- Email: [email protected]
Readiness to adopt a self-management approach to pain: evaluation of the pain stages of change model in a non-pain-clinic sample
Pain 104 (2003) 283–290 www.elsevier.com/locate/pain
Readiness to adopt a self-management approach to pain: evaluation of the pain stages of change model in a non-pain-clinic sample Suzanne Habiba,*, Shirley A. Morrisseyb, Edward Helmesa a
School of Psychology, James Cook University, P.O. Box 6811, Cairns, Queensland 4870, Australia b Psychology Department, Northern Territory University, Darwin, Australia Received 28 March 2002; accepted 13 January 2003
Abstract The Transtheoretical model of stages of behaviour change has stimulated research interest in relation to chronic pain, yet studies using the Pain Stages of Change Questionnaire (PSOCQ; Pain (72) 1997 227) have reported inconsistent findings and have generally utilized painclinic samples. The aims of the current study were to assess the general validity of the PSOCQ with a non-pain-clinic sample of patients with chronic pain, and to examine the utility of the stages of change model as applied to this population. The study employed multi-stage, clustersampling methodology, with 90 participants recruited from 19 medical and allied health clinics and practices. The findings demonstrated a number of limitations of the PSOCQ in terms of its ability to classify individuals into specific stages of change. The stages of change model requires adaptation in order to be useful for treatment planning in a non-pain-clinic sample of patients with chronic pain. q 2003 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. Keywords: Stages of change model; Pain Stages of Change Questionnaire; Self-management approach
1. Introduction Psychologically based programmes for chronic pain have been repeatedly shown to be effective (e.g. Keefe et al., 1992; Compas et al., 1998; Morley et al., 1999). A significant proportion of individuals, however, either fail to engage in, or fail to benefit from this type of treatment (Turk and Rudy, 1990, 1991; Richmond and Carmody, 1999). Although there have been a number of prior research attempts to modify treatment and increase rates of engagement, adherence and maintenance of treatment gains (e.g. King and Snow, 1989; Carosella et al., 1994), these findings have been inconsistent and unreliable, therefore contributing little to improving interventions. Recent attempts have been made to examine as to why there exists a subset of individuals who do not benefit and coworkers (Kerns et al., 1997, 1999; Kerns and Rosenberg, 2000) and Jensen et al. (1996, 2000) considered the relevance and application of the Transtheoretical model of behaviour change (Prochaska and DiClemente, 1982; Prochaska et al., 1994), in conjunction with a cognitive – * Corresponding author. Tel.: þ 7-4991-4443; fax: þ7-4091-4443. E-mail address: [email protected] (S. Habib).
behavioural perspective on chronic pain (Turk et al., 1983). This culminated in the development of the pain stages of change model and the Pain Stages of Change Questionnaire (PSOCQ; Kerns et al., 1997). The pain stages of change model proposes that individuals vary in their readiness to adopt a self-management approach to pain and that individuals may be categorised, on the basis of their beliefs about their pain, into one of the four stages of readiness to change (SRC). These stages are: Pre-contemplation (not considering any change in behaviour), Contemplation (serious consideration of change sometime in the future but no clear plan of intent), Action (concrete activities that will lead to the desired change) and Maintenance (active efforts to sustain the changes made). The model also proposes that an individual’s current stage of change may determine the most effective therapeutic approach. For example, individuals in the Pre-contemplation or Contemplation stage may benefit from strategies, such as cognitive re-structuring and reconceptualization of the pain as being manageable, whereas individuals in the ‘higher’ stages of change would be more likely to benefit from behavioural techniques, such as pacing activities, relaxation and exercise, which ensure readiness for active involvement and responsibility of the
0304-3959/03/$30.00 q 2003 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. doi:10.1016/S0304-3959(03)00016-2
284
S. Habib et al. / Pain 104 (2003) 283–290
individual in the rehabilitation process. Kerns and his colleagues assert that successful demonstration of a confirmatory four-factor analysis is consistent with the stages of change model, and the reliability, criterion-related validity, internal consistency and stability of each of the four scales, provide support for the relevance of the stages of change model to management of pain. Jensen et al. (2000) sought to further explore the psychometric properties of the PSOCQ by examining the generalizability of the findings presented by Kerns et al. (1997). Jensen et al. studied two new sub-samples of patients with chronic pain, a pain-clinic sample and a fibromyalgia sample. Internal consistency co-efficient scores were greater than 0.70 on all the PSOCQ sub-scales for both samples and are consistent with those reported by Kerns et al. (1997). PSOCQ scale score inter-relations were also similar to the associations found by Kerns et al. (1997) in the original scale-development sample. All of the criterion measures were associated with the PSOCQ scales in the directions hypothesized. An attempt to examine the differences among the stages of change groups for each of the PSOCQ scales was only partially successful as none of the group means differed significantly on the Pre-contemplation scale. Participants classified as being in either the Action or Maintenance stage also did not differ significantly on any of the mean PSOCQ scale scores. In both samples, participants classified as being in the Contemplation, Action or Maintenance stage scored higher than three, on average (indicating more agreement than disagreement), on the Contemplation, Action and Maintenance scale items, thus apparently being simultaneously in all three stages concerning readiness to adopt a self-management approach to pain. Jensen et al. (2000) argued that the PSOCQ does not discriminate sufficiently between the stages of change and, therefore, may not be useful in its current form as a tool used to classify painclinic patients into distinct stages of readiness to adopt a self-management approach. Kerns and Rosenberg (2000) conducted a study to assess the predictive validity of the PSOCQ in relation to participation and outcome in treatment programmes. They hypothesized that high Pre-contemplation scores would be predictive of higher rates of dropout and poorer outcomes relative to low Pre-contemplation scores. The results demonstrated significant differences between completers and non-completers, with non-completers having higher scores on the Pre-contemplation scale, and completers generally having low scores on Pre-contemplation and higher scores on the other scales. Whilst the study supported the hypothesis that the PSOCQ (Kerns et al., 1997) could predict engagement in treatment, Kerns and Rosenberg (2000) found no support for the hypothesis that PSOCQ scores could predict outcome subsequent to treatment, as the participants in all the stages of change were shown to be equally likely to achieve improvements in pain management, emotional well-being and functioning. Kerns and
