<%BANNER%>

The Psychological Status of Workers' Compensation Clients and Select Demographic and Forensic Variables

Permanent Link: http://ufdc.ufl.edu/UFE0017262/00001

Material Information

Title: The Psychological Status of Workers' Compensation Clients and Select Demographic and Forensic Variables
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0017262:00001

Permanent Link: http://ufdc.ufl.edu/UFE0017262/00001

Material Information

Title: The Psychological Status of Workers' Compensation Clients and Select Demographic and Forensic Variables
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0017262:00001


This item has the following downloads:


Full Text

PAGE 1

THE PSYCHOLOGICAL STATUS OF WORKERS’ COMPENSATION CLIENTS AND SELECT DEMOGRAPHIC AND FORENSIC VARIABLES By CHAD JAMES BETTERS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Chad James Betters

PAGE 3

This dissertation is dedicated to my wife Jenna, and my parents, Brad and Lori.

PAGE 4

iv ACKNOWLEDGMENTS There are several individuals that I woul d like to acknowledge for their assistance in the completion of this dissertation. I fi rst would like to thank Dr. Linda Shaw, my mentor and committee chair, for all of her guidance and support th roughout the doctoral program. I would also like to thank my other committee members, including Dr. Beth Swett, Dr. Orit Shechtman, and Dr. Russ Bauer, for all of the assi stance they provided during my dissertation research. I would lik e to thank my other mentor, Dr. John Saxon, for allowing me the opportunity to work with him in the Bachelor of Health Science program during my doctoral studies. I woul d like to thank Dr. Steve Pruett for his statistical assistance and SPSS consultation. The Florida Bureau of Rehabilitation and Reemployment Services was more than accomm odating to allow me to collect my data for this dissertation. I esp ecially thank Ann La timer for her assistance with this opportunity. I would like to thank Gale Leme rand for his financial contribution toward this dissertation research. I would like to th ank my parents, Brad and Lori Betters, for their love and support throughout my never-ending college career Lastly, I would like to thank my wife and best friend, Jenna Betters, for all of the love, faith, and patience we share.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... ix CHAPTER 1 INTRODUCTION........................................................................................................1 Background...................................................................................................................1 Research Questions.......................................................................................................4 Variables and Definitions.............................................................................................5 Significance..................................................................................................................5 2 LITERATURE REVIEW.............................................................................................7 The Workers’ Compensation System...........................................................................7 Workers’ Compensation Benefits.................................................................................9 The Coexistence of Medical and Psychological Conditions........................................9 The Interrelationship Between Chronic Pain and Depression.............................10 The Interrelationship Between Chronic Pain and Anxiety..................................14 The Interrelationship Between Ch ronic Pain and Somatization..........................16 The World Health Organization ICF Model...............................................................18 The ICF Model and Chronic Pain...............................................................................21 Psychological Status and Return To Work Outcomes................................................23 Demographic and Forensic Variables.........................................................................25 Return To Work Outcomes in Workers’ Compensation............................................29 Psychological Treatment Cost-Effectiveness.............................................................31 Rationale.....................................................................................................................3 2 3 METHODOLOGY.....................................................................................................34 Research Questions.....................................................................................................34 Hypotheses..................................................................................................................35 Population...................................................................................................................35 Sampling Technique...................................................................................................37 Research Design.........................................................................................................37 Measurement...............................................................................................................38 Betters Injured Worker Inventory (BIWI)...........................................................38 Pain Patient Profile (P-3).....................................................................................40 Statistical Analysis......................................................................................................43

PAGE 6

vi Pilot Study..................................................................................................................44 4 RESULTS...................................................................................................................51 Post-Hoc Power Analyses...........................................................................................51 Descriptive Statistics..................................................................................................51 Independent Variables.........................................................................................52 Dependent Variables...........................................................................................54 Backward Elimination Multiple Regression Analyses...............................................54 Depression...........................................................................................................55 Anxiety................................................................................................................55 Somatization........................................................................................................56 Depression – Supplementary...............................................................................56 Anxiety – Supplementary....................................................................................57 Somatization – Supplementary............................................................................57 Correlations.................................................................................................................57 Independent Variables.........................................................................................57 Dependent Variables...........................................................................................58 5 DISCUSSION.............................................................................................................79 Overview of Significant Findings...............................................................................79 Prediction of Depression.....................................................................................79 Prediction of Anxiety..........................................................................................82 Prediction of Somatization..................................................................................84 Additional Findings....................................................................................................86 Least Predictive Variables...................................................................................86 Order of Satisfaction............................................................................................88 The Presence of the Psychological Conditions...................................................90 The Relationship Between the Psychological Conditions...................................91 Limitations..................................................................................................................91 Future Recommendations...........................................................................................94 Clinical Practice and Policy.................................................................................94 Professional Training...........................................................................................96 Additional Research............................................................................................98 APPENDIX A BUREAU OF REHABILITATION AND REEMPLOYMENT SERVICES PERMISSION LETTER...........................................................................................100 B BETTERS INJURED WORKER INVENTORY (BIWI)........................................101 LIST OF REFERENCES.................................................................................................103 BIOGRAPHICAL SKETCH...........................................................................................110

PAGE 7

vii LIST OF TABLES Table page 3-1. Backwards Elimination Multiple Regression Analysis for Demographic and Forensic Variables Predicti ng Depression – Pilot Study.........................................48 3-2. Backwards Elimination Multiple Regression Analysis for Demographic and Forensic Variables Predicting Anxiety – Pilot Study...............................................49 3-3. Backwards Elimination Multiple Regression Analysis for Demographic and Forensic Variables Predicting Somatization – Pilot Study......................................50 4-1. Backward Elimination Multiple Re gression Analysis for Demographic and Forensic Variables Predic ting Depression (n=111).................................................59 4-2. Backward Elimination Multiple Re gression Analysis for Demographic and Forensic Variables Predicting Anxiety (n=111).......................................................62 4-3. Backward Elimination Multiple Re gression Analysis for Demographic and Forensic Variables Predic ting Somatization (n=111)..............................................65 4-4. Supplementary Backward Elimina tion Multiple Regression Analysis for Demographic and Forensic Variab les Predicting Depression (n=93)......................68 4-5. Supplementary Backward Elimina tion Multiple Regression Analysis for Demographic and Forensic Variab les Predicting Anxiety (n=93)...........................71 4-6. Supplementary Backward Elimina tion Multiple Regression Analysis for Demographic and Forensic Variable s Predicting Somatization (n=93)...................74 4-7. Pearson-product Moment Correla tions of Independent Variables............................77 4-8. Pearson-product Moment Correla tions of Dependent Variables...............................78

PAGE 8

viii FIGURE Figure page 2-1. World Health Organization’s ICF Model..................................................................33

PAGE 9

ix Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE PSYCHOLOGICAL STATUS OF WORKERS’ COMPENSATION CLIENTS AND SELECT DEMOGRAPHIC AND FORENSIC VARIABLES By Chad James Betters December 2006 Chair: Linda R. Shaw Major Department: Rehabilitation Science The research objective of this study is to identify select demographic and forensic predictor variables th at would allow rehabilitation counselors providing Workers’ Compensation rehabilitation services to interv ene earlier in the rehabilitation process to address unmet psychological needs. It is believed that individuals with unmet psychological needs are not as successful in completing vocational rehabilitation directly due to the presence of the psychological condition. This study assessed individuals injured workers in the Workers’ Compensatio n system for three common psychological conditions: depression, anxiety, and somatiza tion. An attempt to identify predictive relationships for these conditions was made us ing four select demogr aphic variables (age, gender, ethnicity, and marital status), as well as six select forensic variables (time since injury, perceived social support, perceived pa in interference, satis faction with former employer, satisfaction with insurance carrier, and attorney retainment status). The data was collected through the Betters Injured Worker Inventory and the Pa in Patient Profile.

PAGE 10

x Backward elimination multiple regression analysis was utilized in order to determine the select demographic variables’ capability of predicting unmet psychological need. Upon statistical analysis, higher levels of perceived pain interferen ce, retaining an attorney, and lower levels of satisfaction with the attorney were predictive of de pression. Higher levels of perceived pain interferen ce and lower levels of satisfaction with the Workers’ Compensation insurance carrier were predictiv e of somatization. No significant predictor variables for anxiety were identified; howev er a noteworthy trend was found, specifically that females demonstrated higher levels of anxiety than males. Given these findings, rehabilitation counselors may be able to utiliz e this information in order to justify early screening and subsequent treatment for unmet psychological need based on acquiring this information during an initial interview.

PAGE 11

1 CHAPTER 1 INTRODUCTION The psychological condition of individuals with work-related injuries is an important factor to consider when providing vocational rehabilitation services. There is literature that suggests the psyc hological well being of an individual with a work-related injury is just as importa nt as the actual physical injury (Boersma & Linton, 2005; Gatchel, 2004; Severeijns, Vlaeyen, van de n Hout, & Weber, 2001; Sullivan & Stanish, 2003; Vowles & Gross, 2003). To date, no resear ch has been able to determine variables that predict psychological status for claimant s in the Workers’ Compensation system that would assist rehabilitation counselors. The identification of thes e predictor variables would greatly benefit the Workers’ Compensa tion rehabilitation field. By determining these predictor variables, re habilitation counselors might be better able to identify individuals with special psyc hological needs for appropriate treatment or referral for additional services. Background In 2003, approximately 4.4 million non-fata l work-related injuries and illnesses occurred in the United States (Bureau of Labor Statistics, n.d.). These figures suggest that a non-fatal work–related injury or illness occu rs at a rate of 5/100 full-time workers. Incidence rates greatly vary among specific industries, with th e top three specific industries for non-fatal workplace injuries in cluding manufacturing (21%), health care and social assistance (15.9%), and construction (9.7%), wher eas specific industries, such as educational services (0.9%) and mining (0.4 %), have lower incide nce rates (Bureau of

PAGE 12

2 Labor Statistics). Out of the 4.4 million cas es in 2003, 1.3 million injuries and illnesses resulted in lost productivity due to missed days from work (Bureau of Labor Statistics). Numerous studies (Blackwell, Leierer, Haupt, & Kampitsis, 2003; McGeary et al., 2003; Selander, Marnetoft, Bergroth, & Elkholm, 2002; Severeijns, Vlaeyen, vanden Hout, & Weber, 2001; Tate, 1992) have at tempted to explain how rehabilitation professionals may be able to predict return to work capability for individuals with various disabilities. These researchers have suggested age, gender, type of occupation, previous job satisfaction, type of injury, and attorn ey involvement, among ot hers, as predictive variables for individuals with varying disabilities in regard to return to work outcomes. With the capability to predict, vocational re habilitation services might be provided more effectively and additional services, such as psychological assessment and therapy, can be procured in a more timely manner. Successful return to work is typically evaluated by considering type of injury, length of time out of work, and previous job satisfaction (Selander, Marnetoft, Be rgroth, & Ekholm, 2002). Psychological conditions, such as depre ssion, anxiety, and somatization, may also act as psychological return to work barri ers, impeding the voca tional rehabilitation process. Without addressing these needs, vo cational rehabilitation may be hindered or futile until the psychological we ll being of the individual is treated (Boersma & Linton, 2005). There is strong evidence that suggests an interrelationship between chronic pain, which is typically found within injured workers in the Workers’ Compensation system, and psychological disorders (Boersma & Li nton, 2005; Butchner, 1985; Gatchel, 2004; Piccinelli, Patterson, Braithwaite, Boot, & Wilkinson, 1999; Severeijns, Vlaeyen, van den

PAGE 13

3 Hout, & Weber, 2001; Sullivan & Stanish, 2003; Vowles & Gross, 2003). While a distinct comorbidity exists, both chroni c pain and psychologi cal disability are independent. Both entities are true hea lth concerns and without addressing each separately, neither will cease (Giesecke et al., 2005). Many hea lth care providers are learning that treating one condition, such as the chronic pain source, does not eliminate the psychological disturbance. In fact, treat ing an individual with chronic pain when comorbid, untreated psychological distress is present typically results in difficult to impossible treatment of the chronic pain a nd vice versa (Sullivan & Stanish, 2003). Only when both entities are treated effectively can the individual expe rience rehabi litation. Evidence suggests that most rehabilitati on providers and facilities focus great attention on physical rehabilitation for in jured workers while providing little to no attention to mental health concerns (D ush, Simons, Platt, Nation, & Ayres, 1994; Fishbain et al., 1993; Gardner, 1991; Sullivan & Stanish, 2003; Tugman & Palmer, 2004; Vowles & Gross, 2003). These researcher s argue that an emphasis on physical rehabilitation has been encouraged by insuranc e carriers, and that there is also a general lack of awareness among health care profe ssionals working in Workers’ Compensation rehabilitation (such as physical and occupa tional therapists) in re gard to psychological problems that accompany physical disability. Given that many Workers’ Compensati on clients experience chronic pain associated with their injuries, and give n the previously discussed comorbid, yet independent relationship between chronic pa in and psychological disorders, it would appear that a failure to address the psyc hological status of Workers’ Compensation clients may result in poorer return to work outcomes. However, the literature discussed

PAGE 14

4 above suggests that unmet psychological need in Workers’ Compensatio n clients is either not being fully addressed to completely being missed. One way to identify unmet psychological need would be to assess all injured workers upon entrance into a vocational reha bilitation program. Unfortunately, this is probably not possible due to cost. However, if easily assessed predictor variables for certain psychological conditions can be established through empirical research, rehabilitation counselors could screen individuals and refer for appropriate psychological assessment. By doing so, the negative imp act of comorbid psychological conditions might be minimized. Furthermore, it has b een argued that trea ting the psychological would allow physical rehabilitation to be more successful (Sullivan & Stanish, 2003), thus providing another reason why identifyi ng predictor variable s would be greatly beneficial. Unfortunately, no research to date has clearly establishe d predictor variables that identify individuals with unmet psychological needs within the Workers’ Compensation system Research Questions The research question that wi ll be addressed in this st udy is: To what extent are select demographic and forensic variables ab le to predict the presence of clinical depression, anxiety, and somatization in Worker s’ Compensation clients? Specifically, the research questions are the following: 1. Can selected demographic and forensic vari ables predict the presence of clinical depression in injured workers within the Workers’ Compensation system? 2. Can selected demographic and forensic vari ables predict the presence of clinical anxiety in injured workers within the Workers’ Compensation system? 3. Can selected demographic and forensic vari ables predict the presence of clinical somatization in injured workers within the Workers’ Compensation system?

PAGE 15

5 Variables and Definitions The three dependent variables of inte rest include depression, anxiety, and somatization. The following represents the operational definitions of the dependent variables to be used in the study: Depression – as defined by Tollison and Langl ey’s (1995) Pain Patient Profile (P3), which utilized the clinical defi nition of major depression found in the Diagnostic and Statistical Manual of Mental Disorder s (American Psychiatric Association, 2000), the presence of, “sl eep, psychomotor activity, energy level, concentration, and decision-making disturban ces, as well as feelings of low selfworth, including affective distress, and a loss of interest, pleasure, and enjoyment in activities that were previously considered pleasurable” (pg. 21). Anxiety – as defined by Tollison and Langley’s Pain Patient Profile (P-3), which utilized the clinical definition of anxi ety found in the Diagnostic and Statistical Manual of Mental Disorders (American Ps ychiatric Association), the presence of, “inner turmoil, anger, worry, nervousness, restlessness, and emotional instability, including fear” (pg. 23). Somatization – as defined by Tollison and Langley’s Pain Patient Profile (P-3), which utilized the clinical definition of somatization found in the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association), the presence of, “physical symptoms that suggest a general medical condition and are not fully explained by a general medical condition, by the direct effects of a substance, or by another mental disorder” (pg. 25). Significance The research objective is to identify potenti al predictor variables that would allow rehabilitation counselors provi ding Workers’ Compensation re habilitation services to intervene earlier in the rehabilitation process to address unmet psychological needs. This might greatly improve the quality of vocatio nal rehabilitation services provided to individuals in the Workers’ Compensation sy stem by allowing the opportunity for more time-efficient, positive return to work outcomes. The establishment of predictor variables, which can be utilized by reha bilitation counselors, will create a more efficacious treatment modality and allow accessibility to needed mental health services.

PAGE 16

6 Identification of psychological return to work barriers al lows appropriate referral and treatment. Both physical and psychologica l complaints would be addressed, allowing a greater chance for rehabilitation and reintegrat ion into the workforce. Rapid return to work would also alleviate systemic probl ems existing in the Workers’ Compensation arena, such as increasing premiums, increa sing costs for treatment, increasing shifts towards pharmacologically treating all prob lems for cost purposes, and decreasing productivity in the workforce due to lengthi er absenteeism from work-related injuries (Blackwell, Leirer, Haupt, & Kampitsis, 2003).

PAGE 17

7 CHAPTER 2 LITERATURE REVIEW There is an abundance of literature focu sing on return to work outcomes and the role of chronic pain in relationship to th ese outcomes. Numerous studies suggest the development of psychological distress at the onset of disability, including work-related injuries (Bair, Robinson, Katon, & Kroenke, 2003; Crombez, Vlaeyen, Heuts, & Lysens, 1999; Kuch, 2001; Sareen, Cox, Clara, & Asmundson, 2005; Vlaeyen & Linton, 2000; Vowles & Gross, 2003). Several researchers have described the interrelationship of chronic pain and psychological distress (Bair, Robinson, Katon, & Kroenke, 2003; Gieseck et al., 2005; Kuch, 2001; Linton & A ndersson, 2000; Sullivan & Stanish, 2003; Vlaeyen & Linton, 2001). However, there is a lack of literature that focuses on the role of unmet psychological needs in conjunction with physical work-related disabili ty in relationship to return to work outcomes. There is no literature spec ifically pinpointing predictor variables for psychological conditions that ma y serve as return to work barriers for individuals within the Workers’ Compensation system. The following literature provides support for the need to identify such a knowledge base. The Workers’ Compensation System Workers’ Compensation is a disability comp ensatory system that provides several forms of benefits to individuals who have sustained work-related injuries (Shaw & Betters, 2004). Initially cr eated after the Industrial Revolution, the Workers’ Compensation system is federally mandated and state regulated. Each state, through

PAGE 18

8 annual legislation, determines the Workers’ Compensation statutes that govern the behavior of all parties involve d: the injured worker, the em ployer at the time of injury, the Workers’ Compensation insurance compa ny (carrier), the Workers’ Compensation attorneys, and all of the medical and health pr ofessionals involved in the treatment of the individual. The ultimate goal for all parties is a quick and successful return-to-work for the injured individual. All parties benefit from a su ccessful return-to-work and all possible solutions to facilitate this obj ective should therefore be implemented. When an individual sustains a work-relate d injury that is deemed compensable, meaning that the injury was indeed a byproduct of work, the individual is warranted three forms of benefits. The first benefit category initiated is medical benefits, which provides medical care for the compensable injury. The second, monetary benefits to accommodate for lost wages due to the injury, usually is in itiated after a specifie d time period, typically two to four weeks post-injury (FWCI, 2003). Th ese benefits continue until the individual reaches a critical point: Maximum Medi cal Improvement, or MMI. The Workers’ Compensation process involves two phases: be fore the individual re aches MMI and after the individual reaches MMI. MMI is a determination made by the authorized treating physician that specifies that the individual wi th the work-related inju ry has attained the greatest level of rehabilitation foreseeable in th e near future since the injury. Ideally, the individual will have reached full recovery with no residual impairment. However, many individuals will have functional impairments attr ibuted to their work-related injury with accompanying physical restrictions. Once place d at MMI, the monetary benefits cease. At this time, the third benef it system activates: vocational rehabilita tion. Rehabilitation now focuses not on impairment, but on disabi lity, specifically work disability (Shaw &

