12
Further Validation of the Chronic Pain Coping Inventory Gabriel Tan,* ,‡ Quang Nguyen, Karen O. Anderson, Mark Jensen, and John Thornby § Abstract: Multidisciplinary treatment programs for chronic pain typically emphasize the importance of decreasing maladaptive and encouraging adaptive coping responses. The Chronic Pain Coping Inventory (CPCI), developed to assess coping strategies targeted for change in multidisciplinary pain treatment, is a 64-item instrument that contains 8 subscales: Guarding, Resting, Asking for Assistance, Relaxation, Task Persistence, Exercising/Stretching, Coping Self-Statements, and Seeking Social Sup- port. A previous validation study with 210 patients in a Canadian academic hospital setting supported an 8-factor structure for the CPCI. The current study was undertaken to validate the CPCI among 564 veterans with a more extended history of chronic pain. Patients completed the study questionnaires before multidisciplinary treatment. A confirmatory factor analysis was used to examine the factor structure of the 64-item CPCI. A series of hierarchical multiple regression analyses were performed with depression, pain interference, general activity level, disability, and pain severity as the criterion variables and the 8 CPCI factors as the predictor variables, controlling for pain severity and demo- graphic variables. The confirmatory factor analysis results strongly supported an 8-factor model, and the regression analyses supported the predictive validity of the CPCI scales, as indicated by their association with measures of patient adjustment to chronic pain. Perspective: This article validated the 8-factor structure of the CPCI by using a confirmatory factor analysis and a series of linear regressions. The results support the applicability and utility of the CPCI in a heterogeneous population of veterans with severe chronic pain treated in a tertiary teaching hospital. The CPCI provides an important clinical and research tool for the assessment of behavioral pain coping strategies that might have an impact on patient outcomes. Key words: Coping with chronic pain, behavioral assessment instrument, validation of CPCI. A cognitive-behavioral model of chronic pain em- phasizes the importance of an individual’s at- tempts to cope with pain and pain-related issues or problems. Coping has been defined as “constantly changing cognitive and behavioral efforts to manage specific external and/or internal demands that are ap- praised as taxing, or exceeding the resources of the per- son.” 34 Coping is an ongoing process that includes ap- praisals of a demand or stressor; cognitive, behavioral, and emotional coping responses; and subsequent reap- praisals of the demand. Most individuals view pain and its impact as stressful and cope according to their unique history, personality, beliefs, biologic characteristics, and social environment. 8 Coping responses can be classified as problem-focused responses directed at changing or managing the problem causing the stress or as emotion- focused responses directed at regulating emotional re- actions to the problem. 34 Coping responses also can be considered adaptive or maladaptive, depending on the usual outcome of the coping behavior, cognition, or emotion. Multidisciplinary treatment programs for chronic pain typically emphasize the importance of stopping the use of maladaptive and teaching and encouraging the use of adaptive coping responses. 37,48 Certain coping re- sponses, such as task persistence, have been associated with better physical and psychological functioning among individuals with chronic pain, whereas other cop- ing responses, such as pain-contingent resting and guarding, have been correlated with impaired func- tion. 12,15,25,31,51 Cross-sectional and longitudinal studies have demon- strated that coping responses are associated with both short- and long-term adjustment to chronic pain. 1,5,8,28,36,47,59 Pain coping responses have been shown to be related to measures of pain intensity, emo- tional distress, physical function, disability status, and other psychosocial variables. For example, the use of multiple pain coping strategies has been correlated neg- atively and significantly with disability 14,38 and with psy- chosocial dysfunction. 23 A composite measure represent- ing cognitive coping responses (labeled “perceived control over pain and rational thinking”) has been sig- Received February 24, 2004; Revised September 17, 2004; Accepted Sep- tember 25, 2004. From the *Pain Section, Anesthesiology and Mental Health Care Line and †Psychosocial Rehabilitation Program, VA Medical Center; ‡Department of Anesthesiology and §Department of Family Medicine, Baylor College of Medicine; and Department of Symptom Research, The University of Texas M.D. Anderson Cancer Center, Houston, Texas; and ¶Department of Rehabilitation Medicine, University of Washington, Seattle, Washington. Address reprint requests to Gabriel Tan, PhD, ABPP, Veterans Affairs Medical Center, Department of Anesthesiology, 2002 Holcombe Blvd (116 MH), Houston, TX 77030. E-mail: [email protected] 1526-5900/$30.00 doi:10.1016/j.jpain.2004.09.006 29 The Journal of Pain, Vol 6, No 1 (January), 2005: pp 29-40

Further validation of the chronic pain coping inventory

Embed Size (px)

Citation preview

Page 1: Further validation of the chronic pain coping inventory

Aocspspapihsam

RtF†ooTRAM(1d

Further Validation of the Chronic Pain Coping Inventory

Gabriel Tan,*,‡ Quang Nguyen,† Karen O. Anderson,� Mark Jensen,¶ andJohn Thornby§

Abstract: Multidisciplinary treatment programs for chronic pain typically emphasize the importanceof decreasing maladaptive and encouraging adaptive coping responses. The Chronic Pain CopingInventory (CPCI), developed to assess coping strategies targeted for change in multidisciplinary paintreatment, is a 64-item instrument that contains 8 subscales: Guarding, Resting, Asking for Assistance,Relaxation, Task Persistence, Exercising/Stretching, Coping Self-Statements, and Seeking Social Sup-port. A previous validation study with 210 patients in a Canadian academic hospital setting supportedan 8-factor structure for the CPCI. The current study was undertaken to validate the CPCI among 564veterans with a more extended history of chronic pain. Patients completed the study questionnairesbefore multidisciplinary treatment. A confirmatory factor analysis was used to examine the factorstructure of the 64-item CPCI. A series of hierarchical multiple regression analyses were performedwith depression, pain interference, general activity level, disability, and pain severity as the criterionvariables and the 8 CPCI factors as the predictor variables, controlling for pain severity and demo-graphic variables. The confirmatory factor analysis results strongly supported an 8-factor model, andthe regression analyses supported the predictive validity of the CPCI scales, as indicated by theirassociation with measures of patient adjustment to chronic pain.Perspective: This article validated the 8-factor structure of the CPCI by using a confirmatory factoranalysis and a series of linear regressions. The results support the applicability and utility of the CPCIin a heterogeneous population of veterans with severe chronic pain treated in a tertiary teachinghospital. The CPCI provides an important clinical and research tool for the assessment of behavioralpain coping strategies that might have an impact on patient outcomes.

Key words: Coping with chronic pain, behavioral assessment instrument, validation of CPCI.

facue

toaswaigt

sbpstomaci

cognitive-behavioral model of chronic pain em-phasizes the importance of an individual’s at-tempts to cope with pain and pain-related issues

r problems. Coping has been defined as “�constantlyhanging cognitive and behavioral efforts to managepecific external and/or internal demands that are ap-raised as taxing, or exceeding the resources of the per-on.”34 Coping is an ongoing process that includes ap-raisals of a demand or stressor; cognitive, behavioral,nd emotional coping responses; and subsequent reap-raisals of the demand. Most individuals view pain and

ts impact as stressful and cope according to their uniqueistory, personality, beliefs, biologic characteristics, andocial environment.8 Coping responses can be classifieds problem-focused responses directed at changing oranaging the problem causing the stress or as emotion-

eceived February 24, 2004; Revised September 17, 2004; Accepted Sep-ember 25, 2004.rom the *Pain Section, Anesthesiology and Mental Health Care Line andPsychosocial Rehabilitation Program, VA Medical Center; ‡Departmentf Anesthesiology and §Department of Family Medicine, Baylor Collegef Medicine; and �Department of Symptom Research, The University ofexas M.D. Anderson Cancer Center, Houston, Texas; and ¶Department ofehabilitation Medicine, University of Washington, Seattle, Washington.ddress reprint requests to Gabriel Tan, PhD, ABPP, Veterans Affairsedical Center, Department of Anesthesiology, 2002 Holcombe Blvd