Rosenberg concluded that replication of the study and further refinements of the pain stages of change model are necessary before the PSOCQ can be confidently utilized for clinical decision-making. In a subsequent study on the predictive validity of the PSOCQ (Kerns et al., 1997), Biller et al. (2000) supported the findings of Kerns and Rosenberg (2000) that high Precontemplation scores and low Contemplation scores were the strongest predictor of completion of a pain-management programme. Although the findings of Biller et al. generally support the predictive validity of the PSOCQ, the authors caution against its use as a screening instrument in its current form. According to Biller et al. (2000), whilst the moderate predictive validity of the mean scale score may be helpful for the practitioners in identifying appropriate intervention targeted to stages of change, it is not sufficient to be used as an inclusion or exclusion criterion for treatment. Since the introduction of the pain stages of change model (Kerns et al., 1997), to date, two published studies have sought to develop or modify other stage-based psychometric measures for specific sub-groups of pain patients. In the first study, Keefe et al. (2000) used a cluster analysis to identify five distinct sub-groups of patients with arthritis pain, based on their responses on a modified version of the University of Rhode Island Change Questionnaire (URICA; McConnaughy et al., 1983). Keefe et al. (2000) concluded that as the sub-groups loosely corresponded with the stages proposed by Prochaska and DiClemente (1998), the Transtheoretical model of stages of change may be applicable to arthritis patients. The second study, subsequent to the development of the pain stages of change model (Kerns et al., 1997), was conducted by Dijkstra et al. (2001). The study sought to explore the psychometric properties of a Dutch version of a questionnaire, assessing readiness to change in a sample of 321 Dutch fibromyalgia patients. The SRC measure was adapted from an unpublished English readiness to change questionnaire (Nielson and Vlaeyen, 1996 in Dijkstra et al., 2001), which in turn was based on McConnaughy et al.’s (1983) measure of readiness to change. The items were adapted to reflect a readiness to adopt a self-management approach to pain. Concerns voiced by Dijkstra et al. (2001) in relation to the application of the stages model to chronic pain are similar to those of Jensen et al. (2000) in that there is lack of clarity as to which specific behaviour the change refers to, and that the theoretical basis of the construct of readiness to adopt a self-management approach to pain needs to be further developed. It can be tentatively concluded from the previous research that the stages of change model has the potential to assist in the assessment and treatment planning for a number of sub-groups of pain patients. Inconsistent research findings indicate, however, that continued investigation and revision of the model is warranted in order to be able to work confidently within this theoretical framework for
S. Habib et al. / Pain 104 (2003) 283–290
assessment and treatment of pain. Clearly, the PSOCQ lacks adequate sensitivity to be used in clinical decision making with regards to offering self-management treatment. However, if we are to develop prescriptive ‘stage-specific’ treatment programmes as recommended by Kerns et al. (1999) and Kerns and Rosenberg (2000), we must first refine our means of assessing stage of readiness to adopt a selfmanagement approach in order to assign patients to appropriate treatment programmes. Other than the study by Keefe et al. (2000), all of the published studies of stages of change and chronic pain have been conducted with pain-clinic samples; therefore, it is not clear that to what extent the findings would generalize to the general population of individuals with chronic pain. A number of earlier studies have demonstrated that pain-clinic patients differ significantly from pain patients treated in the community (Turk, 1990). In studies comparing these two populations, pain-clinic patients were found to complain of more constant pain, suffer higher levels of functional impairment and report greater emotional and psychosocial distress than the pain patients being treated in the community (e.g. Chapman et al., 1997; Crook et al., 1984, 1989). Whilst patients in pain clinics are reported to suffer from higher levels of distress and disability than those treated in the community, the low number of Pre-contemplators in previous studies using pain-clinic samples suggests that by the time pain-clinic patients attend an intake assessment, their motivation may have been influenced by referring practitioners and/or the knowledge that pain-clinic treatment focuses mainly on self-management techniques. It may also be the case that patients who are not interested in this type of approach are screened out of pain-clinic programmes or simply fail to attend the initial assessment. It is important, therefore, to establish whether measures and interventions developed on pain-clinic samples are appropriate for use on the pain sufferers in the general community where the numbers of individuals who are not contemplating a self-management approach may be significantly greater. The aims of the present research were, firstly, to assess the general validity of the PSOCQ (Kerns et al., 1997) with a non-pain-clinic sample in order to determine the generalizability of the measure, and, secondly, to examine the utility of the stages of change model as applied to this population in terms of its ability to assist in treatment planning. The study employed a survey questionnaire design with participants recruited from medical and allied health clinics and practices (Habib, 2002).