PAGE 19

9 Betters, 2004). The goal of vocational rehabi litation is providing services that will facilitate return-to-work, such as counseling and evaluation. Workers’ Compensation Benefits Included in the medical benefits, indivi duals with work-related injuries are generally warranted medical care, surgical care, hospitalization, pr escription coverage, and mental health care (FWCI, 2003). Me ntal health care in cludes psychological treatment, including evaluation, psychothe rapy, and pharmaceutical therapy. However, individuals in the Workers’ Compensation syst em may not be receiving necessary mental health care. This may be attributed to a l ack of awareness on behalf of the medical and health care professionals invol ved with the individuals (D ush, Simons, Platt, Nation, & Ayres, 1994). There also may be financial constr aints, specifically in the “managed carelike” system that Workers’ Compensation has become (Tugman & Palmer, 2004). Mental health care and rehabilitation needs may be viewed as less impor tant in relation to physical rehabilitation needs (Sullivan & Stanish, 2003). Ho wever, there are numerous studies that suggest otherwise, including significant findings that suggest the importance of psychological rehabilitati on over physical rehabilitation during the return-to-work process (Dush, Simons, Platt, Nation, & Ayre s, 1994; Fishbain et al., 1993; Gardner, 1991; Sullivan & Stanish, 2003; Tugman & Palmer, 2004). The Coexistence of Medical and Psychological Conditions Tugman and Palmer (2004) assert that fo r individuals working in the field of rehabilitation, there has alwa ys been an underlying problem: a potential coexistence of not only a physical impairment and accompa nying functional restrictions, but also the psychological malady and subsequent disabilit y. When this combination occurs within certain settings, such as the Workers’ Comp ensation system, rehabilitation professionals

PAGE 20

10 have generally not acknowledged the psyc hological problem, only focusing on the physical impairment (Tugman & Palmer). T ypically, a possible failu re to recognize the coexistence may occur, which in turn ma y seriously compromise the rehabilitation professional and injured employee’s ability to meet the Workers’ Compensation prime objective: a rapid return-to-work (Dush, Si mons, Platt, Nation, & Ayres, 1994). The injured worker, as a result, may be held liable for an unsuccessful return-to-work outcome (Vowles & Gross, 2003). The Interrelationship Between Chronic Pain and Depression Chronic pain, as defined by Block, Kr emer, and Callewart (1999, p.185), simply, “pain lasting six months or more.” An estimated $125 billion are spent annually on chronic pain treatment for the approximately 50 million individuals w ith partial to total disabling chronic pain (Block, Kremer, & Calle wart). Chronic pain is truly universal and unique; it is found around the world, in all types of people, and in ev ery part of the body (Block, Kremer, & Callewart). Depression, a Mood Disorder under the DS M IV-TR classification system, is a change of mood typically invol ving a loss of interest or pl easure (American Psychiatric Association, 2000). Specific de pressive symptoms that ar e included in the diagnostic criteria for depression include the following: depressed mood most of the day, nearly every day markedly diminished interest or pleasure significant weight loss or gain insomnia or hypersomnia nearly every day psychomotor agitation fatigue or loss of en ergy nearly every day

PAGE 21

11 feelings of worthlessness or ex cessive or inappropriate guilt diminished ability to think or concentrate recurrent thoughts of death (p. 356). Chronic pain and depression are often comorbid (Bair, Robinson, Katon, & Kroenke, 2003; Boersma & Linton, 2005; Butc hner, 1985; Gatchel, 2004; Piccinelli, Patterson, Braithwaite, Boot, & Wilkinson, 199 9; Severeijns, Vlaeyen, van den Hout, & Weber, 2001; Sullivan & Stanish, 2003; Vo wles & Gross, 2003). There are numerous documented examples of this relationship in the literature. Bair, Robinson, Katon, and Kroenke (2003) provide an extensive liter ature review focusing on individuals with comorbid pain and depression in all medical se ttings. After analyzing over 170 pieces of literature pertaining to como rbid pain and depression, Ba ir and colleagues observed certain themes, including “1) The prevalence of pain in a ‘depressed sample’ and the prevalence of depression in a pain sample are higher than the prevalence rates when the conditions are individually examined. 2) Depression is most prevalent in pain, psychiatric, and specialty clinics, as opposed to primary care settings. 3) The presence of pain negatively aff ects the recognition and treatment of depression.” (p. 2441) The themes reported by Bair and colleague s, although important and useful for the general population with comorbid pain and depression, does not provide sufficient information for certain populations, specifically populations with high prevalence of pain and depression, such as individuals with wo rk-related injuries within the Workers’ Compensation system.

PAGE 22

12 A longitudinal study conducted by Piccine lli, Patterson, Braithwaite, Boot, and Wilkinson (1999) suggests that individuals with chronic pain demonstrated higher frequencies of developing a ps ychological disorder, specifi cally depression or anxiety disorders, than individuals without chroni c pain complaints. These findings, however, were not differentiated by type of injury or initial severity between patients. As long as chronic pain, which could be pinpointed back to a specific injury, was reported by the patients, psychological distur bances were significant. The degree of psychological di stress, typically depression, is also associated with the amount of pain catastrophizing, or exaggera tion, portrayed by the patient. Severeijns, Vlaeyen, van den Hout, and Weber (2001) report that although patients with chronic pain report depression, the degree of depression is si gnificantly related to the degree of pain catastrophization. When patients vi ew their pain as disabling, or when they catastrophize, they are also more inclined to view their depression as disabling (Severeijns, Vlaeyen, van den Hout, & Weber). This relationship demonstrates covariation; inversely, individuals that do not view th eir pain as disabling are bette r at coping with depressive symptoms. These suggestions were supported by Gatche l’s findings (2004), specifically when examining self-assessment of physical functioni ng. Individuals that believed they were incapable of functioning physica lly during an evaluation were also identified as having psychological conditions, including depressi ve disorders. Although Gatchel provides several other reasons for poor functional performance duri ng an evaluation (secondary gain, somatization, compliance and resistance issu es), there is significa nt data to support

PAGE 23

13 the concept that an inability to cope with chronic pain results in an inability to cope with depression as well. There have been recent interesting findings in the literature that has investigated the connection between chronic pain and depres sion. Giesecke, Gracely, Williams, Geisser, Petzke, and Clauw (2005) report that depressi ve symptoms and chronic pain complaints are, in fact, neurologically separate. These findings do not imply that depression is a physiological or a psychological condition. Giesecke and colleagues do not support a biogenic or a psychogenic mode l of depression. This litera ture does, however, provide evidence that there is a comorbidity occurr ing; chronic pain a nd depression, although coexisting, are independent entities. Depressi on is not necessarily a symptom of chronic pain, just as chronic pain is not necessarily a symptom of depression. One of the most significant and heavily re searched moderator variables discovered upon reviewing the relationship between chronic pain and depression is the role of job satisfaction. Individuals who have reported lower levels of satisfaction in their work have demonstrated more problems post-inju ry than those who found their job to be satisfying. Fishbain, Cutler, Rosomoff, Khalil, and Steele-Rosomoff (1997) focused on how individuals with chronic pain due to a work-related injury pe rceived their job and how this perception influenced the individuals’ desire for return-to-work. The article suggests that individuals’ perceptions about th eir former job significa ntly relate s to the intention of return-to-work at the same jobsite (1997). Indivi duals who did not like their jobs prior to their injuries did not want to return to them, esp ecially if the injury resulted in chronic pain. A previous study conducted by Fishbain, Rosomoff, Cutler, and SteeleRosomoff (1995) examined this phenomenon during rehabilitation, pr ior to receiving a

PAGE 24

14 release to return-to-work. The findings suggest that the intention of returning is already minimized in the chronic pain population even before the focus is on actual return-towork. The level of job dissatisfaction greatly influenced the rehabilitation process. Individuals who demonstrated, through their rehab ilitation practices, a drive to return-towork reported no prior job dissatisfac tion, opposed to those who reported job dissatisfaction and did not participate in their rehabilitation as enthusiastically (Fishbain, Rosomoff, Cutler, & Steele-Rosomoff). A study focusing on job satisfaction as a pred ictor of chronic pain complaints and psychological distress was conducted, specificall y examining work-related low back pain (Williams et al., 1998). Using hierarchial mu ltiple regression analysis, the researchers found that job satisfaction pred icted the level of impairment for both chronic pain and psychological distress, specifically the level of depression was signi ficantly greater than would be expected, given the presence of othe r suggested predictor variables. Job satisfaction has been r ecognized as an important variable in the equation, however it is a variable that, in most s ituations, cannot be manipulated during the rehabilitation process. I ndividuals will either have f ound or not found their employment satisfying prior to their injury. This info rmation is important, however, for screening individuals who may be more prone to higher levels of ps ychological problems, which might or would likely directly a ffect return-to-work outcomes. The Interrelationship Between Chronic Pain and Anxiety There has been more research examining th e interrelationship be tween chronic pain and depression than other psychological diso rders. However, the relationship between anxiety and chronic pain has also been re searched and similar findings have emerged

PAGE 25

15 (Geisser, Robinson, Miller, & Bade, 2003; Kuch, 2001; Sareen, Cox, Clara, & Asmundson, 2005). Generalized anxiety, as defined by th e DSM IV-TR (American Psychiatric Association, 2000), is “excessive worry and apprehensive exp ectation . the anxiety and worry are accompanied by at least three additional symptoms from a list that includes restlessness, being easily fatigued, difficulty concentrating, irritability, muscle tension, and disturbed sleep” (p.472). There are severa l anxiety disorders listed in the DSM IVTR, each slightly different based on the cause of the anxiety or some distinction in either frequency or duration. Kuch (2001) suggests that most individuals with chronic pain with accompanying anxiety demonstrate either genera lized anxiety disorder or posttraumatic stress disorder. The latter is especially prev alent in individuals w ith traumatic accidents that resulted in work-relate d injury and disability. Indi viduals without a “frightening accident,” as described by Kuch (p.33), de monstrate a milder, but equally disabling generalized anxiety that results in co nsistent anxiety symptom presentation. The relationship between chronic pain a nd anxiety was investigated by Sareen, Cox, Clara, and Asmundson ( 2005). Their findings suggest a strong association between physical disorders, especially those with chronic pain, and anxiety disorders. Furthermore, the researchers argue that anxiet y that coexists with chronic pain and other physical disorders greatly increases the level of disability. These findings are nearly identical to those findings by Giesecke et al. (2005) when they examined the interrelationship between chr onic pain and depression. Anot her similarity between these two studies is the emergence of a hypothesi s that physical and ps ychological disorders are comorbid and independent; anxiety and chro nic pain must both be treated in order to

PAGE 26

16 reduce the level of disability cau sed by each, as opposed to just treating one or the other (Sareen, Cox, Clara, & Asmundson, 2005). Geisser, Robinson, Miller, and Bade ( 2003) also concluded that anxiety may influence pain and subsequent disability. They argue that fear related to pain may indeed more be disabling than the act ual pain. Geisser and collea gues also argue that PTSD, a specific anxiety disorder, is increasing among in dividuals with chronic pain, specifically those involved in motor vehicle accid ents and work-related injuries. The Interrelationship Between Ch ronic Pain and Somatization The relationship between chronic pain and somatization also bears similarities to the previously discussed re lationships between physical and psychological conditions (depression and anxiety), including the como rbid/independent relati onship and treatment specifics (Barsky, Orav, & Bates, 2005; George, Dannecker, & Robinson, 2005; Woby, Watson, Roach, & Urmston, 2004). Somatization, as defined by the DSM IV-T R (American Psychiatric Association, 2000), is the “presence of physical symptoms that suggest a general medical condition and are not fully explained by a general medi cal condition, by the di rect effects of a substance, or by another mental disorder ” (p.485). Somatization can occur within individuals with work-related injuries, and is often susp ected when objective medical evidence cannot provide a rationale for the pe rceived chronic pain; the injury does not explain the level or degree of chronic pain suggested by the indivi dual with the injury. These situations are heavily sc rutinized in the Workers’ Compensation system. It is often assumed that somatization is nothing more th an a means to receive additional external gains or incentives, such as benefits. However, somatization differs from malingering and factitious disorders in that the physic al complaints are unintentional and, in the

PAGE 27

17 perception of the individual with the chr onic pain, very real (American Psychiatric Association). Somatoform disorders, according to Bars ky, Orav, and Bates (2005) are on the rise. Consistent with the number of patients with somatoform di sorders is the equally climbing costs associated with the diso rders. Barsky and colleagues s uggest that individuals with somatization utilize two times the medical car e and therefore accumulate twice the cost for medical care (estimated at $256 billion) than individuals without somatoform disorders. Given these findings, individuals in the Workers’ Compensation system with unrecognized and untreated somatization are in severe need of recognition and treatment in order to decrease the likelihood of unacknowledged soma tization within the already financially-constrained system. George, Dannecker, and Robinson (2005) sugge st that fear of future pain is nonpredictive of actual pain toleran ce. Their findings suggest that f ear of pain is therefore not based on the amount of current pain experien ced. These findings are contraindicative of previous findings, including those by Crom bez, Vlaeyen, Heuts, & Lysens (1999) and Linton (2000). George and colleagues’ findi ngs might suggest that individuals with somatoform disorders that complain of unrec ognized pain and subsequent disability are not doing so based on the expected pa in attributed from the injury. Woby, Watson, Roach, and Urmston (2004) claim that proper treatment for individuals with chronic pain and somatizati on is not just physical, but psychological as well. By implementing a cognitive-behavio ral intervention, Woby and colleagues were able to reduce somatization in their research sample, which also decreased the level of complaints of chronic pain. The research samp le also reported that participants’ level of

PAGE 28

18 disability had drastically reduced. This research supports a n eed for psychological treatment to accompany physical rehabilita tion for individuals in the Workers’ Compensation system, especially for those with unr ecognized somatoform disorders. The World Health Organization ICF Model The World Health Organization (WHO) ha s been a recognized source for the maintainence and classification of health-re lated information since its creation in 1947 (Ustun, Chatterji, Bickenbach, Kostanjsek, & Schneider, 2003). In order to better facilitate classification, the WHO has develope d and modified several models to better describe human health states, leading up to the presently recognized International Classification of Functioning, Disability, a nd Health, second edition (ICFDH-2), or commonly known as the ICF Model. This model, adopted in 2001, has attempted to cover all facets of an individual’s health status. Prior models that existed to classify health conditions placed great emphasis on four main concepts: Pathology, Impairment Functional Limitation, and Disability. Across this spectrum, previous models utili zed the traditional medical model approach to describe and classify diseases and disabili ties (Bickenbach, Chatterji, Badley, & Ustun, 1999). These models include the Nagi, Institute of Medicine, and th e National Center for Medical Rehabilitation Resear ch Models (Bickenbach, Chatterji, Badley, & Ustun). These three models, with some minor different iation, viewed these four concepts as such: Pathology (Active Pathology, Pathophysiol ogy) – Physiological and biological disturbances and abnormalities. Impairment – Organ or organ system abnormalities resulting in physiological or anatomical dysfunction. Functional Limitation – Physical restrictions interfering with action performance or range of abilities associated with dysfunctional organ or organ system.

PAGE 29

19 Disability – Limitation in task perfor mance and activities associated with functional limitations. The National Center for Medical Research Model also includes the concept of Societal Limitation, which addresses structural and attitudinal barrie rs associated with disabilities that prevent equa l opportunity (Bickenbach, Chatterji, Badley, & Ustun). In 1980, the WHO enacted the first ICF DH model, which consisted of four concepts: Disease, Impairment, Disability, and Handicap. According to this model, Disease referred to abnormal etiology due to physiological or biological abnormalities; Impairment referred to loss of functioning due to abnormal etiology; Disability referred to activity restrictions due to impairmen t; and Handicap referred to disadvantages encountered in society due to disabilities (Ustun, Chatterji, Bickenbach, Kostanjsek, & Schneider, 2003). However, several flaws were pointed out in the logic of this model when examining individuals with disabiliti es. One of the most apparent was that, according to this model, individuals with di sabilities are disadvantaged solely by their disabilities. The model also suggested that impairments and disa bilities are the only causative factor for handicapping situations (Bickenbach, Chatterji, Badley, & Ustun, 1999). No consideration for external or in ternal factors excluding the disease and impairment was available based on the first ICFDH model. Another motivating factor for the decision to revise was the first-pers on language movement, which frowned upon the use of “handicap” in the model terminology (B ickenbach, Chatterji, Badley, & Ustun). In 2001, the ICFDH-2, or ICF model, was adopted in order to alleviate the problems of the ICFDH model, as well as to better enhance the understanding of disease and disability condition for easier classi fication (Ustun, Chatterji, Bickenbach,

PAGE 30

20 Kostanjsek, & Schneider, 2003). The ICF model, which is more comprehensive in recognizing all aspects regarding health states, is illustrated in Figure 2-1. The following definitions have been provi ded by the WHO for the ICF terms (2003): Health Condition – the me dically-recognized disease state or disability. Body Structures – anatomical parts of the body, including the organ and organ systems levels. Body Functions – physiological or psychological functioning. Activity – capacity for task comp letion involved during daily living. Participation – performance of involvement in a social environment. Environmental Factors – external positive or negative conditions that impact participation. Personal Factors – internal positive or negati ve conditions that impact participation. Any deficiency in body function or struct ure constitutes impairment. A lack in activity suggests limitations, and similarly any incapability in participation presents as a restriction. A key feature of this model is that, unlik e the stage-like presentation of former models, individuals can move across this m odel, suggesting a continuum rather than discrete stages. The previous models allowed for this as well, but not with the clarity of the ICF model, which allows consideration for each characteristic. The most significant feature, however, would be the separation fr om the traditional medical model, which viewed disease and disability as a problem th at needed to be medically resolved. The WHO’s ICF Model incorporates the medical, so cial, and psychological aspects of disease and disability, and therefore can be considered to have a biopsychosocial approach (2003). Because of this incorporation, all health-related domains can be included when classifying a given dise ase or disability.

