Journal of Personality Assessment - Meagher Lab - Texas A&M ...

This article was downloaded by:[Texas A & M University]On: 14 July 2008Access Details: [subscription number 776115308]Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UKJournal of Personality AssessmentPublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t775653663Predicting the Completion of an Integrative andIntensive Outpatient Chronic Pain Treatment With thePersonality Assessment InventoryChristopher J. Hopwood a ; Suzannah K. Creech b ; Timothy S. Clark c ; Mary W.Meagher d ; Leslie C. Morey da Department of Psychiatry, Massachusetts General Hospital and Harvard MedicalSchool,b Department of Psychiatry and Health Behavior, Medical College of Georgia,c Baylor Center for Pain Management, Baylor University Medical Center,d Department of Psychology, Texas A&M University,Online Publication Date: 01 January 2008To cite this Article: Hopwood, Christopher J., Creech, Suzannah K., Clark, Timothy S., Meagher, Mary W. andMorey, Leslie C. (2008) 'Predicting the Completion of an Integrative and Intensive Outpatient Chronic Pain TreatmentWith the Personality Assessment Inventory', Journal of Personality Assessment, 90:1, 76 — 80To link to this article: DOI: 10.1080/00223890701693785URL: http://dx.doi.org/10.1080/00223890701693785PLEASE SCROLL DOWN FOR ARTICLEFull terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdfThis article maybe used for research, teaching and private study purposes. Any substantial or systematic reproduction,re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expresslyforbidden.The publisher does not give any warranty express or implied or make any representation that the contents will becomplete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should beindependently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with orarising out of the use of this material.

Downloaded By: [Texas A & M University] At: 23:22 14 July 2008Journal of Personality Assessment, 90(1), 76–80, 2008Copyright C○ Taylor & Francis Group, LLCISSN: 0022-3891 print / 1532-7752 onlineDOI: 10.1080/00223890701693785Predicting the Completion of an Integrative and IntensiveOutpatient Chronic Pain Treatment With the PersonalityAssessment InventoryCHRISTOPHER J. HOPWOOD, 1 SUZANNAH K. CREECH, 2 TIMOTHY S. CLARK, 3 MARY W. MEAGHER, 4 AND LESLIE C. MOREY 41 Department of Psychiatry, Massachusetts General Hospital and Harvard Medical School2 Department of Psychiatry and Health Behavior, Medical College of Georgia3 Baylor Center for Pain Management, Baylor University Medical Center4 Department of Psychology, Texas A&M UniversityThe effects of intensive, integrative treatments for chronic pain are affected by patient compliance, and in many cases, selecting noncompliantindividuals adversely impacts the cost-effectiveness of such programs. The pretreatment identification of individuals who are at risk for dropoutcould assist clinicians in augmenting treatments with motivational enhancement strategies for high-risk patients or using such information to selectindividuals who are most likely to complete a given intervention program. In this study, we tested the ability of indicators from the PersonalityAssessment Inventory (PAI; Morey, 1991), administered prior to treatment, to identify individuals who dropped out of a 20-day chronic painprogram. Results replicate findings from outpatient psychotherapy research in finding that PAI Mean Clinical Elevation and Treatment ProcessIndex significantly differentiated dropouts from graduates, particularly when the Treatment Rejection scale suggested patients were motivated fortreatment. We discuss these results and offer recommendations for the prediction of treatment dropout in pain settings.The prediction of individuals who are likely to comply withtreatment regimens is an important issue in chronic pain becausecompliance relates to outcome (Turk & Okifuji, 2002).Several factors have been identified as barriers to treatmentcompliance in pain settings including severity of pain, severityof emotional problems/psychopathology, life disruption as aconsequence of pain, motivation for and openness to treatment,and social support (Cutler, Fishbain, Cole, Steele-Rosomoff, &Rosomoff, 