable explanation for difference between patient-provider relationship quality and chronic-pain subgroups. Also no significant relationships were noted between age and pain level, and attachment: F (3, 56) = 1.130, p < .751; F (3, 10) = 2.208, p < .061. Classification of Attachment Style Attachment style was measured using the Experiences in Close Relationships instrument. The internal consistencies for the present study on the anxiety and avoidance scales were excellent: 11 = .94 (n = 133) and .91 (n = 136), respectively. To obtain participant attachment style, the directions provided by the authors were followed (Brennan et al., 1998). First, ten of the questions were reverse scored and recoded using temporary variables. The mean scores for two dimensions—avoidance and anxiety—were calculated using a formula that ignored cases with less than 14 responses out of 18 possible for each scale. In this sample, all of the cases met the minimum requirement for classification. Attachment-style categories were computed using the coefficients listed by the authors based on their sample of n = 1,082. See 92 Table 4.6. Lastly, cases were assigned a variable label of secure, fearful, and preoccupied Table 4.6 Fisher’s Linear Discriminant Functions Provided by Brennan et al. (1998) for Computation of Attachment-Style Categories Compute SEC2 = avoidance*3.2893296 + anxiety*5.4725318 —- 11.5307833 Compute FEAR2 = avoidance*7.2371075 + anxiety*8. 1776446 — 32.3553266 Compute PRE2 = avoidance*3.9246754 + anxiety‘9.7102446 — 28.4573220 Compute DIS2 = avoidance‘7.3654621 + anxiety’4.9392039 — 22.2281088 or dismissing based upon which classification score was greater relative to the other three categories. As expected all 141 cases received an attachment-style category: secured 47 (33.3%); fearfirl 35 (24.8%); preoccupied 36 (25.5%); and dismissing 16 (16.3%). The following three studies are presented for comparison. In a study conducted on female patients, 34% were identified as securely attached; 21% fearfirl; 22% preoccupied, and 23% dismissing (Ciechanowski, Walker et al., 2002). In a sample of young adults, 27% were securely attached, whereas 22, 24, and 27% were categorized as fearful, preoccupied and dismissing, respectively (Ognibene & Collins, 1998). In a sample of patients with type 1 and 2 diabetes, 30, 21, 24, and 25% were secure, fearful, preoccupied and dismissing, respectively (Ciechanowski et al., 2001). 93 Attachment Style and Chronic Pain Subgroup Analysis was guided by the following hypotheses. The research question asking, “Is there a relationship between attachment and chronic pain subgroup?” was addressed with the following hypotheses. Hypothesis 1: Chronic pain patients who are classified as dysfirnctional or interpersonally distressed are more likely to have an insecure attachment style such as preoccupied, dismissing or fearfirl. Hypothesis 2: Chronic pain patients who are classified as adaptive copers are more likely to have a secure attachment style. A chi-squared test of association suggested a relationship between the categories of attachment and chronic pain subgroup, showed significance x2 (6, N = 141) = 24.072, p < .001. The test of association suggested a moderate relationship: Cramer’s V = .292, p < .001. Standardized residuals were used to examine cells with values equal to or greater than two standard deviations deemed significant (std res Z :1: 2). Supporting the first hypothesis, significant relationships were found between fearfirl attachment style and the interpersonally distressed chronic-pain subgroup with 42.9% of individuals with this attachment style also belonging to the interpersonally distressed chronic pain subgroup. Only 14.3% of individuals with a fearful attachment style were also members of the adaptive coper subgroup. No significant relationships were found between preoccupied or dismissing attachment styles and chronic-pain subgroup. Supporting the second hypothesis, significant relationships were found between the adaptive chronic-pain subgroup and both secure and fearful attachment 94 styles as predicted. Individuals with secure attachment were more likely to be in the adaptive chronic pain subgroup (47.9%) and less likely to be in the interpersonally distressed chronic pain subgroup (12.3%). The results of these findings are summarized in Table 4.7. 95 Table 4.7 Attachment Style by Pain Subgroup Coefficient-Based Attachment Category Total Secure Fearfirl Preoccupied Dismissing Chronic- Pain Subgroup Adaptive Count 35 9 9 9 73 %WithinPain 47.9% 12.3% 27.4% 12.3% 100% Group Std. 2.2 -2.1 .3 -.8 Residual Dysfirnctional Count 7 11 10 5 33 %WithinPain 21.2% 33.3% 30.3% 15.2% 100% Group Std. -1.2 1.0 .5 -.2 Residual Interpersonally Distressed Count 5 15 6 9 35 %withinPain Subgroup 14.3% 42.9% 17.1% 25.7% 100% Std. Residual -2.0 2.1 -1.0 1.4 Total Count 47 35 36 23 141 %withinPain Subgroup 33.3% 24.8% 25.5% 16.3% 100% 96 Patient Provider Relationship Quality and Chronic Pain Subgroup The research question asking, “Is there a relationship between patient-provider relationship quality and chronic pain subgroup?” was addressed with the following hypotheses. Hypothesis 3: Chronic pain patients who are classified in the dysfunctional or interpersonally distressed subgroups are more likely to report poor n patient-provider relationship quality. Hypothesis 4: Chronic pain patients who are classified in the adaptive copers subgroup are more likely to report better patient-provider relationship quality. g The internal consistency for the Patient Reactions Assessment was excellent: (1 = .90. The null hypothesis was retained in both of these hypotheses with no differences found among chronic-pain subgroup and patient-provider relationship quality; Madaptive = 79.56; Mdysfunctional = 76.79; Minterpersonally distressed = 73.60; 12(2) = 1.996, p <.14o. Attachment Style and Patient Provider Relationship Quality . The relationship between attachment and patient-provider relationship quality was significant: F(3) = 4.928, p < .003. Individuals with a secure attachment style reported better patient provider relationship quality than any of the insecure attachment styles: fearfirl (p < .005), preoccupied (p < .046), and dismissing (p < .03 8). No significant differences were found among the insecure attachment styles. Details are presented in Table 4.8. The variance of patient provider 97 Table 4.8 Patient-Provider Relationship Quality by Attachment Style (I) Coefficient- (j) Coefficient- Mean Std. 95% Confidence Interval Based Based Difference Error Sig. Attachment Attachment (I—J) Lower Upper Category Category Bound Bound Secure Fearful 10.75" 3.167 .005 2.51 18.99 Preoccupied 8.28‘ 3.142 .046 .l 16.45 Dismissing 9.77“ 3.610 .038 .38 19.15 Fearful Secure -10.75 3.167 .005 -18.99 -2.51 Preoccupied - 2.47 3.368 .883 -11.23 6.29 Dismissing -.99 3.808 .994 -10.89 8.92 Preoccupied Secure -8.28"' 3.142 .046 -l6.45 -.11 Fearful 2.47 3.368 .883 - 6.29 11.23 Dismissing 1.48 3 .787 .979 -8.36 1 1.33 Dismissing Secure -9.77‘ 3 .610 .038 -19.15 -.38 Fearful .99 3.808 .994 -8.92 10.89 Preoccupied -1.48 3.787 .979 -l 1.33 8.37 * The mean difi‘erence is significant at the .05 level. 98 relationship quality accounted for by attachment style was estimated with attachment style accounting for less than 8% of the variance: 032 = .0771. Attachment Style and Patient Provider Relationship Quality as Predictors of Chronic Pain Subgroup Membership Examination of the research question asking, “Are differences in attachment and patient provider relationship quality predictive of membership in a chronic-pain subgroup?” was contingent on both attachment style and patient-provider relationship quality demonstrating significant relationships with the chronic pain subgroups. The following hypothesis was not examined because only attachment style demonstrated significance. Hypothesis 5: Attachment style and patient-provider relationship quality will improve the prediction of members who are classified in the chronic-pain subgroups. Summary This study found support for moderate relationships between attachment-style and chronic-pain subgroup. However, no relationship was found between patient- provider relationship quality and chronic-pain subgroup. A relationship was found, albeit small, between attachment-style and patient-provider relationship quality. Chapter V discusses the implications of the present study’s findings. 99 CHAPTER FIVE DISCUSSIONS AND CONCLUSIONS Introduction This chapter begins with a summary of this research. Based on the data analyses, conclusions are presented, along with a section that integrates the findings of this study with previous studies. Limitations of this study are presented, followed by recommendations for firture research. Thoughts regarding the important contributions counseling psychology can make to the study of attachment and chronic pain conclude this chapter. This chapter is organized around the hypotheses put forth in this research study. Overview Considerable progress has been made toward understanding individual psychosocial variables influencing the interpretation of pain. Research has examined factors, such as personality traits and psychopathology (W eisberg & Keefe, 1999), stress (Melzack, 1999), gender (Miaslowski, 1999), the role of emotion (Robinson & Riley III, 1999), and religious and cultural influences (Morris, 1999). Advancing this line of investigation, researchers have attempted to develop profiles of chronic-pain patients based on these, and other psychosocial characteristics. One promising taxonomy presented by Turk and Rudy (1987, 1988) utilized cluster analysis to derive three subgroups of chronic-pain patients—adaptive, dysfunctional, and interpersonally distressed. Other research has expanded the knowledge of characteristics of these subgroups and the relationship of these 100 subgroups to other variables (Bergstroem et al., 2001; Etscheidt et al., 1995; Kems & Rosenberg, 1995; Lousberg et al., 1996; Strong et al., 1994). Overall, however, the number of variables investigated is small relative to the importance of understanding the unique characteristics of the chronic pain subgroups. The discovery of additional variables delineating subgroup membership would improve the understanding of fundamental differences, and potentially lead to tailored treatments based on patient needs. Extrapolating from the direction of current research, this research attempted to provide information on the relationships between chronic-pain subgroups and attachment style and patient-provider relationship quality. Analysis examined the relationships between attachment style and patient-provider relationship quality. Furthermore, this research was designed to examine attachment style and patient- provider relationship quality for their potential to predict membership into a chronic- pain subgroup if both variables demonstrated significance. For clarity, the presentation of this discussion is organized in the order of the hypotheses tested. Attachment and Chronic Pain Subgroup The research question asking, “Is there a relationship between attachment and chronic pain subgroup?” was addressed with the two following hypotheses. 3390mm: Chronic-pain patients who are classified as dysfunctional or interpersonally distressed are more likely to have an insecure attachment style such as preoccupied, dismissing, or fearful. _I:I_ypgt_h_e_s_is_2: Chroniepain patients who are classified as adaptive copers are more likely to have a secure attachment style. 