116 MH), Houston, TX 77030. E-mail: [email protected]/$30.00

coi:10.1016/j.jpain.2004.09.006

The Journal of Pain, Vol 6, No 1

ocused responses directed at regulating emotional re-ctions to the problem.34 Coping responses also can beonsidered adaptive or maladaptive, depending on thesual outcome of the coping behavior, cognition, ormotion.Multidisciplinary treatment programs for chronic pain

ypically emphasize the importance of stopping the usef maladaptive and teaching and encouraging the use ofdaptive coping responses.37,48 Certain coping re-ponses, such as task persistence, have been associatedith better physical and psychological functioningmong individuals with chronic pain, whereas other cop-ng responses, such as pain-contingent resting anduarding, have been correlated with impaired func-ion.12,15,25,31,51

Cross-sectional and longitudinal studies have demon-trated that coping responses are associated withoth short- and long-term adjustment to chronicain.1,5,8,28,36,47,59 Pain coping responses have beenhown to be related to measures of pain intensity, emo-ional distress, physical function, disability status, andther psychosocial variables. For example, the use ofultiple pain coping strategies has been correlated neg-

tively and significantly with disability14,38 and with psy-hosocial dysfunction.23 A composite measure represent-ng cognitive coping responses (labeled “perceived

ontrol over pain and rational thinking”) has been sig-

29(January), 2005: pp 29-40

Page 2: Further validation of the chronic pain coping inventory

nlcldcstssaa

ttcsfscfidpcc

omMQtVlaa(atpord

ashhpp

smpTetiem

ASSfcsApwiSr

pCtitptisisap

mtmsottswtsspjacsafmdC

tctapwdep

30 Validation of the Chronic Pain Coping Inventory

ificantly associated with less intense pain3,27,46,53 andower levels of distress and disability.3,10,27,52 In contrast,atastrophizing responses have been significantly corre-ated with greater pain intensity,11,12,19,50,51,58 affectiveistress,12,20,23,55,57 and disability.13,34,35,38,42 Studies ofoping responses such as diverting attention,1,49 copingelf-statements,18,23,33 reinterpreting pain sensa-ions,1,27 ignoring pain,22 praying or hoping,2,23,41 andocial support16,29,44 have produced mixed or inconclu-ive results. One reason for the inconsistent findingscross studies might be the variety of measures used tossess coping responses.Accurate assessment of pain coping responses is impor-

ant in research and clinical settings. For example, in mul-idisciplinary pain treatment programs, measurement ofoping responses helps to determine a patient’s copingtrengths and weaknesses, to identify possible targetsor treatment, and to evaluate treatment outcomes. In atudy to specifically identify the coping responses mostlosely linked to treatment outcome, Jensen et al24

ound that decreases in the coping strategies of guard-ng, resting, and catastrophizing were associated withecreased disability during the course of multidisci-linary pain treatment. In this same study, decreases inatastrophizing cognitions were also associated with de-reases in pain intensity and depression.A variety of pain coping measures have been devel-ped to address the need for reliable and valid assess-ent of pain coping responses. The Vanderbilt Painanagement Inventory (VPMI),7 Coping Strategiesuestionnaire (CSQ),45 and Chronic Pain Coping Inven-

ory (CPCI)25 are 3 of the most widely used measures. ThePMI assesses active and passive coping strategies re-

ated to chronic pain. Active strategies are defined asdaptive coping responses (eg, staying busy or active),nd passive strategies are defined as maladaptive oneseg, restricting social activities). Passive coping has beenssociated with high levels of pain, disability, and dis-ress. In contrast, active coping has been correlated withositive affect and higher activity levels.48,56,61 More-ver, longitudinal studies have found that passive copingesponses were predictive of subsequent pain levels andisability.40,56

The CSQ has been the most frequently used tool for thessessment of pain coping strategies in previous re-earch. The 50-item CSQ assesses 6 cognitive and 2 be-avioral coping strategies. Scores on the CSQ subscalesave been significantly associated with measures ofain, distress, and function across a variety of samples ofatients with chronic pain.5,8

The VPMI and CSQ focus largely on cognitive copingtrategies. In contrast, the CPCI was developed to assessany of the behavioral coping strategies that are em-hasized in multidisciplinary pain treatment programs.25

hese coping strategies include ones that are taught andncouraged during treatment (eg, relaxation, exercising,ask persistence), ones that are discouraged (eg, guard-ng, resting, asking for assistance), and one neutral strat-gy (seeking social support). The CPCI contains 8

ultiple-item subscales: Guarding, Resting, Asking for d

ssistance, Relaxation, Task Persistence, Exercising/tretching, Coping Self-Statements, and Seeking Socialupport. In the initial validation study, 176 patients re-erred to a multidisciplinary pain treatment programompleted the CPCI and measures of function. The re-ults indicated that scores on the Guarding, Resting, andsking for Assistance subscales were associated withoor adjustment to pain. Task Persistence was associatedith more favorable adjustment.25 Surprisingly, in the

nitial validation study, the strategy of Coping Self-tatements was associated with a higher level of pain-elated emotional distress.In a recent study of 564 veterans referred to a chronicain program, subjects were administered the CPCI andSQ, as well as measures of pain, distress, and disabili-y.51 Both the CPCI and CSQ subscale scores were signif-cantly associated with concurrent disability, althoughhe CPCI Guarding subscale was the single most powerfulredictor of disability. The CPCI Seeking Social Support,he CSQ Catastrophizing, and the CSQ Increasing Behav-oral Activities subscale scores were also significantly as-ociated with disability. Although the CSQ Catastrophiz-ng subscale score and education level were thetrongest predictors of pain severity, the CPCI Guardingnd Resting subscale scores were also associated withain severity in this Veterans Affairs sample.In a factor analytic study of the CPCI, 210 patients in aultidisciplinary pain treatment program were adminis-

ered multiple measures of pain, coping, and adjust-ent.17 An exploratory factor analysis of the CPCI items

upported an 8-factor solution that corresponded to theriginal 8 CPCI subscales.17 Although the item content ofhe subscales was generally supported, some modifica-ions to subscale scoring were suggested for the patientample in this study. For example, the Guarding subscaleas supported by the factor analysis results, but a few of

he Guarding items had higher factor loadings on othercales. Regression analyses found that scores on the sub-cales of Asking for Assistance, Guarding, and Social Sup-ort were significantly associated with less favorable ad-

ustment to chronic pain. Surprisingly, Relaxation wasssociated with increased emotional distress and de-reased perceptions of control. Task Persistence was as-ociated with more favorable adjustment to pain. Theuthors concluded that the CPCI is a valuable measureor identifying coping strategies and evaluating adjust-ent and treatment outcomes. They recommended ad-

itional studies to confirm the factor structure of thePCI in other patient populations.The purpose of the current study is to confirm the fac-

or structure of the CPCI in a population of patients withhronic pain in a Veterans Administration Medical Cen-er. In the previous factor analytic study of the CPCI citedbove, an exploratory factor analysis was used, and theatient sample was composed of women and men whoere evaluated in a pain treatment program in an aca-emic medical center.17 This program focused on work-rs compensation rehabilitation in which the majority ofatients presented with pain in one site, and the average

uration of pain was 14 months. In contrast, the sample
Page 3: Further validation of the chronic pain coping inventory

ivcorwJjatt

tacoabcgrLim

M

P

tgtiCopaPhmpatntrso14tp

wCt8ts(

wAoa5twwdtp

drfssdrg0dc

M

C

tpoeoftUsCS

eldnwsramstbfcrwr

31ORIGINAL REPORT/Tan et al

n the current study was composed primarily of maleeterans with a more severe and extended history ofhronic pain (more than half of this population had beenn disability for more than 5 years, and a majority, 72%,eported pain at multiple sites). We sought to determinehether the 8-factor structure originally proposed by

ensen et al25 and subsequently supported by Had-istavropoulos et al17 would be confirmed by performing

confirmatory factor analysis in a new sample of pa-ients with chronic pain and also determine how the fac-or scores relate to patient adjustment.A secondary but equally important reason for under-

aking this study was to underscore the importance ofssessing the behavioral domain of coping in addition toognitive coping strategies. Coping with pain dependsn what people do as well as what they think. Pain man-gement programs teach and encourage changes inoth thoughts and behavior. A well-validated behavioraloping measure is, therefore, of immense importance,iven that measures of both cognitive and behavioralesponses to pain are needed to guide interventions.imiting coping assessment to just one potentially limitsnformation needed to guide treatment and therefore

ight ultimately limit treatment efficacy.