PAGE 31

21 The ICF Model and Chronic Pain Chronic pain, like all disease states and di sabilities, can be described using the ICF Model. Within the Body Function and Struct ures domain of the IC F Model, the obvious item is the presence of pain, which would be the individual’s chief complaint. Depending on the individual, other secondary physical a nd psychological complaints may exist, such as deficiencies in range of motion if the pa in is related to the musculoskeletal system. Maximal performance may be reduced due to chronic pain, and therefore strength, endurance, and muscular stamina may be comp romised. Chronic pain may interfere with cognitive functioning, including attenti on, memory, and decision-making. Several psychological disturbances may coexist with chronic pain, includi ng depression, anxiety, and adjustment disorders (Bloc k, Kremer, & Callewart, 1999). When considering how chronic pain may reduce an individual’s capacity to function, and therefore inhibit Activity, there are numerous phys ical limitations that may exist as a result of chronic pain. Depe nding on the pain location, presentation, and severity, any physical function typically m easured in functional capacity evaluation, a systematic assessment of physical functionality used to evaluate physical impairment and subsequent limitations, may be limited, including sitting, standing, walking, running, climbing ladders or stairs, crawling, squatti ng, kneeling, and lifting behaviors (Brouwer et al., 2003). These limitations may pose serious restrictions in societal situations A critical area where this may occur is within the workpl ace, in which the individual may encounter severe barriers in Participation. If th e individual sustained the impairment and subsequent limitations, a major concern is how these limitations will interfere, if at all, with the essential functions of the indivi dual’s job. If the limitations prohibit the

PAGE 32

22 individual from performing in the job, the individual’s Participation is negatively affected. This restriction is detrimental in numerous ways, include socially, financially, and psychologically for the individual (S immonds, Kumar, & Lechelt, 1996). When examining the contextual factors of the ICF Model, including the Environmental and Personal Factors, Witti nk (2005) provides several examples of potential characteristics to consider. Envir onmental Factors include external items that can positively or negatively impact the othe r domains in the ICF Model, such as Body Function and Structures, Activity, and Partic ipation. Examples of Environmental Factors that may positively impact an indivi dual with chronic pain, specifically a Workers’ Compensation claimant, include: st rong social support system, accommodating employer, cooperative insurance carrier, e fficacious rehabilitation program post-injury, attention to psychological needs, successful pharmacology for pain management, assistive technology, and limited phys ical barriers in the environment. When presented with these positive factors, the impairment, limitation, and restricti on are decreased, thus allowing individual enablement. Negative impacts may include a lack of support, poor employer and insurance carrier relationships, unsuccessf ul medical treatment, ignored psychological needs, and physical environmental barriers. These fact ors increase the impairment, limitation, and restriction, thus instilling individu al disablement. When considering Personal Factors that may positively or negatively impact the other domains as well, special factors to c onsider include the claimant’s age, gender, ethnicity, marital status, education level, c oping mechanisms, motiva tion, pain threshold, and intention in regards to returning to work (Wittink, 2005). All of these Personal

PAGE 33

23 Factors may either enable or disable the in dividual with chroni c pain, depending on how they relate to the other ICF domains. The ICF Model does an excellent job identi fying the key areas of concern regarding disease and disability. However, the IC F Model truly excels in recognizing the contextual factors (Environmental and Pers onal). These two facets can drastically influence the other domains, which completely support the need to move away from a traditional medical model and accept the need to focus on psychological and social concerns. Only when examining a disease or disability while including these matters might one be able to accurately use the ICF Model to describe and classify a health condition at an individual level. For this reason, it might be valuable to identify predictor variables for health conditions. By identif ying consistent and predictable demographic and forensic variables, which would fall into the contextual fact ors (Environmental and Personal) of the ICF model, one might be ab le to predict the leve l of enablement or disablement. For example, if certain dem ographic variables were identified to be predictive of greater levels of depression fo r individuals in the Workers’ Compensation system, rehabilitation professionals may be able to intervene earlier in the rehabilitation process. Psychological Status and Return To Work Outcomes As previously suggested, there are seve ral studies that suggest psychological return-to-work barriers exist. Gardner ( 1991) reported that the l onger the amount of time between injury and return-to-work, the less su ccess found in return-to-work. Suggested in the discussion, Gardner provides several expl anations for why this time period can drastically impact return-t o-work, including the devel opment and persistence of

PAGE 34

24 psychological concerns. Depression and disability have a profound correlational relationship, reported as high as r = .86 (Butchner, 1985). Another study (Fishbain et al., 1993) sugge sts that psychological distress after work-related disability is highly prevalent. Depression, anxiety, and fear of re-injury were all pinpointed as psychological impairme nts that directly affected the success of return-to-work. Sullivan and Stanish (2003) suggest a “Pai n Disability Prevention Program,” which would facilitate successful return-to-work following work-related injury. Surprisingly, this program does not implement physical or occupational therapy interventions, but a cognitive-behavioral treatment intervention focused on reducing psychological return-towork barriers and increasing goal-directed beliefs. This was due to overwhelmingly consistent reports from rehabilitation counsel ors that the psychologica l needs of the pain patients were not being met. Post-partici pation in the program, Sullivan and Stanish reported a 60% increase in return-to-work among their study’s population, which they directly attributed to participati on in the psychological intervention. Depression and anxiety are commonly recognized ps ychological conditions. However, somatization and fear of re-injury are not. Vowles and Gross (2003) examined the concept of “fear-avoidance,” a theoretical model that suggests that individuals with injuries or disability maintain an irrationa l fear of re-injury should physical activity be reinstated. This model portrays a form of an “activity phobia,” particularly for an activity that is identical or similar to the activity resul ting in the injury. It is easy to understand how this psychological condition impairs return-to-work; the individual is simply afraid to perform the job-related functions for fear of re-exacerbating the previous injury.

PAGE 35

25 Given the presence of psychological conditions that act as return-to-work barriers, it is necessary to identify individuals that are at greater risk for given psychological impairments. It is possible to review the general population’s pr evalence and incidence rates for psychological conditions, including the demographic estimates, provided in the DSM IV-TR (American Psychiat ric Association, 2000). Howe ver, neither the DSM IVTR, nor any previously conducted study, has suggested predic tor variables for psychological distress for individuals in the Workers’ Compensation system. These predictor variables have not ye t been examined, even though ther e is significant literature that reports difficulty in return-to-work dire ctly attributed to psychological distress. Demographic and Forensic Variables There are several demographic variables that have been identified as risk variables for certain psychological conditions. The DSM IV-TR suggests that females are more at risk for developing clinical depression than males, 12-25% and 8-15%, respectively (American Psychiatric Association, 2000). This differentiation may be based on cultural norms; males are less inclined to report de pression in order to maintain masculine standards (Dixon & Gatchel, 1999). Age is al so a risk factor for depression, which increases significantly as individuals grow older. A ccording to the DSM IV-TR, depression is unrelated to marital status (American Psychiatric Association, 2000). Several risk factors also exist for gene ralized anxiety disorder. Females are at greater risk for developing an anxiety disord er than males. Approximately 66% of the population with anxiety disorder s is female (American Psychi atric Association). Age has not been found to have a definitive relationshi p with anxiety disorders, as it has been found across the age spectrum (Geisser, R obinson, Miller, & Ba de, 2003; Piccinelli, Patterson, Braithwaite, Boot & Wilkinson, 1999).

PAGE 36

26 Somatization, the belief of having physical problems with no clinical evidence, or the fear of developing pain and other physical problems (American Psychiatric Association, 2000), has very spec ific risk factors. Gender has been found to be unrelated in the general population, however age and mar ital status are profound risk factors. The typical age of onset is in the twenties and thir ties. Somatization is also reported in greater frequency among individuals that are not marri ed, particularly in women (Nakao et al., 2001). Given that risk factors have been id entified with in the general population, an examination of the predictive power for thes e three demographic variables (age, gender, and marital status) in this subpopulation is warranted. In 1986, Hester, Decelles, & Gaddis conducted the Menninger Study, which attempted to identify characteristics that may potentially predict whether or not an individual injured on the job and receiv ing long term disability benefits would successfully return to work or not. Three de mographic variables that they identified included: Age – individuals younger in age were most likely to return to work. Gender – females were most likely to return to work. Marital status – individuals who were single at the time of injury were most likely to return to work. Regarding ethnicity and psychological distre ss prevalence, the scientific literature provides conflicting information. Accordi ng to Zhang and Snowden (1999), African Americans are less likely than Caucasians to experience major depressive disorder, dysthymic disorder, obsessive-compulsive di sorder, although they are more likely to experience phobias and somatization. Accord ing to the researchers, Asian Americans also report fewer mood, anxiety, and somatofo rm disorders than Caucasians. However,

PAGE 37

27 Plant and Sachs-Ericsson (2004) suggest differe nt findings. Accordi ng to their research, minority groups experience depressive symp toms more frequently and have a higher prevalence of major depression disorder th an Caucasians. The DSM IV-TR (American Psychiatric Association (2000) does not specify any differences in disorder prevalence for major depression disorder and generali zed anxiety disorder. It does provide a cautionary alert for professionals to consid er “ethnic and cultural specificity” while diagnosing major depressive disorder, partic ularly in symptoms description (p. 353). Only two specific ethnicities (Greek and Pu erto Rican men) are mentioned as ethic groups that evidence significan t epidemiological differences in somatization disorder prevalence. Williams, Takeuchi, & Adair (1992) provide an explanation for why there are inconsistent findings regard ing the relationship between ethnicity and psychological distress. They argue that certain moderator variables, such as socioeconomic status and substance abuse, need to be considered wh en examining this relationship. This is especially true when looking at low socioec onomic status populations. According to the researchers, Caucasian males with low soci oeconomic status experience significantly higher rates of psychiatric disorders th an African American males with low socioeconomic status. When considering substance abuse as a moderator variable, African American women with low socioec onomic status have the highest rates of psychiatric illness. Because of situationa l contexts such as socioeconomic status, substance abuse, and other unidentified potenti al moderator variables, establishing a clear relationship between ethnicity and ps ychological distress is difficult.

PAGE 38

28 The context of the situation, namely the Workers’ Compensation system, must also be considered when determining predictor vari ables. Three forensic variables, or legal system variables, are known to be associ ated with successful versus non-successful return-to-work outcomes (Blackwell, Leiere r, Haupt, & Kampitsis, 2003; McGeary et al., 2003; Selander, Marnetoft, Bergroth, & E kholm, 2002; Severeijns, Vlaeyen, van den Hout, & Weber, 2001; Tate, 1992). These three variables include the injured individual’s satisfaction with his or her former employer, insurance carrier, and attorney (if retained) since the time of injury. Thes e three entities are directly i nvolved with the individual’s overall perception of the Workers’ Compensatio n system, their level of participation in the system, and the overall process (Tate, 1992). Selander, Marnetoft, Bergroth, and E kholm (2002) stated that the injured individual’s perception of how the former employer viewed them in the Workers’ Compensation system has a significant impact on successful return-t o-work. Individuals that thought the employer believed them to be malingering or cheating the system showed greater difficulty in returning to work after rehabilitation. Selander and colleagues also report that a good working relationship with the Workers’ Compensation insurance carrier is indicativ e of a successful return-to-wo rk. It was not determined, however, whether the psychologica l health of the individuals was affected by the level of satisfaction with either party. Satisfaction with one’s attorney was established as a significant predictor variable for return-to-work as an outcome variable by Blackwell, Leirer, Haupt, and Kamptisis (2003). Individuals that reporte d satisfaction with the attorn ey’s work, participating in

PAGE 39

29 the decision-making process, and a feeling of control over their legal situation had a greater frequency of return-towork than those who did not. Although all of these demographic variable s have been found to either a) be predictive of clinical psychological conditions, or b) be predictive of successful return-towork, none of them have been evaluated to be predictive of clinical psychological conditions in a return-to-work system. It seems possible that that presence of psychological distress may influence the predic tive power of certain variables, which in turn suggests that psychological impairment may indeed be a determining factor for return-to-work. It is because of this possibility that an ex amination of the predictability of select variables for psychol ogical conditions is needed. Return To Work Outcomes in Workers’ Compensation There have been several studi es examining predictor variables and “red flags” that may be useful for rehabilitation professiona ls when focusing on return-to-work as a successful outcome. However, the studies ar e limited in scope within the medicolegal context of the Workers’ Compensation syst em. Most studies focus on chronic pain patients and do not specify any disability co mpensatory system. Demographic variables have been suggested, including the patient’s ag e, gender, and educat ion level (Blackwell, Leirer, Haupt, & Kampitsis, 2003). Type of employment and job satisfaction has also been reviewed. In regards to the presence and coexistence of chronic pain and depression, a couple of studies s uggest that a) both drastically inhibit a successful returnto-work, and b) rehabilitation is not succe ssful unless both are tr eated appropriately (Adams & Williams, 2003; Fishbain et al ., 1993; Haslam-Hapwood, 2003; Selander, Marnetoft, Bergroth, & Elkholm, 2002; Watson, Booker, Moores, & Main, 2004).

PAGE 40

30 Fishbain et al. (1993) examined pain tr eatment centers and their efficacy in returning injured individuals to work. Upon revi ew of the findings, it was determined that pain clinics are not focusing on all needed aspe cts, including the psyc hological status of the patients. Difficulty in return-to-work was noted in individuals with depressive symptoms, which was not initially identified during the rehabilitation at the pain clinic. Selander, Marnehoft, Bergroth, and Ekholm (2002) reported that chronic pain and psychological distress dramatically impair re turn-to-work. Both factors were greater barriers for returning to work when they we re coexistent and not recognized as such. Adams and Williams (2003), while examini ng individuals with chronic upper limb pain, realized that “many work rehabilita tion programs neglect ps ychological aspects” (p.103). The researchers also c oncluded that “many work-based health services may also be poorly informed about chronic pain and its best management, and their decisions may serve the employers’ needs rather than the wo rkers’” (p. 103). Clear ly, according to this study, there may be discrepancies regardi ng appropriate rehabilitation for injured workers. Watson, Booker, Moores, and Main (2004), after studying differences in return-towork outcomes for individuals with chronic low back pain, determined that one of the most significant variables that prevented indi viduals from successful return-to-work was psychological distress, specifically depres sion. The researchers also noted that individuals who were out of work longer due to chronic pain also reported higher levels of depression. Haslam-Hapwood (2003), of the Menninger Clinic, reports th at successful rehabilitation should involve a comprehensiv e treatment strategy focusing on all aspects

PAGE 41

31 of the individual. By following a biopsyc hosocial approach, the individual receiving rehabilitation would have the greatest chan ce of having all needs met, as all aspects pertaining to the individual’s he alth concern would be consider ed. This is consistent with how the ICF Model’s contextual factors, both the personal an d the environmental, impact the medical condition of intere st, thus supporting the fore mentioned use of the model for chronic pain. Psychological Treatment Cost-Effectiveness The cost-effectiveness of psychological treatment for chronic pain has been evaluated to see if psychological thera py should be included in the traditional rehabilitation plan (de Boer, Wijker, & de Haes, 1997; Turk, 2002; Turk & Burwinkle, 2005; Turk & Okifuji, 2002). Turk and Burwinkl e report that this li ne of thought is not new, as the Commission on the Accreditation of Rehabilitation Faci lities (CARF) has required a psychological treatment component in order for facilities and programs to become certified. However, the question still remains as to wh ether larger entities, such as disability compensatory systems (Workers ’ Compensation, Long-term disability, etc.) would benefit from this inclusion. Tu rk and Burwinkle suggest the following considerations: Chronic pain involves psychosocial and beha vioral components, as well as physical ones. Patients with complex chronic pain problems are best treated within a rehabilitation model and by a team of rehabilitation pr ofessionals (including psychological). The treatment must address the pain itself a nd not just be a search for hidden causes and specific remedies for these causes. The treatment must address the restorati on of well-being and not just aim at the alleviation of symptoms.

PAGE 42

32 Emphasis needs to be given to strategies that will facilitate patients’ ability to selfmanage their situations for extended periods (p. 608). Given these considerations, it is easy to see how the two most common forms of chronic pain management, which are pharm acology and surgery, do not address all of these considerations (Turk & Burwinkle). The researchers su ggest that by developing an interdisciplinary rehabilitati on plan, including psychological treatment, treatment costs for individuals with chronic pain will be mi nimized while treatment benefits will be maximized. The benefits, based upon self-re ported measures from patients, include improved functionality and quality of life, while the research ers’ objective measurements found additional benefits, including faster retu rn to work, a decrease in pain medication use and dependency, and a more efficient utilization of health care resources. Fiscally, Turk (2002) reports remarkab le potential financ ial incentives for implementing psychological treatment into re habilitation programs for individuals with chronic pain. Turk estimated that health ca re savings per chronic pain patient annually could exceed $78,000. This is more impressive when looking at the situation globally. Although there are varying statis tics for the actual number of individuals with chronic pain, the health care savings for all chronic pain patients, after th e implementation of an interdisciplinary pain rehabilitation progr am including psychological treatment, could conservatively exceed $45 billion (Turk). Rationale Given the lack of literature pertaining to unmet psychological need and prediction of such need in individuals in the Workers’ Compensation system, it is critical to conduct an analysis to identify predictor variab les for potential unmet psychological distress. Within the Workers’ Compensation system, a ll parties would benef it from acquiring this

PAGE 43

33 information. Individuals with work-related injuries might be more likely to receive needed psychological care, coexisting chronic pain might be reduced, and return to work might be more successfully achieved. Insurance carriers, although having to initially invest by funding mental health care, might increase profits by expending fewer total dollars when claimants return to wo rk in shorter periods of time. This study aims to identify key predictor variables in hope of assisting individuals with work-related injuries and the professiona ls that work with them to facilitate a successful return-to-work. Figure 2-1. World Health Or ganization’s ICF Model. Health Condition Body Structure & Body Function Activity Participation Personal Factors Environmental Factors

PAGE 44

34 CHAPTER 3 METHODOLOGY Individuals in the Florida Workers’ Co mpensation system may not be receiving appropriate mental health care benefits. In order to demonstrate the unmet need, the presence of mental health concerns must be identified within the Workers’ Compensation claimant population. This chapter presents the research questions, hypotheses, and methodology, including the sampling techniqu e, research design, measurements, and statistical analysis. Research Questions The rationale for conducting this study was to determine to what extent select demographic and forensic variables are able to predict the presence of clinical depression, anxiety, and somatization, specifically in Work ers’ Compensation clients. In order to accomplish this, a representative sample of Workers’ Compensation clients was surveyed in order to examine the ten demographic and fore nsic variables of inte rest, including: age, gender, ethnicity, marital status amount of time since injur y, perceived level of social support, perceived level of pa in interference, satisfaction level with former employer, satisfaction level with Workers’ Compensati on insurance carrier, a nd attorney retention status. The satisfaction leve l with attorney was also examined in a supplementary analysis for those who have re tained attorneys. The same clients were also evaluated in order to determine the present level of clin ical depression, anxiet y, and somatization.