101 This research found support for moderate relationships among attachment style and chronic-pain subgroups. Supporting the first hypothesis, individuals who were members of the interpersonally distressed subgroup were more likely to have a fearfirl attachment style. Further support for the relationship between chronic pain subgroup and fearful attachment style was demonstrated in the paucity of these individuals in the adaptive coper chronic-pain subgroup. In contrast to the findings of relationships between fearfirl attachment style and chronic-pain subgroup, relationships between members of the dysfunctional chronic-pain subgroup and attachment style were not supported. Supporting the second hypothesis, individuals who were classified as adaptive copers were more likely to be securely attached, and less likely to have a fearful attachment style. Interestingly, no support was found for relationships between chronic pain subgroup and either preoccupied or dismissive attachment styles. Characteristics of Attachment Consistent with a multitude of research on adult attachment and health, this study supported the general premise that secure attachment is related to better health and adjustment, while insecure attachment is related to poor health and adjustment (e.g., Feeney & Ryan, 1994; Gerlsma & Luteijn, 2000; Hemesath, 1997; Kaati, Vester, Sjostrom, & Bygren, 2001; Mikulincer, Florian, & Tolmacz, 1990). In a study of symptom perception and reporting, attachment style was related to symptom reporting with preoccupied and fearful attachment styles associated with greater symptom reporting than secure attachment (Ciechanowski, Walker et al., 2002). Kotler, Buzwell, Romeo and Bowland’s study (1994) provided support for avoidant 102 attachment style as a link to physical ill-health. (Avoidant attachment is the name given to one of the attachment types in a classification scheme with three adult styles (Hazan & Shaver, 1987)). This study’s findings were similar to a study of patients with Hepatitis C in which patients with fearful attachment were linked to more symptoms than those with secure attachment (Ciechanowski, Katon, Russo, & Dwight-Johnson, 2002) and also to a study conducted with HIV -positive individuals in which attachment style was associated with adjustment to chronic illness (Tumer- Cobb et al., 2002). The findings of this research suggested that, similar to other chronic medical conditions, one’s ability to cope and function, despite chronic pain is related to the ability to establish and maintain significant relationships. Of interest are this study’s findings that only secure and fearful-attachment styles are associated with chronic-pain subgroups. Examining the nature of each of the attachment styles ofl‘ers potential explanations. In fearful adults the internal working model is a combination of negative self- view and negative view of others. Not only do these individuals experience anxiety and feelings of inadequacy, but their interpretation of the motives and opinions of others makes them feel vulnerable with increasing afiiliation. Bowlby contended that the internal working models of individuals were heightened during periods of stress. In other words, the pattern of interpretation of self and others is roused or activated with chronic pain. In individuals who have a fearful attachment style, chronic pain may elicit a tendency to avoid proximity maintenance or affiliation because instead of feeling support, safety, and protection in the presence of others, they feel discomfort. 103 Compounding the problem, Fonagy (1998) suggested that these individuals lack self- reflection and the ability to interpret intimate relationships realistically. The theoretical opposite of individuals with fearful attachment style is individuals'who are securely attached. While the former have negative thoughts of self and others, the latter have more favorable thoughts that translate into views of themselves as worthy of care and others as capable of providing care. During times of stress, securely attached individuals are comfortable with closeness and seek it out in a sincere manner. They are capable of the give-and-take necessary to sustain a healthy relationship and are more trusting and self-confident than insecurely attached individuals. In addition, qualities of the securely attached individual include the ability to appraise accurately personal needs, as well as the needs of others (Fonagy, 1998) The preoccupied-attachment style has been associated with less self- awareness and more of a dependency on others to assuage their distress (Feeney & Noller, 1990). They are likely to be more other-focused and emotionally reactive to perceived criticism or disapproval of others (Fuendeling, 1998). The significant other in a relationship is likely to feel drained or overtaxed in response to the preoccupied adult’s needs (Pistole, 1994). In comparison, the dismissing-attachment style has been associated with a proclivity for less affiliation and more isolation during times of stress (F eeney & Noller, 1990). Instead of seeking out significant others during times of stress, individuals with this attachment style are more likely to engage in task-related activities, such as work or recreation (Hazan & Shaver, 1990). A significant other in a 104 relationship with an individual with a dismissing attachment style is likely to feel devalued, excluded or dismissed during times of stress. The dismissive’s appraisal is likely to be distorted with regard to their needs, rather than the perceptions of others. To decrease personal vulnerability, they are likely to deny or minimize their stress, situation, or need for affiliation. Neither the preoccupied nor dismissing-attachment styles were associated with any of the chronic pain subgroups. Several self-report attachment scales, including the one used in this study, were factor analyzed yielding support for two dimensions—avoidance and anxiety (Brennan et al., 1998; Collins & Read, 1990; Simpson, 1990). The first dimension represents differences along a secure-avoidant continuum, reflecting thoughts and feelings about others, while the second dimension represents differences along a comfort-anxiety continuum, reflecting thoughts and feeling about self. The four- group taxonomy of adult attachment uses categories based on the valence assigned to each of the two dimensions: secure (+ self, + other); dismissing (+ self, - other); preoccupied (- self, + other); and fearful (- self, - other). Avoidance has been related to negative evaluations of significant others (Feeney & Noller, 1991), poor socioemotional functioning, and rejection of the need for affiliation (Rom & Mikulincer, 2003), while attachment anxiety has been linked to negative emotions, including a tendency to experience anger, hostility, and other less functional emotions (Mikulincer, 1998; Rholes, Simpson, & Orina, 1999). It seems that securely-attached individuals experience better emotional regulation and the ability to utilize social support in a healthy manner, whereas fearfirlly-attached individuals experience poor emotional regulation and are unable to 105 benefit fi'om social support during times of stress. These two categories of attachment represent polar extremes on both dimensions of attachment—avoidance and anxiety. In contrast, preoccupied and dismissing-attachment styles are reciprocal with each having a buffer afforded by a positive valence on either the avoidant (preoccupied) or anxious (dismissing) dimension. From this perspective, it is possible to think of the benefits of attachment in a hierarchical manner, with secure attachment afl‘ording ' individuals the most benefit, fearfirl attachment the least, and preoccupied and I dismissing somewhere in between. One potential explanation for the present study’s inability to support relationships between preoccupied or dismissing-attachment styles and chronic-pain subgroup may be that the characteristics of these individuals’ attachment styles are neither beneficial enough nor detrimental enough to be detected. Similar to the finding of attachment research, studies on social support and anxiety affirm the importance of relationships. Research using the MMPI as a predictor of pain-coping strategy demonstrated that the use of social support was related to good adjustment, while the use of a nonsocial coping strategy (medical remedy) was related to poor adjustment (Kleinke, 1994). A study assessing predictors of back injury report found social anxiety to be as much or more predictive than bodily complaints (F ordyce, Bigos, Batti'e, & Fisher, 1992). Future research may uncover common mechanisms accounting for the present study’s findings and those in the attachment and social support literature. Characteristics of Chronic Pain Subgroups Further support for the relationship between fearfirl-attachment style and ability to cope with chronic pain was demonstrated by the paucity of individuals with 106 fearful-attachment style in the adaptive coper chronic pain subgroup. Adaptive copers report higher levels of daily activity and the perception of more life control. They also report lower levels of pain severity and affective distress. And, as the name suggests, they appear to be adapting or coping better with pain. The negative relationship between fearful-attachment style and the adaptive chronic-pain group, in combination with the positive relationship between fearful attachment and interpersonally distressed chronic-pain group, suggests that those who have a fearfirl-attachment style have a higher likelihood of perceiving significant others as being less supportive and responsive to their needs. It was also associated with a lower likelihood of coping well with their condition. These characteristics are hallmarks of the interpersonally distressed, chronic-pain subgroup. This study was not designed to establish cause and effect; however, the chronology of attachment suggests that individuals with fearfirl attachment are more likely to have difficulties coping with chronic pain, rather than the other way around. It is left to future research to affirm this conjecture. Individuals labeled as adaptive copers were more likely to be securely attached. The present study supported a relationship between secure attachment and the ability to com with chronic pain more effectively, to experience less affective distress, and to function at a higher level under conditions of chronic pain. It also affirmed that characteristics of the interpersonally distressed chronic pain group— e.g., perceiving others to be less supportive and attentive—are less compatible with a secure attachment style. The dysfunctional chronic-pain subgroup reported higher levels of pain severity in combination with lower levels of daily activity, more effective distress, 107 and less perceived control over their life. Although it was expected that relationships would exist between this chronic-pain subgroup and insecure attachment, this was not the case. It may be that the characteristics of this subgroup—affective distress in the absence of reports of interpersonal distress—exclude qualities necessary to support a relationship between it and an attachment style. Patient-Provider Relationship Quality The research question asking, “Is there a relationship between patient-provider relationship quality and chronic-pain subgroup?” was addressed with the following hypotheses. Hyppthesis 3: Chronic-pain patients who are classified in the dysfirnctional or interpersonally distressed subgroups are more likely to report poor patient-provider relationship quality. Hyppthesis 4: Chronic-pain patients who are classified in the adaptive copers subgroup are more likely to report better patient-provider relationship quality. Perceptions of patient-provider relationship have been shown to be an important influence on health-care outcome, regardless of whether outcome was measured physiologically, behaviorally, or subjectively (Kaplin et al., 1989). Extrapolating from this, it seems plausible that individuals who are better adjusted and adapting to chronic pain would also report a better patient-provider relationship. Furthermore, both theoretical (Goldberg, 2000) and empirical research suggested potential relationships between these two variables. However, the present research did 108 not support a relationship between chronic-pain subgroup and patient-provider relationship quality. The findings of the present study appear contrary to past research, which found that pain and social firnctioning were related to levels of satisfaction (Cohen, 1996). The present research also appears to contradict findings that health status was related to health-care use, satisfaction, and physician-patient communication (Kaplin et al., 1989; Zastowny et al., 1989). The findings of the present research differ fi'om the findings of Greenley, Young, and Schoenherr (1982) who found support for a relationship between psychological distress and dissatisfaction with health-care services. More carefully comparing the findings of this present study with other studies, it is possible to see that, although there are similarities between facets of the earlier studies and this study, none of them are direct comparisons. For example, the comparison studies examining patient satisfaction are examining a broader construct than patient-provider relationship. Several studies reported the association of psychological distress with patient-provider relationship quality (Abram, 1969; Balint, 1957). Aspects of the chronic-pain subgroups include facets of psychological distress (e.g., adaptive copers report less affective distress than members of the dysfunctional group). However, it may be the case that these features alone (or in combination with other features) are not enough to support a significant relationship between chronic-pain subgroup and patient-provider relationship quality. Furthermore, while these studies were conducted on patient populations, they were not limited to chronic pain patients. It may be the case that, despite adjunct research findings suggesting a relationship between chronic-pain subgroups and patient- 109 provider relationship quality, the essence of what makes these other relationships significant is absence fi'om the narrower parameters found in the research questions addressed in this study. While it is possible that the findings of this research are reflective of the true nature of the relationship between chronic-pain subgroup and patient-provider relationship quality, it is more likely that factors unique to this study are responsible for this finding. Problems