ethods

articipantsThe sample was obtained from the population of pa-

ients with chronic noncancer pain referred to the Inte-rated Pain Management Program (IPMP) of the Hous-on VA Medical Center, a tertiary teaching hospital. Thiss the same sample used in a previous study comparingPCI and CSQ responses.51 The IPMP is a multidisciplinaryutpatient pain assessment, consulting, and treatmentrogram that receives referrals from surgical, medical,nd psychiatric departments within the medical center.atients referred to the IPMP typically have an extendedistory of chronic pain and have received prior treat-ents for pain. Patients were required to complete aacket of self-report questionnaires before initial IPMPssessment. The questionnaire packet was mailed to pa-ients with a cover letter explaining that the question-aires were to be used primarily for clinical purposes, buthe data might also be used for program evaluation andesearch. Before data were analyzed for the currenttudy, approval from the Institutional Review Board wasbtained. Of the 1265 packets mailed from 1995 through998, 564 packets were returned, with a return rate of4.6%. Thus, slightly less than half of all patients referredo the IPMP followed through on their referral and com-leted the necessary paperwork to receive services.The demographic characteristics of the subject sampleere reported in our previous study comparing CPCI andSQ responses.51 The mean age of patients completinghe questionnaires was 50.8 years (SD, 11.4; range, 22 to2 years). Most (84.5%) had at least a high school educa-ion; 12% were college graduates. The majority of theubjects were male (90.3%). Although most were white

62.4%), 22.6% were black, 4.6% were Hispanic, 0.5% t

ere “Other,” and 9.9% did not respond to this item.pproximately half (50.2%) were married. Almost halff the participants (48.0%) were already receiving dis-bility compensation for a pain-related condition, and8.0% indicated disability as a result of pain for morehan 5 years. Information about specific pain diagnosisas not collected, but breakdowns by primary pain sitesere as follows: back (39.0%), limbs (32.0%), neck/shoul-ers (19.0%), head (6.0%), and “all over” (4.0%). Fur-hermore, 72% of the patients reported that they hadain at multiple sites.To evaluate possible nonresponse bias, we collectedemographic and disability information on all patientseferred between January 1996 and December 1996. In-ormation was obtained on 94% of the patients and re-ulted in a sample of 126 responders and 168 nonre-ponders. The responders and nonresponders did notiffer with regard to age (mean, 54.5 and 53.9 years,espectively). There was also no significant difference inender (�2 � 1.92, not significant), marital status (�2 �.03, not significant), or disability status (ie, receiving noisability, service-connected disability, or non-service-onnected disability) (�2 � 0.68, not significant).

easures

PCIThe CPCI has 65 items that measure 11 coping strategies

hat patients might use to cope with or manage chronicain.25 The CPCI was specifically developed to assess manyf the coping strategies targeted for change (either to bencouraged or discouraged) in multidisciplinary treatmentf chronic pain.25 The strategies assessed include illness-ocused strategies (Guarding, Resting, Asking for Assis-ance, Opioid Medication Use, Nonsteroidal Medicationse, Sedative-Hypnotic Medication Use), wellness-focused

trategies (Relaxation, Task Persistence, Exercise/Stretch,oping Self-Statements), and the strategy of Seeking Socialupport.In the initial validation and cross-validation studies,25

ach of the CPCI scales demonstrated adequate to excel-ent test-retest stability (ranging from .65 to .90). In ad-ition, the internal consistency of the multiple-item (ie,ot medication) scales ranged from .74 to .91. Validityas demonstrated by the scales’ ability to predict

pouse-rated patient functioning. In particular, patient-eported use of Guarding, Resting, Asking for Assistance,nd Sedative-Hypnotic Medication Use demonstratedoderate to strong associations (rs � .30 or � �.30) with

pouse-rated patient disability, spouse-rated patient ac-ivity level, or both. Less strong, but still substantial (rsetween .20 and .30 or �.20 and �.30), associations wereound between Coping Self-Statements and Seeking So-ial Support and spouse-reported disability. For the cur-ent study, subscales relating to medication consumptionere excluded because of difficulty getting accurate and

eliable self-reports on this measure from the subjects;

herefore, only 8 of the 11 CPCI scales were used.
Page 4: Further validation of the chronic pain coping inventory

M

isDIswAeW

losim(TaCCh

aSsaidepi

Rchap

atcibbcsus

idwaScfh

rcdtowearw.

P

(letacsahT0

D

ioawvmecwatsftmamwliiuaOltsatc

32 Validation of the Chronic Pain Coping Inventory

easures of Function and AdjustmentMeasures of pain and functioning available for analysis

ncluded the Center for Epidemiological Studies Depres-ion Scale (CES-D) to assess depression, the Roland-Morrisisability Questionnaire (RMDQ) to assess disability, the

nterference scale of the West Haven–Yale Multidimen-ional Pain Inventory (WHYMPI) to assess the extent tohich pain interferes with the patient’s life, the Generalctivity scale of the WHYMPI to assess the patient’s gen-ral activity level, and the Pain Severity scale of theHYMPI.CES-D. The CES-D was used as a measure of psycho-

ogical functioning in this study. The CES-D was devel-ped to assess the presence and severity of depressiveymptoms in a general population. The CES-D includes 20tems that are answered on a 4-point scale (0, rarely; 3,

ost of the time), resulting in scores ranging from 0 to 60higher scores indicating greater depressive symptoms).he CES-D has been widely used in pain research6,22,39

nd has adequate reliability and convergent validity.41

riterion validity has also been established because theES-D scores of depressed and nondepressed subjectsave been found to be significantly different.41

RMDQ. The RMDQ is used to assess disability associ-ted with chronic pain. The RMDQ, derived from theickness Impact Profile, was originally developed to as-ess disability associated with back pain.43 Items focuslmost exclusively on the physical dimensions of disabil-ty.9 Subjects indicate which, if any, of 24 statementsescribe them today and are related to their pain (forxample, “I stay at home most of the time because of myain”). Scores range from 0 to 24, with higher scores

ndicating greater disability.Chronic pain treatment centers have found that theMDQ is simple to use and provides a great deal of usefullinical information about patients’ disability.60 Researchas supported the validity and reliability of the scale forssessing disability among persons with mixed chronicain problems23 as well as low back pain.9,32

WHYMPI. The WHYMPI30 is a 56-item measure thatssesses the impact of pain on the patient’s life, the pa-ient’s view of how significant others respond to theirommunication of pain, and the patient’s general activ-ty level. The validity and reliability of the WHYMPI haveeen well established.30 The validity of the WHYMPI haseen further supported by the results of exploratory andonfirmatory factor-analytic procedures.54 Two of thecales, Interference and General Activity Level, weresed as measures of function and adjustment for thistudy.The Interference scale of the WHYMPI consists of 11

tems that assess how pain has interfered with day-to-ay activities and functioning, including the ability toork, to enjoy family, to participate in social and recre-tional activities, and to perform household chores.30

cale scores range from 0 to 66, with higher scores indi-ating greater perceived interference over one’s dailyunctioning. The Interference scale has been reported to

ave an excellent internal consistency of 0.90 and a test- t

etest stability of 0.86.30 The General Activity Level scaleonsists of 18 items representing household chores, out-oor work, activities away from home, and social activi-ies. The scale score is computed by averaging the scoresf these 4 activities. Scale scores range from 0 to 108,ith higher scores indicating higher levels of activitiesngaged by the patients from self-report. In the originalrticle,30 the Cronbach � of the 4 individual activity scalesanged from .70 to .86; however, the � for the total scaleas not reported. A subsequent study reported an � of