PAGE 45

35 Hypotheses The following research hypotheses, based on the aforementioned literature, were tested in this study: Hypothesis #1: It is hypothesized that the following variables will be significant in predicting clinical depression: older age a nd female gender, as well as interactions between these variables. Hypothesis #2: It is hypothesized that the following variables will be significant in predicting clinical anxiety: female gender, as well as interactions between these variables. Hypothesis #3: It is hypothesized that the following variables will be significant in predicting clinical somatization: younger ag e and single marital status, as well as interactions between these variables. Population The research population included injured workers entering the State of Florida Workers’ Compensation system, the Bureau of Rehabilitation and Reemployment Services. Because Workers’ Compensation is state regulated, every state has differences in their designated programs (FWCI, 2003). Th is study examined indi viduals specifically in the Florida Workers’ Compensation syst em. Both genders and all ethnicities are represented in this resear ch population. The research popul ation’s age was considered “working age,” ranging from approximately sixteen to the upper sixties (FWCI). The individuals were recruite d to participate prior to the beginning of the Bureau of Rehabilitation and Reemployment Services mandatory orientation program for new clients entering the program. The rehabi litation counselor conduc ting the orientation informed the injured individuals that if interested and eligible, they may voluntarily participate in a research study. The rehabil itation counselor exited the conference room,

PAGE 46

36 ensuring that the staff at the Bureau of Rehabilitation and Reemployment Services remained unaware of which clients chose to participate in the study. The principal investigator entered the room to meet the in terested individuals a nd discuss the study’s purpose and eligibility require ments in more detail. The injured individuals were informed of the eligibility requirements, which include 1) having a compensable injur y, as defined by the Florida Workers’ Compensation statutes (FWC I, 2003); and 2) placement at Maximum Medical Improvement (MMI) status by a treating Workers’ Compensation physician. The MMI status serves to screen ou t individuals with future me dical interventions, such as upcoming surgeries and physical therapy. Any potential medical interventions may alter the permanent restrictions provided by the tr eating physician, which in turn may act as a moderator variable, as it may affect the psychological condition of the clients. All of the participants were actual, curre nt Workers’ Compensation clients and at the time had reached medical stability accord ing to the treating phys icians. All of the participants were medically ready for return to work and at the time had been determined suitable for vocational rehabilitation. For their voluntary participat ion, each individual was compensated ten dollars for their time. Participants were directed not to discuss the study with others after completing the assessments in order to ensure that the participants remain anonymous to the Bureau of Rehabilitation and Reemployment Services, as well as to the Principal Investigator. The participants were instructed not to provide any identifying information to the Principal Investigator, such as name, addre ss, or social security number to ensure anonymity. The participants were compensated in cash as opposed to check in order to

PAGE 47

37 ensure compliance with this as well. Docume ntation of the Bureau of Rehabilitation and Reemployment Services’ permission for participant recruitment was obtained in accordance to IRB protocol (see Appendix A). Sampling Technique The research sample was acquired fro m two Bureau of Rehabilitation and Reemployment Services offices, located in two North Florida ci ties. Both cities are vastly different in population per capita, urban ge ography, and prominent industry types. A convenience sample was obtained from each of the Bureau of Rehabilitation and Reemployment Services offices. Convenien ce sampling was utilized, as participant recruitment was based on orientation atte ndance and therefore wa s beyond the Principal Investigator’s control. Research Design The research design for this study was a quasi-experimental one-group design, as only one group of participants was evalua ted without the use of random assignment (Dooley, 2001). All participants were compar ed together in order to determine the predictive capability of the te n select demographic and forens ic variables in relation to the participants’ psychological status. The four demographic variables analyzed included the injured individual’s age, gender, ethnic ity, and martial status. The six forensic variables analyzed included th e amount of time out of work due to injury, level of social support, level of perceived pain interference, the individual’s satisfa ction with his or her employer at the time of the injury onset, satisfaction with his or her Workers’ Compensation insurance carrier, and whether or not the indivi dual retained an attorney. The injured workers’ satisfaction with his or her attorney was also collected for a

PAGE 48

38 supplementary analysis. All ten of these demographic and forens ic variables served as the independent variables in the design. The three psychological conditions of inte rest included depression, anxiety, and somatization. These three conditions served as dependent variables and were measured to examine the predictive capabilit y of the independent variables. Measurement All participants completed two instruments: The Betters Injured Worker Inventory (BIWI) and the Pain Patient Profile (P-3). Half of the sample completed the two assessments in this manner, while the other half completed them in a reverse fashion to eliminate order effects. Completion of bot h instruments took approximately thirty minutes for the participants. Betters Injured Worker Inventory (BIWI) The BIWI, created by the principal investig ator, is a ten-question survey that has been developed to measure the ten inde pendent variables (s ee Appendix B). The independent variables for th is study included the select demographic and forensic variables that were measured for predictive capability of psychological distress. All of the independent variables were collected th rough administration of the Betters Injured Worker Inventory (BIWI). Th e ten variables included: Age – measured on a continuous scale, in which the participants provide their age in years. Gender – measured on a categorical scale representing either male or female gender. Ethnicity – measured on a categorical scale representing Caucasian, African American, Hispanic/Latino, As ian American, or Other.

PAGE 49

39 Marital Status – measured on a categorical scale representing current marital status, including single – never marri ed, living with partner, married, separated, divorced, or widowed. Amount of Time out of Work due to Inju ry – measured on a continuous scale, in which the participants provide the amount of time in months. Amount of Perceived Social Support – m easured on a Likert scale representing perceived amount of social support received while in jured with the following choices: very low, low, neutral, high, and very high. Level of Pain Interference – measured on a Likert scale representing perceived level of pain interference with daily activ ities with the following choices: very low, low, neutral, high, and very high. Satisfaction with Employer – measured on a Likert scale repres enting current level of satisfaction with the fo llowing choices: very low, lo w, neutral, high, and very high. Satisfaction with Carrier – m easured on a Likert scale re presenting current level of satisfaction with the following choices: ve ry low, low, neutral, high, and very high. Satisfaction with Attorney (if retained) – measured on a Likert scale representing current level of satisfaction with the fo llowing choices: very low, low, neutral, high, and very high, as well as the presence of a “no attorney retained” option for individuals who are not legally represented. During its creation and utilization in a pilot study, the BIWI underwent expert review by eight individuals, including tw o academicians in research methodology, as well as six vocational rehabilitation profe ssionals working in Workers’ Compensation rehabilitation to ensure validity and prope r survey design. Minor modifications were made after the review, including adding four new independent variables, survey design format, and wording. Ethnicity, amount of time out of work due to injury, social support, and pain interference were not initially on the BIWI, but were added after the review in order to analyze the predictive capability of each. Age was modified into a continuous variable rather than a categor ical variable for greater va riance. Format changes made after the review included rearranging the demographic question sequence and the

PAGE 50

40 inclusion of an explanatory introduction. Wo rding changes involved clarifying “former employer,” as the purpose of this question wa s to assess satisfaction with the employer at the time of the work-related injury and not another former employer. Each BIWI was given an identification number that will be paired with a Pain Patient Profile to ensure accurate data entry and analys is. Face validity has been determined through a pilot study described herein, as well as through the expert review. Pain Patient Profile (P-3) The dependent variables for this study, wh ich were measured by the Pain Patient Profile (P-3), included the pr esence of three select psychol ogical conditions: depression, anxiety, and somatization. Clinical presence was established through administration of the P-3, a valid and reliable measure for the three variables, specifically normed for injured populations such as Workers’ Compensation (Tollison & Langley, 1995). The P-3 is a forty-four item self-report, multiple-choice instrument created to determine how psychological factors, specifi cally depression, anxiety, and somatization, correlate with physiological pain measuremen ts (Tollison & Langley). The P-3 provides valid and reliable measures of these variab les, which are highly useful for comparison with physical evaluative measuremen ts, such as functional capacity. The P-3 has many beneficial uses in the field of rehabilitatio n. It is highly valuable for this study in that it has been validly and reliably tested to measure psychological variables associated with pain and the direct influences towards the demonstration of pain symptoms (Tollison & Langley). The P-3 has also been recognized as an appropriate instrument for assessment of pain patients suffering from accident-related pain, including Workers’ Compensation claims resulting in impaired functionality (Tollison & Langley).

PAGE 51

41 The P-3 has three scales: Depression, A nxiety, and Somatization. A Validity Index has also been created to monitor patient symptom magnification (T ollison & Langley). The Depression Scale consists of fourteen items that measure components associated with depressed mood, including helplessness, hopelessness, and self -worth. Consistency between this scale and the DSM-IV TR cr iteria for depression ha s been established (Tollison & Langley). The T-Scores on this sc ale for pain patients range from 29-71, with 29-45 representing below-average depressed mood for pain patients. This suggests positive psychological health, including optimism and confidence. The next mini-range includes 47-54, which represents average depr essed mood for pain patients. Individuals within this range demonstrate dissatisfacti on with life and frustration. The last minirange, 55-71, is recognized as above-average depressed pain patients, suggesting lack of motivation, sadness, and co-existing physiologi cal problems, including sleep and appetite imbalances. Individuals who score within this range are potentially appropriate candidates for mental health considerations (Tollison & Langley). The Anxiety Scale consists of 12 items and analyzes factors associated with emotional instability, such as nervousness and restlessness (Tollison & Langley). The possible T-Scores attainable for this s cale range from 29-71, w ith 29-44 representing below-average anxiety for pain patients. These individuals dem onstrate relaxation and security. T-Scores ranging from 46-55 are clas sified as average anxiety measurements for this population, implying tension and difficulty in relaxing. The final mini-range, 56-71, is considered above-average anxious indi viduals. These patients experience apprehension and a sense of control loss. In dividuals with above-ave rage scores on this

PAGE 52

42 scale are typically found to have difficult ies in completing physical rehabilitation (Tollison & Langley). The Somatization Scale consists of 13 ite ms focusing on the patient’s concern for his or her physical health st atus. T-Scores range from 2169, with below-average scores including 21-44. These individuals do not demonstrate any major concerns for their current pain. The average T-Scores, 46-55, re present concerns for physical health, but a lack of an obsession over the pain expe rienced. The above-average scores, 56-69, suggest unhealthy distribution of attention towards pain and health status, which can be considered obsessive. Individua ls with these scores have ma jor difficulties participating in rehabilitation (Tollison & Langley). The Validity Index consists of five items and was implemented to pinpoint symptom magnification. It can also detect testing error, such as comprehension and response biases (Tollison & Langley). Scores range from 5 to 15. A cutoff value of 11 has been established, suggesting scores higher than 11 represent invalidity. Any omitted items occurring in any of the scales or the Va lidity Index also repres ents an invalid test. Scores of 9 or 10 should be considered as questionable and require further examination by the administrator. Scores below nine suggest validity (Tollison & Langley). During the development of the P-3, both the validity and reliability characteristics were assessed. Validity was determined by utilizing scale interc orrelations (r = .58 to .73), as well as through comparison with sel ect scales (Depression = .82, Hysteria = .62, and Hypochondriasis = .65) of the Minnesota Multiphasic Personality Inventory (MMPI) to establish significant convergent validity (Tollison & Langley). Willoughby, Hailey, & Wheeler (1999) examined the construct validity of the P-3, finding strong positive

PAGE 53

43 correlations with three other instruments ev aluating the same constructs as the P-3 subscales. These three instruments included the Beck Depression Inventory, the anxiety scale of the State-Trait Anxiety Inventor y, and the somatization scale of the Brief Symptom Inventory. Reliability was established by using both test-retest reliability for each scale (Depression = .99, Anxiety = .98, Somatization = .98), as well as Cronbach’s alpha, for which the alphas ranged from .70 to .91, sugge sting significance (Tollison & Langley). The P-3 has been normed for both pain and non-pain populations. This is done by taking the raw scores for each scale and conve rting them to T-scores based on a normal curve. A normal curve has been establishe d for pain and non-pain populations to allow for comparison between populati ons (Tollison & Langley). Statistical Analysis After scoring both instruments for each partic ipant, all of the data was entered into SPSS v.13 for statistical an alysis. In order to determine th e predictive capability of the select demographic and forensic variables in regards to the psychol ogical condition of the injured individuals, backwards elimination multiple regression analyses (p < 0.10) was conducted for each of the three dependent variab les. A total of six backwards elimination multiple regression analyses were conducted. Th e first three analyses examined the ten independent variables for each of the three dependent variables. The second three analyses examined the same independent va riables and dependent variables, except substituting the satisfaction le vel with the retained attorney in place of the attorney retention status. These supplementary analyses therefore, served to provide predictive capability of the ten variables for the partic ipants in the sample who had retained an attorney.

PAGE 54

44 According to Mertler and Vannatta (2005), “Regression analysis procedures have as their primary purpose the development of an equation that can be used for predicting values on some dependent variable for all members of the population.” (p.165). Backwards multiple regression analysis, one of the several forms of stepwise regression analysis, was used, as backwards multiple regression systematically removes the least significant variable after computing a partia l F-test until only statistically significant independent variables remain and can be id entified (Mertler & Vannatta). Backwards multiple regression analysis, therefore, allows for a more conservative approach toward identifying prediction (Stevens, 1992). A Pearson-product moment correlation matrix was also conducted for the independent variables, as well as a second corr elation matrix for the dependent variables, to determine the relationship of the variab les within the research population (Mertler and Vannatta, 2005). In order to determine an adequate samp le size, a power analysis was conducted using G*POWER, a power calculator so ftware program. G* POWER has been determined to be statistically precise for pr oviding a priori power analyses (Erdfelder, Faul, & Butchner, 1996; Thomas & Krebs, 1997). According to a power analysis with = .05, power = .80, and a desired medium effect size of f2 = .15, an n = 118 is an appropriate target sample size. The power analysis is based on all ten independent variables, thereby allowing a comprehensive analysis of the variables and potential interactions. Pilot Study Twelve participants completed the BIWI and P-3 as part of a pilot study. The twelve participants were conveniently samp led from the two Bureau of Rehabilitation

PAGE 55

45 and Reemployment Services offices. They were recruited from the Bureau of Rehabilitation and Reemployment Services prior to the program orientation. All participants had compensable injuries and ha d been placed at MMI at the time of the administration. The participants were verbally provided informed consent, as no personal identifying information, including names and si gnatures, were collected by the principal investigator. Upon the completion of consent for participation, the participants were instructed on how to complete the BIWI a nd P-3. Any clarification questions were answered prior to the group administration. On ce the individual participants completed the instruments, they were compensated twenty dollars for their participation. The instruments were paired based on an identi fication number to ensure accurate data analysis. The twelve pa rticipants included nine males a nd three females, ranging in age from the 19-26 to 57-64 age categories on th e BIWI. As fore mentioned, age was modified into a continuous variab le to achieve greater variance. The results for the backward elimination multiple regression analyses for the six initial independent variables (participant’s ag e, gender, marital status, satisfaction with former employer, satisfaction with insurance carrier, and satisfaction with attorney) and each of the three dependent variables (depre ssion, anxiety, and somatization) are shown in Tables 3-1, 3-2, and 3-3. The age and mar ital status categorical data were collapsed for analysis, with age being collapsed into two categories: 36 years of age and younger, and 37 years of age and older. This was done since 36 years of age was the median age attainable on the age scale. The marital status data was collapsed into two categories: no marital history and marital history.

PAGE 56

46 The backward multiple regression analysis conducted to determine the predictability of depression i ndicated that the select demogr aphic and forensic variables did not account for a significant am ount of the depression variability, R2 = .095, F (1,11) = 1.04, p = .331. This suggests that, within the pilot sample, there was no predictive relationship between the independent vari ables and depression (see Table 3-1). The backward multiple regression analysis conducted to determine the predictability of anxiety indi cated that the select demogra phic and forensic variables did not account for a significant amount of the anxiety variability, R2 = .204, F (1,11) = 2.57, p = .140. This suggests that, within the pilot sample, there was no predictive relationship between the independent variable s and anxiety (see Table 3-2). The backward multiple regression analysis conducted to determine the predictability of somatization indicated that the select demo graphic and forensic variables did not account for a significant am ount of the somatization variability, R2 = .181, F (1,11) = 2.22, p = .168. This suggests that, within the pilot sample there was no predictive relationship between the independent vari ables and somatization (see Table 3-3). However, when examining the level of signi ficance, the coefficients suggest values moving towards significance. It is probable that the variables will become significantly predictive once the complete research sample of n = 118 participants is in the study, as opposed to the sample size of n = 12 used in the pilot study. Ther e is also a clear indication that, although not pres ently at a significant level du e to low sample size, there are certain independent variab les that are more predictiv e of depression, anxiety, and somatization than the other independent variab les. Additionally, the modification of the age variable, as well as the inclusion of th e new variables, may also provide a better

PAGE 57

47 indication of significance. It is hypothesized that more of th e variables will prove to be statistically and clinically significant with an appropriate research sample size. Upon the completion of the instruments, the participants were asked for their opinions regarding the BIWI. Only one sugge stion was made, whic h involved enlarging the font size from 10 to 12 for easier reading.

PAGE 58

48 Table 3-1. Backwards Elimination Multiple Regression Analysis for Demographic and Forensic Variables Predic ting Depression – Pilot Study Regression Models B SE B Model 1a Age 3.896 6.490 .241 Gender -6.877 7.820 -.373 Marital Status .104 1.649 .031 Employer Satisfaction 1.430 2.612 .217 Carrier Satisfaction -2.129 4.119 -.255 Attorney Satisfaction -2.217 2.092 -.393 Model 2b Age 4.029 5.601 .249 Gender -7.090 6.435 -.385 Employer Satisfaction 1.382 2.283 .210 Carrier Satisfaction -1.970 2.968 -.236 Attorney Satisfaction -2.229 1.902 -.395 Model 3c Age 4.891 5.166 .302 Gender -7.036 6.136 -.382 Carrier Satisfaction -1.582 2.764 -.190 Attorney Satisfaction -2.115 1.805 -.375 Model 4d Age 4.148 4.785 .256 Gender -8.052 5.621 -.437 Attorney Satisfaction -2.165 1.725 -.384 Model 5e Gender -7.720 5.530 -.419 Attorney Satisfaction -2.310 1.693 -.409 Model 6f Gender -5.667 5.546 -.307 Note: aR2 = .383, Adjusted R2 = -.357 bR2 = .383, Adjusted R2 = -.132 cR2 = .345, Adjusted R2 = -.030 dR2 = .314, Adjusted R2 = .057 eR2 = .250, Adjusted R2 = .083 fR2 = .095, Adjusted R2 = .004

PAGE 59

49 Table 3-2. Backwards Elimination Multiple Regression Analysis for Demographic and Forensic Variables Predic ting Anxiety – Pilot Study Regression Models B SE B Model 1a Age 5.158 5.010 .358 Gender 3.542 6.037 .216 Marital Status 1.036 1.273 .344 Employer Satisfaction 2.747 2.016 .469 Carrier Satisfaction -5.425 3.179 -.707 Attorney Satisfaction -.153 1.615 -.031 Model 2b Age 5.200 4.559 .361 Gender 3.702 5.297 .226 Marital Status 1.047 1.158 .347 Employer Satisfaction 2.734 1.838 .467 Carrier Satisfaction -5.268 2.896 -.710 Model 3c Age 5.719 4.331 .397 Marital Status .702 1.008 .233 Employer Satisfaction 2.556 1.753 .436 Carrier Satisfaction -4.330 2.471 -.583 Model 4d Age 6.627 3.994 .460 Employer Satisfaction 2.231 1.634 .381 Carrier Satisfaction -3.442 2.047 -.464 Model 5e Age 7.961 4.055 .552 Carrier Satisfaction -2.813 2.088 -.379 Model 6f Age 6.514 4.066 .452 Note: aR2 = .536, Adjusted R2 = -.020 bR2 = .536, Adjusted R2 = .148 cR2 = .498, Adjusted R2 = .211 dR2 = .463, Adjusted R2 = .262 eR2 = .338, Adjusted R2 = .191 fR2 = .204, Adjusted R2 = .125

PAGE 60

50 Table 3-3. Backwards Elimination Multiple Regression Analysis for Demographic and Forensic Variables Predicti ng Somatization – Pilot Study Regression Models B SE B Model 1a Age -1.555 8.161 -.084 Gender -1.814 9.834 -.086 Marital Status -8.232 2.074 -.021 Employer Satisfaction 2.087 3.285 .277 Carrier Satisfaction -.429 5.179 -.045 Attorney Satisfaction -3.078 2.630 -.477 Model 2b Age -1.661 7.041 -.090 Gender -1.644 8.090 -.078 Employer Satisfaction 2.125 2.870 .282 Carrier Satisfaction -.555 3.732 -.058 Attorney Satisfaction -3.068 2.391 -.475 Model 3c Age -1.852 6.421 -.100 Gender -1.981 7.204 -.094 Employer Satisfaction 2.033 2.599 .270 Attorney Satisfaction -3.077 2.217 -.476 Model 4d Age -1.924 6.034 -.104 Employer Satisfaction 1.976 2.437 .263 Attorney Satisfaction -2.913 2.008 -.451 Model 5e Employer Satisfaction 1.735 2.198 .231 Attorney Satisfaction -2.822 1.886 -.437 Model 6f Attorney Satisfaction -2.750 1.848 -.426 Note: aR2 = .255, Adjusted R2 = -.639 bR2 = .255, Adjusted R2 = -.366 cR2 = .252, Adjusted R2 = -.175 dR2 = .244, Adjusted R2 = -.040 eR2 = .234, Adjusted R2 = .064 fR2 = .181, Adjusted R2 = .099

PAGE 61

51 CHAPTER 4 RESULTS Post-Hoc Power Analyses Post-hoc power analyses were conducted in order to determine statistical power for each of the three dependent variables. This wa s done because the resear ch sample size (n = 111) did not fulfill the target sample size (n = 118), as indicated by the a priori power analysis. Upon completion of the analys es, significant power was obtained for somatization in both the first and supplementary backward elimination multiple regression analyses (.995 and .987, respectively). Significant power was also established for depression in the first analysis, but not for the supplementary analysis (.962 and .732, respectively). Significant power was not met for either of the analyses for anxiety (.503 and .431, respectively). Descriptive Statistics Descriptive statistics were conducted for the independent and dependent variables. The four demographic independent variables included age, gender, ethnicity, and marital status. The six forensic indepe ndent variables included time si nce injury, perceived social support, perceived pain interfer ence, satisfaction with employer, satisfaction with carrier, and attorney retainment status. Satisfaction levels with attorn ey were also measured for a supplementary analysis. The three depende nt variables included three psychological conditions depression, anxiety, and somatization. Descriptive data is reported as the mean the standard deviation.