that occurred during patient recruitment may have contributed to the finding of a lack of support for a relationship between chronic-pain subgroup and patient-provider relationship quality. This study’s population sample was supposed to be collected fiom two clinics in order to obtain a wider sample of providers. Unfortunately, only one clinic participated in data collection. Neurology of Battle Creek operates with three medical doctors and three nurses, all of whom have been trained in a patient-centered model. This introduces the possibility that the care the patients received is qualitatively different than what is typical at other clinics. It is recommended that future research, examining patient-provider relationship quality in chronic pain patients, be conducted in settings that operate under a variety of models of patient-provider interaction. Other less likely possibilities include the influence of age. Other studies have found tlmt older patients report greater levels of satisfaction with their health care than younger patients (Williams & Calnan, 1991; Zahr, Williams, & El-Hadad, 1991). The literature also suggests that race is a factor in satisfaction evaluations, with minorities reporting greater levels of dissatisfaction (LaVeist, Nickerson, & Bowie, 110 2000; Pascoe & Attkisson, 1983). The mature, White patients in this study’s sample may have responded differently than other chronic-pain patient populations. Attachment and Patient-Provider Relationship Quality Attachment theory postulates a relationship between attachment style and the quality of other relationships. This theoretical link supported the testing of patient- provider relationship quality and attachment. This study found a small, but significant, relationship with members of the secure-attachment style group reporting better patient-provider relationship quality than members of any of the insecure styles. No significant differences existed among the insecure styles. A review of the research found that attachment theory has been applied to an array of adult relationships including professor-student (Lopez, 1997), therapist-client (Pistole & Watkins, 1995), and college students in romantic relationships (Pistole, 1995). The findings of this research suggest that medical-care providers and chronic-pain patients form bonds in manners that are theoretically consistent with attachment theory. This study did not examine the amount of time patients spent with their provider. It is possible that the strength of the relationship is dependent on this and other unexamined factors. Additional research would have to be conducted prior to affirming this assumption. Limitations of the Study The ex post facto (“after the fact”) design examines the association between two or more variables. In this study, the variables measured were intrinsic to the individual, either through genetic contributions, environmental factors, or a combination. In contrast, with experimental designs in which the independent 111 variable can be controlled and subjects randomly assigned, criterion-group designs rely on the formation of discrete groups based on a variable or set of variables. Analyses of chronic-pain subgroup and attachment and patient-provider relationship quality suggest relationships among the variables, but are not proof of cause-and- efl‘ect relationships. It is not possible from this study to know whether attachment style precedes, complements, or is an antecedent of the characteristics that create the chronic-pain subgroups. This study relied on responses to specific questions on the demographic questionnaire to identify the sample population. It is possible that important factors affecting the variables under analysis were omitted. For example, neither social nor economic status was ascertained because very little has been written about them in relation to chronic pain. However, they have been studied in other chronic conditions and were found to influence risk factors and treatment outcomes (Gallo & Matthews, 2003; Whitefield, Weidner, Clark, & Anderson, 2002). Cultural context is another variable that was outside the scope of this study, which has been shown to affect the treatment of and response to chronic conditions including chronic pain (Bates, Rankin-Hill, & Sanchez-Ayendez, 1997). Attitudes about pain, along with expectations and goals for treatment, have been shown to impact patient dropout fi'om treatment (Richmond & Carmody, 1999). Again, measurement of these factors was not practical given the scope of this research. Finally, there are undoubtedly factors yet to be discovered that impact the dependent variables. The racial and gender constitution of this study was primarily White females. The disproportionate focus on the White race has been criticized in previous medical 112 research (Park, Adams, & Lynch, 1998) and is a limitation to the generalizability of these findings. Selection bias in the sample population was noted in comparison to the residents of Calhoun County, Michigan, whose demographics differed: male (48.6%), female (51.4%); racial composition: White (83.9%), Black (10.9%); Hispanic or Latino (3.2%), some other race or combination (2%) (Bureau, 2000). The nature of participant selection involved volunteers and reimbursement. Both of these factors could have influenced the selection of participants and are potential threats to generalizability. Another limitation of this study is the use of self-report instruments. This limitation is common to survey studies. In the study of attachment, in particular, debate continues regarding whether self-reported measures assess the same constructs as the Adult Attachment Interview (AAI) (George, Solomon, & Main, 1985). Some researchers purport that self-report measures of attachment involve a type of censorship because of the conscious effort involved in answering explicit questions (Jacobvitz, Curran, & Moller, 2002). A similar argument can be made regarding the completion of the Patient Reactions Assessment (Galassi et al., 1992). Even though patients were informed that their doctors and medical-care providers would not have access to their responses, the fact that patients were solicited for participation through their neurologist’s practice may have influenced their responses. It is also possible that examining patient- provider relationship quality as a multifaceted construct, instead of examining the separate components, limited the findings. 113 Lastly, it is inevitable that there will be imprecision in measuring human response. The nature of being human opens the door to error given the dynamics of conscious and unconscious factors. Rather than eliminating error, the goal of this study has been to reduce it. Recommendations for Future Research A patient’s ability to endure and function, in spite of a chronic pain condition, involves more than just the sum of their tissue damage and nociception. Social and psychological factors influence one’s conscious experience and interpretation of pain. The effect of these influences occur on a continuum from minimal to major determinant of their complaint (Portenoy & Kanner, 1996). The development of instruments, such as the Multidimensional Pain Inventory (Kems et al., 1985), emphasizes the importance of psychological and social factors in patients’ lives. Current research has moved beyond demonstrating that psychosocial factors impact the experience of pain. Today much of the research seeks to identify various combinations of factors associated with level of functioning and treatment outcome. Future research identifying contributory psychosocial factors may answer the question of why individuals with identical diagnoses respond differently to identical interventions. A multitude of intermediate research is needed to facilitate the development of treatments that are individualized to match patient needs. It is not just the limits of the research conducted in the realm of psychology that limits the understanding of chronic-pain treatment. Very little is known about the differences in physiological mechanisms activated by chronic pain or how chronic pain interacts with attachment 114 style. Given parallel findings in medical and psychological research on attachment, the development of techniques for integrating findings from biochemical, medical, social and psychological data would advance the study of attachment in chronic-pain patients. It is also necessary to understand differences relative to race and social context. Therefore, it is recommended that additional research focus on underserved and understudied populations, such as those with low socioeconomic status and those of color. Given what is known about the differences between chronic pain subgroups, treatment outcome studies comparing response differences to a variety of treatments is recommended. It is possible that, rather than waiting to understand all of the differences prior to initiating outcomes studies, conducting outcome studies may contribute to understanding differences between the subgroups. Given the psychosocial and biomedical components that vary in the subgroups, outcome studies involving multidisciplinary interventions are warranted. F onagy (1998) discusses an attachment-theory approach to the treatment of the difficult patient. He suggests that patients with insecure-attachment styles are difficult because of their inability to understand the mental processes of themselves and others. This inability sets up a vulnerability that is compounded under times of stress (e.g., poor health, chronic pain). In accordance with this refiame, the level of intervention should be aimed at helping chronic-pains patients gain circumspection on their internal working models of self and other. 115 The complete nature of attachment style is not fully understood. Kirpatrick and Hazen (1994) state that their research supports the conceptualization of attachment styles as having both trait-like characteristics and being reflective of an individual’s current relationship. They firrther state that attachment styles appear to be modifiable by “relationship-related situations” (p. 138). Extrapolating fi'om their hypothesis, it may be possible to improve an insecurely attached individual’s relationships through the therapeutic relationship or other psychological intervention. Studies that evaluate interventions for chronic-pain patients aimed at addressing attachment seem warranted in light of the findings of this study. Importance to Counseling Psychology Counseling psychology focuses on understanding the impact of psychological and environmental factors on the adjustment of individuals to stressfirl situations. As researchers and practitioners, counseling psychologists are in a unique position to help those with chronic pain cope with the adversities of their aflliction. Primary concerns of chronic-pain patients include both decreasing their perception of pain and improving their ability to firnction. The integrated training received by counseling psychologists instills the skills necessary to intervene in both areas. Research on attachment has provided empirical support for the importance of Bowlby’s tenet that participation in mutually satisfying relationships contributes to an individual’s well-being over a lifetime. The enduring nature of chronic pain and the findings of this study support firrther research on the impact of attachment in patients with chronic pain. Understanding the relationship between attachment and the 116 experience of pain has the potential of leading to customized interventions that address the unique needs of the individual and their environment. 117 Appendix DOCUMENTATION AND LETTERS 118 Provider Assistance Requested for Subject Recruitment For Chronic-Pain Study You have received this request for assistance because you have been identified as a provider who works with chronic-pain patients. This research seeks to identify psychosocial characteristics of chronic-pain patients that may increase our current understanding and lead to improved treatments. This study is being conducted by Camille A. Hood, a doctorial candidate in Counseling Psychology at Michigan State University and is supervised by Dr. Nancy Crewe, a professor in the Department of Counseling, Educational Psychology, and Special Education at Michigan State University. Enclosed are materials that will be given to potential participants should you elect to assist. Your assistance is requested to distribute the survey packets to patients who are at least 18 years old, have chronic pain existing for at least three months, and who you feel would not be adversely impacted by participating. The participants are asked to fill out a questionnaire regarding the nature of their pain, its impact in their life, and the quality of their relationships. Completion of the package takes an average of 30 minutes and participants will be compensated for their time with $10 gift certificates. Some participants may benefit from completing the survey through an increased understanding of themselves and the nature of their pain and the quality of their relationships. A small risk exists that participants may activate negative affect, concerns about their pain, or the realization that problems exist in one or more of their relationships. As a precaution, please inform potential participants of your availability to discuss any adverse reactions or answer questions that may arise as a consequence of participation or refer them to the researcher. As an additional safeguard, participants will receive a list of professional mental health referrals in their survey packets. Should you have questions please contact me at (269) 986-9844 or by e-mail carpillehpg@comgag.net. You may also contact Dr. Nancy Crewe at (517) 432-0606 or by e-mail at ncrewe@msp.