76 for the total general activity scale.4

ain SeverityBecause pain severity could influence the predictors

CPCI scales) as well as the criterion measures of psycho-ogical and physical functioning (depression, interfer-nce, general activity level, and disability) in the study,he Pain Severity subscale of the WHYMPI was used tossess pain severity and was entered in the analyses toontrol for this potential confound. The Pain Severityubscale consists of 3 items that assess both pain intensitynd suffering. Subscale scores range from 0 to 18, withigher scores indicating greater intensity and suffering.he pain severity subscale has an internal consistency of.72 and a test-retest stability of 0.82.30

ata AnalysisThe confirmatory factor analysis was performed by us-

ng version 8.30 of LISREL.26 The initial model was basedn the 8-factor model obtained from a previous explor-tory factor analysis.17 Each of the 64 observed variablesas assumed initially to be associated with the factor

ariable having its largest factor loading from the Vari-ax rotation result of exploratory factor analysis. Thus

ach observed variable was assumed initially to be asso-iated with 1 and only 1 of the 8 factor variables,hereas each factor variable was assumed to be associ-ted with the observed variables having their largest fac-or loadings. The correlation matrix between all 64 ob-erved variables served as input data to the analysis. Theactor variables were specified as standardized so thathe covariance matrix among factor variables was esti-ated with the restriction of having 1’s on the diagonal

nd correlation coefficients on the off-diagonal ele-ents. In succeeding iterations of the model, changesere made on the basis of the sizes and t values of factor

oadings (�) on the observed variables currently includedn the model and on diagnostic indicators (modificationndices) of improvements as indicated by decreases innexplained �2 values that could be realized by addingdditional loadings not included in the current model.utput from each model included the matrix of factor

oadings, the correlation matrix between factors,he correlation matrix between measurement errors,quared multiple correlations between observed vari-bles, various measures of goodness of fit and modifica-ion indices. In addition, the output from each modelontained the regression coefficients for calculating fac-

or scores from values of the observed variables.
Page 5: Further validation of the chronic pain coping inventory

rtovitwat

R

F

sctrtfiltmmwmT

aasf

iA

0

tanftnfcitwsn

DV

ipI

sdPisdtm14mbp9stc(am�plpc

IC

r(sstfi

CDV

spdepccwobcattwn

g

33ORIGINAL REPORT/Tan et al

Factor scores were obtained by applying the factor scoreegression coefficients to the rescaled input variables. Al-hough there were regression coefficients for each factorn all 64 variables, only those coefficients associated withariables in the model for each factor were actually used, ass customary in exploratory factor analysis. The difference ishat factor scores from exploratory factor analysis used uniteights, whereas factor scores from confirmatory factornalysis used weights that were approximately propor-ional to the factor loadings.

esults

actor Analysis of the CPCIThe initial confirmatory factor analysis model provided

trong support for the exploratory factor analysis modelonducted by the Canadian researchers.17 Loadings ofhe 8 factors on the 64 variables were essentially compa-able in magnitude to those obtained previously forhose associations assumed to exist. However, the modi-cation indices, along with magnitudes of the estimated

oadings, indicated that improvements could be made inhe initial and succeeding confirmatory factor analysisodels, either by adding loadings not part of the currentodel or by eliminating loadings considered to be tooeak for inclusion in the model. The results of the finalodel of the confirmatory facto analysis are shown in

able 1.As a measure of similarity between the initial explor-

tory factors and those derived from confirmatory factornalysis, we computed the correlation between the 2ets of factors when applied to our input data, with theollowing results:Factor: Exercise/Stretch, Coping Self-Statements, Guard-

ng, Seeking Social Support, Task Persistence, Relaxation,sking for Assistance, RestingCorrelation: 0.993, 0.990, 0.979, 0.994, 0.996, 0.937,

.999, 0.995This analysis, showing very high correlations be-

ween factor scores calculated from either the explor-tory or confirmatory factor analysis, provided ratio-ale for using either the unit weights (exploratory) oractor regression coefficients (confirmatory) when es-imating factor scores and also showed that it is notecessary to use all variables when calculating valuesor each factor because the factor regression coeffi-ients not used in the calculations were all negligiblen size and would have had minimal effect on the ac-ual scores. Because of this finding, the tables to followere based on unit weights rather than factor regres-

ion coefficients for ease of comparisons to the origi-al CPCI validation studies.

escriptive Statistics for PrimaryariablesMeans, SDs, and ranges were computed for the empir-

cally derived CPCI factors and the depression (CES-D),ain interference (Multidimensional Pain Inventory–Pain

nterference [MPIIF]), general activity level (Multidimen- t

ional Pain Inventory–General Activity Level [MPIGL]),isability (RMDQ), and pain severity (Multidimensionalain Inventory–Pain Severity [MPIPS]) scales (Table 2). It is

mportant to note that the mean CES-D score of 28.29 isubstantially above the cutoff score of 23 for probableepression as suggested by Husaini et al21 and largerhan that of a sample of persons with chronic pain beforeultidisciplinary pain treatment (mean, 25.11; SD,

2.86).24 Disability scores in this sample (mean, 16.32; SD,.91) are similar to those of a chronic pain sample beforeultidisciplinary pain treatment (mean, 15.10; SD, 4.94)

ut significantly higher than the postmultidisciplinaryain treatment scores from this same sample (mean,.12; SD, 5.70; P � .01).24 Finally, pain severity in thisample (mean, 5.10; SD, .86) is significantly higher thanhe original published data from a heterogeneoushronic pain population at the University of Pittsburghmean, 4.53; SD, 1.04; t � 8.76; P �. 001) but lower than

sample of patients with chronic pain about to enterultidisciplinary pain treatment (mean, 6.15; SD, 1.51; P.001).24 Taken as a whole, our sample appears to re-

ort pain severity, depressive symptoms, and disability ateast as great as, and perhaps greater than, other sam-les of patients with chronic pain that were available foromparison.

ntercorrelations of Empirically DerivedPCI FactorsA correlation matrix was computed to examine the

elationships among the empirically derived CPCI factorsbased on outcome from the confirmatory factor analy-is). As can be seen in Table 3, there are a number oftatistically significant correlations among the CPCI fac-ors, and the Pearson product-moment correlation coef-cients tended to be in the low to moderate range.

orrelations between Empiricallyerived CPCI Scales and CriterionariablesThe relation between the empirically derived CPCI

cales and the criterion variables of depression (CES-D),ain interference (MPIIF), general activity level (MPIGL),isability (RMDQ), and pain severity (MPIPS) were initiallyxamined via zero-order correlations, yielding Pearsonroduct-moment correlation coefficients (Table 4). Be-ause of the large number of correlation coefficientsomputed and examined for these analyses, the � levelas set at P less than .01 to help control for the possibilityf Type I error when interpreting these findings.24 As cane seen in Table 4, depression had significant, positiveorrelations with the Guarding, Asking for Assistance,nd Resting factors and a significant, negative correla-ion with the Coping Self-Statements and Task Persis-ence factors. Pain interference was positively correlatedith Guarding, Asking for Assistance, and Resting andegatively correlated with Task Persistence.Six of the factors were significantly correlated witheneral level of activity: Exercise/Stretch, Task Persis-

ence, and Seeking Social Support had positive associa-
Page 6: Further validation of the chronic pain coping inventory