2. Method 2.1. Sampling methodology Participants for the survey were recruited by means of a
285
multi-stage cluster-sampling methodology. The clusters or self-formed groups comprised physiotherapists, chiropractors, medical practices, bulk-billing medical practices, CRS Australia rehabilitation units and public hospital outpatient clinics. Within each group, every second practice or clinic was selected from the telephone directory. Within each selected practice that agreed to participate, every second patient that arrived at the clinic and who met participation criteria was asked to complete the survey questionnaire. To generalize the research findings to a wide range of treatment-seeking patients with pain, fee-paying and nonfee paying (bulk billing) practices were sampled, participants were recruited both during and after working hours and both rural and city practices were sampled. 2.2. Participants The sample that agreed to participate in the survey and met eligibility criteria comprised 90 participants drawn from 19 practices (see Table 1). Inclusion criteria were that the patient was over the age of 18 and reporting pain for 3 months or more; exclusion criteria were inadequate literacy to read and comprehend the research questionnaire, visibly affected by drugs or alcohol or actively psychotic. The mean age of the participants was 43 years (standard deviation ðSDÞ ¼ 13:83, range ¼ 19 – 77). Fiftytwo participants (57.8%) were male. The majority of the participants (60%) were married, 22% were single, 8% were divorced and 8% were in de-facto relationships. The majority of the participants (42%) had 3 years of secondary education. Seventeen percent of the participants had completed a course at college, and 8% had university degree. Forty-one percent of the participants were unemployed, 28% percent were blue-collar workers, 9% were self-employed, 4% of the participants were engaged in managerial positions, 11% held professional positions, 6% were retired and 6% of the participants were students. Eightfive percent of the sample was of Australian or New Zealander descent, the remaining 15% was identified as British or European. Pain site varied, though 28% of the sample presented with low back pain. Cause of pain Table 1 Proportions of participants from each practice type Practice type
Fee/non-fee Participants Percentage of sample
Physiotherapist Chiropractor Medical centre Bulk-billing centre Orthopedic clinic Hand clinic Fracture clinic Public physiotherapist Rehabilitation unit
Fee Fee Fee Non-fee Non-fee Non-fee Non-fee Non-fee Non-fee
Total
10 12 7 8 17 1 9 6 20
11.1 13.3 7.8 8.9 18.9 1.1 10.0 6.7 22.2
90
100.00
286
S. Habib et al. / Pain 104 (2003) 283–290
included disc pathology (28%), neuropathic pain (18%), osteoarthritis (17%), rheumatoid arthritis (11%), fibromyalgia (6%), ligament damage (9%) and carpal tunnel syndrome (2%). The remaining 8% of the participants had no clear diagnosis. Pain duration ranged from 4 months to 30 years (M ¼ 6:6 years, SD ¼ 13:85). Average pain severity was measured on a numerical rating scale where 0 referred to no pain and 100, the worst pain ever experienced. The mean average pain rating was 53 (SD ¼ 21:81, range ¼ 10 – 100). Seven percent of the participants reported that they were currently receiving compensation, 9% were in the process of applying for compensation or disability payment and 12% were currently litigating in relation to their pain. The participants did not receive any reward or incentive for completing the questionnaire. Non-respondents: Within the practices sampled, nine individuals declined to participate in the questionnaire survey. The majority of the individual refusals came from fee-paying practices and were male. The mean age was 54 years (SD ¼ 12:5, range ¼ 28 – 76). Non-respondents had an average of 2 years of secondary education. Three of the refusing individuals were retired, two were unemployed, one was self-employed, two were blue-collar workers and one was engaged in home duties. Five of the nonrespondents reported a pain duration of greater than 5 years. The remaining four non-respondents reported a pain duration of more than 2 years but less than 5 years. The reasons for non-participation were pain too great, not enough time and not interested. Non-respondents tended to be older and had lower levels of education than respondents; however, the differences between respondents and nonrespondents were not statistically significant.
the practice or clinic, prior to consultations with medical practitioners, physiotherapists and chiropractors.
2.3. Questionnaire
3.3. Differences between individuals in each stage of change
The questionnaire was designed to elicit demographic and medical information, pain intensity, duration and coping strategies and compensation and litigation status. Stage of readiness to adopt a self-management approach to chronic pain was assessed using the PSOCQ (Kerns et al., 1997). Two additional measures of affect were used; however, these findings are reported elsewhere (Habib, 2002). The final page on the questionnaire provided brief information regarding pain-management workshops based on a self-management approach that might be offered as a later part of the research. Participants were informed that if they wished to provide their name and contact phone number they would be contacted in due course with further information.
One-way analyses of variance were performed to examine predicted differences among the classified stage of change groups in the four PSOCQ scales. To control for alpha inflation associated with multiple comparisons, a family-wise Bonferroni alpha was set at 0.0125 (0.05/4) and the data are presented in Table 5. The results demonstrate significant differences among the stages of change for each of the PSOCQ scales, with the results identifying four specific stages. However, participants classified as being in either the Contemplation, Action
2.4. Procedure Every second patient who attended each clinic and met participation criteria was requested to complete a questionnaire. The questionnaires were given to individuals in
3. Results 3.1. Internal consistency scores Using the scoring method outlined by Kerns et al. (1997), each study participant was classified into one of the four stages based on their highest PSOCQ scale score as illustrated in Table 2. The scale score means, SD and the internal consistency co-efficients of the PSOCQ scale scores are presented in Table 3. Internal consistency co-efficients were greater than 0.75 for each of the four scales and are consistent with those reported by Kerns et al. (1997) in the original scaledevelopment sample and Jensen et al. (2000) in a pain-clinic and fibromyalgia sample. 3.2. Associations among the PSOCQ scales Inter-relations of the scale scores for the current sample are presented in Table 4. As illustrated in Table 4, there was no significant association between Pre-contemplation and Contemplation. Pre-contemplation was significantly, negatively associated with Action and Maintenance. Significant positive associations were demonstrated between Contemplation and Action; however, there was no significant association between Contemplation and Maintenance. Significant positive associations were demonstrated between Action and Maintenance.