@. If you are concerned about any aspect of this study, or if you have questions about the rights of participants, you may contact Peter Vasilenko, Ph.D., Chair of the University Committee on Research Involving Human Subjects (UCRIHS) at (517) 355-2180, fax: (517) 432-4503, e-mail: ucrihs@ms1&d_u, or mail: 202 Olds Hall, East Lansing, MI 48824-1047. Contact may be done anonymously, if desired. If you would like results of the study or a copy of the completed study, please contact me. Although, your participation in this study may require a small amount of your time, it is through efforts like this that knowledge that chronic pain is increasing. Thank you for your consideration of this research project. Camille A. Hood, Doctoral Student in Counseling Psychology Michigan State University 119 Recruitment Letter Sent on Office Letterhead Current date inserted here Dear Patient: Neurology of Battle Creek is participating in data collection for a chronic-pain study conducted by Camille A Hood, PhD. candidate at Michigan State University. The goal of the research is to increase the understanding of characteristics of individuals with chronic pain in order to improve treatment. Whp may ppm’cippte? The research is seeking volunteers of either gender, who are I at least 18 years old, and who have chronic pain that has existed for at least three months. What gill 1 pg pgkgg to 1199, Participants will be asked to share information about the nature of their pain and ways in which it impacts their life, including their relationships. The questionnaires take approximately 30 minutes to complete. 5' Participants will return completed questionnaire using a self-addressed, stamped envelope provided. fling; pMp; Privggy? All of the information you provide is strictly confidential. Identification numbers, rather than names, will be used on all of the surveys, and no identifying information will be used when describing our research findings. Your care provider(s) will not know about your decision to participate, and no one will have access to your responses other than the researcher. Additional information regarding precautions to protect your confidentiality is on the Informed Consent Form, which will be enclosed in the survey packet. A; a thank ypp; You will be compensated for your time. Participants who return completed questionnaires will receive a $10 gift certificate of their choice fiom Blockbuster Video, Barnes & Nobles, Schlotzsky’s, or Applebee’s Restaurant. W It is hoped you will agree to participate in this study. To have a survey packet sent to you return the enclosed post card. You may also call Camille Hood at (269-986-9844), or email her at Wet should you have questions. Sincerely, Paula Royer Office Manager Neurology of Battle Creek 120 Recruitment Advertisement Research Participants Needed for Chronic Pain Study My name is Camille Hood, and I am currently in a PhD. program at Michigan State University. As part of my program I am given the opportunity to contribute to current understandings of chronic pain through a research project. The goal of this research is to increase our understanding of characteristics of individuals with chronic pain in order to improve treatment. Egg may pgrtigipgte? I am seeking volunteers of either gender, who are at least 18 years old, and who have chronic pain that has existed for at least three months. What gill I he ppm to do? Participants will be asked to share information about the nature of their pain and ways in which it impacts their life including their relationships. The questionnaires take approximately 30 minutes to complete. We ask that you mail the completed questionnaire back to us in the stamped envelope provided. What aMut Privpg? All of the information you provide is strictly confidential. Identification numbers rather than names will be used on all of the surveys and no identifying information will be used when describing our research findings. Your primary care provider will not know about your decision to participate and no one will have access to your responses other than me. Additional information regarding precautions to protect your confidentiality is included on the Informed Consent Form. mm: You will be compensated for your time. Participants who return completed questionnaires will receive a $10 gift certificate of their choice fiom Blockbuster Video, Barnes & Nobles, Schlotzsky’s, or Applebee’s Restaurant. W I hope you will agree to participate in this study, and that it will help us learn more about treating patients with chronic pain. Please don’t hesitate to call me at (269-986-9844), or email at camillghood@cpmcast.net if you have any questions. Additional information is provided on the Informed Consent Form which you will be asked to read and sign prior to answering the survey questions. 121 Participant Recruitment Form Please read the following information if you are considering participating. My name is Camille Hood, and I am currently in a PhD. program at Michigan State University. As part of my program I am given the opportunity to contribute to current understandings of chronic pain through a research project. The goal of this research is to increase our understanding of characteristics of individuals with chronic pain in order to improve treatment. Who may pgrtigipgtg’, I am seeking volunteers of either gender, who are at least 18 years old, and who have chronic pain that has existed for at least three months. What w 1 hp psked to go? Participants will be asked to share information about the nature of their pain and ways in which it impacts their life including their relationships. The questionnaires take approximately 30 minutes to complete. We ask that you mail the completed questionnaire back to us in the stamped envelope provided. 2!th gMut Privggy", All of the information you provide is strictly confidential. Identification numbers rather than names will be used on all of the surveys and no identifying information will be used when describing our research findings. Your primary care provider will not know about your decision to participate and no one will have access to your responses other than me. Additional information regarding precautions to protect your confidentiality is included on the Informed Consent Form. a 5 thank yep: You will be compensated for your time. Participants who return completed questionnaires will receive a $10 gift certificate of their choice from Blockbuster Video, Barnes & Nobles, Schlotzsky’s, or Applebee’s Restaurant. To participate: I hope you will agree to participate in this study, and that it will help us learn more about treating patients with chronic pain Please don’t hesitate to call me at (269-986-9844), or email at camillehood@comcast.net if you have any questions. Additional information is provided on the Informed Consent Form that you will be asked to read and sign prior to answering the survey questions. 122 INFORMED CONSENT 1. This research examines individuals with chronic pain and how it affects various aspects of their life including qualities of their close relationships. It is hoped that this research may increase current understandings and contribute to improving treatment for future patients. 2. Participation is voluntary. If you choose to participate in this study, you will be asked to complete questionnaires about the nature of your pain and how it affects different aspects of your life including your relationships. The questionnaires take approximately 30 minutes. A $10 gift certificate will be provided for completed surveys. 3. Refusal to participate or withdrawal from participation will not in any way penalize you. You may have the results of the participation, to the extent that they can be identified as yours, returned to you, removed fiom the research records, or destroyed at any time prior to the end of the study. 4. This is the only form with your name and other identifying information on it. (The only reason for the mailing address is so you may receive compensation). This form will be securely stored separately from the questionnaires you complete. The questionnaires have a code number, but no identifying information. All information will be stored in a locked file and no individual data will be shared. Given the nature of some of the questions, I want to emphasize that the individuals closest to you and your health care providers will pg; see your responses. Your privacy will be protected to the maximum extent allowable by law. 5. No discomforts or stresses are foreseen. Some participants may benefit fiom participating in this study through increased understanding of themselves, the nature of their pain, and the quality of their relationships. A small risk exists that participation in this study may activate negative feelings, concerns about your pain, or the realization that problems exist in one or more of your relationships. Your care provider and the researchers are aware of the nature of this study and are available to discuss concerns or reactions you have as a result of participation. A list of mental health resources is provided in the survey packet. If you have questions or concerns, now or during the course of the project, you are invited to contact any of the following individuals: (a) Camille A Hood, Secondary Investigator, phone (269) 986-9844, email: camillehood@comcast.n§t_, or (b) Dr. Nancy Crewe, Primary Investigator and Dissertation Chair, (517) 432-0606, email: ngewe@msu.gu. Research at Michigan State University that involves human participants is overseen by the University Committee on Research Involving Human Subjects (UCRIHS). If you have any questions or concerns regarding your rights as a study participant, or are dissatisfied at any time with any aspect of this study, you may contact—anonymously, if you wish--Peter Vasilenko, Ph.D., Chair, 202 Olds Hall, Michigan State University, East Lansing, MI 48824- 1047, PHONE (517) 355-2180, FAX (517) 432-4503, E-Mail - UQRIH§@msu.ed_u 123 If you agree to participate in the research please sign below and return this form with your survey packet in the enclosed stamped envelope. Signature: Date: Name (Please print clearly) Street or PO: City, State: , zip code Phone: May we contact you by phone? YES NO Signature of Investigator Date 124 Gift Certificate Form As compensation for my time, I am interested in receiving a gift $10 gift certificate to (Circle only one): The gift certificate will be mailed within a week of receiving your Qmpletfl Survey Packet. Don’t forget to provide your address including your zip code. Please contact the researcher if you have questions or have not received your gift certificate within two weeks. Blockbuster Video Schlotzsky’ 5 Restaurant Barnes & Noble Applebee’s Restaurant 125 Mental Health Referral List Referral List Gryphon Place (A 24-hour crisis and referral service) 1104 South Westnedge Kalamazoo, MI 49008 (269) 381-1510 or (269) 381-4357 Services include suicide prevention, substance abuse assistance, and counseling and mental health treatment, dispute resolution services, assessment and referral. Borgess Mental Health Services 1521 Gull Road Kalamazoo, MI 49048 (269) 226-8000 Emergency (269) 226-5793 Adult Partial Hospitalization Kalamazoo Community Mental Health Services 3299 Gull Road Kalamazoo, MI 49048 (269) 553-8000 TDD (269) 382-0847 Services: adults, children with severe emotional disturbances, persons with developmental disabilities, school-based prevention services Summit Pointe (24-Hour Crisis Services) 140 W. Michigan Ave. Battle Creek, MI 49017 (269) 996-1460 Services include: individual, family and group counseling, supportive living, psychological testing, case management. Website: www.summitpointe.com Battle Creek Health System; Dept. of Psychiatry and Behavioral Health Services (24—hour assessment and evaluation) Fieldstone Center 165 North Washington Avenue Battle Creek, MI 49017 (269) 964-7121 Services include: crisis intervention, child/adolescent/adult inpatient programs 126 Patient Information Questionnaire Please answer the following questions either by filling in the blank or circling the most appropriate answer. Gmeral Informatjpn; Age: Gender: (Circle one) Male or Female Race or Ethnicity: Highest Level of Education Completed: (fill in) Marital Status: Single Married Divorced Widowed Rate Marriage or Relationship Satisfaction with 10 being the best: (0-10) Employment: None 1-31 hours/week 32 or more hours/week History of Chronic Pain How long have you experienced this pain: Circle: Years or Months What Caused Original Pain: Pain-related Diagnosis: Location of Pain: Type of Pain: Intermittent Constant Other: Describe (i.e., Sharp, Burning, Aching, etc.) Pain Intensity: Circle the number that best describes your average level of pain this week 0 1 2 3 4 5 6 7 8 9 10 No Pain Worst Pain 127 Psychiatric Histog Depression Yes No Bipolar Disorder Yes No Anxiety Yes No Schizophrenia Yes No Substance Abuse Yes No Drug(s): Substance Dependency Yes No Drug(s): Alcohol Abuse Yes No Alcohol Dependency Yes No Other Please use the back of this form if you have additional comments. 