T

62641532531345122213143434145314364258152663242515132359

34 Validation of the Chronic Pain Coping Inventory

able 1. Eight-Factor Model of the CPCI: Factor Loadings by ItemFACTORS

ITEM (ORIGINAL CPCI SCALE)I II III IV V VI VII VIII

(EXE) (COP) (GUA) (SEE) (TAS) (REL) (ASK) (RES)

0. (Exercise/Stretch) .799. (Exercise/Stretch) .781. (Exercise/Stretch) .763. (Exercise/Stretch) .747. (Exercise/Stretch) .735. (Exercise/Stretch) .718. (Exercise/Stretch) .706. (Exercise/Stretch) .692. (Exercise/Stretch) .67. (Exercise/Stretch) .673. (Exercise/Stretch) .592. (Exercise/Stretch) .315. (Coping Self-Statements) .774. (Coping Self-Statements) .729. (Coping Self-Statements) .717. (Coping Self-Statements) .673. (Coping Self-Statements) .641. (Coping Self-Statements) .630. (Coping Self-Statements) .627. (Coping Self-Statements) .624. (Coping Self-Statements) .619. (Coping Self-Statements) .600. (Coping Self-Statements) .596. (Guarding) .713. (Guarding) .678. (Guarding) .655. (Guarding) .571. (Guarding) .566. (Guarding) .545. (Guarding) .501. (Guarding) .460. (Resting) .399. (Guarding) .32 .384. (Resting) .324. (Seeking Social Support) .720. (Seeking Social Support) .707. (Seeking Social Support) .64. (Seeking Social Support) .636. (Seeking Social Support) .623. (Seeking Social Support) .482. (Seeking Social Support) .48. (Seeking Social Support) .423. (Task Persistence) .714. (Task Persistence) .648. (Task Persistence) .64. (Task Persistence) .64. (Task Persistence) .591. (Task Persistence) .54. (Relaxation) .639. (Relaxation) .622. (Relaxation) .616. (Relaxation) .534. (Relaxation) .511. (Relaxation) .490. (Relaxation) .49

. (Asking for Assistance) .77
Page 7: Further validation of the chronic pain coping inventory

ticiwcsT

atStar

M

thwa((Pibweaaifbtnc

taaj

R4mei2rvw

cgrava

wTApvttdftswWapPcGptpsfa

T

26441575

AA

35ORIGINAL REPORT/Tan et al

ions, whereas Guarding, Asking for Assistance, and Rest-ng had negative associations. Disability was positivelyorrelated with Guarding, Seeking Social Support, Ask-ng for Assistance, and Resting and negatively correlatedith Task Persistence. Finally, pain severity had signifi-

ant, positive correlations with Guarding, Asking for As-istance, and Resting and a negative correlation withask Persistence.In general, the correlations between the CPCI factors

nd the dependent measures were in the expected direc-ions, assuming that the CPCI Task Persistence, Exercise/tretch, and Coping Self-Statements scales assess “adap-ive” coping responses, and the CPCI Guarding, Resting,nd Asking for Assistance assess “maladaptive” copingesponses.

ultiple Regression AnalysesTo explore the ability of the empirically derived CPCI fac-

ors to predict important pain-related outcomes, a series ofierarchical multiple regression analyses were conductedith depression (CES-D), pain interference (MPIIF), generalctivity level (MPIGL), disability (RMDQ), and pain severityMPIPS) as the criterion variables and the 8 CPCI factorsGuarding, Resting, Asking for Assistance, Relaxation, Taskersistence, Exercise/Stretch, Seeking Social Support, Cop-ng Self-Statements) as the primary predictor variables (Ta-le 5). Pain severity (when not used as a criterion variable)as entered in the first step. Age, sex, and education werentered in the second step. We controlled for these vari-bles because they often correlate with adjustment, as wells coping strategy use. The CPCI factors were entered lastnto the equations to determine whether they accountedor unique variance in the criterion variables, above andeyond what is already accounted for by pain severity andhe demographic variables. Again, because of the largeumber of analyses, we set the � level at P less than .01 toontrol for � inflation associated with multiple analyses.24

As can been seen in Table 5, in each regression model,he control variables and the CPCI factors as a groupccounted for a statistically significant amount of vari-nce in the criterion variable, as indicated by the ad-

able 1. Continued

ITEM (ORIGINAL CPCI SCALE)I II

(EXE) (COP) (G

5. (Asking for Assistance)2. (Asking for Assistance)2. (Asking for Assistance)7. (Resting)8. (Resting). (Resting). (Resting)8. (Resting)

bbreviations: EXE, Exercise/Stretch; COP, Coping Self-Statements; GUA, Guarsking for Assistance; RES, Resting.

usted coefficient of multiple determination (adjusted

2), which ranged from 10.50% for pain severity to0.70% for disability (P � .001 in all models). Further-ore, the addition of the CPCI factors in the last step in

ach of the equations yielded a statistically significantncrease in R2, ranging from 9.80% for depression to3.20% for disability (P � .001 in all models). In eachegression model, the CPCI factors accounted for uniqueariance in the criterion variable, above and beyondhat is already accounted for by the control variables.Table 6 presents the standardized, partial regression

oefficients (�) for all predictor variables in the final re-ression equations. The � indicates the nature of theelationship between the predictor and criterion vari-bles, specifically the association between each predictorariable and the criterion when all other predictor vari-bles are controlled.With regard to depression (CES-D), the Resting factoras found to be a significant, positive predictor, whereasask Persistence was found to be a negative predictor.mong the control variables, pain severity was a positiveredictor, and age was a negative predictor, when otherariables are taken into account. For pain interference,he Guarding and Resting factors were significant, posi-ive predictors, and Task Persistence was a negative pre-ictor. Among the control variables, pain severity wasound to be a significant, positive predictor of pain in-erference. Furthermore, the size of the � (.37) and theignificance level (P � .001) suggest that pain severityas the most powerful predictor of pain interference.ith respect to general activity level, Task Persistence

nd Seeking Social Support were significant, positiveredictors, whereas Guarding was a negative predictor.ain severity was the only control variable to signifi-antly predict general activity level. Regarding disability,uarding and Asking for Assistance were significant,ositive predictors, whereas Task Persistence was foundo be a negative predictor. Among the control variables,ain severity was a positive predictor. Finally, when paineverity served as the dependent variable, Resting wasound to be a positive predictor. No demographic vari-bles were significant predictors of pain severity.

FACTORS

IV V VI VII VIII(SEE) (TAS) (REL) (ASK) (RES)

.75

.74

.71.70.68.66.61.59

EE, Seeking Social Support; TAS, Task Persistence; REL, Relaxation; ASK,

IIIUA)

ding; S

In summary, the results of the regression analyses indi-

Page 8: Further validation of the chronic pain coping inventory

cafimsCeShmtcittam

D

fo

tCftanmw

Cbnpssesdwiio

TR

C

CMMRM

T

ECGSTRAR

AA

*

36 Validation of the Chronic Pain Coping Inventory

ate that as a group, the empirically derived CPCI factorsccounted for unique variance in depression, pain inter-erence, general activity level, disability, and pain sever-ty above and beyond what is accounted for by the de-

ographic variables (age, sex, education) and paineverity (when inserted as a predictor variable). SeveralPCI factors were found to be significant predictors ofach of the criterion variables. The Coping Self-tatements, Relaxation, and Exercise/Stretch factors,owever, were not significant predictors in any of theodels. Task Persistence was the most consistent predic-

or among the CPCI factors. Pain severity was a signifi-ant predictor in all models, which lends support to itsnclusion as a control variable when assessing the predic-ive ability of the coping scales. Finally, with the excep-ion of age predicting depression, the demographic vari-bles were not significant predictors in any of theodels.

iscussionThe purpose of the present study was to confirm the

actor structure and validity of the CPCI in a population

able 2. Means, SDs, and Ranges for EmpiricallMDQ, and MPIPS

SCALE MEAN

PCI factorsExercise/Stretch 2.20Coping Self-Statements 3.33Guarding 4.43Seeking Social Support 2.35Task Persistence 3.13Relaxation 2.20Asking for Assistance 2.82Resting 4.92

ES-D 28.29PIIF 5.04PIGL 1.64

MDQ 16.32PIPS 5.10

able 3. Intercorrelations of the Empirically DeFACTOR EXE COP GUA

xercise/Stretch — .41* .21*oping Self-Statements — .24*uarding —eeking Social Supportask Persistenceelaxationsking for Assistanceesting

bbreviations: EXE, Exercise/Stretch; COP, Coping Self-Statements; GUA, Guarsking for Assistance; RES, Resting.