PAGE 62

52 Independent Variables The research sample size recruited include d 111 participants. The mean age for the sample was 41.32 8.60 years. The range was 19 to 62 years, although nearly half (48.6%) of the individuals were between the ages of 40-49. Se venty-two of the participants were male, representi ng 64.9% of the research sample. Two variables, ethnicity and marital stat us, were collapsed for data analysis purposes in order to create dich otomous variables. Ethnicity, pr ior to the variable collapse for the regression analyses, included 72 Ca ucasians (64.9%), 29 African Americans (26.1%), 3 Hispanics (2.7%), 1 Asian Ameri can (.9%), and 6 individuals (5.4%) who indicated Other. Once collaps ed into Caucasian and NonCaucasian categories, 64.9% were Caucasian and 35.1% were Non-Caucasian. Regarding current marital status, prior to the variable collapse for the regression analyses, 57 participants we re currently married (51.4%), 23 were single with no marital history (20.7%), 21 were separa ted or divorced (18.9%), 6 were living with a partner (5.4%), and 4 were widowed (3.6%). Once collapsed into Presently Married and Not Presently Married categories, 51.4% were presently married and 48.6% were not. Time since injury was measured to dete rmine how long the individual had been out of work and in the Workers’ Compensati on system. The mean time since injury was 11.68 7.18 months. The mode time was 6 months and the range was 2 to 48 months. Perceived social support leve ls were reported as neutra l by 30 participants (27%), low by 29 participants (26.1%), high by 28 participants (25.2%), very low by 16 participants (14.4%), and very high by 8 participants (7.2%).

PAGE 63

53 Regarding perceived pain interference, 67 participants (60.4%) re ported that their current pain levels highly interfered with their daily activities. Other measures of perceived pain interference included 27 participants (24.3%) reporting very high interference, 11 participants (9.9%) reporting neutral interference, and 6 participants (5.4%) reporting low inte rference. No participants repor ted very low levels of pain interference. The participants’ satisfaction levels with their former employers indicated that 51 participants (45.9%) reported ve ry low satisfaction since their injury. Other satisfaction levels included 38 claims of low satisfacti on (34.2%), 20 claims of neutral satisfaction (18%), and 2 claims of very high satisfact ion (1.8%). No participants reported high satisfaction. Forty-four participants (39.6 %) reported very low satisf action with their Workers’ Compensation insurance carriers since their injury. Other carrier satisfaction levels included 40 claims (36%) of low satisfaction, 21 claims (18.9%) of neutral satisfaction, 5 claims (4.5%) of high satisfaction, and 1 claim (.9%) of very high satisfaction. Regarding Workers’ Compensation attorney retention, 93 participants had retained an attorney, representing 83.8% of the research sample. Of the 93 that had attorney representation, 35 participants (31.5%) reported a neutral leve l of satisfaction with their attorney. Other satisfaction levels include d 24 claims of high satisfaction (21.6%), 23 claims of low satisfaction (20.7%), 7 claims of very high satisfaction (6.3%), and 4 claims of very low satisfaction (3.6%).

PAGE 64

54 Dependent Variables Descriptive statistics we re conducted on the P-3 measurements of depression, anxiety, and somatization. These measurements in clude the t-scores and clinical levels of each psychological condition as determined by the P-3. Regarding depression, the mean P-3 t-score measured within th e research sample was 47.64 7.06. The range included t-scores of 31 through 61. Using these t-scores, 51 participants (45.9%) demonstrat ed average levels of depres sion for pain patients. Other levels of depression included 32 particip ants (28.8%) demonstrating below average depression and 28 participants (25.2%) demonstr ating above average levels of depression for pain patients, as determined by the P-3. When examining anxiety, the mean P-3 t-score measured was 43.17 6.47. The range was 31 to 56. Given these t-scores, 60 participants (54.1%) demonstrated below average levels of anxiety for pain patien ts. Forty participants (36%) demonstrated average levels of anxiety and 11 participants (9.9%) demonstrated above average levels of anxiety for pain patients. A mean t-score of 48.80 7.50 was obtained by participants on the P-3, which measured somatization. The range was 30 to 63. Using these t-scores, 51 participants (45.9%) demonstrated average levels of somatization for pain patients. Twenty-six participants (23.4%) demonstrated below average levels of somatization and 34 participants (30.6%) demonstrated above average levels of somatization for pa in patients. Backward Elimination Multip le Regression Analyses For inferential statistical analysis, backward elimination multiple regression analysis was used in order to determine wh ich variables were most predictive of the

PAGE 65

55 psychological conditions. Analyses were c onducted for all three dependent variables using the ten demographic and forensic vari ables. Supplementary analyses were also conducted using only the data from the 93 part icipants who have at torney representation in order to include their satisfaction with thei r attorney into the analyses. All statistical data is available on Tables 4-1 through 4-8. Depression The backward elimination multiple regressi on analysis conducted to determine the predictability of depression indicated that two variables did account for a significant amount of depression variability, R2 = .203, F (2,110) = 13.76, p .001. The two independent variables, perceived pain interference ( p .001) and attorney retainment status ( p = .005), were predictive of depressi on. Specifically, individuals with higher levels of perceived pain inte rference and individuals who ha ve attorney representation reported higher levels of depression. The ot her eight variables were not significantly predictive of depression (see Table 4-1). Anxiety The backward elimination multiple regressi on analysis conducted to determine the predictability of anxiety indi cated that the select demogra phic and forensic variables did account for a significant amount of anxiety variability, R2 = .084, F (2,110) = 3.27, p = .024. No individual variables were significantly predictive of anxiety. However, within a regression model including female gender, high level of perceived pa in interference, and having attorney representation, anxiet y was predictable. While gender ( p = .054) was not significantly predictive of anxiety, a trend was noted that females reported higher levels

PAGE 66

56 of anxiety. The other nine variables were also not significantly predic tive of anxiety (see Table 4-2). Somatization The backward elimination multiple regressi on analysis conducted to determine the predictability of somatizati on indicated that two variable s did account for a significant amount of somatization variability, R2 = .258, F (2,110) = 18.74, p .001. The two independent variables, perceived pain interference ( p .001) and satisfaction with carrier ( p = .029), were predictive of somatization. Spec ifically, individuals wi th higher levels of perceived pain interference and individuals wi th lower levels of satisfaction with their carrier reported higher levels of somatiza tion. The other eight variables were not significantly predictive of so matization (see Table 4-3). Depression – Supplementary A supplementary backward elimination mu ltiple regression analysis was conducted for each of the three dependent variables in or der to examine the inclusion of the level of satisfaction with a retained attorney as an independent variable. The supplementary backward elimination multiple regressi on analysis conducted to determine the predictability of depression indicated that two variables did account for a significant amount of depression variability, R2 = .123, F (2,92) = 6.31, p = .003. The two independent variables, perceived pain interference ( p = .021) and satisfaction with attorney ( p = .026), were predictive of depression. Specifically, individuals with higher levels of perceived pain in terference and individuals with lower levels of attorney satisfaction reported higher levels of depr ession. The other eight variables were not significantly predictive of depression (see Table 4-4).

PAGE 67

57 Anxiety – Supplementary The supplementary backward elimination multiple regression analysis conducted to determine the predictability of anxiety indica ted that the select demographic and forensic variables did not account for a signif icant amount of anxiety variability, R2 = .073, F (2,92) = 3.53, p = .034. As in the first regression anal ysis, no individual variables were significantly predictive of anxiety. Gender ( p = .053) was not signi ficantly predictive of anxiety, although a trend was noted that female s reported higher levels of anxiety. The other nine variables were not significantly predictive of anxiety (see Table 4-5). Somatization – Supplementary The supplementary backward elimination multiple regression analysis conducted to determine the predictability of somatization indicated that one vari able did account for a significant amount of so matization variability, R2 = .233, F (2,92) = 13.63, p .001. Perceived pain interference ( p .001) was predictive of somatization. Specifically, individuals with higher levels of perceived pain interferen ce reported higher levels of somatization. The other nine variables were not significantly predictive of somatization (see Table 4-6). Correlations Pearson-product moment correlation matri ces were conducted in order to determine the relationships between the independent and dependent variables. All significance testing was done using two-tailed tests. Independent Variables Pearson-product moment correlation matr ices were conducted to determine the relationships between the ten demographic and forensic variables, as well as including

PAGE 68

58 the supplementary satisfacti on with attorney variable (see Table 4-7). Significant correlations between the variables include: Age and satisfaction with carrier (r = .209) – more individuals older in age reported higher levels of satisfact ion with their carriers. Gender and collapsed ethnicity (r = .249) – more females than males reported a non-Caucasian ethnicity. Gender and collapsed marital status (r = .190) – more females than males reported a marital history. Gender and time since injury (r = .202) – more males than females reported longer time periods since their Workers’ Compensation injury. Collapsed ethnicity and collapsed marital st atus (r = -.303) – more individuals that have a non-Caucasian ethnicity reported no marital history. Collapsed ethnicity and perceived social support (r = -.276) – more individuals that have a non-Caucasian ethnicity reported lo wer levels of perceived social support. Collapsed ethnicity and satisfaction with attorney (r = -.216) – more individuals that have a non-Caucasian ethnicity reporte d lower levels of satisfaction with their attorneys. Collapsed marital status and perceived soci al support (r = .398) – more individuals with a marital history reported higher levels of perceived social support. Perceived social support a nd satisfaction with atto rney (r = .238) – more individuals with higher levels of so cial support reported higher levels of satisfaction with their attorneys. Perceived pain interference and satisf action with carrier (r = -.272) – more individuals with higher levels of perceived pain interf erence reported lower levels of satisfaction with their carriers. Dependent Variables Pearson-product moment correlation matr ices were conducted to determine the relationships between the three psychologica l conditions. Significant correlations were found between all three of the conditions, suggesting a high potential for comorbidity between depression, anxiety, and somatization (see Table 4-8).

PAGE 69

59 Table 4-1. Backward Elimination Multiple Regression Analysis for Demographic and Forensic Variables Pred icting Depression (n=111) Regression Models B SE B Model 1a Age 5.558 .077 .068 Gender 1.467 1.426 .100 Collapsed Ethnicity .846 1.432 .057 Collapsed Marital Status .387 1.441 .027 Time Since Injury -4.769 .093 -.048 Social Support 1.447 .604 .002 Pain Interference 2.855 .879 .303 Employer Satisfaction -.644 .722 -.079 Carrier Satisfaction -.973 .736 -.127 Attorney Retainment 4.523 1.776 .237 Model 2b Age 5.518 .075 .067 Gender 1.467 1.419 .100 Collapsed Ethnicity .839 1.402 .057 Collapsed Marital Status .398 1.352 .028 Time Since Injury -4.753 .092 -.048 Pain Interference 2.856 .873 .303 Employer Satisfaction -.642 .716 -.079 Carrier Satisfaction -.971 .728 -.127 Attorney Retainment 4.525 1.765 .237 Model 3c Age 5.836 .074 .071 Gender 1.388 1.387 .094 Collapsed Ethnicity .747 1.361 .051 Time Since Injury -4.163 .090 -.042

PAGE 70

60 Table 4.1. Continued Regression Models B SE B Pain Interference 2.843 .868 .302 Employer Satisfaction -.626 .711 -.077 Carrier Satisfaction -.951 .722 -.124 Attorney Retainment 4.596 1.741 .241 Model 4d Age 5.853 .074 .071 Gender 1.240 1.344 .084 Collapsed Ethnicity .864 1.332 .059 Pain Interference 2.782 .855 .295 Employer Satisfaction -.608 .707 -.075 Carrier Satisfaction -.954 .719 -.124 Attorney Retainment 4.677 1.726 .245 Model 5e Age 5.438 .073 .066 Gender 1.468 1.294 .100 Pain Interference 2.831 .849 .301 Employer Satisfaction -.653 .702 -.080 Carrier Satisfaction -.937 .717 -.122 Attorney Retainment 4.768 1.715 .250 Model 6f Gender 1.552 1.286 .105 Pain Interference 2.880 .845 .306 Employer Satisfaction -.665 .700 -.082 Carrier Satisfaction -.842 .703 -.110 Attorney Retainment 4.596 1.696 .241 Model 7g Gender 1.625 1.283 .110 Pain Interference 2.947 .842 .313

PAGE 71

61 Table 4.1. Continued Regression Models B SE B Carrier Satisfaction -.872 .702 -.114 Attorney Retainment 4.618 1.695 .242 Model 8h Gender 1.421 1.276 .096 Pain Interference 3.204 .818 .340 Attorney Retainment 5.001 1.671 .262 Model 9i Pain Interference 3.190 .819 .339 Attorney Retainment 4.753 1.658 .249 aR2 = .239, Adjusted R2 = .163 bR2 = .239, Adjusted R2 = .171 cR2 = .238, Adjusted R2 = .179 dR2 = .237, Adjusted R2 = .185 eR2 = .234, Adjusted R2 = .190 fR2 = .230, Adjusted R2 = .193 gR2 = .223, Adjusted R2 = .194 hR2 = .212, Adjusted R2 = .190 iR2 = .203, Adjusted R2 = .188

PAGE 72

62 Table 4-2. Backward Elimination Multiple Regression Analysis for Demographic and Forensic Variables Predicting Anxiety (n=111) Regression Models B SE B Model 1a Age .000 .076 .184 Gender 2.479 1.405 .096 Collapsed Ethnicity 1.300 1.411 .921 Collapsed Marital Status .927 1.420 .072 Time Since Injury -.125 .092 -.139 Social Support -.247 .596 -.045 Pain Interference 1.542 .866 .179 Employer Satisfaction -.287 .712 -.039 Carrier Satisfaction .292 .725 .042 Attorney Retainment 2.598 1.751 .149 Model 2b Gender 2.478 1.388 .184 Collapsed Ethnicity 1.300 1.399 .096 Collapsed Marital Status .926 1.383 .072 Time Since Injury -.125 .091 -.139 Social Support -.247 .579 -.045 Pain Interference 1.542 .858 .179 Employer Satisfaction -.287 .708 -.039 Carrier Satisfaction .292 .709 .042 Attorney Retainment 2.599 1.727 .149 Model 3c Gender 2.480 1.382 .184 Collapsed Ethnicity 1.338 1.390 .099 Collapsed Marital Status .901 1.376 .070

PAGE 73

63 Table 4.2. Continued Regression Models B SE B Time Since Injury -.122 .091 -.136 Social Support -.267 .574 -.048 Pain Interference 1.564 .852 .181 Carrier Satisfaction .283 .706 .040 Attorney Retainment 2.617 1.720 .150 Model 4d Gender 2.556 1.363 .190 Collapsed Ethnicity 1.369 1.382 .102 Collapsed Marital Status .954 1.365 .074 Time Since Injury -.123 .090 -.137 Social Support -.251 .570 -.045 Pain Interference 1.483 .824 .172 Attorney Retainment 2.481 1.679 .142 Model 5e Gender 2.576 1.357 .191 Collapsed Ethnicity 1.471 1.357 .109 Collapsed Marital Status .763 1.289 .059 Time Since Injury -.126 .090 -.140 Pain Interference 1.480 .821 .172 Attorney Retainment 2.432 1.669 .139 Model 6f Gender 2.446 1.335 .181 Collapsed Ethnicity 1.281 1.315 .095 Time Since Injury -.115 .087 -.128 Pain Interference 1.443 .816 .167 Attorney Retainment 2.526 1.656 .145 Model 7g Gender 2.828 1.276 .210

PAGE 74

64 Table 4.2. Continued Regression Models B SE B Time Since Injury -.130 .086 -.145 Pain Interference 1.535 .811 .178 Attorney Retainment 2.642 1.651 .151 Model 8h Gender 2.455 1.260 .182 Pain Interference 1.360 .807 .158 Attorney Retainment 2.939 1.650 .168 aR2 = .119, Adjusted R2 = .031 bR2 = .119, Adjusted R2 = .040 cR2 = .118, Adjusted R2 = .048 dR2 = .116, Adjusted R2 = .056 eR2 = .115, Adjusted R2 = .063 fR2 = .112, Adjusted R2 = .069 gR2 = .104, Adjusted R2 = .070 hR2 = .084, Adjusted R2 = .058

PAGE 75

65 Table 4-3. Backward Elimination Multiple Regression Analysis for Demographic and Forensic Variables Predicting Somatization (n=111) Regression Models B SE B Model 1a Age -2.434 .078 -.003 Gender .944 1.445 .060 Collapsed Ethnicity 1.231 1.451 .079 Collapsed Marital Status 1.921 1.460 .129 Time Since Injury 9.797 .094 .094 Social Support -.173 .612 -.027 Pain Interference 3.888 .891 .389 Employer Satisfaction -.245 .732 -.028 Carrier Satisfaction -1.513 .746 -.186 Attorney Retainment 2.883 1.800 .142 Model 2b Gender .939 1.427 .060 Collapsed Ethnicity 1.235 1.438 .079 Collapsed Marital Status 1.912 1.423 .128 Time Since Injury 9.802 .094 .094 Social Support -.169 .595 -.026 Pain Interference 3.886 .882 .388 Employer Satisfaction -.245 .728 -.028 Carrier Satisfaction -1.517 .729 -.186 Attorney Retainment 2.890 1.776 .143 Model 3c Gender .956 1.419 .061 Collapsed Ethnicity 1.302 1.413 .083 Collapsed Marital Status 1.790 1.350 .120

PAGE 76

66 Table 4.3. Continued Regression Models B SE B Time Since Injury 9.589 .093 .092 Pain Interference 3.878 .877 .388 Employer Satisfaction -.263 .722 -.030 Carrier Satisfaction -1.531 .725 -.188 Attorney Retainment 2.849 1.762 .141 Model 4d Gender .959 1.413 .061 Collapsed Ethnicity 1.344 1.402 .086 Collapsed Marital Status 1.753 1.340 .117 Time Since Injury 9.816 .092 .094 Pain Interference 3.898 .872 .390 Carrier Satisfaction -1.541 .721 -.189 Attorney Retainment 2.861 1.755 .141 Model 5e Collapsed Ethnicity 1.568 1.359 .100 Collapsed Marital Status 1.594 1.316 .107 Time Since Injury .114 .089 .110 Pain Interference 3.868 .869 .387 Carrier Satisfaction -1.475 .713 -.181 Attorney Retainment 2.770 1.745 .137 Model 6f Collapsed Marital Status 1.146 1.260 .077 Time Since Injury .108 .089 .103 Pain Interference 3.961 .866 .396 Carrier Satisfaction -1.410 .711 -.173

PAGE 77

67 Table 4.3. Continued Regression Models B SE B Attorney Retainment 2.938 1.741 .145 Model 7g Time Since Injury .124 .087 .119 Pain Interference 3.909 .864 .391 Carrier Satisfaction -1.356 .708 -.166 Attorney Retainment 3.118 1.729 .154 Model 8h Pain Interference 4.081 .860 .408 Carrier Satisfaction -1.329 .712 -.163 Attorney Retainment 2.785 1.721 .137 Model 9i Pain Interference 4.214 .862 .421 Carrier Satisfaction -1.558 .703 -.191 aR2 = .308, Adjusted R2 = .239 bR2 = .308, Adjusted R2 = .246 cR2 = .307, Adjusted R2 = .253 dR2 = .307, Adjusted R2 = .259 eR2 = .303, Adjusted R2 = .263 fR2 = .295, Adjusted R2 = .261 gR2 = .289, Adjusted R2 = .262 hR2 = .275, Adjusted R2 = .255 iR2 = .258, Adjusted R2 = .244