Table 2 Proportions of participants in each stage of change Stage
n
Percentage
Total sample (N ¼ 90) Pre-contemplation Contemplation Action Maintenance
25 16 18 29
27.8 18.9 20.0 33.3
S. Habib et al. / Pain 104 (2003) 283–290 Table 3 Means, SD and internal consistency scores for PSOCQ scales PSOCQ scale
Mean
SD
Cronbach’s alpha
Total sample (N ¼ 90) Pre-contemplation Contemplation Action Maintenance
3.06 3.23 3.25 3.44
0.80 0.62 0.87 0.81
0.76 0.79 0.82 0.86
or Maintenance stage also had mean scale scores of three or over for Contemplation, Action and Maintenance, indicating more agreement than disagreement with items on each of the other scales. 3.4. Associations between stages of change and demographic variables Pearson correlations were calculated in order to explore relationships between PSOCQ scale scores and demographic variables. In order to correct for the large number of planned comparisons, results were considered significant if they were at, or below, the significance level of 0.01. The analyses revealed no significant relationships for age, ethnicity, marital status or average pain rating and any of the PSOCQ scale scores. Pain duration was negatively associated with Pre-contemplation (r ¼ 20:270, P , 0:01) and positively associated with Action (r ¼ 0:258, P , 0:01) and Maintenance (r ¼ 0:233, P , 0:01). The t-tests revealed no significant differences for feepaying versus non-fee-paying, married versus non-married or compensating versus non-compensating. The male participants had significantly higher mean Pre-contemplation scores than the female participants (male Precontemplation score, M ¼ 3:23, female Pre-contemplation score, M ¼ 2:82, P ¼ 0:02), and this fact is consistent with the findings of Kerns et al. (1997) in the original scaledevelopment sample. 3.5. Associations between stages of change and current coping strategies The second aim of the study was to determine the utility of the pain stages of change model (Kerns et al., 1997) in relation to its usefulness for treatment planning in a nonTable 4 PSOCQ inter-scale correlation co-efficients PSOCQ scale
Contemplation
Non-pain-clinic sample (N ¼ 90) Pre-contemplation 0.09 Contemplation Action Maintenance * P , 0:05, * * P , 0:01, * * * P , 0:001:
Action
Maintenance
0.45*** 0.32**
20.46*** 20.07 0.68***
287
pain-clinic sample. Univariate statistics were used to examine the characteristics of the participants in each stage of change, particularly in relation to coping strategies, including medication use and willingness to participate in a self-management pain programme. A number of inconsistencies were discovered in relation to the predictions based on the model and reported behaviour. 3.5.1. Pre-contemplation According to the pain stages of change model (Kerns et al., 1997), individuals in the Pre-contemplation stage believe that their pain is purely a medical problem, thus, would not use self-management strategies and would not be interested in this type of approach. Contrary to the model, 44% of the participants in the Pre-contemplation stage indicated an interest in participating in a pain programme based on self-management strategies with no medical treatment involved. As it was not the intention of the researchers to examine the predictive ability of the PSOCQ, actual participation in treatment was not examined in this study; however, the large number of Pre-contemplators indicating an interest in this type of approach is thought to be inconsistent with the definition of Pre-contemplation provided by Kerns et al. (1997) in the scale development study. Although it could be argued that subsequent studies (e.g. Kerns and Rosenberg, 2000) have demonstrated that Pre-contemplators can be successfully engaged in treatment with a self-management focus (though they are less likely to complete treatment than Contemplators), this shows that either stages are less stable than originally thought to be, or the wording on the PSOCQ is not specific enough to accurately classify individuals into appropriate stages of readiness to adopt a self-management approach to pain. Further speculation about other possible reasons for Precontemplators expressing an interest in self-management programmes raised three possibilities. Firstly, it could be that completing the PSOCQ may have had a motivational effect, thus moving participant closer to considering a selfmanagement approach; secondly, individuals in these settings (i.e. where no pain clinic is available) may be willing to try any type of treatment; and thirdly, participants may have appreciated the attention from researchers and felt that participating in research validated their pain to others. 3.5.2. Contemplation The Contemplation stage is characterized by a consideration of the potential value of adopting a self-management approach, but a reluctance to relinquish pursuit of a medical solution (Kerns et al., 1997). Fifty-nine percent of the participants in the Contemplation stage indicated that they would like to participate in a pain-treatment programme based on self-management strategies with no medical treatment involved. 3.5.3. Action The Action stage is characterized by acceptance of a self-
288
S. Habib et al. / Pain 104 (2003) 283–290
Table 5 Mean scale scores for classified stage of change groups Classified stage Pre-contemplation (n ¼ 25) PSOCQ scale score mean (SD) Pre-contemplation 4.00 (0.40) Contemplation 3.07 (0.58) Action 2.52 (0.70) Maintenance 2.87 (0.64)
Contemplation (n ¼ 16)
Action (n ¼ 18)
Maintenance (n ¼ 29)
F-value
2.92 (0.74) 3.78 (0.77) 3.10 (1.0) 3.00 (0.76)
2.61 (0.47) 3.33 (0.41) 3.85 (0.51) 3.72 (0.57)
2.65 (0.56) 3.00 (0.50) 3.52 (0.66) 4.02 (0.47)
36.06* 7.92* 14.22* 21.30*
* P , 0:001:
management approach and active engagement in attempts to improve self-management skills (Kerns et al., 1997). In this study, 33% of the participants classified by the PSOCQ as being in the Action stage were using daily analgesics as the primary pain-coping strategy, and 11.1% were using daily narcotics to manage their pain. These participants were taking their medication on a pain-contingent basis, reported few active coping strategies and did not express an interest in participating in a pain-treatment programme based on self-management strategies. 3.5.4. Maintenance According to the pain stages of change model (Kerns et al., 1997), the Maintenance stage is characterized by beliefs reflecting an established self-management perspective and a desire to continue to acquire and maintain adaptive self-management strategies, which can include the appropriate time-contingent use of medication. In this study, however, 36.7% of the participants in the Maintenance stage reported using analgesic medication as a primary coping strategy on a daily, pain-contingent basis and 10% reported using daily narcotic medication on a pain-contingent basis to manage their pain. This dependence on medication and inappropriate use is inconsistent with a self-management approach. Also at odds with the model is the fact that 27% of the participants in the Maintenance stage indicated that they had further surgery planned. Sixteen percent of the participant in the Maintenance stage used marijuana to manage their pain (compared with 10% of individuals in the Pre-contemplation stage and 16.7% of individuals in the Contemplation stage).