P � .001.

P � .01.

f patients with chronic pain in a Veterans Administra- s

ion Medical Center. The factor structure of the original 8PCI subscales was generally confirmed. A confirmatoryactor analysis strongly supported the 8-factor model. Ofhe original 64 items in the 8 subscales, only 2 items (#40nd #64) did not fit in the original Guarding subscale andeeded to be moved to the Resting subscale. The factorodel contained only 4 loadings between 0.30 and 0.40,ith the remaining loadings above 0.40.The findings also support the predictive validity of thePCI scales and were consistent with a cognitive-ehavioral model of chronic pain that hypothesizes sig-ificant associations between coping and functioning inersons with chronic pain. As predicted, the CPCI factorcores were significantly related to a number of mea-ures of patient adjustment to chronic pain; with thexception of the Relaxation factor, all of the factors wereignificantly correlated with at least one of the depen-ent variables. Although the significant correlationsere generally modest and also variable across the cop-

ng factors, not all coping factors are necessarily equallymportant to the management of chronic pain. In fact,ne of the goals of the CPCI is to identify those coping

erived CPCI Factors, CES-D, MPIIF, MPIGL,

SD RANGE N

1.96 0-7 5081.96 0-7 5081.63 0-7 5101.68 0-7 5091.92 0-7 5091.68 0-7 5082.24 0-7 5081.84 0-7 510

12.14 0-60 526.89 1.33-6.00 564.98 0-5.19 561

4.91 1-24 561.86 1-6 564

d CPCI FactorsSEE TAS REL ASK RES

.28* .28* .54* .11 .21*

.47* .34* .58* .22* .31*

.22* �.02 .28* .38* .55*— .10 .39* .51* .28*

— .26* �.12† .01— .17* .32*

— .32*—

EE, Seeking Social Support; TAS, Task Persistence; REL, Relaxation; ASK,

y D

rive

ding; S

trategies that are most important to patient function-

Page 9: Further validation of the chronic pain coping inventory

iitcaEtA2amtJpaat

igcsaivDesCcpil

TM

ECGSTRAR

*

TG

C

C

C

C

C

*

37ORIGINAL REPORT/Tan et al

ng. The current validity data demonstrate that the CPCIs useful for this purpose, given the variability found inhe correlates of the CPCI factor scores. The zero-orderorrelations between the CPCI factors and self-reporteddjustment to chronic pain suggest that Task Persistence,xercise/Stretch, and Coping Self-Statements are adap-ive coping responses, and that Guarding, Resting, andsking for Assistance are maladaptive in this sample. Theremaining factors, Seeking Social Support and Relax-

tion, are less clear as to whether they are adaptive oraladaptive. Our results are generally consistent with

hose from previous studies of the CPCI. For example,ensen et al25 found that Guarding and Resting wereositively associated with depression and pain-relatedffective distress and negatively correlated with generalctivity. Task Persistence demonstrated significant nega-

able 4. Correlations of the Empirically DerivedPIPS

FACTOR CES-D MP

xercise/Stretch �.06 �.0oping Self-Statements �.13* .0uarding .20† .3eeking Social Support �.04 .1ask Persistence �.24† �.1elaxation .01 .0sking for Assistance .12* .2esting .22† .3

P �.01.

P � .001.

able 5. Hierarchical Multiple Regression Analyeneral Activity Level, Disability, and Pain Sev

STEP R2/ADJUST

riterion variable: Depression (CES-D)1. Pain severity (MPIPS) .064/.02. Age, sex, education .084/.03. CPCI subscales .182/.1

riterion variable: Pain interference (MPIIF)1. Pain severity (MPIPS) .238/.22. Age, sex, education .243/.23. CPCI subscales .357/.3

riterion variable: General activity level (MPIGL)1. Pain severity (MPIPS) .063/.02. Age, sex, education .096/.03. CPCI subscales .280/.2

riterion variable: Disability (RMDQ)1. Pain severity (MPIPS) .183/.12. Age, sex, education .190/.13. CPCI subscales .422/.4

riterion variable: Pain severity (MPIPS)1. Age, sex, education .018/.02. CPCI subscales .125/.1

P � .001.

P � .01.

ive associations with depression and affective distress. s

The empirically derived CPCI factors also demonstratedmpressive predictive ability. After controlling for demo-raphic variables and pain intensity, the CPCI factors ac-ounted for a significant amount of variance in patients’elf-reported depression, disability, pain interference,nd general activity level. The CPCI factors also made anndependent contribution to the prediction of pain se-erity. The factors that were most highly associated withisability, Pain Interference, and General Activity Levelxplained 41.0%, 35.0%, and 28.0% of the variance, re-pectively, in these variables. In a previous study of thePCI, the empirically derived factors also made signifi-ant contributions to the prediction of pain severity andain-related interference.17 However, unlike the find-

ngs of the previous research, this study did not find Re-axation to be associated with emotional distress. In-

CI Factors with CES-D, MPIIF, MPIGL, RMDQ,

MPIGL RMDQ MPIPS

.15* .00 �.05

.10 .08 �.01�.23† .50† .22†

.12* .21† .10

.36† �.21† �.15*

.04 .10 .04�.16† .37† .17†�.21† .33† .26†

Predicting Depression, Pain Interference,ty with CPCI Empirically Derived Factors2 F R2 CHANGE F CHANGE

33.09* .064 33.09*10.96* .020 3.428.77* .098 7.12*

151.36* .238 151.36*38.60* .005 1.0221.92* .114 10.52*

32.82* .063 32.82*12.74* .032 5.73†15.35* .184 15.15*

108.47* .183 108.47*28.19* .007 1.3528.77* .232 23.73*

3.02 .018 3.026.16* .107 7.22

CP

IIF

238†19†90†7†

seseri

ED R

627661

373741

628862

818307

1205

tead, our finding suggests that Relaxation appears to be

Page 10: Further validation of the chronic pain coping inventory

mcAnp

ptGdtAiatTGSscp

psiSatauahwipauii

crisstrtt

cfwppaacgocGpwtntacto

uaoaitI

TR

PASEC

*

38 Validation of the Chronic Pain Coping Inventory

ore adaptive, a finding more consistent with currentonception of the role of relaxation. We also found thatsking for Assistance and Seeking Social Support wereot associated with less favorable adjustment to chronicain.With regard to the individual CPCI factors in theresent study, Task Persistence was a significant predic-or of depression, interference, activity, and disability.uarding predicted all of the criterion variables exceptepression and pain severity, and Resting predicted all ofhe variables except disability and general activity level.sking for Assistance, Seeking Social Support, and Cop-

ng Self-Statements each predicted one criterion vari-ble. Overall, these results confirm the clinical impor-ance of teaching and reinforcing the coping strategy ofask Persistence and discouraging the regular use ofuarding and Resting. The results suggest that Copingelf-Statements is a positive strategy, but this copingtrategy was significantly associated with only one out-ome variable. Relaxation and Exercise/Stretch did notredict any outcome variable.The Seeking Social Support factor was a significantositive predictor of general activity level. These resultsuggest that Seeking Social Support is an adaptive cop-ng strategy. In the zero-order correlations, however,eeking Social Support was positively associated with dis-bility. Our findings demonstrate the complicated na-ure of the association between Seeking Social Supportnd functioning. The social support coping strategy isndoubtedly multidimensional and might include bothdaptive (eg, obtaining emotional support that en-ances functioning) and maladaptive (eg, obtaining helpith tasks that the patient can perform) components. It

s also possible that this strategy is adaptive, but thatatients use it more when they experience greater dis-bility or distress (eg, effective medications would besed more often by people who are ill, thereby produc-

ng a positive association between medication use andllness, even though the medications are effective).

able 6. Standardized Regression Coefficients (egression Equation

PREDICTOR VARIABLES CES-D MP

ain severity (MPIPS) .15* .3ge �.13* �.0ex �.01 .0ducation �.07 �.0PCI subscalesExercise/Stretch �.03 �.0Coping Self-Statement �.14 �.0Guarding .07 .2Seeking Social Support �.10 .0Task Persistence �.16* �.1Relaxation .09 .0Asking for Assistance .07 �.0Resting .18* .1

P � .01.