128 West Haven-Yale Multidimensional Pain Inventory BEFORE YOU BEGIN, PLEASE ANSWER 2 PRE-EVALUATION QUESTIONS BELOW: 1. Some of the questions in this questionnaire refer to your “significant other.” A significant other is a person with whom you feel closest. This includes anyone that you relate to on a regular or infrequent basis. It is very important that you identify someone as your “significant other.” Please indicate below who your significant other is (check one): E1 Spouse DPartner/ Companion [Housemate/Roommate DFriend CINeighbor DParent/Child/Other relative UOther (please describe): 2. Do you currently live with this person DYES EINO When you answer questions in the following pages about “your significant other,” always respond in reference to the specific person you just indicated above. A In the following 20 questions, you will be asked to describe your pain and how it afl‘ects your life. Under each question is a scale to record your answer. Read each question carefully and then 93;le a number on the scale under that question to indicate how that specific question applies to you. 1. Rate the level of your pain at the present moment. 112L3 4 I 5 16 Nopain Very intense pain 2. In general, how much does your pain problem interfere with your day to day activities? 11213 41516 No interference Extreme interference 3. Since the time you developed a pain problem, how much has your pain changed our ability to work? 1 I 2 l 3 4 I 5 1 6 No change Extreme change __ Check here, if you have retired for reasons other than your pain problem. 129 4. How much has your pain changed the amount of satisfaction or enjoyment you et from participating in social and recreational activities? 1 I 2 I 3 4 I 5 T 6 No change Extreme change 5. How supportive or helpful is your spouse (significant other) to you in relation to our pain? 1 l 2 | 3 4 l 5 J 6 Not at all supportive Extremely supportive 6. Rate your overall mood during the g: week. 1 1 2 1 3 4 l 5 ] 6 Extremely low mood Extremely high mood 7. On the average, how severe has your pain been during the last week? 11213 4 I Ti 6 Not at all severe l I Extremely severe 8. How much has your pain changed your ability to participate in recreational and other social activities? 1 J 2 1 3 4 1 5 J 6 No change Extreme change 9. How much has your pain changed the amount of satisfaction you get from family- related activities? 1 [ 2 I 3 4 I 5 1 6 No change Extreme change 10. How worried is your spouse (significant other) about you in relation to your pain ‘ roblem? 1 I 2 l 3 l 4 | 5 ] 6 Not at all worried L Extremely worried 11. During the peg week, how much control do you feel that you have had over your life? 1 [ 2 1 3 4 L 5 l 6 Not at all in control 1 Extremely in control 12. How much suffering do 1 Iou experience because of your pain? 1 1 2 3 4 l 5 l 6 No suffering Extreme suffering 13. How much has your pain changed your marriage and other family relationships? 1 l 2 3 4 I 5 j 6 No change - Extreme change 130 14. How much has your pain changed the amount of satisfaction or enjoyment you get from work? 1 I 2 I 3 4 I 5 I 6 No change Extreme change __Check here, if you are not presently working 15. How attentive is your spouse (significant other) to your pain problem? 1 I 2 l 3 4 I 5 I 6 Not at all attentive Extremely attentive 16. During the pg week, how much do you feel that you’ve been able to deal with our problems? 1 I 2 I 3 I 4 I 5 I 6 Not at all 7 Extremely well 17. How much has your pain changed your ability to do household chores? 1 I 2 I 3 4 I 5 I 6 No change Extreme change 18. During the past wee; how irritable have you been? 1 I 2 I 3 4 I 5 I 6 Not at all irritable Extremely irritable 19. How much has your pain changed your fiiendships with people other than your family? 1 I 2 I 3 4 I 5 I 6 No change Extreme change 20. During the pg weeg how tense or anxious have you been? 1 I 2 I 3 4 I 5 I 6 Not at all tense or anxious Extremely tense or anxious 131 B. In this section, we are interested in knowing how your significant other (this refers to the person you indicated above) responds to you when he or she knows that you are in pain. On the scale listed below each question, circle a number to indicate how often your significant other generally responds to you in that particular way when you are in in. l. Ignores me 1 | 2 I 3 I 6 Never Very often 2. Asks me what he/she can do to help. 1 I 2 I 3 I 6 Never ’ Very often 3. Reads to me. 1 I 2 I 3 F 6 Never Vgry often 4. Eyresses irritation at me. 1 I 2 I 3 I 6 Never Very often 5. Takes over my jobs or duties. 1 I 2 I 3 I 6 Never Very often 6. Talks to me about something else to take 13y mind off the pain. 1 I 2 I 3 I 6 Never Very often 7. Expresses fi'ustration at me. 1 I 2 I 3 I 6 Never Veg often 8. Tries ttLget me to rest. 1I2I3 I 6 Never Very often 9. Tries to involve me in some activity 1 I 2 I 3 I 6 Never Very often 132 10. Expresses anger at me. 1 I 2 I 3 4 6 Never Very often 11. Gets me some pain medications. 1 I 2 I 3 4 6 Never Very often 12. Encourages me to work on a hobby. 1 I 2 I 3 4 6 Never Very often 13. Gets me somethingto eat or drink. 1 I 2 I 3 4 6 Never Veg often 14. Turns on the TV. to take my mind off my pain. 1 I 2 I 3 4 6 Never Veg often C Listed below are 18 common daily activities. Please indicate how pften you do each of these activities by circlipg a number on the scale listed below each activity. Please complete afl 18 questions. 1. Wash dishes 1 I 2 I 3 4 6 Never Veg often 2. Mow the lawn. 1 I 2 I 3 I 4 6 Never Very often 3. Go out to eat. 1 I 2 I 3 4 6 Never Very often 4. Play cards or other games. 1 I 2 I 3 4 6 Never Veg often 133 5. Go grocery shopping. 1 2 3 6 Never Veg often 6. Work in the garden. 1 I 2 3 6 Never Very often 7. Go to a movie. 1 I 2 3 6 Never Very often 8. Visit fiiends. 1 I 2 3 6 Never Veg often 9. Help with the house cleaning. 1 I 2 3 6 Never Very often 10. Work on the car. 1 I 2 3 6 Never Very often 11. Take a ride in a car. 1 I 2 3 6 Never Very often 12. Visit relatives. 1 I 2 3 6 Never Very often 13. Prepare a meal. 1 I 2 3 6 Never Very often 14. Wash the car. 1 I 2 3 6 Never Veg often 15. Take a trip. 1 I 2 3 6 Never Very often 134 16. Go to a park or beach. 1 I 2 I 3 4 I 5 I 6 Never Veg often 17. Do a load of laundry. 1 J 2 I 3 4 I 5 I 6 Never Very often 18. Work on a needed house repair. 1 I 2 I 3 4 I 5 I 6 Never Very often 135 Patient Reactions Assessment In filling out this survey please think about your recent contact with the medical professional (a specific attending physician, nurse, resident, or physician assistant) who primarily treats your pain. Circle the number (1 — 7) that best describes how you felt about your recent contact with that person. 1 2 3 4 5 6 7 Very Strongly Disagree Neither Agree Strongly Very Strongly Disagree Agree or Agree Strongly Disagree Disagree Agree 1. I have an understandingof the treatment side effects. I 1 I 2 I 3 I 4 I 5 I 6 I 7 I 2. It is difficult for me to get conflicting information straightened out. I 1 I 2 I 3 I 4 I 5 I 6 L 7 I 3. My health care provider is warm and caring toward me. I 1 I 2 I 3 I 4 I 5 I 6 I 7 I 4. It is difficult for me to ask about something I don’t understand. I l I 2 I 3 I 4 I 5 I 6 I 7 I 5. My health care provider told me what the treatment would do. I 1 I 2 I 3 I 4 I 5 I 6 I 7 I 6. My provider makes me feel comfortable discussing personal issues. I 1 I 2 I 3 I 4 I 5 I 6 I 7 I 7. It is hard for me to tell my health care provider about new symptoms. I 1 I 2 I 3 I 4 I 5 I 6 I 7 I 8. It is hard for me to ask about how treatment is going. I 1 I 2 I I 3 I 4 I 5 I 6 I 7 I 9. My health cmgprovider really respects me. I 1 I 2 I 3 I 4 I 5 I 6 I 7 I 10. I understand my medical treatment plan. I 1 I 2 I 3 I 4 I 5 I 6 I 7 I 11. I have a good idea about the changes to expect in my health. I 1 I 2 I 3 I 4 I 5 I 6 I 7 J 136 12. Sometimes I feel insulted when talking to my health care provider. I 1 I 2 I 3 I 4 I 5 I 6 I 7 13. It is difficult for me to ask my health care provider questions. I l I 2 7 3 T 4 I 5 I 6 I 14. Treatment procedures are clearly eglained to me. I 1 I 2 L 3 I 4 I 5 I 6 I 15. My health care provider doesn’t seem interested in me as a person. I 1 I 2 I 3 I 4 I 5 I 6 I 7 Please answers to the following questions. 16. How many visits have you had with your care provider? 17. Other types of contact: Type: Number of times: 137 Experiences in Close Relationships Instructions: The following statements concern how you feel in close relationships. We are interested in how you generally experience relationships, not just in what is happening in a current relationship. Respond to each statement by indicating how much you agree or disagree with it. Write the number in the space provided, using the following scale: Disagree Strongly I Neutral/Mixed I Agree Strongly 1 I 2 I 3 I 4 I 5 I 6 I 7 1. __ I prefer not to show a partner how I feel deep down. 2. _ I worry about being abandoned. 3. _ I am very comfortable being close to my partners. 4. __ I worry a lot about my relationships. 5. _ Just when my partner starts to get close to me I find myself pulling away. 6. __ I worry that my partners won’t care about me as much as I care about them. 7. _I get uncomfortable when my partner wants to be very close. 8. __ I worry a fair amount about losing my partner. 9. __ I don’t feel comfortable opening up to close partners. 10. __I often wish that my partner’s feelings for me were as strong as my feelings for him/her. 11. __ I want to get close to my partner, but I keep pulling back. 12. __ I often want to merge completely with close partners, and this sometimes scares them away. 13. _ I am nervous when partners get too close to me. 14. _ I worry about being alone. 15. _I feel comfortable sharing my private thoughts and feelings with my partner. 16. __ My desire to be very close sometimes scares people away. 17. _ I try to avoid getting to close to my partner. 18. __ I need a lot of reassurance that I am loved by my partner. 19. _ I find it relatively easy to get close to my partner. 138 Disagree Strongly I Neutral/Mixed I Agree Strongly 1I 2I I 4 I 5 I 6 7 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. __ Sometimes I feel that I force my partners to show more feeling, more commitment. __ I find it difiicult to allow myself to depend on close partners. _ I do not often worry about being abandoned. __ I prefer not to be too close to my partners. __ IfI can’t get my partner to show interest in me, I get upset or angry. __ I tell my partner just about everything. _ I find that my partner(s) don’t want to get as close as I would like. _ I usually discuss my problems and concerns with my partner. _ When I’m not involved in a relationship, I feel somewhat anxious and insecure. _ I feel comfortable depending on close partners. _ I get frustrated when my partner is not around as much as I would like. __ I don’t mind asking close partners for comfort, advice, or help. _ I get fi'ustrated if close partners are not available when I need them. __ It helps to turn to my close partner in times of need. _ When close partners disapprove of me, I feel really bad about myself. __ I turn to my partner for many things, including comfort and reassurance. _ I resent it when my partner spends time away from me. 139 Debriefing / Thank You Letter Accompanying Payment Date: Dear Thank you for participating in this research study. Although participation in this study required some of your time, it is through efforts like yours that knowledgement about chronic pain is increased. Enclosed you will frnd your $10 gift certificate. Most individuals reported that this survey was interesting or that it stimulated thinking about their health, situation, or relationships. There is a very small chance that participating in this survey brought up issues or concerns that you would like to discuss firrther. Ifyou are concerned about your medical condition and feel comfortable talking to your care provider, he or she should be the first person you contact. Ifyour concerns are more appropriately addressed by a mental health care provider, you should contact your current mental health professional. If you are not currently in treatment or are not sure to whom to contact, please refer to the mental health resource list enclosed in the survey packet. (This list is not exhaustive and you may prefer to receive services fi'om a