2001; Lutz, Silbret, & Olshan, 1983; Richmond &Carmody, 1999).Recent research on dropout prediction has made important advances,with much attention paid to the stages of change modelof therapy readiness (Kerns & Habib, 2004; Kerns, Rosenberg,Jamison, Caudill, & Haythornwaite, 1997; Nielsen, Jensen, &Kerns, 2003; Prochaska & Velicer, 1997). This model was basedon a theoretical integration of several schools of psychotherapyand is based on the assumption that various therapeutictechniques emphasized by different schools could be optimallymatched to different stages of client readiness for change. Forexample, the theory suggests that individuals who are just beginningto contemplate personal changes would be most helped bytechniques designed to raise their consciousness regarding howtheir behavior could change to improve their lives. Conversely,individuals who are ready for therapeutic action should be providedwith activities that would allow them to make use of thismotivation to make personal change (Kerns & Habib, 2004). Intheory, people who are in the “precontemplation stage” shouldbe the least amenable to an action-oriented treatment such asan integrative and intensive chronic pain treatment. So far, re-Received October 17, 2006; Revised February 8, 2007Address correspondence to Christopher Hopwood, Texas A&M University,Department of Psychology, College Station, TX 88843-4235; Email:chopwood@partners.org76search using this approach to predict completion of pain treatmentshas generally yielded positive results. For example, severalstudies have demonstrated moderate relationships betweenpatients’ readiness as assessed by the Pain Stages of ChangeQuestionnaire (PSOCQ; Kerns et al., 1997) and the completionof self-management pain programs (Biller, Arnstein, Caudill,Federman, & Guberman, 2000; Kerns & Rosenberg, 2005).Another approach to the prediction of premature terminationinvolves the use of personality assessment instruments thatare commonly administered in integrative chronic pain treatmentsettings (Piotrowski, 1998). For example, research usingthe Minnesota Multiphasic Personality Inventory (MMPI;Hathaway & McKinley, 1943) in pain settings has generallyfound that overall severity is associated with a greater likelihoodof premature termination (Evans & Blanchard, 1988; Tsushima,Stoddard, Tsushima, & Daly, 1991). These findings are consistentwith psychotherapy research that has indicated that generalizedpsychopathology is the most robust indicator of therapydiscontinuation on the MMPI (e.g., Minnix et al., 2005).The Personality Assessment Inventory (PAI; Morey, 1991)represents a broadband personality instrument with severalindexes designed to assist clinicians in predicting treatmentdropout. The PAI has been found to be reliable in a pain setting(Karlin et al., 2005), and recent research suggests that it candistinguish individuals who complete outpatient psychotherapyfrom those who do not (Baity, 2003; Hopwood, Ambwani, &Morey, 2007). Hopwood et al. reported that the Treatment ProcessIndex (TPI), a configuration of several PAI scales theoreticallylinked to treatment amenability, with higher scores representinglower amenability (Morey, 1996; Stone, 1985; Strupp &Binder, 1984), was significantly lower for individuals who completedtherapy (as defined by a therapist rating) than for thosewho terminated prematurely. In addition, results demonstratedthat among individuals who were motivated for treatment as

PAIN TREATMENT COMPLETION 77Downloaded By: [Texas A & M University] At: 23:22 14 July 2008defined by the PAI Treatment Rejection (RXR; higher levelsindicate lower motivation) scale, the Mean Clinical Elevation(MCE; i.e., generalized psychiatric severity) significantly differentiatedthese groups, whereas generalized severity was ineffectiveat differentiating treatment completers from noncompleterswhen motivation was low. Hopwood et al. suggested that thisfinding may relate to the fact that among resistant individuals,a variety of external (e.g., finances, moving) or therapy dyadic(e.g., therapeutic alliance) factors may be more likely to interferewith the treatment process, making severity a less salient factor.Morey (1996) suggested that both high and low scores onRXR may be