P � .001.

The Asking for Assistance coping strategy was signifi- v

antly associated with Seeking Social Support and mightepresent a negative component of social support seek-ng. In the regression models, the Asking for Assistancetrategy might have controlled for the negative aspect ofocial support and left the positive aspects of supporthat are associated with adaptation to chronic pain. Ouresults emphasize the clinical importance of helping pa-ients distinguish between the adaptive and maladap-ive aspects of seeking social support.Given that our study design was cross-sectional and

orrelational, it is not possible to draw causal conclusionsrom the findings. For example, we cannot determinehether the coping strategy of Task Persistence helps toromote activity and prevent depression, disability, andain-related interference. It is possible that patients whore active, not depressed, and functioning relatively wellre more apt to develop and use Task Persistence as aoping strategy. In addition, it should be noted that, as aroup, the 8 CPCI factors accounted for only 10% to 18%f the variance in each of the criterion variables, afterontrolling for pain severity and demographic variables.iven the difficulties associated with treating chronicain, however, identifying coping strategies associatedith positive outcomes is clinically meaningful, even if

he associations found are modest. Additional research iseeded to identify other factors that help to predict pa-ient outcomes. In particular, prospective longitudinalnd experimental studies are needed to assess patientoping strategies and other potential predictors overime and to determine their causal relation to multipleutcome variables.Additional limitations of the current study include the

nique characteristics of the patient sample and the reli-nce on self-report measures. Because our sample consistedf primarily male veterans, we were not able to evaluateny possible gender differences. The results of recent stud-

es suggest that men and women might differ somewhat inheir preferred strategies for coping with chronic pain.8,58

n addition, the patients in our sample reported more se-

or All Predictor Variables in the Final

MPIGL RMDQ MPIPS

�.13* .28† –�.07 .06 �.02

.08 .00 .05

.08 �.07 �.12

.11 �.05 �.08�.01 �.02 �.08�.15* .41† .12

.23† .07 .05

.28† �.13* �.11�.09 .02 .03�.11 .13* .05�.11 �.02 .19*

�) f

IIF

7†435

522†12*307†

ere pain, pain duration, depression, and disability than did

Page 11: Further validation of the chronic pain coping inventory

stcpdvvo

ibp

fcTdiwppiim

R

1wP

2c

3cp2

4tS

5wi1

6s

7fc

8cs

9ft

1Pc

1r3

1su

1pa1

1K

39ORIGINAL REPORT/Tan et al

amples of patients with chronic pain from other painreatment settings. Also, only 45% of the eligible patientshose to participate in the study by returning the com-leted questionnaires. Although the responders did notiffer significantly from nonresponders on demographicariables, they might have differed on other pain-relatedariables that were not assessed. Thus, the generalizabilityf the current findings is limited.Despite the limitations of the current study, the find-

ngs provide additional support for the cognitive-ehavioral model of chronic pain, for the potential im-

ortance of coping in adjustment to chronic pain, and a

inney TR: Sickle cell disease pain in children and adoles-

ct

1caP

1sp1

1V4

1PP

1p

2m

2RS

2a4

2td

2cmc

2Cn

2C

2Wtr

2T

or the reliability and validity of the CPCI for assessingoping strategies important to chronic pain adjustment.he findings of this study also provide additional evi-ence for the utility of the CPCI by demonstrating that it

s equally valid and reliable when used for pain sufferersith a more chronic, severe, and extended history ofain. Additional studies of the CPCI with other patientopulations are needed to determine the generalizabil-

ty of the current findings. Future research also shouldnvestigate the relation of coping strategies to outcome

easures such as family function, employment status,

nd utilization of health care services.

eferences

. Affleck G, Urrows S, Tennen H, Higgins P: Daily copingith pain from rheumatoid arthritis: patterns and correlatesain 51:221-229, 1992

. Ashby JS, Lenhart RS: Prayer as a coping strategy forhronic pain patients. Rehab Psychol 39:205-209, 1994

. Beckham JC, Keefe FJ, Caldwell DS, Roodman AA: Painoping strategies in rheumatoid arthritis: relationships toain, disability, depression, and daily hassles. Behav Ther2:113-124, 1991

. Bernstein IH, Jaremko ME, Hinkley BS: On the utility ofhe West-Haven-Yale Multidimensional Pain Inventory.pine 20:956-963, 1995

. Boothby JL, Thorn BE, Stroud MW, Jensen MP: Copingith pain, in Getchel RJ, Turk DC (eds): Psychosocial Factors

n Pain: Critical Perspectives. New York, NY, Guildford Press,999, pp 345-359

. Brown GK: A causal analysis of chronic pain and depres-ion. J Abnorm Psychol 99:127-137, 1990

. Brown GK, Nicassio PM: Development of a questionnaireor the assessment of active and passive coping strategies inhronic pain patients. Pain 31:53-64, 1987

. DeGood DE, Tait RC: Assessment of pain beliefs and painoping, in Turk DC, Melzack R (eds): Handbook of Pain As-essment. New York, NY, Guilford, 2001, pp 320-345

. Deyo RA, Centor RM: Assessing the responsiveness ofunctional scales to clinical change: an analogy to diagnosticest performance. J Chron Dis 39:897-906, 1986

0. Dozois DJA, Dobson KS, Wong M, Hughes D, Long A:redictive utility of the CSQ in low back pain: individual vsomposite measures. Pain 66:171-180, 1996

1. Flor H, Behle DJ, Birbaumer N: Assessment of pain-elated cognitions in chronic pain patients. Behav Res Ther1:63-73, 1993

2. Geisser ME, Robinson ME, Keefe FJ, Weiner ML: Cata-trophizing, depression and the sensory, affective and eval-ative aspects of chronic pain. Pain 59:79-83, 1994

3. Geisser ME, Roth RS: Knowledge of and agreement withain diagnosis: Relation to pain beliefs, pain severity, dis-bility, and psychological distress. J Occup Rehabil 8: 73-88,998

4. Gil KM, Thompson RJ, Keith BR, Tota-Faucette M, Noll S,

ents: change in pain frequency and coping strategies overime. J Pediatr Psychol 18:621-637, 1993

5. Gil KM, Williams DA, Thompson RJJ, Kinney TR: Sickleell disease in children and adolescents: the relation of childnd parent pain coping strategies to adjustment. J Pediatrsychol 16:643-663, 1991

6. Goldberg GM, Kerns RD, Rosenberg R: Pain-relevantupport as a buffer from depression among chronic painatients low in instrumental activity. Clin J Pain 9:34-40,993

7. Hadjistavropoulos HD, MacLeod FK, Asmundson GJG:alidation of the Chronic Pain Coping Inventory. Pain 80:71-481,1999

8. Haythornwaite JA, Menefee LA, Heinberg LJ, Clark MR:ain coping strategies predict perceived control over pain.ain 77:33-39, 1998