4. Discussion The results of this study provide empirical support for the findings presented by Kerns et al. (1997) and Jensen et al. (2000) in terms of the internal consistency of the PSOCQ scales, with Cronbach’s alpha scores being greater than 0.75 for each of the scales in a sample of people with chronic pain in the community. Associations between the PSOCQ scale scores other than Pre-contemplation were generally consistent with the findings of Kerns et al. (1997) andJensen et al. (2000).
The results of this study support the assertion by Jensen et al. (2000) that individuals may present as agreeing with statements in a number of different stages for separate selfmanagement behaviours, so that, for example, someone may be in the Pre-contemplation stage for one behaviour (e.g. thinking that they will never be able to manage their pain without analgesic or opioid medication), the Contemplation stage concerning other behaviours (e.g. considering commencement of an exercise programme), the Action stage for other behaviours (e.g. in the process of learning relaxation strategies for pain) and the Maintenance stage for yet other behaviours (e.g. regular use of coping self-statements). As noted by Dijkstra et al. (2001), the items do not appear to be specific enough regarding the particular activity to which the statement refers and may, therefore, cause respondents to endorse items that do not relate to the type of self-management intended by the developers of the scale. For example, a number of participants in the Maintenance stage indicated that the ‘new way’ they had learned to manage their pain was the use of marijuana. In addition to the wording on the questionnaire items, the scoring method may also be contributing to the overlap in stages. The method of scoring in the PSOCQ has been discussed in a number of earlier studies (e.g. Keefe et al., 2000; Dijkstra et al., 2001). Whilst in the present study, the scoring method employed was the one suggested by Kerns et al. (1997) in the scale development study, it is possible that different ‘staging’ techniques would yield different results by providing a profile or pattern of responses rather than a score on an individual scale. However, given the lack of clarity regarding the self-management activities to which the items refer, it is unlikely that a different scoring method would significantly enhance the clarity or validity of the findings. In the original scale-development sample (Kerns et al., 1997), the authors noted an unexpectedly high correlation between Action and Maintenance and argued that greater discrimination may be found in a community-based sample, such as the one described in the present study. It is evident, however, that Action and Maintenance are also highly correlated in this sample and this may be due to failure of the PSOCQ Maintenance stage items to include a temporal dimension, thus distinguishing Action from Maintenance. Use of a time-frame within the items may help to elicit more
S. Habib et al. / Pain 104 (2003) 283–290
accurate responses and would increase the validity of the measure, as the temporal dimensions of the stages of change are theoretically important. 4.1. Study limitations The study may have been strengthened in a number of ways. Firstly, the relatively small sample size in this study and the small number of participants from each recruitment site limit the generalizability of these findings. A larger sample may also have provided stronger support for the stability of the correlations and the reliability of the findings. Secondly, the research participants were heterogeneous in terms of diagnosis and included patients with osteoarthritis, rheumatoid arthritis and fibromyalgia. These patients could be different in that they may only be motivated to self-manage when the symptoms are present (Turk and Rudy, 1990). Given that the participants were seeking treatment and relief of symptoms at the time of the study, it is likely that they were in an acute phase during this period. However, further research may be strengthened by excluding individuals who have chronic conditions that are characterized by fluctuations between remission and recurrent acute phases or designing specific studies to examine these sub-groups, such as the study of Keefe et al. (2000). Finally, the limited number of criterion measures restricts the investigation of the psychometric properties of the PSOCQ in this non-pain-clinic sample. Obviously, further studies are required to validate these preliminary findings with the PSOCQ in a non-pain-clinic population. Despite these limitations, the following conclusions can be drawn. Whilst the stages of change model may be a useful conceptual framework for understanding some types of behaviour change, the theoretical basis of the construct of readiness to adopt a self-management approach to pain needs to be further developed. This will require a more unified understanding and definition of what a self-management approach means to individuals who have chronic pain; as clearly, our current conceptualization of this type of approach is somewhat different from that of our patients. Future research efforts should also be directed towards the development of a valid means of assessing readiness to change in this population. A useful measure would discriminate clearly between specific self-management activities with the inclusion of a temporal dimension to the items, and would assess both belief and behaviour in relation to self-management activities, thereby enabling us to develop more appropriate and effective treatment strategies that are better suited to the motivational needs of our patients.