problematic. For example, high scores may suggesttreatment resistance and very low scores might suggest desperation,with both of these potentially resulting in decreasedmotivation and personal responsibility for change. Given thefindings discussed previously, it may be that motivation interactswith psychiatric severity in this fashion in pain settings. Forthe resistant, psychiatric severity may not be a salient predictoras discussed previously. Conversely, hypermotivated (i.e.,“desperate,” “crying for help”) patients may expect a great dealof change that does not occur and may easily become disappointed,precipitating dropout. For the appropriately motivated,however, severity would be anticipated to be a stronger predictorof likelihood of graduation.The prediction of treatment compliance with the PAI has notbeen tested in a pain setting. Given that the PAI is commonlygiven for a variety of purposes in pain settings, using it to predictdropout would streamline the assessment process, whether datawere used in lieu of or as a screen for measures like the PSOCQ.The purpose of this study was to test the ability of PAI indicatorsto differentiate individuals who did versus did not complete a20-day outpatient chronic pain treatment.METHODParticipantsParticipants were 309 individuals sampled from the files ofa hospital-based, integrative outpatient treatment program, 242of whom graduated and 67 of whom did not. To be eligible forthis program, individuals must have been referred by physiciansfor treatment-resistant chronic pain. Participants completed apacket of questionnaires including the PAI as part of eligibilityconsideration. The average age of sampled individuals was48.58 years (SD = 10.96); 72.0% were women; 79.5% wereWhite, 13.2% African American, 5.0% Hispanic, and the restdid not report ethnicity. All participants consented to participatein research, which was approved by the Institutional ReviewBoard at Texas A&M University.The program is a Commission on Accreditation of RehabilitationFacilities accredited, 4-week long, day treatment programthat includes physical therapy, aquatics therapy, cognitivebehavioralgroup psychotherapy, occupational therapy, individualbiofeedback and counseling, and vocational services asneeded. The average duration of pain in the clinic where datawere sampled is approximately 9 years. The type of pain experiencedby participants in this study varied; the most commonprimary diagnoses involved lumbar spine with radicular(i.e., related to pressure on nerve roots) symptoms (46.4%),cervical pain (16.1%), or fibromyalgia (6.0%). Patients werefunded primarily through commercial carriers (53.0%), secondarilythrough Medicare (20.8%), and least frequently throughWorkers Compensation (18.0%) or other means (8.2%).InstrumentScales and indicators from the PAI (Morey, 1991) were testedfor their ability to differentiate individuals who graduated fromthose who did not. The PAI is a 344-item, self-report instrumentwith 22 nonoverlapping full scales as well as a varietyof other indicators. The median internal consistency across allPAI full scales in the clinical normative sample was .86, and themedian 1-month test–retest reliability in a community samplewas .85 (Morey, 1991). A previous report that used data from510 individuals from the setting where this data were sampledand including approximately half of the individuals comprisingthis data reported a median internal consistency across PAI fullscales of .80 (Karlin et al., 2005).AnalysesAnalyses corresponded to the three levels of effects, two ofwhich were guided by hypotheses based on previous research,and the third of which was exploratory. First, the ability of threespecific PAI indicators to distinguish graduates from nongraduateswas tested based on their development for that purposeand previous results: MCE (severity), TPI (amenability), andRXR (motivation). Specifically, we hypothesized that means onthese variables would be higher (indicating greater severity andlower amenability and motivation) for nongraduates than forgraduates.Second, RXR was used to test hypotheses regarding the relativesalience of psychiatric severity across resistant, motivated,and hypermotivated patients. In pain settings, RXR scores tendto