9. Hill A: The use of pain coping strategies by patients withhantom limb pain. Pain 55:347-353, 1993

0. Hill A, Niven CA, Knussen C: The role of coping in adjust-ent to phantom limb pain. Pain 62:79-86, 1995

1. Husaini BA, Neff JA, Harrington JB, Hughes MD, StoneH. Depression in rural communities: validating the CES-Dcale. J Comm Psychol 8:20-27, 1980

2. Jensen MP, Karoly P. Control beliefs, coping efforts, anddjustment to chronic pain. J Consult Clin Psychol 59:431-38, 1991

3. Jensen MP, Strom SE, Turner JA, Romano JM: Validity ofhe Sickness Impact Profile Roland scale as a measure ofysfunction in chronic pain patients. Pain 50:157-162, 1992

4. Jensen MP, Turner JA, Romano JM: Changes in beliefs,atastrophizing and coping are associated with improve-ent in multidisciplinary pain treatment. J Consult Clin Psy-

hol 69:655-662, 2001

5. Jensen MP, Turner JA, Romano JM, Strom SE: Thehronic Pain Coping Inventory: development and prelimi-ary validation. Pain 60:203-216, 1995

6. Joreskog KG, Sorbom D: Lisrel 8 user’s reference guide.hicago, IL, Scientific Software International, 1993

7. Keefe FJ, Caldwell DS, Martinez S, Nunley J, Beckham J,illiams DA: Analyzing pain in rheumatoid arthritis pa-

ients: pain coping strategies in patients who have had kneeeplacement surgery. Pain 46:153-160, 1991

8. Keefe FJ, Affleck G, Lefebvre JC, Starr K, Caldwell DS,

ennen H: Pain coping strategies and coping efficacy in
Page 12: Further validation of the chronic pain coping inventory

r1

2se2

3t

3Rc

3a1

3w7

3N

3ra

3pH

3ta

3ap6

3Sp

4as

4AMD1

4Gl1

4bs

4pa

4ca

4Kj

4Fp

4c1

4sP

5S7

5Mm

5MP

5PJ

5r1

5sdc

5DtP

5pa

5a1

5l

6f2

6Kcn

40 Validation of the Chronic Pain Coping Inventory

heumatoid arthritis: a daily process analysis. Pain 69:35-42,997

9. Kerns RD, Rosenberg R, Otis JD: Self-appraised problemolving and pain-relevant social support as predictors of thexperience of chronic pain. Ann Behav Med 24:100-105,002

0. Kerns RD, Turk DC, Rudy TE: The West Haven-Yale Mul-idimensional Pain Inventory. Pain 23:345-356, 1985

1. Kleinke CL: How chronic pain patients cope with pain:elation to treatment outcome in a multidisciplinary painlinic. Cogn Ther Res 16:669-685, 1992

2. Lanier DC, Stockton P: Clinical predictors of outcomes ofcute episodes of low back pain. J Fam Pract 27:483-489,988

3. Large RG, Strong J: The personal constructs of copingith chronic low back pain: is coping a necessary evil? Pain3:245-252, 1997

4. Lazarus RS, Folkman S: Stress, appraisal, and coping.ew York, NY, Springer, 1984, p 445

5. Lester N, Lefebvre JC, Keefe FJ: Pain in young adults, III:elationships of three pain coping measures to pain andctivity interference. Clin J Pain 12:291-300, 1996

6. Lin C, Ward SE: Perceived self-efficacy and outcome ex-ectancies in coping with chronic low back pain. Res Nursealth 19:299-310, 1996

7. Mahoney ND, Devine JE, Angres D: Multidisciplinaryreatment of benign chronic pain syndrome in substancebusing patients. Curr Rev Pain 3:321-331, 1999

8. Martin MY, Bradley LA, Alexander RW, Alarcon GS, Tri-na-Alexander M, Aaron LA, Alberts KR: Coping strategiesredict disability in patients with primary fibromyalgia. Pain8:45-53, 1996

9. Monga TN, Tan G, Ostermann HJ, Monga U, Grabois M:exuality and sexual adjustment of patients with chronicain. Disabil Sexual 20:317-329, 1998

0. Potter RG, Jones JM: The evolution of chronic painmong patients with musculoskeletal problems: a pilottudy in primary care. Br J Gen Pract 42:462-464, 1992

1. Radloff LS, Locke BZ. The Community Mental Healthssessment Survey and the CES-D Scale, in Weissman MM,yers JK, Ross CE (eds): Community Surveys of Psychiatricisorders. New Brunswick, NJ, Rutgers University Press,986, pp 177-189

2. Robinson ME, Riley JL, Myers CD, Sadler IJ, Kvaal SA,eiser ME, Keefe FJ: The Coping Strategies Questionnaire: a

arge sample, item level factor analysis. Clin J Pain 13:43-49,997

3. Roland M, Morris R: A study of the natural history ofack pain, I: development of a reliable and sensitive mea-ure of disability in low back pain. Spine 8:141-144, 1983

4. Romano JM, Turner JA, Jensen MP: Chronic painatient-spouse behavioral interactions predict patient dis-

bility. Pain 63:353-360, 1995 c

5. Rosenstiel AK, Keefe FJ: The use of coping strategies inhronic low back pain patients: relationship to patient char-cteristics and current adjustment. Pain 17:33-44, 1993

6. Schanberg LE, Lefebvre JC, Keefe FJ, Kredich DW, GilM: Pain coping and the pain experience in children with

uvenile chronic arthritis. Pain 73:181-189, 1997

7. Schmitz U, Saile H, Nilges P: Coping with chronic pain:lexible goal adjustment as an interactive buffer againstain-related distress. Pain 67:41-51, 1996

8. Snow-Turek AL, Norris MP, Tan G: Active and passiveoping strategies in chronic pain patients. Pain 64:455-462,996

9. Stevens JJ: Interaction of coping style and cognitivetrategies in the management of acute pain. Imag Cognerson 11:225-232, 1992

0. Sullivan MJL, Bishop SR, Pivik J: The Pain Catastrophizingcale: development and validation. Psychol Assess:524-532, 1995

1. Tan G, Jensen MP, Robinson-Whelen S, Thornby JI,onga TN: Coping with chronic pain: a comparison of twoeasures. Pain 90:127-133, 2001

2. Tan G, Jensen MP, Robinson-Whelen S, Thornby J I,onga TN: Measuring control appraisals in chronic pain. J

ain 3:385-393, 2002

3. Tota-Faucette ME, Gil KM, Williams DA, Keefe FJ, Goli V:redictors of response to pain management treatment. ClinPain 9:115-123, 1993

4. Turk DC, Rudy TE: The robustness of an empirically de-ived taxonomy of chronic pain patients. Pain 68:423-430,990

5. Turner JA, Jensen MP, Warms CA, Cardenas DD: Cata-trophizing is associated with pain intensity, psychologicalistress, and pain-related disability among individuals withhronic pain after spinal cord injury. Pain 98:127-134, 2002

6. Turner JA, Whitney C, Dworkin SF, Massoth D, Wilson L:o changes in patient beliefs and coping strategies predict

emporomandibular disorder treatment outcomes? Clin Jain 11:177-188, 1995

7. Ulmer FJ: An exploratory study of pain, coping, and de-ressed mood following burn injury. J Pain Symptom Man-ge 13:148-157, 1997

8. Unruh AM, Ritchie J, Merskey H: Does gender affectppraisal of pain and pain coping strategies? Clin J Pain5:31-40, 1999

9. Wilkie DJ, Keefe JF: Coping strategies of patients withung cancer-related pain. Clin J Pain 7: 292-299, 1991

0. Williams RC: Toward a set of reliable and valid measuresor chronic pain assessment and outcome research. Pain 35:39-251,1988

1. Zautra AJ, Burleson MH, Smith CA, Blalock SJ, WallstonA, DeVellis RF, DeVellis BM, Smith TW: Arthritis and per-eptions of quality of life: an examination of positive andegative affect in rheumatoid arthritis patients. Health Psy-

hol 14:399-408, 1995