References Biller N, Arnstein P, Caudill M, Federman C, Guberman C. Predicting completion of a cognitive-behavioural pain management program by
289
initial measures of a chronic pain patient’s readiness for change. Clin J Pain 2000;16:352 –9. Carosella AM, Lackner JM, Feurstein M. Factors associated with early discharge from a multi-disciplinary work rehabilitation program for chronic low back pain. Pain 1994;57:69–76. Chapman CR, Sola AE, Bonica JJ. Illness behaviour and depression in pain center and private pain practice patients. Pain 1997;6:1–7. Compas BE, Haaga DAF, Keefe FJ, Leitenberg H, Wiliams DA. Sampling of empirically supported psychological treatments from health psychology: smoking, chronic pain, cancer and bulimia nervosa. J Consult Clin Psychol 1998;66:89– 112. Crook J, Rideout E, Browne G. The prevalence of pain complaints in a general population. Pain 1984;18:299– 314. Crook J, Weir R, Tunks E. An epidemiological follow-up survey of persistent pain sufferers in a group family practice and specialty pain clinic. Pain 1989;36:49–61. Dijkstra A, Vlaeyen J, Rinjen H, Nielsen W. Readiness to adopt the selfmanagement approach to cope with chronic pain in fibromyalgic patients. Pain 2001;90:37–45. Habib S. Chronic pain: assessing and enhancing readiness to adopt a selfmanagement approach. Doctoral dissertation, James Cook University, Queensland, Australia; 2002 (unpublished). Jensen MP. Enhancing motivation to change in pain treatment. In: Gatchel RJ, Turk DC, editors. Psychological approaches to pain management: a practitioners handbook. New York, NY: Guilford Press, 1996. p. 78– 111. Jensen MP, Nielson W, Romano J, Hill M, Turner J. Further evaluation of the Pain Stages of Change Questionnaire: is the transtheoretical model of change useful for patients with chronic pain? Pain 2000;86:255– 64. Keefe FJ, Lefebvre JC, Kerns RD, Rosenberg R, Beaupre P, Prochaska J, Prochaska JO, Caldwell DS. Understanding the adoption of arthritis self-management: stages of change profiles among arthritis patients. Pain 2000;87:303 –13. Keefe FJ, Salley AN, Lafebvre JC. Coping with pain: conceptual concerns and future directions. Pain 1992;51:131–4. Kerns RD, Bayer LA, Findley JC. Motivation and adherence in the management of chronic pain. In: Block AR, Kremer EF, Fernandez E, editors. Handbook of pain syndromes: biopsychosocial perspectives. Hillsdale, NJ: Lawrence Erlbaum, 1999. p. 99–121. Kerns RD, Rosenberg R. Predicting responses to self-management treatments for chronic pain: application of the pain stages of change model. Pain 2000;84:49–55. Kerns RD, Rosenberg R, Jamison RN, Caudill MA, Haythornwaite J. Readiness to adopt a self-management approach to chronic pain: the pain stages of change questionnaire (PSOCQ). Pain 1997;72:227–34. King SA, Snow BR. Factors for predicting premature termination from a multidisciplinary inpatient chronic pain program. Pain 1989;39:281 –7. McConnaughy EA, Prochaska JO, Velicer WF. Stages of change in psychotherapy: measurement and sample profiles. Psychother Theory Res Pract 1983;20:368 –75. Morley S, Eccleston C, Williams A. A systematic review and meta-analysis of randomized controlled trials of cognitive-behaviour therapy and behaviour therapy for chronic pain in adults, excluding headache. Pain 1999;80:1–13. Prochaska JO, DiClemente CC. Transtheoretical therapy: towards a more integrated model of change. Psychother Theory Res Pract 1982;19: 276– 88. Prochaska JO, DiClemente CC. Towards a comprehensive, transtheoretical model of change: states of change and addictive behaviours. In: Miller WR, Heather N, editors. Applied clinical psychology, 2nd ed. Treating addictive behaviours, New York, NY: Plenum Press, 1998. p. 3–24. Prochaska JO, Velicer WF, Rossi JS, Goldstein MG, Marcus BH, Rakowski W, Fiore C, Harlow LL, Redding CA, Rosenbloom D, Rossi SR. Stages of change and decisional balance for twelve problem behaviours. Health Psychol 1994;13:39–46. Richmond RL, Carmody TP. Dropout from treatment for chronic low back pain. Prof Psychol Res Pract 1999;30:51–5.
290
S. Habib et al. / Pain 104 (2003) 283–290
Turk DC. Customising treatment for chronic pain patients: who, what, and why. Clin J Pain 1990;6:255 –70. Turk DC, Rudy TE. Neglected factors in chronic pain treatment outcome studies – referral patterns, failure to enter treatment and attrition. Pain 1990;43:7–25. Turk DC, Rudy TE. Neglected topics in the treatment of chronic pain
patients – relapse, non-compliance, adherence enhancement. Pain 1991;44:5–28. Turk DC, Miechenbaum D, Genest M. Pain and behavioural medicine: a cognitive–behavioural perspective. New York, NY: Guilford Press, 1983.