be somewhat higher than in other clinical settings and somewhatlower than in community settings (Karlin et al., 2005),suggesting that individuals in pain settings are more motivatedfor personal change than those in the community but less sothan those in psychotherapy clinics or inpatient settings. Thus,for our study, scores above the community mean (50T) suggestedresistance, scores below the clinical sample mean (39T)suggested hypermotivation, and scores in between these levelssuggested appropriate levels of motivation. However, it isimportant to recognize that such cut scores are likely to varyacross settings (e.g., outpatient psychotherapy clinic, inpatientpsychiatric hospital).The third set of analyses was exploratory in nature, owingto limited previous research using the PAI to predict treatmentcompliance in a pain setting. Independent sample t tests wereconducted to investigate for differences between individualswho did and did not graduate from the pain treatment programacross all PAI scales, and we computed effect sizes.RESULTSThree PAI indicators were predicted to be higher for treatmentdropouts than completers: MCE, TPI, and RXR. The effects ofMCE and TPI were significant and moderate in this discrimination,whereas the effect of RXR was not significant (Table 1).receiver operating characteristics (ROC) analyses suggested thatboth MCE (area under the curve [AUC] = .60 p

78 HOPWOOD ET AL.Downloaded By: [Texas A & M University] At: 23:22 14 July 2008TABLE 1.—Mean differences on hypothesized PAI indicators between graduatesand nongraduates of a 20-day outpatient chronic pain treatment program.Scale Graduate M (SD) Nongraduate M (SD) |t| p |d|Treatment 45.24 (9.95) 43.52 (10.95) 1.23 ns .18RejectionMean Clinical 55.41 (7.30) 58.54 (8.76) 2.97

PAIN TREATMENT COMPLETION 79Downloaded By: [Texas A & M University] At: 23:22 14 July 2008TABLE 3.—Pretreatment means (standard deviations) on PAI scales for graduates and nongraduates of a 20-day outpatient chronic pain treatment program.Scale Graduate M (SD) Nongraduate M (SD) |t| p |d|Validity scalesInconsistency 53.90 (9.99) 54.52 (10.67) 0.43 ns .06Infrequency 48.87 (7.59) 51.70 (10.25) 2.49

80 HOPWOOD ET AL.Downloaded By: [Texas A & M University] At: 23:22 14 July 2008Meyer, G. J., McGrath, R. E., & Rosenthal, R. (2003). Basic effect sizeguide with SPSS and SAS syntax. Retrieved March 1, 2007, fromhttp://www.erlbaum.com/Media/LEADocs/JPA/Basic%20ES%20Guide%2001-13-03.rtfMinnix, J. A., Reitzel, L. R., Repper, K. A., Burns, A. B., Williams, F., Lima,E. N., et al. (2005). Total number of MMPI–2 clinical scale elevations predictspremature termination after controlling for intake symptom severity andpersonality disorder diagnosis. Personality and Individual Differences, 38,1745–1755.Morey, L. C. (1991). The Personality Assessment Inventory professional manual.Odessa, FL: Psychological Assessment Resources.Morey, L. C. (1996). An interpretive manual for the Personality AssessmentInventory. Odessa, FL: Psychological Assessment Resources.Nielson, W. R., Jensen, M. P., & Kerns, R. D. (2003). Initial development andvalidation of a Multidimensional Pain Readiness to Change Questionnaire(MPRCQ). Journal of Pain, 104, 529–537.Piotrowski, C. (1998). Assessment of pain: A survey of practicing clinicians.Perceptual and Motor Skills, 86, 181–182.Prochaska, J., & Velicer, W. (1997). The transtheoretical model of healthbehavior change. American Journal of Health Promotion, 12, 38–48.Richmond, R. L., & Carmody, T. P. (1999). Dropout from treatment for chroniclow-back pain. Professional Psychology: Research and Practice, 30(10), 51–55.Stone, M. H. (1985). Analytically oriented psychotherapy in schizotypal andborderline patients: At the border of treatability. Yale Journal of BiologicalMedicine, 58, 275–288.Strupp, H. H., & Binder, J. L. (1984). Psychotherapy in a new key. New York:Basic Books.Tsushima, W. T., Stoddard, V. M., Tsushima, V. G., & Daly, J.(1991). Characteristics of treatment drop-outs among two samples ofchronic headache patients. Journal of Clinical Psychology, 47, 199–205.Turk, D. C., & Okifuji, A. (2002). Psychological factors in chronic pain: Evolutionand Revolution. Journal of Consulting and Clinical Psychology, 70,678–690.