PAGE 78

68 Table 4-4. Supplementary Backward Elimination Multiple Regression Analysis for Demographic and Forensic Variab les Predicting Depression (n=93) Regression Models B SE B Model 1a Age .085 .085 .105 Gender 1.594 1.594 .150 Collapsed Ethnicity 1.609 1.609 -.025 Collapsed Marital Status 1.583 1.583 -.043 Time Since Injury .110 .110 .001 Social Support .655 .655 .003 Pain Interference 1.078 1.078 .169 Employer Satisfaction .781 .781 -.066 Carrier Satisfaction .879 .879 -.184 Attorney Satisfaction .728 .728 -.213 Model 2b Age 8.404 .084 .105 Gender 2.150 1.519 .150 Collapsed Ethnicity -.352 1.589 -.025 Collapsed Marital Status -.575 1.553 -.043 Social Support 1.451 .649 .003 Pain Interference 1.699 1.064 .169 Employer Satisfaction -.503 .775 -.066 Carrier Satisfaction -1.489 .872 -.184 Attorney Satisfaction -1.449 .720 .213 Model 3c Age 8.366 .082 .104 Gender 2.151 1.509 .150 Collapsed Ethnicity -.357 1.566 -.026 Collapsed Marital Status -.563 1.448 -.042 Pain Interference 1.699 1.057 .169

PAGE 79

69 Table 4.4. Continued Regression Models B SE B Employer Satisfaction -.501 .766 -.066 Carrier Satisfaction -1.490 .866 -.184 Attorney Satisfaction -1.446 .704 -.213 Model 4d Age 8.420 .081 .105 Gender 2.060 1.446 .143 Collapsed Marital Status -.461 1.369 -.034 Pain Interference 1.709 1.050 .170 Employer Satisfaction -.485 .759 -.064 Carrier Satisfaction -1.491 .861 -.184 Attorney Satisfaction -1.414 .686 -.208 Model 5e Age 7.999 .080 .100 Gender 2.130 1.424 .148 Pain Interference 1.731 1.043 .172 Employer Satisfaction -.499 .754 -.066 Carrier Satisfaction -1.515 .854 -.187 Attorney Satisfaction -1.430 .680 .210 Model 6f Age 7.897 .080 .099 Gender 2.216 1.413 .154 Pain Interference 1.710 1.039 .170 Carrier Satisfaction -1.564 .848 -.193 Attorney Satisfaction -1.494 .671 .220 Model 7g Gender 2.250 1.413 .157 Pain Interference 1.803 1.035 .179 Carrier Satisfaction -1.397 .831 -.173

PAGE 80

70 Table 4.4. Continued Regression Models B SE B Attorney Satisfaction -1.505 .671 -.221 Model 8h Pain Interference 1.952 1.039 .194 Carrier Satisfaction -1.207 .829 -.149 Attorney Satisfaction -1.486 .677 -.218 Model 9i Pain Interference 2.364 1.006 .235 Attorney Satisfaction -1.540 .680 -.226 aR2 = .183, Adjusted R2 = .083 bR2 = .183, Adjusted R2 = .094 cR2 = .183, Adjusted R2 = .105 dR2 = .182, Adjusted R2 = .115 eR2 = .181, Adjusted R2 = .124 fR2 = .177, Adjusted R2 = .129 gR2 = .167, Adjusted R2 = .130 hR2 = .143, Adjusted R2 = .115 iR2 = .123, Adjusted R2 = .104

PAGE 81

71 Table 4-5. Supplementary Backward Elimination Multiple Regression Analysis for Demographic and Forensic Variab les Predicting Anxiety (n=93) Regression Models B SE B Model 1a Age 7.2 .082 .098 Gender 2.441 1.546 .184 Collapsed Ethnicity .482 1.561 .037 Collapsed Marital Status .666 1.536 .054 Time Since Injury 3.103 .107 .003 Social Support -.304 .635 -.058 Pain Interference .604 1.046 .065 Employer Satisfaction -.280 .758 -.040 Carrier Satisfaction -.370 .853 -.049 Attorney Satisfaction -.924 .706 -.147 Model 2b Age 7.267 .081 .098 Gender 2.453 1.474 .185 Collapsed Ethnicity .477 1.542 .037 Collapsed Marital Status .673 1.507 .054 Social Support -.303 .630 -.058 Pain Interference .607 1.032 .065 Employer Satisfaction -.282 .752 -.040 Carrier Satisfaction -.371 .846 -.050 Attorney Satisfaction -.922 .699 -.147 Model 3c Age 7.130 .081 .096 Gender 2.576 1.412 .194 Collapsed Marital Status .559 1.454 .045 Social Support -.329 .621 -.063 Pain Interference .596 1.026 .064

PAGE 82

72 Table 4.5. Continued Regression Models B SE B Employer Satisfaction -.298 .746 -.043 Carrier Satisfaction -.372 .842 -.050 Attorney Satisfaction -.960 .684 -.153 Model 4d Age 7.802 .079 .106 Gender 2.503 1.392 .189 Social Support -.234 .568 -.045 Pain Interference .569 1.018 .061 Employer Satisfaction -.298 .742 -.043 Carrier Satisfaction -.340 .834 -.045 Attorney Satisfaction -.964 .681 -.154 Model 5e Age 7.692 .104 .104 Gender 2.550 .192 .192 Social Support -.264 -.050 -.050 Pain Interference .556 .060 .060 Carrier Satisfaction -.369 -.049 -.049 Attorney Satisfaction -.994 -.158 -.158 Model 6f Age 7.004 .095 .095 Gender 2.467 .186 .186 Social Support -.263 -.050 -.050 Pain Interference .689 .074 .074 Attorney Satisfaction -1.010 -.161 -.161 Model 7g Age 7.445 .101 .101 Gender 2.510 .189 .189 Pain Interference .687 .074 .074 Attorney Satisfaction -1.085 -.173 -.173 Model 8h Age 7.630 .103 .103 Gender 2.559 .193 .193

PAGE 83

73 Table 4.5. Continued Regression Models B SE B Attorney Satisfaction -1.158 -.184 -.184 Model 9i Gender 2.633 .199 .199 Attorney Satisfaction -1.166 -.186 -.186 aR2 = .097, Adjusted R2 = -.013 bR2 = .097, Adjusted R2 = -.010 cR2 = .096, Adjusted R2 = .010 dR2 = .095, Adjusted R2 = .020 eR2 = .093, Adjusted R2 = .030 fR2 = .091, Adjusted R2 = .039 gR2 = .089, Adjusted R2 = .047 hR2 = .083, Adjusted R2 = .052 iR2 = .073, Adjusted R2 = .052

PAGE 84

74 Table 4-6. Supplementary Backward Elimination Multiple Regression Analysis for Demographic and Forensic Variable s Predicting Somatization (n=93) Regression Models B SE B Model 1a Age 1.020 .085 .012 Gender .398 1.607 .026 Collapsed Ethnicity .482 1.623 .032 Collapsed Marital Status 1.616 1.596 .112 Time Since Injury .107 .111 .098 Social Support -3.225 .660 -.001 Pain Interference 4.020 1.087 .373 Employer Satisfaction -.391 .788 -.048 Carrier Satisfaction -1.630 .886 -.188 Attorney Satisfaction -.978 .734 -.134 Model 2b Age 1.028 .083 .012 Gender .398 1.597 .026 Collapsed Ethnicity .483 1.601 .032 Collapsed Marital Status 1.613 1.497 .112 Time Since Injury .107 .110 .098 Pain Interference 4.020 1.080 .373 Employer Satisfaction -.392 .778 -.048 Carrier Satisfaction -1.630 .880 -.188 Attorney Satisfaction -.978 .718 -.134 Model 3c Gender .411 1.585 .027 Collapsed Ethnicity .476 1.590 .032 Collapsed Marital Status 1.640 1.473 .114

PAGE 85

75 Table 4.6. Continued Regression Models B SE B Time Since Injury .106 .110 .098 Pain Interference 4.033 1.068 .374 Employer Satisfaction -.391 .774 -.048 Carrier Satisfaction -1.611 .861 -.186 Attorney Satisfaction -.981 .714 -.135 Model 4d Collapsed Ethnicity .594 1.515 .040 Collapsed Marital Status 1.604 1.459 .111 Time Since Injury .114 .105 .105 Pain Interference 4.048 1.061 .375 Employer Satisfaction -.398 .769 -.049 Carrier Satisfaction -1.572 .844 -.181 Attorney Satisfaction -.970 .709 -.133 Model 5e Collapsed Marital Status 1.416 1.371 .098 Time Since Injury .113 .104 .103 Pain Interference 4.044 1.055 .375 Employer Satisfaction -.432 .760 -.053 Carrier Satisfaction -1.558 .839 -.180 Attorney Satisfaction -1.020 .694 .140 Model 6f Collapsed Marital Status 1.353 1.361 .094 Time Since Injury .117 .103 .107 Pain Interference 4.022 1.050 .373

PAGE 86

76 Table 4.6. Continued Regression Models B SE B Carrier Satisfaction -1.591 .833 -.183 Attorney Satisfaction -1.076 .684 .148 Model 7g Time Since Injury .137 .101 .125 Pain Interference 3.933 1.047 .365 Carrier Satisfaction -1.505 .829 -.174 Attorney Satisfaction -1.044 .683 -.143 Model 8h Pain Interference 4.104 1.044 .381 Carrier Satisfaction -1.515 .833 -.175 Attorney Satisfaction -.913 .680 -.125 Model 9i Pain Interference 4.295 1.039 .398 Carrier Satisfaction -1.576 .835 -.182 aR2 = .276, Adjusted R2 = .188 bR2 = .276, Adjusted R2 = .197 cR2 = .276, Adjusted R2 = .207 dR2 = .275, Adjusted R2 = .216 eR2 = .274, Adjusted R2 = .223 fR2 = .271, Adjusted R2 = .229 gR2 = .263, Adjusted R2 = .229 hR2 = .248, Adjusted R2 = .222 iR2 = .233, Adjusted R2 = .215

PAGE 87

77Table 4-7. Pearson-product Moment Corre lations of Independent Variables. Age Gender Ethnicity Marital Time Social Pain Employer Carrier Retain Attorney Age .132 -.056 .131 .058 -.134 .007 -.018 .209* -.178 -.011 Gender .132 .249** -.190* .202* -.130 -.036 -.048 .155 -.137 .017 Ethnicity -.056 .249** -.303** -.114 -.276** .091 -.114 .011 .068 -.216* Marital .131 -.190* -.303** .183 .398** -.050 .101 .081 .061 .094 Time .058 .202* -.114 .183 .136 .115 -.053 .020 -.123 .122 Social -.134 -.130 -.276** .398** .136 -.004 .127 .063 .068 .238* Pain .007 -.036 .091 -.050 .115 -.004 -.099 -.272** .152 -.157 Employer -.018 -.048 -.114 .101 -.053 .127 -.099 .065 -.030 .144 Carrier .209* .155 .011 .081 .020 .063 -.272** .065 -.230* .096 Retain -.178 -.137 .068 .061 -.123 .068 .152 -.030 -.230* Attorney -.011 .017 -.216* .094 .122 .238* -.157 .144 .096 Note: Correlation is significant at 0.05 (two-tailed) **Correlation is significan t at 0.01 (two-tailed)

PAGE 88

78 Table 4-8. Pearson-product Moment Corre lations of Depende nt Variables. Depression T-Score Anxiety T-Score Somatization T-Score Depression T-Score .558 .489 Anxiety T-Score .558 .297 Somatization T-Score .489 .297 Note: Correlations are significan t at the .01 level (two-tailed).

PAGE 89

79 CHAPTER 5 DISCUSSION Overview of Significant Findings This section will provide a review of the statistical results, as well as possible interpretation and explanati on of the findings, in relations hip to the three research questions. The three res earch questions included: Can selected demographic and forensic vari ables predict the presence of clinical depression in injured workers within the Workers’ Compensation system? Can selected demographic and forensic vari ables predict the presence of clinical anxiety in injured workers within the Workers’ Compensation system? Can selected demographic and forensic vari ables predict the presence of clinical somatization in injured workers within the Workers’ Compensation system? Prediction of Depression The backward elimination multiple regr ession analysis suggested that two variables, perceived pain inte rference and whether or not an individual had retained an attorney, were predictive of depression within the research sample. Individuals that reported higher levels of perceived pain interference while completing daily activities demonstrated higher levels of depression on the P-3. These findings are consistent with previously mentioned litera ture (Bair, Robinson, Kat on, & Kroenke, 2003; Boersma & Linton, 2005; Butchner, 1985; Gatchel, 2004; Piccinelli, Patterson, Braithwaite, Boot, & Wilkinson, 1999; Severeijns, Vlaeyen, van den Hout, & Weber, 2001; Sullivan & Stanish, 2003; Vowles & Gross, 2003), which in dicated that pain a nd depression are, in fact, comorbid conditions. This presence of th is relationship would s uggest that pain is

PAGE 90

80 not causative of depression, nor is depressi on causative of pain. Th is study documents that this comorbid relationship also ex ists for individuals within a disability compensatory system. The findings of this study may be due to the participants’ inability to engage in activities that they would find pleasurable, as according to the DS M IV-TR’s (American Psychiatric Association, 2000) definition of depression, a lack of pleasure and interest must be noted. It could be that the interferen ce in daily activities, not the presence of the physical pain itself, may be attributing to the development of clinical depressive features. These findings would also suggest that if an individual with a work-related injury in the Workers’ Compensation system was to self -report high levels of perceived pain interference in the completion of daily activities, it may be warranted for the rehabilitation professional wo rking with the individual with an injury to screen for the presence of depression or depr essive features. This is espe cially true given Bair and colleagues’ (2003) findings; indivi duals with chronic pain typi cally are not recognized as having depression, nor are they appr opriately treated for depression. It also cannot be assumed that individuals may develop ways to cope with their chronic pain issues as time goes on, ther eby becoming less at risk of experiencing depressive symptoms. Gatchel (2004) reminds us that individuals with poor pain coping skills are typically unable to cope with depres sion. This would sugge st that individuals with chronic pain complaints and unsuccessful coping strategies fo r pain may also be unsuccessfully managing depression or other psychological issues. As previously mentioned, individuals who had retained attorneys while in the Workers’ Compensation system also reported hi gher levels of depr ession on the P-3. At

PAGE 91

81 first thought this might seem counterintuitive, as the act of retaini ng legal representation may be considered to provide a sense of security for individuals in the Workers’ Compensation system. However, it may also be possible that individuals who retain attorneys lose a sense of wh at little control they wielde d while participating in the Workers’ Compensation system. Typicall y, once an injured worker seeks legal representation, the insuran ce carrier handling the clai m will also seek legal representation. The relationship between th e injured worker and insurance company, particularly the insurance claims adjuster, may cease to exist, as both opposing attorneys now handle the claim’s specifics. This may l eave the injured worker feeling powerless and no longer in charge of hi s or her situation. Such an occurrence may be an example of the social rank theory of depression, whic h stipulates that ex periencing involuntary decreased social ranking is highl y associated to the development of depressive symptoms (Gilbert, 2005). Essentially, once the injured individual is removed from the activity pertaining to the claim, the individual may se nse a feeling of subordination. This may be especially true since the enti re situation is built upon the individual and his or her workrelated injury, yet the indivi dual has now become completely passive due to attorney involvement. The supplementary backward elimination multiple regression analysis suggested that two variables, perceived pain interference and satisfaction with a retained attorney, were predictive of depression within the research sample. As in the primary analysis, greater perceived pain resulted in higher levels of depre ssion. Individuals that had retained attorneys and reported lower levels of satisfaction de monstrated higher levels of depression on the P-3. This relationship ma y further strengthen the argument regarding

PAGE 92

82 regret and hopelessness once the case is out of the injured workers’ hands. Blackwell, Leirer, Haupt, and Kamptisis (2003) suggest that higher le vels of satisfaction with attorney representation are predictive of su ccessful return to work outcomes. They stipulated that these individuals reported a feeling of control, especially when available to participate in the decision-making processe s. It may be argued that this type of participation and subsequent feelings of cont rol may not be present in all attorney-client relationships, including t hose found within the Workers’ Compensation system. It was hypothesized that age and gender woul d be significant predictor variables for depression. These hypotheses were made base d on the literature, which suggests that individuals older in age, as well as females, typically report higher levels of depression (American Psychiatric Association, 2000; Dixon & Gatchel, 1999; Gatchel, 2004; Sullivan & Stanish, 2003; Vowles & Gross, 2003). According to the research sample, age and gender were not predictive of depr ession found in individuals in the Workers’ compensation system. However, this may be explained by the participants’ demographics. Of the 111 participants, 97 (87.4%) were under the age of 50 years, which may justify how age was not predictive as hypo thesized. A bias may also be present for gender, as only 39 females (35.1%) were incl uded in the research sample. Therefore, rehabilitation professionals working in th e Workers’ Compensation system should keep in mind that individuals within a disability compensatory system, such as Workers’ Compensation, may not abide to recognized pred ictor variables previ ously established for certain psychological conditions, such as depression. Prediction of Anxiety Both the primary and supplementary backward elimination multiple regression analyses were unable to determine a signifi cantly predictive variable for anxiety. Three

PAGE 93

83 phenomena may have occurred, either individually or collectively, in this study to explain this. The first possibility is that an insuffici ent sample size may have resulted in failure to not produce enough power to determine significan ce within the independent variables. The second possibility is that anxiety may not have been properly measured by the P-3, suggesting an issue with content validity within the instrument The specific items used to measure anxiety within the P-3, which may support appropriate construct validity, may not have been the best measures of anxiet y for this unique population. An example item would be asking participants to choose between the followi ng three responses: “I am comfortable enough in a group of people; In a group of people I sometimes feel a little nervous; In a group of people I often feel like I don’t really belong” (Tollison & Langley, 1995, p.24). A similar item attempts to measure anxiety by providing the following responses: “I enjoy being around other people; I can mix with others but would rather be alone; I avoid having to be around others ” (Tollison & Langley, p.24). Although these items may be appropriate measures of anxi ety, they seem to address fear and excess worry of social situations, such as that found in panic disord er or social phobia (American Psychiatric Association, 2000). Also no studies examining the P-3’s content validity have been conducted. Other items meas uring anxiety may be more appropriate to gauge anxiety found in individuals within the Workers’ Compensation system. The third possible explanation is that there may have been a lack of clinical anxiety present in the research sample. This, howev er, is probably unlike ly given the observed presence of the other two psychological condi tions. It would be difficult to determine which, if any, of these phenomena was res ponsible. Given that the sample size was sufficient, in terms of statistical power, to identify predictor variables of the other two

PAGE 94

84 dependent variables, the anxiet y results of the research samp le are more likely due to either a sample size with insufficient power or P-3 content validity issues. It was hypothesized that gender would be found to be a significant predictor variable of anxiety, specifi cally being female. Upon completion of the regression analyses, female gender was the independent va riable closest to statistical significance. This finding may suggest a potential trend even though significance was not established. Prediction of Somatization The backward elimination multiple regr ession analysis suggested that two variables, perceived pain in terference and satisfaction w ith the Workers’ Compensation insurance carrier, were predictive of somatization within the research sample. Individuals with reports of higher percei ved pain interference while completing daily activities reported higher levels of somatization on the P-3. Several studies examining individuals with chronic pain who also demonstrated comorbid somatization issues support these findings (Barsky, Orav, & Bates, 2005; Ge orge, Dannecker, & Robinson, 2005; Woby, Watson, Roach, & Urmston, 2004). Somatization may be a serious issue in Workers’ Compensation because of the prevalence of low back pain associated with work injury that often cannot be objectified by medical evidence. Some pain complaints may be easily interpreted as somatiza tion, especially when there is insufficient methodology for assessing such complaints. Examples of such complaints might include undiagnosed chronic lumbar back pain, headaches, and gast rointestinal issues. These complaints may lead to chronic pain interfer ence if treatment is not provi ded due to an inability to determine pain etiology, which in turn may result in the di agnosis of chronic somatization disorders.