provider recommended by someone you trust) If you have questions or concerns you are invited to contact any of the following individuals: (a) Camille A Hood, Secondary Investigator, phone (269) 986-9844, email: millehood@comcast.net, (b) Dr. Nancy Crewe, Primary Investigator and Dissertation Chair, (517)432-0606, email: ncrewe@msu.ed_u, (0) University Committee on Research Involving Human Subjects (UCRIHS), Peter Vasilneko, Ph.D., Chair, 202 Olds Hall, Michigan State University, East Lansing, MI 48824- 1047, PHONE (517) 355-2180, FAX (517) 432-4503, E-Mail - UCRIHS@msu.Qu. Again, thank you for your time and interest. Sincerely, Camille A. Hood PhD. Candidate Michigan State University Enclosure: $10 gift certificate 140 REFERENCES 141 REFERENCES Abram, H. S. (1969). Psychological responses to illness. Psychosomatics, 10(10), 218-224. Adler, R, & Grota, L. J. (1969). Effects of early experience on adrenocortical reactivity. Physiology card Behavior, 4, 303 -305. Aldrich, J. H., & Nelson, F. D. (1984). Linear probability, Iogit, and probit models. Beverly Hills, CA: Sage. Anonymous. (2000). Open physician-patient dialogue paves the way for effective treatment of depression. Geriatrics, 55(7), 27. Armsden, G. C., & Greenberg, M. T. (1987). The Inventory of Parent and Peer Attachment: Individual differences and their relationship to psychological well-being in adolescence. Journal of Youth and Adolescence, I 6(5), 427-454. Arnold, L. M., Witzeman, K. A, Swank, M. L., McElroy, S. L., & Keck Jr., P. E. (2000). Health-related quality of life using the SF -36 in patients with bipolar disorder compared with patients with chronic back pain and the general population. Journal of Aflective Disorders, 5 7(23 5-23 9). Aronoff, G. M. (2000). Pain medicine: hope is a powerful analgesic. Geriatrics, 55(9), 38-39. Bacon, N. M., Bacon, S. R, Atkinson, J. H., Slater, M. A, Patterson, T. L., Grant, I., et al. (1994). Somatization symptoms in chronic low back pain patients. Psychosomatic Medicine, 56(2), 118-127. Balint, M. (1957). Ihe Doctor, the Patient and the Illness. New York: International University Press. Bartholomew, K. (1997). Adult attachment processes: Individual and couple perspectives. British Journal of Medical Psychology, 70, 249-263. Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61, 226-244. 142 Basler, S. C., Grzesiak, R C., & Dworkin, R. H. (2002). Integrating relational psychodynamics and action-oriented psychotherapies: Treating pain and suffering. In D. C. Turk & R. J. Gatchel (Eds), Psychological Approaches to Pain Management: A Practitioner's Handbook (Second ed.). New York: Guilford Press. Bates, M. S., Rankin-Hill, L., & Sanchez-Ayendez, M. (1997). The effects of the cultural context of health care on treatment of and response to chronic pain and illness. Social Science and Medicine, 45(9), 1433-1447. Becker, M. H., & Maiman, L. A. (1980). Strategies for enhancing patient compliance. Journal of Community Health, 6(2), 113 - 135. Bell, 1. R., Caspi, 0., Schwartz, G. E., Grant, K. L., Gaudet, T. W., Rychener, D., et al. (2002). Integrative medicine and systemic outcomes research: Issues in the emergence of a new model for primary health care. Achieves of Internal Medicine, 162(2), 133-140. Bendelow, G. A, & Williams, S. J. (1996). The end of the road? Lay views on a pain-relief clinic. Social Science and Medicine, 43(7), 1127-1136. Bennett, R. M., Burckhardt, C. S., Clark, S. R., O'Reilly, C. A, Wiens, A N., & Campbell, S. M. (1996). Group treatment of fibromyalgia: a 6 month outpatient program. Journal of Rheumatology, 23, 521-528. Bergstroem, G., Bodin, L., Jensen, I. B., Linton, S. J., & Nygren, A L. (2001). Long- terrn, non-specific spinal pain: Reliable and valid subgroups of patients. Behavior Research and Therapy, 39(1), 75-87. Bergstrom, G., Jensen, 1. B., Bodin, L., Linton, S. J., & Nygren, A L. (2001). The impact of psychologically different patient groups on outcome after a vocational rehabilitation program for long-term spinal pain patients. Pain, 93, 229-237. Bernstein, I. H., Jaremko, M. E., & Hinkley, B. S. (1995). On the utility of the West Haven-Yale Multidimensional Pain Inventory. Spine, 20, 956-963. Bowlby, J. (1969). Attachment and Loss: Attachment (Vol. 1). New York: Basic Books. Bowlby, J. (1973). Attachment and Loss: Separation: anxiety and anger (Vol. 2). New York: Basic Books. Bowlby, J. (1988). A secure base: parent-child attachment and healthy hummr development. New York: Basic Books. 143 Brennan, K. A, Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment. In J. A. Simpson & W. S. Rholes (Eds), Attachment Theory and Close Relationships (pp. 46-76). New York: Guilford Press. Brennan, K. A, & Shaver, P. R. (1995). Dimensions of adult attachment, affect regulation, and romantic relationship functioning. Personality and Social Psychology Bulletin, 21, 267-283. Bretherton, I. (1992). The origins of attachment theory: John Bowlby and Mary Ainsworth. Developmental Psychology, 28, 759-775. Buller, M. K., & Buller, D. B. (1987). Physicians' communication style and patient satisfaction. Journal of Health and Social Behavior, 28, 375-388. Bureau, U. S. C. (2000). Fast facts for congress: Calhoun County, Michigan - Fact Sheet. Retrieved March 23, 2005, fi'om http:/[fastfactscensusgov/mlet/CWSFM Burns, J. W., Kubilus, A, Bruchl, S., & Harden, R N. (2001). A fourth empirically derived cluster of chronic pain patients based on the Multidimensional Pain Inventory: Evidence for repression within the dysfunctional group. Journal of Consulting and Clinical Psychology, 69(4), 663-673. Camelley, K., Pietromonaco, P., & Jaffe, K. (1994). Depression, working models of others, and relationship functioning. Journal of Personality and Social Psychology, 66, 127-140. Carver, C. S. (1994). Personality and adult attachmentUnpublished manuscript, University of Miami, Coral Gables, FL. Cassileth, B. R, Zupkis, R. V., & Sutton-Smith, K. (1980). Information and participation preferences among cancer patients. Annals of Internal Medicine, 92, 83 2-83 6. Chaitckil, S., Kreitler, S., Shaked, S., Schwartz, 1., & Rosin, R (1992). Doctor-patient communication in a cancer ward. Journal of Cancer Education, 7, 41-54. Chambless, D. L., & Hollen, S. D. (1998). Defining empirically supported therapies. Journal of Consulting and Clinical Psychology, 66, 7 -18. Ciechanowski, P. S., Katon, W. J., Russo, J. E., & Dwight-Johnson, M. M. (2002). Association of attachment style to lifetime medically unexplained symptoms in patients with hepatitis C. Psychosomatics, 43(3), 206-212. 144 Ciechanowski, P. S., Katon, W. J ., Russo, J. E., & Walker, E. A. (2001). The patient- provider relationship: Attachment theory and adherence to treatment in diabetes. American Journal of Psychiatry, 158(1), 29-35. Ciechanowski, P. S., Walker, E. A, Katon, W. J., & Russo, J. E. (2002). Attachment theory: A model for health care utilization and somatization. Psychosomatic Medicine, 64, 660-667. Clark, J. D. (2002). Chronic pain prevalence and analgesic prescribing in a general medical population. Journal of Pain and Symptom Management, 23(2), 131- 137. Cline, R. J. W. (2003). At the intersection of micro and macro: opportunities and challenges for physician-patient communication research. Patient Education and Counseling, 50, 13-16. Cohen, G. (1996). Age and health status in a patient satisfaction survey. Social Science and Medicine, 42(7), 1085-1093. Collins, N. L., & Read, S. J. (1990). Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology, 58, 644-663. Compas, B. E., Haaga, D. A. F ., Keefe, F. J ., Leitenberg, H., & Williams, D. A. (1998). Sampling of empirically supported psychological treatments from health psychology: Smoking, chronic pain, cancer, and bulimia nervosa. Journal of Consulting and Clinical Psychology, 66, 89-112. Crook, J., Weir, R., & Turk, D. C. (1989). An epidemiological follow-up survey of persistent pain suffers in a group family practice and specialty pain clinic. Pain, 36, 49-61. Cruz, M., & Pincus, H. A. (2002). Research on the influence that communication in psychiatric encounters has on treatment. Retrieved May 15, 2003, from ht_tp://psychservices.psychiagonlineorg. Davis, M. (1968). Physiological, psychological and demographic factors in patient compliance with doctor's orders. Medical Care, 6, 115-122. Deale, A, & Wessely, S. (2001). Patients' perceptions of medical care in chronic fatigue syndrome. Social Science and Medicine, 52(12), 1859 - 1864. 145 Derogatis, L. R., & Spencer, P. M. (1982). The Brief Symptom Inventory (BS1): Administration, Scoring, and Procedures Manual-II. Baltimore: Johns Hopkins University School of Medicine, Clinical Psychometrics Research Unit. DiMatteo, M. R. (1979). A social psychological analysis of physician-patient rapport: toward a science of the art of medicine. Journal of Social Issues, 35 , 12-33. DiMatteo, M. R., & DiNicola, C. (1978). Sources of assessment of the physician performance: a study of comparative reliability and patterns of intercorrelations. Medical Care, 16, 829-839. Dorfman-Botens, D. (1994). Adult attachment: Individual diflerences in defensive strategies and implications for physical health. Unpublished Dissertation, State U New York, Buffalo, NY. Dozier, M. (1990). Attachment organization and treatment use for adults with serious psychopathological disorders. Development and Psychopathology, 2, 47-60. Egeland, B., Jacobvitz, D. B., & Sroufe, A. (1988). Breaking the cycle of abuse. Child Development, 59, 1080-1088. Engel, G. L. (1977). The need for a new medical model: A challenge for biomedicine. Science, 1961, 129-136. Engel, G. L. (1988). How much longer must medicine be bound by a seventeenth century world view? In K. White (Ed), The Task of Medicine. Menlo l'ark, CA: The Henry J. Kaiser Family Foundation. Etscheidt, M. A., Steger, H. G., & Braverman, B. (1995). Multidimensional Pain Inventory profile classifications and psychopathology. Journal of Clinical Psychology, 51(1), 29-36. Ettner, S. L. (1999). The relationship between continuity of care and the health behaviors of patients: Does having a usual physician make a difference? Medical Care, 37(6), 547-555. Evers, A. W. M., Kraaimaat, F. W., van Riel, P. L. C. M., & de Jong, A. J. L. (2002). Tailored cognitive-behavioral therapy in early rheumatoid arthritis for patients at risk: a randomized controlled trial. Pain, 100, 141-153. F alvo, D., Woehlke, P., & Deichmann, J. (1980). Relationship of physician behavior to patient compliance. Patient Counseling and Health Education, Fourth Quarter, 185-187. 146 Feeney, J. A. (2000). Implications of attachment style for patterns of health and illness. Child: Care, Health & Development, 26(4), 277-288. F eeney, J. A., & Noller, P. (1990). Attachment style as a predictor of adult romantic relationships. Journal of Personality and Social Psychology, 58, 281-291. Feeney, J. A., & Noller, P. (1991). Attachment style and verbal descriptions of romantic partners. Journal of Social and Personal Relationships, 8, 187-215. Feeney, J. A., Noller, P., & Hanrahan, M. (1994). Assessing adult attachment. In M. B. Sperling & W. H. Berman (Eds.), Attachment in adults: Clinical and developmental perspectives (pp. 128-152). New York: Guilford Press. Feeney, J. A., & Ryan, S. M. (1994). Attachment style and affect regulation: Relationships with health behavior and family experiences of illness in a student sample. Health Psychology, 13(4), 334-345. Fonagy, P. (1998). An attachment theory approach to treatment of the difficult patient. Bulletin of the Menninger Clinic, 62(6), 147-169. Fordyce, W. E., Bigos, S. J ., Batti'e, M. C., & Fisher, L. D. (1992). MMPI scale 3 as a predictor of back injury report: What does it tell us? The Clinical Journal of Pain, 8, 222-226. Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology, 78, 350-365. Freeman, B., Goldman, V. F., & Cecil, R. R. (1980). Gaps in doctor-patient communication: doctor-patient interaction analysis. Pediatric Research, 5, 298-311. F uendeling, J. M. (1998). Adult attachment and affect regulation: a test of a stylistic model. Michigan State University, East Lansing, MI. Galassi, J. P., Schanber, R., & Ware, W. B. (1992). The Patient Reactions Assessment: a brief measure of the quality of the patient-provider medical relationship. Psychological Assessment, 4(3), 346-351. Gallo, L. C., & Matthews, K. A. (2003). Understanding the association between socioeconomic status and physical health: Do negative emotions play a role? Psychological Bulletin, 129(1 ), 10-5 1 . Garralda, M. E. (1992). A selective review of child psychiatric syndromes with a somatic presentation. British Journal of Psychiatry, 161, 759-773. 