Readiness to adopt a self-management approach to pain: evaluation of the pain stages of change model in a non-pain-clinic sample Suzanne Habiba,*, Shirley A. Morrisseyb, Edward Helmesa a
School of Psychology, James Cook University, P.O. Box 6811, Cairns, Queensland 4870, Australia b Psychology Department, Northern Territory University, Darwin, Australia Received 28 March 2002; accepted 13 January 2003
Abstract The Transtheoretical model of stages of behaviour change has stimulated research interest in relation to chronic pain, yet studies using the Pain Stages of Change Questionnaire (PSOCQ; Pain (72) 1997 227) have reported inconsistent findings and have generally utilized painclinic samples. The aims of the current study were to assess the general validity of the PSOCQ with a non-pain-clinic sample of patients with chronic pain, and to examine the utility of the stages of change model as applied to this population. The study employed multi-stage, clustersampling methodology, with 90 participants recruited from 19 medical and allied health clinics and practices. The findings demonstrated a number of limitations of the PSOCQ in terms of its ability to classify individuals into specific stages of change. The stages of change model requires adaptation in order to be useful for treatment planning in a non-pain-clinic sample of patients with chronic pain. q 2003 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. Keywords: Stages of change model; Pain Stages of Change Questionnaire; Self-management approach
1. Introduction Psychologically based programmes for chronic pain have been repeatedly shown to be effective (e.g. Keefe et al., 1992; Compas et al., 1998; Morley et al., 1999). A significant proportion of individuals, however, either fail to engage in, or fail to benefit from this type of treatment (Turk and Rudy, 1990, 1991; Richmond and Carmody, 1999). Although there have been a number of prior research attempts to modify treatment and increase rates of engagement, adherence and maintenance of treatment gains (e.g. King and Snow, 1989; Carosella et al., 1994), these findings have been inconsistent and unreliable, therefore contributing little to improving interventions. Recent attempts have been made to examine as to why there exists a subset of individuals who do not benefit and coworkers (Kerns et al., 1997, 1999; Kerns and Rosenberg, 2000) and Jensen et al. (1996, 2000) considered the relevance and application of the Transtheoretical model of behaviour change (Prochaska and DiClemente, 1982; Prochaska et al., 1994), in conjunction with a cognitive – * Corresponding author. Tel.: þ 7-4991-4443; fax: þ7-4091-4443. E-mail address: [email protected] (S. Habib).
behavioural perspective on chronic pain (Turk et al., 1983). This culminated in the development of the pain stages of change model and the Pain Stages of Change Questionnaire (PSOCQ; Kerns et al., 1997). The pain stages of change model proposes that individuals vary in their readiness to adopt a self-management approach to pain and that individuals may be categorised, on the basis of their beliefs about their pain, into one of the four stages of readiness to change (SRC). These stages are: Pre-contemplation (not considering any change in behaviour), Contemplation (serious consideration of change sometime in the future but no clear plan of intent), Action (concrete activities that will lead to the desired change) and Maintenance (active efforts to sustain the changes made). The model also proposes that an individual’s current stage of change may determine the most effective therapeutic approach. For example, individuals in the Pre-contemplation or Contemplation stage may benefit from strategies, such as cognitive re-structuring and reconceptualization of the pain as being manageable, whereas individuals in the ‘higher’ stages of change would be more likely to benefit from behavioural techniques, such as pacing activities, relaxation and exercise, which ensure readiness for active involvement and responsibility of the
0304-3959/03/$30.00 q 2003 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. doi:10.1016/S0304-3959(03)00016-2
284
S. Habib et al. / Pain 104 (2003) 283–290
individual in the rehabilitation process. Kerns and his colleagues assert that successful demonstration of a confirmatory four-factor analysis is consistent with the stages of change model, and the reliability, criterion-related validity, internal consistency and stability of each of the four scales, provide support for the relevance of the stages of change model to management of pain. Jensen et al. (2000) sought to further explore the psychometric properties of the PSOCQ by examining the generalizability of the findings presented by Kerns et al. (1997). Jensen et al. studied two new sub-samples of patients with chronic pain, a pain-clinic sample and a fibromyalgia sample. Internal consistency co-efficient scores were greater than 0.70 on all the PSOCQ sub-scales for both samples and are consistent with those reported by Kerns et al. (1997). PSOCQ scale score inter-relations were also similar to the associations found by Kerns et al. (1997) in the original scale-development sample. All of the criterion measures were associated with the PSOCQ scales in the directions hypothesized. An attempt to examine the differences among the stages of change groups for each of the PSOCQ scales was only partially successful as none of the group means differed significantly on the Pre-contemplation scale. Participants classified as being in either the Action or Maintenance stage also did not differ significantly on any of the mean PSOCQ scale scores. In both samples, participants classified as being in the Contemplation, Action or Maintenance stage scored higher than three, on average (indicating more agreement than disagreement), on the Contemplation, Action and Maintenance scale items, thus apparently being simultaneously in all three stages concerning readiness to adopt a self-management approach to pain. Jensen et al. (2000) argued that the PSOCQ does not discriminate sufficiently between the stages of change and, therefore, may not be useful in its current form as a tool used to classify painclinic patients into distinct stages of readiness to adopt a self-management approach. Kerns and Rosenberg (2000) conducted a study to assess the predictive validity of the PSOCQ in relation to participation and outcome in treatment programmes. They hypothesized that high Pre-contemplation scores would be predictive of higher rates of dropout and poorer outcomes relative to low Pre-contemplation scores. The results demonstrated significant differences between completers and non-completers, with non-completers having higher scores on the Pre-contemplation scale, and completers generally having low scores on Pre-contemplation and higher scores on the other scales. Whilst the study supported the hypothesis that the PSOCQ (Kerns et al., 1997) could predict engagement in treatment, Kerns and Rosenberg (2000) found no support for the hypothesis that PSOCQ scores could predict outcome subsequent to treatment, as the participants in all the stages of change were shown to be equally likely to achieve improvements in pain management, emotional well-being and functioning. Kerns and