PAGE 95

85 Individuals with reported lower levels of satisfaction with their Workers’ Compensation insurance carrier demonstrated higher levels of somatization on the P-3. Carriers are consistently cons ervative when authorizing medical treatment and have been selective when determining what they belie ve is necessary (Sullivan & Stanish, 2003; Fishbain et al., 1993). This is partly due to th e rising costs of health care and a heightened awareness of insurance fraud, specifically malingering. For example, Suter (2002) reports that disability litigation, such as that found in Workers’ Compensation, is possibly inflated due to individuals seeking monetary gain in excess to en titled wage replacement benefits. Carriers, therefore, are cognizant of this and may be, as a cost-minimizing defensive strategy, overestimating malinge ring occurrence. Because carriers are responsible for allowing or refusing access to medical treatment, individuals with somatic complaints may easily develop negative views toward the carriers. This is especially possible given that individuals may believe that appropriate tr eatment exists, but is being denied to them on the basis of their non-objective, or a lack of medically evident, pain complaints. Somatic complaints may perpetuate beliefs that the “managed care-like” cost containment practices of the insurance car riers are denying appropriate treatment and rehabilitation, thereby intens ifying the somatization. The supplementary backward elimination multiple regression analysis also suggested that perceived pain in terference was predictive of so matization. This finding is not surprising, as individuals with somatization are likely to have pain interference. There may have not been a sufficient attempt to alleviate the pain through appropriate rehabilitation due to a lack of medical evid ence, thereby allowing pain interference to

PAGE 96

86 persist.. No other variables, however, were significantly predictive of somatization in the supplementary analysis. It was hypothesized that age and marital stat us would be significant predictors of somatization. Both variables were documented in the literature (American Psychiatric Association, 2000; Nakao et al., 2001) as pr edictive of somatizat ion among individuals with disabilities. Specifically, younger indi viduals and those not ma rried reported higher levels of somatization. In this study, age a nd marital status were not predictive of somatization in the Workers’ Compensation system. A likely explanation for age may be that, according to Nakao a nd colleagues and the DSM IV-T R, younger age is typically predictive of somatization. However, a bias in the Workers’ Compensation demographics, as for depression, may explain the inconsistency. Of the 111 participants, 43 (38.7%) were under the age of 40 years. Beca use of the these findings, rehabilitation professionals working in the Workers’ Compensation system should keep in mind that clients may not present recognized predictor va riables previously es tablished for certain psychological conditions, such as somatization. Additional Findings There were four additional findings that are notable in regards to this study. These four findings include the identification of va riables found to be l east predictive of the psychological conditions; the participants’ or der of satisfaction with their former employers, insurance carriers, and attorney s; the presence of the psychological conditions; and the relationships be tween the psychological conditions. Least Predictive Variables The two backward elimination multiple regression analyses for depression indicated that the least predictive variable for depression in each was perceived social

PAGE 97

87 support and the amount of time since injury, re spectively. These results suggest that the presence of a support system was unrelated to the presence of depre ssive features found in the research sample. Secondly, the supplem entary analysis sugge sted that depression was not related to the amount of time an indi vidual had spent out of work due to the injury. This seems contradictory to the li terature that suggests a strong relationship between length of time out of work and difficu lty in achieving a successful return to work outcome (Blackwell, Leirer, Haupt, & Kamptitsis, 2003; Gardner, 1991; Mason, Wardrope, Turpin, & Rowlands, 2002; Sc houten & Williams, 2000; Stice & Moore, 2006; Tate, 1992; Weed & Field, 2001). It may also be that extended time out of work due to injury is misconstrued to be a cause, or even a moderating variable, for depression. The difficulty in successful return to work mi ght be attributed to factors other than the presence of depression, or any other psychological compla int, thereby explaining the contradictory nature of this finding. The least predictive variable s of anxiety in the research sample for both analyses were age and amount of time si nce injury, respectively. A ccording to the DSM IV-TR (American Psychiatric Association, 2000), the diagnostic criteria fo r recognized mental disorders, individuals that are younger in age de monstrate a higher prevalence of anxiety disorders. However, age was unrelated to the complaints of a nxiety reported by the research sample. Again, there may be an age bias within the Workers’ Compensation population, as the research sample’s average age was 41 years old. Time since injury, like in depression previously men tioned, was also unrelated to an xiety. However, just as in depression, factors other than psychological status may be at fault when considering unsuccessful return to work rates. Regardle ss, rehabilitation profe ssionals may not want

PAGE 98

88 to jump to conclusions regarding a client ba sed on the amount of time the client has been injured. The least predictive variables of somatization in the research sample for both analyses were age and ethnicity once co llapsed into Caucasian and non-Caucasian, respectively. This was very surprising, as typically somatization is noted in younger adults, including ages between twenty a nd thirty years old (Nakao et al., 2001). According to the research sample, age was not a factor regarding so matization within the Workers’ Compensation system. Regardi ng somatization and et hnicity, there is no conclusive relationship found in the literatu re. However, there is a strong correlation noted between somatization and marital stat us, where non-married individuals report greater amounts of somatic complaints than married individuals (American Psychiatric Association, 2000; Nakao et al., 2001). The major ity of the participants in this study who were not currently married were non-Caucas ian. Nakao and colleagues and the DSM IVTR suggest that individuals not married ar e more susceptible to somatization issues. However, this does not account for the poten tial moderating effect of ethnicity, which may explain why ethnicity was unrelated to somatization regardless of the associated high numbers of non-married individuals. Order of Satisfaction The participants were asked to rate thei r current level of satisfaction with the former employer at the time of injury, thei r Workers’ Compensation insurance carrier, and if retained, their Workers’ Compensation attorney. Interesti ngly, the participants reported the greatest degree of satisfaction wi th their attorney, followed by the carrier, and finally the former employer. When speci fically looking at thos e who reported “very low” and “low” satisfaction levels, the lowest levels available for the participant to rate,

PAGE 99

89 89 participants (80.2%) reported very low or low satisfaction with their former employers. Although there were 84 indi viduals (75.6%) who reported negative satisfaction (very low and low) with their in surance carriers, there were fewer very low satisfaction levels indicated for insuran ce carriers (39.6%) compared to employers (45.9%). These are dramatically different th an that found when examining very low and low levels of satisfaction with the retained attorneys, as only 11 out of the 93 individuals (11.8%) who reported attorney retention suggested negative satisfaction levels. One possible explanation for these findings may be attributed to preconceived expectations. Individuals ma y expect satisfaction from their Workers’ Compensation attorneys, as they are hired to serve as an advocate during thei r rehabilitation. This explanation may be especially true if th e individuals are entering the system with expectations that the insurance carrier will be heavily scrutinizing them and looking for any reason to suggest malingeri ng. It may be possible that in dividuals working in certain industries, especially those with frequently associated work-rel ated injury, may be informed of the carrier s’ actions by former co-workers who were injured. Individuals may be expecting a negative relationship with the carrier, which may influence the reports of lower satisfaction levels. If one is expecting a difficult relationship, satisfaction levels reported by the injured individuals ma y be skewed unless an unpredicted positively viewed behavior on the part of the carrier arises. An exampl e of an unpredicted positive behavior may include the carrier’s complete lack of hesitance toward the authorized physician’s suggested treatment. Following this explanation, individuals may initially expect support and assistance from their former employer during their rehabilitation. Individuals may not be finding their expecta tions fulfilled, especially if the employer

PAGE 100

90 does not appear to want to assist the injured employer. An example of this may be if the employer appears to be disinterested in th e injured worker’s situation, unwilling to provide light duty work, or fail to retain a position for the injured worker. This may result in minimal levels of satisfaction, as the inju red worker may develop a sense of rejection or betrayal given the preconceived unmet e xpectations (Al-Darmaki, & Kivlighan, 1993; Betters & Shaw, 2006; Chan, McMahon, Shaw, & Lee, 2004; Koch, 2001; Shaw, McMahon, Chan, & Hannold, 2004). A possible re medy, if this explanation is valid, might be promoting sustained working relati onships between the employers and injured employees to alleviate the unmet expectations. The Presence of the Psychological Conditions The findings of this study suggest that there may be unmet psychological need among individuals in the Workers’ Compen sation system. Although this study did not specifically assess unmet psychological need, it can be deduced based on the participants’ scores on the P-3. Approximately 25% of th e research sample scored above average depression for pain patients, as indicate d by the P-3. Although onl y approximately 10% of the research sample scored above average on anxiety, this is stil l important, given the considerations previously mentioned for the measurement of anxiet y during this study. These considerations may include an insuffici ent sample size or potentially poor content validity within the asse ssment. Either way, it is possible that a greater amount of anxiety is actually present. This also means, shoul d there, in fact, not be a greater amount of anxiety present, that eleven i ndividuals within the research sample demonstrated clinical anxiety features. Their identified anxiety ma y be hindering their rehabilitation. Finally, approximately 46% of the research sample scored above average somatization, as indicated by the P-3. Although these number s are not staggering, nor does it represent a

PAGE 101

91 majority, they are still clinically important in relationship to the potential interference with the rehabilitation process of the indi viduals. And although the research sample findings cannot be generalized to the enti re Workers’ Compensation population, this study raises important questions about the degr ee to which such findings might exist in the Workers’ Compensation system in genera l. Future research with the Workers’ Compensation population that produces findings remotely similar would be tremendously noteworthy and would reflect obvious unmet psychological need, which may drastically impede the vocational rehabilitation process. The Relationship Between the Psychological Conditions A correlation matrix was conducted on the th ree dependent variables to determine if the independent variables were actually measuring the same construct during each measure of the dependent variables. The co rrelations were all si gnificant between the three variables, suggesting that they may be co morbid. However, they were all distinct in regards to the potential pred ictor variables. With the ex ception of perceived pain interference, all thr ee dependent variables suggested di fferent significantly predictive variables. The dependent variables also diffe red regarding the least predictive variables noted. Given these two findings, an argument can be made that the dependent variables, although correlated, were distinct entiti es within the research sample. Limitations There are several limitations to this study that must be considered when examining the results. The first limitati on is the sampling technique. Convenience sampling was used to acquire the resear ch sample, which consequently imposes limitations to the external validity of the study. Convenience sampling was required, however, as there was no conceivable wa y of developing a sampling frame for all

PAGE 102

92 individuals with Workers’ Compensation cl aims in the Northeast Florida region, who would also be seeking assistance from the Bureau of Rehabilitation and Reemployment Services. Therefore, due to a non-probabalist ic sampling technique, a few considerations are warranted. The potential limitations to external validity include the lack of generalization to other indivi duals within the Workers’ Compensation system. Although it was known that generalization could not occur outside of Florida due to state differences in Workers’ Compensation law, the study results also cannot be generalized to other individuals with Workers’ Compensation claims in Florida, especially if they are not in the Northeast Florida region. Several considerations are not ed, including potential regional differences in the demographic variable s, such as ethnicity and marital status. If there are regional differences within the state of Florida in regards to these variables, then the predictive nature of the variables may be different than that which was found in the research sample. Differences in the forensic variables may exist as well, such as social support and satisfaction levels with the othe r parties. Overall, generalization of the findings may be limited, which shoul d be appropriately considered. The research design used for this st udy was a quasi-exp erimental one-group posttest-only design. There are potential limita tions associated with this design choice. Since it is quasi-experimental and therefore lacking random assignment, internal validity is threatened. However, given that the fo cus of the study was aimed at prediction, not causality, this may not be considered a true lim itation. It must be mentioned that most of the recognized threats to internal validity were manage d, including testing, mortality, and statistical regression. Two threats potent ially limit the study, in cluding history and

PAGE 103

93 instrumentation. It is not possible to rule out historical events occurring prior to the assessment, and therefore conf ounding variables may exist. There may be limitations to the measurem ent of the study, specifically regarding the instrumentation. Although construct validity for the BIWI and P-3 appear sound, content validity may be a limitation. First, the BIWI may not have included the best variables for predicting depression, anxi ety, and somatization in the Workers’ Compensation system. Other uni dentified variables may, in fact, be better predictor variables for each of the three psychological conditions. Such variables may include type of injury, type of former employment, current financial situation, and previous psychological history. Also, although statisti cally significant predic tor variables were established for the research sample, these va riables may not be clinically significant given the low coefficients of determina tion. An example would include the initial backward elimination multiple regression analysis for depression, whereas the coefficient of determination was found to be R2 = .203. This would suggest the suggested predictor variables account for approximately 20% of the variance, meaning 80% is still unaccounted. Future research examining a dditional variables should be conducted. Secondly, as previously mentioned, the content validity of the P-3’s anxiety subscale may not be adequate, specifically for Workers’ Compensation clients. There may be other unidentified inst ruments available that would be more appropriate for anxiety in Workers’ Compensation clients. The last limitation noted for the study was the insufficient power for the statistical analyses of anxiety. It may, in fact, be a sample size issue and not a measurement issue

PAGE 104

94 as previously mentioned. If a sufficient sa mple size with consequently sufficient power was obtained, it is possible that some of th e variables may be predictive of anxiety. Future Recommendations The following section provides future recommendations based on this study’s findings in the following areas: clinical practice and policy, prof essional training, and additional research. Clinical Practice and Policy It is apparent that the Workers’ Compensation system needs to recognize the interrelationship betw een physical and psychological di sability. Selander, Marnetoft, Bergroth, and Ekholm (2002, p.704) report the following: “People with greater chances of job retu rn after vocational rehabilitation are younger, native, highly educated, find a st eady job with high income, married and have stable social networks, are self-c onfident, happy with life, not depressed, have low level of disease severity and no pain.” Fortunately, Selander and colleague s were quick to admit that th is description is far from the typical depiction of an injured worker It would be obviously clear to see why individuals that “fit” this pr ofile are successfully returningto-work. The problem is that most of these conditions are not realistic. Indi viduals in the Workers’ Compensation system who are unable to return-to-work to the previous job are usually unable because of residual functional impairment, typically resulting in permanent physical work restrictions. Exceeding these restrictions ma y place the individual at risk for inducing or increasing pain or injury. Individuals who have sustained injuries resulting in permanent impairment and accompanying chronic pain are very suscep tible to also developing psychological

PAGE 105

95 disorders. Major depression has been por trayed as a perfect example of this phenomenon. Rehabilitation prof essionals also need to rea lize that not only does the psychological condition warrant treatment, th e presence of the psychological disorder greatly amplifies any physical impairment Without providing treatment for both the physical and psychological concerns, return-to-work outcomes will more than likely not prove as successful as one would like to see in the Workers’ Compensation system. How can the Workers’ Compensation sy stem more effectively provide the necessary treatment to those at risk? On e possible solution is screening for early intervention. Boersma and Linton (2005) sugge st assessing individuals for psychological symptoms so that appropriate intervening steps may be taken, including psychotherapy treatment or necessary referral. However, sc reening is not always considered a viable option, as arguments of cost and a lack of necessity ofte n arise. Boersma and Linton indicate that initial psychological screening is occasionally viewed as being too costly to provide to all individuals. This is com pounded by the secondary argument. The other hurdle often encountered is that because some individuals who ente r the system will be symptom-free, psychological screening of a ll individuals is unnecessary. Despite these arguments, a compromise could be cr eated which would minimize unnecessary assessment and associated cost. Individuals wi th Workers’ Compensation claims could be tested to treat psychological problems as well as their physical problems, specifically when they present recognizable predicto r variables. Individuals who report ongoing chronic pain should especially be assessed fo r comorbid psychological symptoms as early as possible. By identifying and treating both concerns appropriately, return-to-work may be a more frequent outcome. According to a recent study conducte d by Della-Posta and

PAGE 106

96 Drummond (2006), individuals in the Workers’ Compensation system were more likely to demonstrate improvement in their psychologi cal health, as well as be more successful in returning to work once they included a cognitive behavioral th erapy component into their rehabilitation. By addressing bo th physical and psychological well being, individuals may be more likely to experience more successful rehabilitation outcomes. Given the findings of this study, individuals within the Workers’ Compensation system who report the following should be initially screened for the corresponding potential psychological concern: Initial assessment of psychological distre ss should be conducted once individuals enter the Workers’ Compensation system, especially if predictor variables and ongoing chronic pain complaints are present. Early intervention, specifica lly psychological screening, should be implemented into the common practice of Workers’ Compensation rehabilitation counseling, especially if predictor vari ables and ongoing chronic pain complaints are present. Individuals who report high levels of pe rceived pain interference should be screened for depressive and somatic features. Individuals who report attorn ey representation should be screened for depressive features, especially if the i ndividuals also report low leve ls of satisfaction with their attorneys. Individuals who report low levels of sa tisfaction with their insurance carriers should be screened for somatic features. Females should be screened for anxiety features. Follow-up assessment of psychological distress may need to be considered. Professional Training Early identification of unmet psychologi cal need should be included in the professional training of all parties involved in the Workers’ Compensation system. This should include rehabilitation counselors, cas e managers, authorized treating physicians, Workers’ Compensation insurance adjustors, al l allied health care professionals that may

PAGE 107

97 encounter individuals with the Workers’ Compensation system, and even Workers’ Compensation attorneys. Obviously, there shou ld be differences between the focus of such training, consistent with the role and functions of the profession. Obviously, all of the parties should not be re sponsible for diagnosing psyc hological conditions. This should be placed in the hands of the aut horized treating physicians and any referred mental health professional, such as psychi atrists, psychologists, and mental health counselors. Although rehabilitati on counselors who are not licen sed to provide a clinical mental health diagnosis are excluded, thei r training should provide them the knowledge and skills to identify potentia l psychological distress for appr opriate referral. This would also be true for rehabilitati on counselors who perform in roles in which diagnosis and treatment are not required, such as case management. For the other professionals mentioned who would not be responsible for diagnosing individuals with psychologica l conditions, education pertaini ng to awareness of potential psychological distress should be implemented. This may include coursework or seminars on psychological symptoms and risk factors, specifically identified predictor variables. Training should also include how to appr opriately document psychological complaints and how to appropriately repor t potential psychological distress in the event that a referral would be necessary. Th e following represents recommendations that would allow earlier identification to occur: Include material pertaining to predictor variables of psychological conditions for individuals within the Work ers’ Compensation system in the educational programs for professionals that work in the Workers’ Compensation arena. Provide specific training focusing on ps ychological distress screening, including the use of clinical interviewing and a ppropriate instrumentation usage, to rehabilitation counselors practicing in the Workers’ Compensation system.