147 Gatchel, R. J ., Polatin, P. B., Noe, C., Gardea, M., Pulliam, C., & Thompson, J. (2003). Treatment- and cost effectiveness of early intervention for acute low- back pain patients: a one-year prospective study. Journal of Occupational Rehabilitation, 13(1), 1-9. Geisser, M. E., Pema, R., Kirsch, N. L., & Bachman, J. E. (1998). Classification of chronic pain patients with the Brief Symptom Inventory: Patient characteristics of cluster profiles. Rehabilitation Psychology, 43(4), 313-326. George, C., Solomon, J ., & Main, M. (1985). The adult attachment interview.Unpublished manuscript, University of California, Berkeley. Gerlsma, C., & Luteijn, F. (2000). Attachment style in the context of clinical and health psychology: a proposal for the assessment of valence, incongruence, and accessibility of attachment representations in various working models. British Journal of Medical Psychology, 73(1), 15-34. Goldberg, P. E. (2000). The physician-patient relationship: Three psychodynamic concepts that can be applied to primary care. Achieves of Family Medicine, 9, 1 164-1 168. Golden, J. S., & Johnston, G. D. (1970). Problems of distortion in doctor-patient communication. Psychiatry in Medicine, I, 127-149. Goldring, A. B., Taylor, S. E., Kemeny, M. E., & Anton, P. A. (2002). Impacts of health beliefs, quality of life, and the physician-patient relationship on the treatment intentions of inflammatory bowl disease patients. Health Psychology, 21(3), 219-228. Grandinetti, D. (1998). Turning no into yes: How to motivate the reluctant patient. Medical Economics, 75(1 1), 107-1 1 1. Greenley, J. R., Young, T. B., & Schoenherr, R. A. (1982). Psychological distress and patient satisfaction. Medical Care, XX(4), 373-383. Griffin, D. W., & Bartholomew, K. (1994). The metaphysics of measurement: The ease of adult attachment. Advances in Personal Relationships, 5, 15-52. Grimm, L. G., & Yarnold, P. R. (1995). Introduction to multivariate statistics. In L. G. Grimm & P. R. Yamold (Eds), Reading and Understanding Multivariate Statistics (pp. 1-18). Washington, D. C.: American Psychological Association. Hahn, S. R., Thompson, K. 8., Wills, T. A., Stern, V., & Budner, N. S. (1994). The difficult doctor-patient relationship: Somatization, personality and psychopathology. Journal of Clinical Epidemiology, 4 7(6), 647 - 657. 148 Hall, J. A., Horgan, T. G., Stein, T. S., & Roter, D. L. (2002). Liking in the physician- patient relationship. Patient Education and Counseling, 48, 69-77. Hall, M. A., Dugan, E., Zheng, B., & Mishra, A. K. (2001). Trust in physicians and medical institutions: What is it, can it be measured, and does it matter? The Milbank Quarterly, 79, 613-639. Harstall, C., & Ospina, M. (2003). How prevalent is chronic pain? Pain Clinical Updates, X1(2), 1-4. Hausman, A. (2001). Taking your medicine: Relational steps to improving patient compliance. Health Marketing Quarterly, 19(2), 49-71. Hazan, C., & Shaver, P. (1987). Conceptualizing romantic love as an attachment process. Journal of PersOnality and Social Psychology, 52, 51 1-524. Hazan, C., & Shaver, P. (1990). Love and work: An attachment-theoretical perspective. Journal of Personality and Social Psychology, 5 9, 270-280. Health, B. A. (1997). Compliance: Big Problem, Bigger Opportunity. 43. Hellstrom, C., Jansson, B., & Carlsson, S. G. (2000). Perceived future in chronic pain: the relationship between outlook on future and empirically derived psychological patient profiles. European Journal of Pain, 4, 283-290. Helme, R. D. (2001). Chronic pain management in older people. European Journal of Pain, 5(Supp1. A), 31-36. Hemesath, K. K. (1997). Relationships between attachment styles, perceived family expressed emotion and somatization in primary care patients. Kansas State University, Kansas City, MO. Heszen-Klemens, 1., & Lapinska, E. (1984). Doctor-patient interaction, patients' ' health behavior and effects of treatment. Social Science and Medicine, 19(1), 9-1 8. Hill, R. D., Rigdon, M., & Johnson, S. (1993). Behavioral smoking cessation treatment for older chronic smokers. Behavior Therapy, 24(2), 321-329. (Hindy, C. G., Schwartz, J. C., & Brodsky, A. (1989). If this is love, why do I feel so insecure? New York: Atlantic Monthly Press. Hoffer, M. A. (1994). Hidden regulators in attachment, separation, and loss. Monographs for the Society for Research in Child Development, 5 9, 192-207. 149 Horvath, A., & Symonds, B. (1991). Relationship between the working alliance and outcome in psychotherapy: a meta-analysis. Journal of Counseling Psychology, 38, 139-149. Hotopf, M., Mayou, R., & Wadsworth, M. (1999). Childhood risk factors for adults with medically unexplained symptoms: results from a national birth cohort study. American Journal of Psychiatry, I 56, 1796-1800. Jacobvitz, D., Curran, M., & Moller, N. (2002). Measurement of adult attachment: The place of self-report and interview methodologies. Attachment and Human Development, 4, 207-215. Jaffe, J. (1958). Language of the dyad: A method of interaction analysis in psychiatric interviews. Psychiatry, 21 (3), 249-258. Jamison, R. N. (1996). Learning to Master Your Chronic Pain (V 01. 2). Sarasota, FL: Professional Research Press. Jamison, R. N., Rock, D. L., & Parris, W. C. W. (1988). Empirically derived Symptom Checklist-9O subgroups of chronic pain patients: a cluster analysis. Journal of Behavioral Medicine, I I, 147-157. Jamison, R. N., Rudy, T. E., Penzien, D. B., & Mosley Jr., T. H. (1994). Cognitive- behavioral classifications of chronic pain: replication and extension of empirically derived patient profiles. Pain, 5 7, 277-292. Jarvik, J. J ., Hollingworth, W., Heagerty, P., R., H. D., & Deyo, R. A. (2001). The Longitudinal assessment of imaging and disability of the back (LAIDBack) study: Baseline data. Spine, 26(10), 1158-1166. Jensen, M. B., Turner, J. A., & Romano, J. M. (1994). Correlates of improvement in multidisciplinary treatment of chronic pain. Journal of Consulting and Clinical Psychology, 62, 172-179. Johansson, C., Dahl, J ., Jannert, M., Melin, L., & Andersson, G. (1998). Effects of a cognitive-behavioral pain-management program. Behavior Research and Therapy, 36, 915-930. Johnson, M. (1994). Current trends in health psychology. The Psychologist, 7, 114- 118. Kaati, G., Vester, M., Sjostrom, M., & Bygren, L. O. (2001). Patient attachment patterns and response to a program for alleviating cardiovascular risk factors: a 1-year follow-up of patients in a patient education program. Patient Education and Counseling, 44, 235-241. 150 Kalb, C. (2003, May 19). Taking a new look at pain. Newsweek, 45-52. Kaplin, S. 11., Greenfield, S., & Ware, J. E. (1989). Assessing the effects of physician-patient interactions on the outcomes of chronic disease. Medical Care, 27(3), 8110-8127. Kems, R. D., & Rosenberg, R. (1995). Pain-relevant responses from significant others: development of a significant-other version of the WHYMPI scales. Pain, 61, 245-249. Kems, R. D., Turk, D. C., & Rudy, T. E. (1985). The West Haven-Yale Multidimensional Pain Inventory (WHYMPI). Pain, 23, 345-346. Kirkpatrick, L. A., & Hazan, C. (1994). Attachment styles and close relationships: A four-year prospective study. Personal Relationships, 1, 123-142. Kleinke, C. L. (1994). MMPI scales as predictors of pain-coping strategies preferred by patients with chronic pain. Rehabilitation Psychology, 39(2), 123-129. Kolb, L. C. (1982). Attachment behavior and pain complaints. Psychosomatics, 23(4), 413-425. Korman, M., Stubblefield, R. L., & Martin, L. W. (1964). Faculty and student perceptions of medical roles. Journal of Medical Education, 39, 197-202. Kotler, T., Buzwell, 8., Romeo, Y., & Bowland, J. (1994). Avoidant attachment as a risk factor for health. British Journal of Medical Psychology, 67, 237-245. Kraytman, M. (1979). The complete patient history. New York: McGraw-Hill. Kroenke, K., & Mangelsdorfi‘, A. D. (1989). Common symptoms in ambulatory care: Incidence, evaluation, therapy, and outcome. American Journal of Medicine, 86, 262-266. LaPlante, M. P., & Carlson, D. (1996). Disability in the United States: Prevalence and causes. Retrieved April 2, 2002, from htto:l/www.d_sgucsf.edu/UCSF/pub.taf?grow=3. LaVeist, T. A., Nickerson, K. J ., & Bowie, J. V. (2000). Attitudes about racism, medical mistrust, and satisfaction with care among Afi'ican American and White cardiac patients. Medical Care Research and Review, 5 7(1), 146-161. Lee, H. B., Cheung, F. B., Man, H., & Hsu, S., Y. C. (1992). Psychological characteristics of Chinese low back pain patients: an exploratory study. Psychology and Health, 6, 119-128. 151 Levendusky, P., & Pankratz, L. (1975). Self-control techniques as an alternative to pain medication. Journal of Abnormal Psychology, 84, 165-168. Levine, S. (1957). Infantile experiences and resistance to physiological stress. Science, 126, 405-406. Levine, S. (1962). Plasma-free corticosteroid response to electric shock in rats stimulated in infancy. Science, 135, 795-796. Levine, S., Haltmeyer, G. C., Kaas, G. G., & Penenberg, V. H. (1967). Physiological and behavioral effects of infantile stimulation. Physiology and Behavior, 2, 55-63. Levinson, W., Roter, D. L., Mullooly, J. P., Dull, V. T., & Frankel, R. M. (1997). Physician-patient communication: The relationship with malpractice claims among primary care physicians and surgeons. Journal of the American Medical Association, 277(7), 553-559. Lipman, A. G. (2004). Practical Pain Management. Retrieved April 27, 2005, from http://www.acofp.org/memberjublieations/OS03t3.htrnl. Lister, B. J. (1996). Dilemmas in the treatment of chronic pain. American Journal of Medicine, 101(Suppl. 1A), 1A-ZS-1A-SS. Long, C. J ., & Webb, W. L. (1980). Improving diagnostic strategies for pain patients. Neurosurgery, 7(3), 225-229. Lopez, F. G. (1997). Student-professor relationship styles, childhood attachment bonds and current academic orientations. Journal of Social and Personal Relationships, 14, 271-282. Lousberg, R., Greonman, N., & Schmidt, A. (1996). Profile characteristics of the MPI-DLV clusters of pain patients. Journal of Clinical Psychology, 52(2), 161-167. Magora, A., & Schwartz, A. (1980). Relation between the low back pain syndrome and x-ray findings. Scandinavian Journal of Rehabilitation Medicine, I 2, 9- 15. Main, Wood, P. L. R., Hollis, S., Spanswick, C. C., & Waddell, G. (1992). The distress and risk assessment method: A simple patient classification to identify distress and evaluate the risk of poor outcome. Spine, 1 7, 42-52. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. Monographs for the Society of Research in Child Development, 50, 66-104. 152 Maruta, T., Malinchoc, M., Offord, K. P., & Colligan, R. C. (1998). Status of patients with chronic pain 13 years after treatment in a pain management center. Pain, 74, 199-204. Maruta, T., Swanson, D. W., & Swenson, W. M. (1979). Which patients may a pain management program help? Pain, 7, 321-329. McCracken, L. M., Spertus, I. L., Janeck, A. S., Sinclair, D., & Wetzel, F. T. (1999). Behavioral dimensions of adjustment in persons with chronic pain: pain- related anxiety and acceptance. Pain, 80, 283-289. McEwen, B. S., & Sapolsky, R. M. (1995). Stress and cognitive function. Current Opinion in Neurobiology, 5, 205-216. Mechanic, D., & Meyer, S. (2000). Concepts of trust among patients with serious illness. Social Science and Medicine, 51, 657-668. Melzack, R. (1999). Pain and stress: A new perspective. In R. J. Gatchel & D. C. Turk (Eds), Psychosocial Factors in Pain: Critical Perspectives (pp. 89-106). New York: Guilford Press. Meredith, L. 8., Orlando, M., Humphrey, N., Camp, P., & Sherboume, C. D. (2001). Are better ratings of the patient-provider relationship associated with higher quality care for depression? Medical Care, 39(4), 349 - 360. Miaslowski, C. (1999). The role of sex and gender in pain perception and responses to treatment. In R. J. Gatchel & D. C. Turk (Eds), Psychosocial Factors in Pain: Critical Perspectives (pp. 401-411). New York: Guilford Press. Mikail, S. F ., DuBreuil, S., & D'Eon, J. L. (1993). A comparative analysis of measures used in the assessment of chronic pain patients. Psychological Assessment, 5(1), 117-120. Mikulincer, M. (1998). Adult attachment style and individual differences in functional versus dysfunctional experiences of anger. Journal of Personality and Social Psychology, 74, 513-524. Mikulincer, M., & Florian, V. (1995). Appraisal and coping with a real-life stressful situation: The contribution of attachment styles. Personality and Social Psychology Bulletin, 21, 406-414. Mikulincer, M., Florian, V., & Tolmacz, R. (1990). Attachment styles and fear of personal death: A case study of affect regulation. Journal of Personality and Social Psychology, 58, 273-280. 153 Molleman, E., Krabbendarn, P. J ., & Annyas, A. A. (1984). The significance of the doctor-patient relationship in coping with cancer. Social Science and Medicine, 18(6), 475-480. Morris, D. B. (1999). Sociocultural and religious meanings of pain. In R. J. Gatchel & D. C. Turk (Eds), Psychosocial Factors in Pain: Critical Perspectives (pp. 1 18-131). New York: Guilford Press. Murphy, J ., & Coster, G. (1997). Issues in patient compliance. Drugs, 54(6), 797-800. Nachmias, M., Gunnar, M., Mangelsdorf, S., Parritz, R. H., & Buss, K. (1996). Behavioral inhibition and stress reactivity: The moderating role of attachment security. Child Development, 6 7, 508-522. Noyes Jr., R., Stuart, 8., Langbehn, D. R., Happel, R. L., Longley, S. L., & Yagla, S. J. (2002). Childhood Antecedents of Hypochondriasis. Psychosomatic Medicine, 43(3), 282-289. Noyes Jr., R., Stuart, S. P., Langbehn, D. R., Happel, R. L., Longley, S. L., Muller, B. A., et al. (2003). Test of an Interpersonal Model of Hypochondriasis. Psychosomatic Medicine, 65, 292-300. Ognibene, T. C., & Collins, N. L. (1998). Adult attachment styles, perceived social support and coping strategies. Journal of Social and Personal Relationships, 15(3), 323-345. Olshaker, J. S., Barish, R. A., Naradzay, J. F., A., J. D., Safir, E., & Campbell, L. (1999). Prescription noncompliance: contribution to emergency department visits and cost. Journal of Emergency Medicine, 1 7(5), 909-912. Onishi, M., & Gjerde, P. F. (1994). Adult attachment styles: A multi-method examination of personal ity characteristics. Paper presented at the International Society for the Study of Behavioral Development, Amsterdam, The Netherlands. Park, T. L., Adams, S. G., & Lynch, J. (1998). Sociodemographic factors in health psychology research: 12 years in review. Health Psychology, 1 7, 381-383. Pascoe, G. C., & Attkisson, C. (1983). The evaluation ranking scale: a new methodology for assessing satisfaction. Evaluation and Program Planning, 6, 335-347. Pearson, J. L., Cohen, D. A., Cowan, P. A., & Cowan, C. P. (1994). Earned— and continuous-security in adult attachment: Relation to depressive symptomatology and parenting style. Development and Psychopathology, 6, 359-3 73. ' 154 Peck, C. L., & King, N. J. (1982). Increasing patient compliance with prescriptions. Journal of the American Medical Association, 248(21), 2874-2877. Physicians, A. A. o. F. (2005). 2005 FA C TS about family practice. Retrieved April 26, 2005, from http://www.aafo.org1x25059.xml Pistole, M. C. (1994). Adult attachment styles: Some thoughts on closeness-distance struggles. Family Process, 33, 147-159. Pistole, M. C. (1995). College students' ended love relationships: Attachment style and emotion. Journal of College Student Development, 36, 53-60. Pistole, M. C. (1999). Caregiving in attachment relationships: a perspective for counselors. Journal of Counseling and Development, 77(4), 437-446. Pistole, M. C., & Watkins, C. E., Jr. (1995). Attachment theory, counseling process, and supervision. The Counseling Psychologist, 23, 457-478. Portenoy, R. K., & Kanner, R. M. (1996). Definition and Assessment of Pain. In R. K. Portenoy & R. M. Kanner (Eds), Pain Management: Theory and Practice (pp. 3-18). Philadelphia: F. A. Davis Company. Regestine, Q. R. (2000). Psychiatrists' view of managed care and the future of psychiatry. General Hospital Psychiatry, 22, 97-106. Rholes, W. 8., Simpson, J. A., & Orina, M. M. (1999). Attachment and anger in an anxiety-provoking situation. Journal of Personality and Social Psychology, 76, 940-957. Richmond, R. L., & Carmody, T. P. (1999). Dropout from treatment for chronic low- back pain. Professional Psychology: Research and Practice, 30(1), 51-55. Robins, L. S., & Wolf, F. M. (1988). Confrontation and politeness strategies in physician-patient interactions. Social Science and Medicine, 27(3), 217-221. Robinson, M. E., & Riley III, J. L. (1999). The role of emotion in pain. In R. J Gatchel & D. C. Turk (Eds), Psychosocial Factors in Pain: Critical Perspectives (pp. 74-88). New York: Guilford Press. Rogers, C. R. (1951). Client-centered therapy: its current practice, implications, and theory. Boston: Houghton Mifflin. 155 Rom, E., & Mikulincer, M. (2003). Attachment theory and group processes: The association between attachment style and group-related representations, goals, memories, and functioning. Journal of Personality and Social Psychology, 84, 1220-1235. Ross, C. K., Steward, C. A., & Sinacore, J. M. (1995). A comparative study of seven measures of patient satisfaction. Medical Care, 33(4), 392-406. Roter, D. L. (2000). The Roter Method of Interaction Process Analysis Manual. Baltimore: Johns Hopkins School of Public Health, Department of Health Policy and Management. Rothbard, J. C., Roberts, L. J., Leonard, K. E., & Eiden, R. D. (1993). Attachment styles and marital interaction behavior. Paper presented at the American Psychological Association, Toronto, Ontario, Canada. Rudy, T. E. (1989a). MPI Microcomputer Program (Version 2.0) [PC-DOS V2.0]. Pittsburgh, PA: University of Pittsburgh School of Medicine. Rudy, T. E. (1989b). Multiaxial assessment of pain: computer program and user's manual (Version 2.1). Pittsburgh, PA: University of Pittsburgh. Rudy, T. E. (2004). Multidimensional Pain Inventory Software (Version 3). Pittsburgh: Univeristy of Pittsburgh, UPMC Pain Medicine. Rudy, T. E., Turk, D. C., Kubinski, J. A., & Zaki, H. S. ( 1995). Differential treatment responses of TMD patients as a function of psychological characteristics. Pain, 61, 103-112. Rudy, T. E., Turk, D. C., Zaki, H. S., & Curtin, H. D. (1989). An empirical taxometric alternative to traditional classification of temporomandibular disorders. Pain, 36, 311-320. Rutter, M. (1997). Clinical implications of attachment concepts: retrospect and prospect. In L. Atkinson & K. J. Aucker (Eds), Attachment and PsychOpathology (pp. 17-46). New York: Guilford Press. Rybstein-Blinchik, E. (1979). Effects of different cognitive strategies on chronic pain experience. Journal of Behavioral Medicine, 2, 93-101. Safran, D. G., Montgomery, J. E., Chang, H., Murphy, J ., & Rogers, W. H. (2001). Switching doctors: Predictors of voluntary disenrollment from a primary physician's practice. Journal of Family Practice, 50(2), 131-142. Sapolsky, R. M. (1996). Why stress is bad for your brain. Science, 273, 749-750. 156 Schofferman, J ., Anderson, D., Hines, R., Smith, G., & Keane, G. (1993). Childhood psychological trauma and chronic refractory low-back pain. Clinical Journal of Pain, 9, 260-265. Schofferman, J ., Anderson, D., Hines, R., Smith, G., & White, A. (1992). Childhood psychological trauma correlates with unsuccessful lumbar spine surgery. Spine, 17 (Suppl), Sl38-Sl44. Schwenk, T. L., Marquez, J. T., Lefever, R. D., & Cohen, M. (1989). Physician and patient determinants of difficult physician-patient relationships. The Journal of Family Practice, 28(1), 59 - 63. Seville, J. L. (1996). Mastering Chronic Pain: A Professional 's Guide to Behavioral Treatment (Vol. 1). Sarasota, FL: Professional Research Press. Shaver, P. R. (1995). Additional items to measure dismissing avoidance. Davis, California: Dept. of Psychology, University of California, Davis. Shaver, P. R., & Mikulincer, M. (2002). Attachment-related psychodynamics. Attachment and Human Development, 4, 133-161. Sihvonen, M., Kayhdo, K., & Kekki, P. (1989). Factors influencing the patient- provider relationship. Vard Nord Utveckl F orsk, 9(2), 4-9. Simpson, J. A. (1990). Influence of attachment styles on romantic relationships. Journal of Personality and Social Psychology, 5 9, 971 -980. Singer, B. H., & Ryff, C. D. (Eds). (2001). National Research Counsel: New Horizons in Health: An Integrative Approach. Washington D. C.: National Academy Press. Stevens, V. T., & Hollis, J. F. (1989). Preventing smoking relapse using an individually tailored skills-training technique. Journal of Consulting and Clinical Psychology, 57(3), 420-424. Stimson, G. V. (1974). Obeying doctor's orders: a view from the other side. Social Science and Medicine, 8(2), 97-104. Strong, J ., Ashton, R., & Stewart, A. (1994). Chronic low back pain: Toward an integrated psychosocial assessment model. Journal of Consulting and Clinical Psychology, 62(5), 1058- l 063. Stump, T. E., Dexter, P. R., Tierney, W. M., & Wolinsky, F. D. (1995). Measuring patient satisfaction with physicians among older and diseased adults in a primary care municipal outpatient setting: an examination of three instruments. Medical Care, 33(9), 985-972. 157 Sullivan, D. A., & Beeman, R. (1982). Satisfaction with maternity care: A matter of communication and choice. Medical Care, XX(3), 321-330. Swanson, D. W., Floreen, A. C., & Swenson, W. M. (19763). Program for managing chronic pain. 11. Short-term results. Mayo Clinic Proceedings, 51(7), 409- 411. Swanson, D. W., Swenson, W. M., Maruta, T., & McPhee, M. C. (1976b). Program for managing chronic pain. I. Program description and characteristics of patients. Mayo Clinic Proceedings, 51 (7), 401-408. Tabachnick, B. G., & F idell, L. S. (2001). Logistic regression. In Using Multivariate statistics (4th ed., pp. 517-581). Needham Heights, MA: Pearson Education Company. Talo, S., Rytokoski, U., & Puukka, P. (1992). Patient classification, a key to evaluate pain treatment: a psychological study in chronic low back pain patients. Spine, 17, 998-1010. Turk, D. C. (1990). Customizing treatment for chronic pain patients: who, what, and why. Clinical Journal of Pain, 6(4), 255-270. Turk, D. C. (1996). Biopsychosocial perspective on chronic pain. In R. J. Gatchel & D. C. Turk (Eds), Psychological approaches to pain management: a practioner's guide (pp. 3-32). New York: Guilford Publications. Turk, D. C., & Flor, H. (1984). Etiological theories and treatments for chronic back pain. 11. Psychological models and interventions. Pain, 19, 209-233. Turk, D. C., & Gatchel, R. J. (2000). Psychosocial assessment of chronic occupational musculoskeletal disorders. In T. G. Mayer, R. J. Gatchel & P. B. Polatin (Eds), Occupational Musculoskeletal Disorders (pp. 587-608). Philadelphia: Lippincott Williams & Wilkins. Turk, D. C., & Nash, J. M. (1996). Psychologic Issues in Chronic Pain. In R. K. Portenoy & R. M. Kanner (Eds), Pain Management: Theory and Practice (pp. 323-335). Philadelphia: F. A. Davis Company. Turk, D. C., & Rudy, T. E. (1987). Towards a comprehensive assessment of chronic pain patients. Behavior Research and Therapy, 25, 23 7-249. Turk, D. C., & Rudy, T. E. (1988). Toward an empirically derived taxonomy of chronic pain patients: integration of psychological assessment data. Journal of Consulting and Clinical Psychology, 56, 233 238. 158 Turk, D. C., & Rudy, T. E. (1990). The robustness of an empirically derived taxonomy of chronic pain patients. Pain, 43, 27-35. Tumer-Cobb, J. M., Gore-Felton, C., Marouf, F., Koopman, C., Kim, B, Israelski, D., et al. (2002). Coping, social support, and attachment style as psychosocial correlates of adjustment in men and women with HIV/AIDS. Journal of Behavioral Medicine, 25(4), 337-353. Vidal, A. M. (2002). The politics of pain: The controversy surrounding chronic pain and opiods. Retrieved April 7, 2005, from mud/WWW.immunesupport.com/librarv/print.cfm?ID=4073&t=CFIDS FM. Von Korff, M., Gruman, J ., Schaefer, J., Cmry, S. J ., & Wagner, E. H. (1997). Collaborative management of chronic illness. Annals of Internal Medicine, 127(12),1097-1102. Wagner, L., & Vaux, A. (1994). The Wagner Attachment Scale (WAS): Reliability and validity of a self-report measure of adult attachment. Paper presented at the Midwestern Psychological Association, Chicago. Ware, J. E., Snyder, M. K., & Wright, W. R. (1976). Development and validation of scales to measure patient satisfaction with health care services: Volume I of a final report Part A: Review of literature, overview of methods, and results regarding construction of scales. Springfield: National Technical Information Service. Weinberger, D. A., Schwartz, G. E., & Davidson, R. J. (1979). Low-anxious, high- anxious and repressive styles: psychometric patterns and behavioral and physiological responses to stress. Journal of Abnormal Psychology, 88(4), 369-380. Weisberg, J. N., & Keefe, F. J. (1999). Personality, individual differences, and psychopathology in chronic pain. In R. J. Gatchel & D. C. Turk (Eds), Psychosocial Factors in Pain: Critical Perspectives (pp. 56-73). New York: Guilford Press. West, M. L., & Sheldon-Keller, A. E. (1994). Patterns of relating: An adult attachment perspective. New York: Guilford Press. Wetzler, H. P., & Radosevich, D. M. (1992). Health Status Questionnaire (SF -36). (Technical Report, Interstudy). Whitefield, K. E., Weidner, G., Clark, R., & Anderson, N. B. (2002). Sociodemographic diversity and behavioral medicine. Journal of Consulting and Clinical Psychology, 70, 463-481. 159 Williams, S. J ., & Calnan, M. (1991). Key determinants of consumer satisfaction with general practice. Journal of Family Practice, 8, 23 7-242. Willson, P., & McNamara, J. R. (1982). How perceptions of a simulated physician- patient interaction influence intended satisfaction and compliance. Social Science and Medicine, 16, 1966 - 1704. Wolf, M. H., Putnam, S. M., James, S. A., & Stiles, W. B. (1978). The Medical Interview Satisfaction Scale; Development of a scale to measure patient perceptions of physician behavior. Journal of Behavioral Medicine, 1, 391 - 401. Zahr, L. K., Williams, S. G., & El-Hadad, A. (1991). Patient satisfaction with nursing care in Alexandria. International Journal of Nursing Studies, 28, 337-342. Zastowny, T. R., Roghmann, K. J ., & Cafferata, G. L. (1989). Patient satisfaction and the use of health services. Explorations in causality. Medical Care, 27(7), 705-723. 160 MICHIGAII SIAIF UNWFRSITV AIBRA 3 1293 02736 05