PAGE 108

98 Implement opportunities for approved contin uing education units (CEUs) focusing on psychological distress awareness for all professionals that work in the Workers’ Compensation arena. Additional Research Future research focusing on the iden tification of predictor variables for psychological conditions is needed. Indivi duals within the Workers’ Compensation system are experiencing psychol ogical distress and may not be receiving treatment, either in a timely manner or at all. More conclusive and consistent research findings will be needed to implement the change needed to alleviate this problem. The following recommendations are provided in or der to accomplish this objective: Change the sampling technique for a futu re study to a random sampling technique from all thirteen Bureau of Rehabilit ation and Reemployment Services office locations in Florida to provide a better argument for external validity. Develop a larger list of possible predictor variables that may better predict the three psychological conditions than those used in this study, such as t ype of injury and type of employment. Increase the sample size per office loca tion to maximize power and accommodate for an increased number of inde pendent variables for analysis. Utilize a different instrument to assess th e psychological conditions, specifically anxiety. If necessary, develop a new instrument to assess the psychological conditions, especially for individuals within Workers’ Compensation and other similar disability compensatory systems. Utilize a different statistical analysis, such as discriminant or path analysis, in order to determine causality as opposed to prediction. Complete a study to identify why early iden tification of psychological conditions is not occurring within Workers’ Compensation and similar disability compensatory systems. Complete a study that examines a cost-b enefit analysis for implementing early screening for psychological conditions based on established predictor variables, should this be determined to be the pre dominant reason why early identification of psychological conditions is not occurring.

PAGE 109

99 By improving the knowledge pertaining to psychological considerations for individuals within the Workers’ Compensatio n system, rehabilitation outcomes may be greatly enhanced due to fulfilling a previously unmet need. It is likely that individuals with work-related injuries may return to work more successfully due to improved psychological health and diminished ongoing ch ronic pain complaints and interference. Although insurance carriers may have to in itially invest additional money by funding mental health care, this may be justifie d by an increased frequency of timely and successful rehabilitation outcomes. The end result may include a more efficient and costeffective Workers’ Compensation rehabilitation system.

PAGE 110

99 APPENDIX A BUREAU OF REHABILITATION AND REEMPLOYMENT SERVICES PERMISSION LETTER

PAGE 111

101 APPENDIX B BETTERS INJURED WORKER INVENTORY (BIWI)

PAGE 112

102

PAGE 113

103 LIST OF REFERENCES Adams, J. H. & Williams, A. C. (2003). What affects return to work for graduates of a pain management program with chronic upper limb pain? Journal of Occupational Rehabilitation, 13, 91-106. Al-Damarki, F. & Kivlinghan, D. M. (1993). C ongruence in client-couns elor expectations for relationship and th e working alliance. Journal of Counseling Psychology, 40, 379-384. American Psychiatri c Association (2000). Diagnostic and statistical manual of mental disorders Text revised. Washington D.C.: APA. Bair, M. J., Robinson, R. L., Katon, W., & Kroenke, K. (2003). Depression and pain comorbidity. Archives of Internal Medicine, 163, 2433-2445. Baril, R., Clarke, J., Frisen, M., Stock, S., & Cole, D. (2003). Management of return-towork programs for workers with musculoske letal disorders: A qualitative study in three canadian provinces. Social Science and Medicine, 57, 2101-2114. Betters, C. J., & Shaw, L. R. (2006). Work ing alliance and workers’ compensation: Implications on the worki ng alliance model for rehabi litation counselors in the workers’ compensation system. Rehab Pro, 14 33-37. Blackwell, T. L., Leierer, S. J., Haupt, S., & Kampitsis, A. (2003). Predictors of vocational rehabilitation return-to-work outcomes in workers' compensation. Rehabilitation Counseling Bulletin, 46, 108-114. Boersma, K. & Linton, S. J. (2005). Screening to identify patients at risk: Profiles of psychological risk factor s for early intervention. The Clinical Journal of Pain, 21, 38-43. Brouwer, S., Reneman, M. F., Dijkstra, J. W., Schellekens, J. M., & Goeken, L. N. (2003). Test-retest reliability of the isernhagen work systems functional capacity evaluation in patients with chronic low back pain. Journal of Occupational Rehabilitation, 13, 207-218. Bureau of Labor Statistics. ( n.d.) Bureau of labor statistics – Injuries, illnesses, and fatalities. Retrieved on September 22, 2005 at http://www.bls.gov/iif/home.htm

PAGE 114

104 Chan, F., McMahon, B. T., Shaw, L. R., & L ee, G. (2004). Psychometric validiation of the expectations about rehabilitation counseling scale: A preliminary study. Journal of Vocational Rehabilitation, 20, 127-135. Choppa, A. J., Cutler, F., Gann, C ., Moreland, T., & Olson, L. (1992). Vocational evaluation in private sector rehabilitation. Menomonie, WI: Materials Development Center. Crombez, G., Vlaeyen, J. W., Heuts, P. H., & Lysens, R. (1999). Pain-related fear is more disbaling than pain itself: Evidence on the role of pain-re lated fear in chronic back pain disability. Pain, 80, 329-339. Daiker, B. (1995). Managed care in workers' compensation. AAOHN Journal, 43, 422427. de Boer, A. G., Wijker, W., & de Haes, H. C. (1997). Predictors of health care utilization in the chronically ill: A re view of the literature. Health Policy, 42, 101-115. Della-Posta, C., & Drummond, P. D. (2006). C ognitive behavioral therapy increases reemployment of job seeking wo rkers’ compensation clients. Journal of Occupational Rehabilitation, 16 223-230. Dixon, A. N. & Gatchel, R. J. (1999). Gender and parental status as predictors of chronic low back pain disability: A prospective study. Journal of Occupational Rehabilitation, 9, 195-200. Dooley, D. (2001). Social research methods. Upper Saddle River, NJ: Prentice Hall. Dush, D. M., Simons, L. E., Platt, M. Nation, P. C., & Ayres, S. Y. (1994). Psychological profiles distinguishing litig ating and nonlitigating pain patients: Subtle, and not so subtle. Journal of Personality Assessment, 62, 299-313. Elders, L. & Burdorf, A. ( 2004). Workplace interventions. Occupational and Environmental Medicine, 61, 287-288. Erdfelder, E., Faul, F., & Buchner, A. (1996). GPOWER: A general power analysis program. Behavior Research Methods, Instruments, & Computers, 28, 1-11. Evanoff, B., Abedin, S., Grayson, D., Dale, A., Wolf, L., & Bohr, P. (2002). Is disability underreported following work injury? Journal of Occupational Rehabilitation, 12, 139-150. Fishbain, D. A., Rosomoff, H. L., Goldberg, M., Cutler, R. B., Abdel-Moty, E., & Khalil, T. M. (1993). The prediction of return to the workplace after mu ltidisciplinary pain center treatment. The Clinical Journal of Pain, 9, 3-15.

PAGE 115

105 Fishbain, D. A., Rosomoff, H. L., Cutler, R. B., & Steele-Rosomoff, R. (1995). Do chronic pain patients' perceptions about th eir preinjury jobs determine their intent to return to the same type of job post-pain facility treatment? The Clinical Journal of Pain, 11, 267-278. Fishbain, D. A., Cutler, R. B., Rosomoff, H. L., Khalil, T. M., & Steele-Rosomoff, R. (1997). Impact of chronic pain patients' j ob perception variables on actual return to work. The Clinical Journal of Pain, 13, 197-206. FWCI. (2003). Florida Workers’ Compensation Institute Handbook Tallahassee, FL: FWCI. Gardner, J. (1991). Early referral and other factors affecting vocational rehabilitation outcome for the workers' compensation client. Rehabilitation Counseling Bulletin, 34, 197-207. Gatchel, R. J. (2004). Psychosocial factors that can influence the self-assessment of function. Journal of Occupational Rehabilitation, 14, 197-206. Geisser, M. E., Robinson, M. E., Miller, Q. L., & Bade, S. M. (2003). Psychosocial factors and functional capacity evalua tion among persons with chronic pain. Journal of Occupational Rehabilitation, 13, 259-273. George, S. Z., Dannecker, E. A., & Robins on, M. E. (2005). Fear of pain, not pain catastrophizing, predicts acute pain intensity, but neither f actor predicts tolerance or blood pressure reactivity: An experimental investigation in pain-free individuals. European Journal of Pain, 17, 133-142. Giesecke, T., Gracely, R. H., Williams, D. A., Ge isser, M. E., Petzke, F. W., & Clauw, D. J. (2005). The relationship between depressi on, clinical pain, and experimental pain in a chronic pain cohort. Arthritis and Rheumatism, 52, 1577-1584. Gilbert, P. (2005). Evolution and de pression: Issues and implications. Psychological Medicine, 35 1-11. Hall, H. V. & Poirier, J. G. (2001). Detecting malingering and deception. (2nd ed.) New York, NY: CRC Press. Haslam-Hopwood, G. T. (2003). The role of th e primary clincian in the multidisciplinary team. Bulletin of the Menninger Clinic, 67, 5-17. Katz, J. N., Losina, E., Amick, B. C., Fossel, A. H., Besette, L., & Keller, R. B. (2001). Predictors of outcomes of carpal tunnel release. Arthritis and Rheumatism, 44, 1184-1193. Keogh, E., McCracken, L. M., & Eccleston, C. (2005). Gender moderates the association between depressiona nd disability in chronic pain patients. European Journal of Pain

PAGE 116

106 Koch, L. C. (2001). The prefer ences and anticipations of pe ople referred for vocational rehabilitation. Rehabilitation Counseling Bulletin, 44, 76-86. Kuch, K. (2001). Psychological factors and the development of chronic pain. The Clinical Journal of Pain, 17, 33-37. Leigh, J. P. & Robbins, J. A. (2004). Occupa tional disease and workers' compensation: Coverage, costs, and consequences. The Milbank Quarterly, 82, 689-721. Linton, S. J. & Andersson, T. (2000). Can chr onic pain be prevented? A randomized trial of a cognitive-behavior intervention and two forms of information for patients with spinal pain. Spine, 25, 2825-2831. MacDonald, H. A., Colotla, V., Flamer, S., & Karlinsky, H. (2003). Posttraumatic stress disorder in the workplace: A descript ive study of workers experiencing PTSD resulting from work injury. Journal of Occupational Rehabilitation, 13, 63-77. Mason, S., Wardrope, J., Turpin, G., & Rowl ands, A. (2002). The psychological burden of injury: An eighteen m onth prospective cohort study. Journal of Trauma, 53, 98103. McGeary, D. D., Mayer, T. G., Gathcel, R. J., Anagnostis, C., & Proctor, T. J. (2003). Gender-related differences in treatment outcomes for patients with musculoskeletal disorders. The Spine Journal, 3, 197-203. McGuire, B. E. & Shores, E. A. (2001). Pa in patient profile a nd the assessment of malingered pain. Journal of Clinical Psychology, 57, 401-409. McGuire, B. E., Harvey, A. G., & Shores, E. A. (2001). Simulated malingering in pain patients: A study with the pain patient profile. The British Journal of Clinical Psychology, 40, 71-79. Mertler, C. A. & Vannatta, R. A. (2005). Advanced and multivariate statistical methods. (3rd ed.) Glendale, CA: Pyrczak Publishing. Nakao, M., Fricchione, G., Zuttermeister, P. C., Myers, P., Barsky, A. J., & Benson, H. (2001). Effects of gender and marital st atus on somatic symptoms of patients attending a mind/body medicine clinic. Behavioral Medicine, 26, 159-168. Piccinelli, M., Patterson, M., Braithwaite, I ., Boot, D., & Wilkinson, G. (1999). Anxiety and depression disorders five years after severe injuries: A prospective follow-up study. Journal of Psychosomatic Research, 46, 455-464. Plant, E. A. & Sachs-Ericsson, N. (2004). R acial and ethnic differences in depression: The roles of social support and meeting basic needs. Journal of Consulting and Clinical Psychology, 72, 41-52.

PAGE 117

107 Reneman, M. F., Brouwer, S., Meinema, A ., Dijkstra, J. W., Geertzen, J. H., & Groothoff, J. W. (2004). Test-retest reliabi lity of the isernhagen work systemes functional capacity evalua tion in healthy adults. Journal of Occupational Rehabilitation, 14, 295-305. Resnick, P. J. (1995). Posttraumatic stress disord er in litigation: Guidelines for forensic assessment. In R.I.Simon (Ed.), Guidelines for the evaluation of malingering in postraumatic stress disorder. (pp. 117-134). Washington D.C.: American Psychiatric Association. Rogers, R. (1997). Clinical assessment of malingering and deception. New York, NY: Guilford Press. Rosomoff, H. L., Fishbain, D. A., Cutler, R. B., & Steele-Rosomoff, R. (1995). Do chronic pain patients' perceptions about th eir preinjury jobs differ as a function of worker compensation and non-compensation status? The Clinical Journal of Pain, 11, 279-286. Sareen, J., Cox, B. J., Clara, I., & As mundson, G. (2005). The relationship between anxiety disorders and physic al disorders in the US national comorbidity survey. Depression and Anxiety, 21, 193-202. Schouten, R. & Williams, C. (2000). Psychiatric assessment and management of chronic disability syndromes in psychiatr ic care of the medical patient London: Oxford University Press. Schultz, I. Z., Crook, J., Meloche, G. R., Be rkowitz, J., Milner, R., & Zuberier, O. A. (2004). Psychosocial factors predictive of occupational low back disability: Towards development of a return-to-work model. Pain, 107, 77-85. Selander, J., Marnetoft, S., Bergroth, A., & Ekholm, J. (2002). Return to work following vocational rehabilitation for neck, back, and shoulder problems: Risk factors reviewed. Disability and Rehabilitation, 24, 704-712. Severeijns, R., Vlaeyen, J. W., van den H out, M., & Weber, W. E. (2001). Pain catastrophizing predicts pain intensit y, disability, and psychological distress independent of the level of physical impairment. The Clinical Journal of Pain, 17, 165-172. Shaw, L. R. & Betters, C. J. (2004). Private se ctor rehabilitation. In T.F.Riggar & D. R. Maki (Eds.), Handbook of rehabilitation counseling (pp. 236-252). New York, NY: Springer Series. Shaw, L. R., McMahon, B. T., Chan, F., & Hannold, L. (2004). Enhancement of the working alliance: A training program to align counselor and consumer expectations. Journal of Vocational Rehabilitation, 20, 107-126.

PAGE 118

108 Stevens, J. (2005). Multiple regression in behavioral research: Exploration and prediction. (2nd ed.) Hillsdale, NJ: Lawrence Erlbaum Associates. Stice, D. C. & Moore, C. L. (2006). A study of the relationship of th e characteristics of injured workers receiving vocational rehabi litation services and their depression levels. Journal of Rehabilitation, 71, 12-22. Sullivan, M. L. & Stanish, W. D. (20 03). Psychologically based occupational rehabilitation: The pain disability prevention program. The Clinical Journal of Pain, 19, 97-104. Suter, P. B. (2002). Employment and litigati on: Improved by work, assisted by verdict. Pain, 100 249-257. Tate, D. G. (1992). Workers' di sability and return to work. American Journal of Physical Medicine and Rehabilitation, 71, 92-96. Thomas, L. & Krebs, C. J. (1997). A review of statistical power analysis software. Bulletin of the Ecological Society of America, 78, 126-139. Tollison, C. D. & Satterthwaite, J. R. ( 1991). Chronic benign pain: Diagnosis and behavioral management. Journal of Musculos keletal Medicine, 8, 55-66. Tollison, C. D. & Langley, J. C. (1995). Pain patient profile manual. Minneapolis, MN: NCS Assessments. Tugman, K. & Palmer, C. (2004). The fear of returning to work: A model to enhance return to work outcomes. Rehab Pro, 48-54. Turk, D. C. (2002). Clinical effectiveness and cost-effec tiveness of tr eatments for patients with chronic pain. Clinical Journal of Pain, 18, 355-365. Turk, D. C. & Okifuji, A. (2002). Psychologi cal factors in chroni c pain: Evolution and revolution. Journal of Consulting and Clinical Psychology, 70, 678-690. Turk, D. C. & Burwinkle, T. M. (2005). C linical outcomes, cost-effectiveness, and the role of psychology in treatments for chronic pain sufferers. Professional Psychology: Research and Practice, 36, 602-610. Turner, J. B. & Turner, R. J. (2004). Phys ical disability, unemployment, and mental health. Rehabilitation Psychology, 49, 241-249. Ustun, T. B., Chatterji, S., Bickenbach, J ., Kostanjsek, N., & Schneider, M. (2003). The international classification of functioning, disability, a nd health: A new tool for understanding disabi lity and health. Disability and Rehabilitation, 25, 565-571.

PAGE 119

109 van Duijn, M., Miedema, H., Elders, L., & Bu rdorf, A. (2004). Barriers for early returnto-work of workers with musculoskeleta l disorders accordin g to occupational health physicians and human resource managers. Journal of Occupational Rehabilitation, 14, 31-39. Vowles, K. E. & Gross, R. T. (2003). Work -related beliefs about injury and physical capability for work in individuals with chronic pain. Pain, 101, 291-298. Waeher, G., Leigh, J. P., Cassady, D., & Mill er, T. R. (2004). Costs of occupational injury and illness across states. Journal of Occupational and Environmental Medicine, 46, 1084-1095. Watson, P. J., Booker, C. K., Moores, L., & Ma in, C. J. (2004). Returning the chronically unemployed with low back pain to employment. European Journal of Pain, 8, 359369. Weed, R. O. & Field, T. F. (2001). Rehabilitation consultant's handbook Athens, GA: Elliot & Fitzpatrick. Williams, D. R., Takeuchi, D. T., & Adair, R. K. (1992). Socioeconomic status and psychiatric disorder among blacks and whites. Social Forces, 71, 179-194. Williams, R. A., Pruitt, S. D., Doctor, J. N ., Epping-Jordan, J. A., Wahlgren, D. R., & Grant, I. (1998). The contribution of job sa tisfaction to the transition from acute to chronic low back pain. Archives of Physical Medicine and Rehabilitation, 79, 366373. Willoughby, S. G., Hailey, B. J., & Wheeler, L. C. (1999). Pain patient profile: A scale to measure psychological distress. Archives of Physical Medicine and Rehabilitation, 80, 1300-1302. Wittink, H. (2005). Functional capacity te sting in patients with chronic pain. The Clinical Journal of Pain, 21, 197-199. Woby, S. R., Watson, P. J., Roach, N. K., & Urmston, M. (2004). Are changes in fearavoidance beliefs, catastrophizing, and apprai sals of control, predictive of changes in chronic low back pain and disability? European Journal of Pain, 8, 201-210. Zatzick, D. F., Kang, M. A., Hinton, L., Kelly, R. H., Hilty, D. M., & Franz, C. E. (2001). Posttraumatic concerns: A patient-centered approach to outcome assessment after traumatic physical injury. Medical Care, 39, 327-339. Zhang, A. Y. & Snowden, L. R. (1999). Ethnic ch aracteristics of mental disorders in five U.S. communities. Cultural Diversity and Ethni c Minority Psychology, 5, 134-146.

PAGE 120

110 BIOGRAPHICAL SKETCH Chad Betters received his Bachelor of Arts degree in psychology and health science from the University of North Florida, Jack sonville, Florida, in 2001. Chad also received his Master of Health Science degree in reha bilitation counseling from the University of Florida, Gainesville, Florida, in 2003. He received his admission into the Rehabilitation Science Doctoral (RSD) Program at the Univ ersity of Florida in 2003. Specializing in rehabilitation counseling, Chad has been working towards the completion of his Doctor of Philosophy degree in rehabilitation science. Chad has become nationally certified in several areas of Rehabilitati on Counseling, including Certif ied Rehabilitation Counselor (CRC), Certified Disability Management Sp ecialist (CDMS), and Certified Vocational Evaluator (CVE). He has also become license d within the State of Florida as a Qualified Rehabilitation Provider (QRP) for Workers’ Compensation. Chad has maintained membership in the International Association of Rehabilitation Professionals (IARP), as well as the National Council of Rehabilitation Education (NCRE). He has worked within private sector rehabilitation counseling, provi ding Workers’ Compensation rehabilitation consultation services and vocational evaluations since 2003. Chad has also worked as an adjunct faculty, instructor, and teaching assist ant for the College of Public Health and Health Professions at the Univ ersity of Florida since 2003.