Rosenberg concluded that replication of the study and further refinements of the pain stages of change model are necessary before the PSOCQ can be confidently utilized for clinical decision-making. In a subsequent study on the predictive validity of the PSOCQ (Kerns et al., 1997), Biller et al. (2000) supported the findings of Kerns and Rosenberg (2000) that high Precontemplation scores and low Contemplation scores were the strongest predictor of completion of a pain-management programme. Although the findings of Biller et al. generally support the predictive validity of the PSOCQ, the authors caution against its use as a screening instrument in its current form. According to Biller et al. (2000), whilst the moderate predictive validity of the mean scale score may be helpful for the practitioners in identifying appropriate intervention targeted to stages of change, it is not sufficient to be used as an inclusion or exclusion criterion for treatment. Since the introduction of the pain stages of change model (Kerns et al., 1997), to date, two published studies have sought to develop or modify other stage-based psychometric measures for specific sub-groups of pain patients. In the first study, Keefe et al. (2000) used a cluster analysis to identify five distinct sub-groups of patients with arthritis pain, based on their responses on a modified version of the University of Rhode Island Change Questionnaire (URICA; McConnaughy et al., 1983). Keefe et al. (2000) concluded that as the sub-groups loosely corresponded with the stages proposed by Prochaska and DiClemente (1998), the Transtheoretical model of stages of change may be applicable to arthritis patients. The second study, subsequent to the development of the pain stages of change model (Kerns et al., 1997), was conducted by Dijkstra et al. (2001). The study sought to explore the psychometric properties of a Dutch version of a questionnaire, assessing readiness to change in a sample of 321 Dutch fibromyalgia patients. The SRC measure was adapted from an unpublished English readiness to change questionnaire (Nielson and Vlaeyen, 1996 in Dijkstra et al., 2001), which in turn was based on McConnaughy et al.’s (1983) measure of readiness to change. The items were adapted to reflect a readiness to adopt a self-management approach to pain. Concerns voiced by Dijkstra et al. (2001) in relation to the application of the stages model to chronic pain are similar to those of Jensen et al. (2000) in that there is lack of clarity as to which specific behaviour the change refers to, and that the theoretical basis of the construct of readiness to adopt a self-management approach to pain needs to be further developed. It can be tentatively concluded from the previous research that the stages of change model has the potential to assist in the assessment and treatment planning for a number of sub-groups of pain patients. Inconsistent research findings indicate, however, that continued investigation and revision of the model is warranted in order to be able to work confidently within this theoretical framework for
S. Habib et al. / Pain 104 (2003) 283–290
assessment and treatment of pain. Clearly, the PSOCQ lacks adequate sensitivity to be used in clinical decision making with regards to offering self-management treatment. However, if we are to develop prescriptive ‘stage-specific’ treatment programmes as recommended by Kerns et al. (1999) and Kerns and Rosenberg (2000), we must first refine our means of assessing stage of readiness to adopt a selfmanagement approach in order to assign patients to appropriate treatment programmes. Other than the study by Keefe et al. (2000), all of the published studies of stages of change and chronic pain have been conducted with pain-clinic samples; therefore, it is not clear that to what extent the findings would generalize to the general population of individuals with chronic pain. A number of earlier studies have demonstrated that pain-clinic patients differ significantly from pain patients treated in the community (Turk, 1990). In studies comparing these two populations, pain-clinic patients were found to complain of more constant pain, suffer higher levels of functional impairment and report greater emotional and psychosocial distress than the pain patients being treated in the community (e.g. Chapman et al., 1997; Crook et al., 1984, 1989). Whilst patients in pain clinics are reported to suffer from higher levels of distress and disability than those treated in the community, the low number of Pre-contemplators in previous studies using pain-clinic samples suggests that by the time pain-clinic patients attend an intake assessment, their motivation may have been influenced by referring practitioners and/or the knowledge that pain-clinic treatment focuses mainly on self-management techniques. It may also be the case that patients who are not interested in this type of approach are screened out of pain-clinic programmes or simply fail to attend the initial assessment. It is important, therefore, to establish whether measures and interventions developed on pain-clinic samples are appropriate for use on the pain sufferers in the general community where the numbers of individuals who are not contemplating a self-management approach may be significantly greater. The aims of the present research were, firstly, to assess the general validity of the PSOCQ (Kerns et al., 1997) with a non-pain-clinic sample in order to determine the generalizability of the measure, and, secondly, to examine the utility of the stages of change model as applied to this population in terms of its ability to assist in treatment planning. The study employed a survey questionnaire design with participants recruited from medical and allied health clinics and practices (Habib, 2002).
2. Method 2.1. Sampling methodology Participants for the survey were recruited by means of a
285
multi-stage cluster-sampling methodology. The clusters or self-formed groups comprised physiotherapists, chiropractors, medical practices, bulk-billing medical practices, CRS Australia rehabilitation units and public hospital outpatient clinics. Within each group, every second practice or clinic was selected from the telephone directory. Within each selected practice that agreed to participate, every second patient that arrived at the clinic and who met participation criteria was asked to complete the survey questionnaire. To generalize the research findings to a wide range of treatment-seeking patients with pain, fee-paying and nonfee paying (bulk billing) practices were sampled, participants were recruited both during and after working hours and both rural and city practices were sampled. 2.2. Participants The sample that agreed to participate in the survey and met eligibility criteria comprised 90 participants drawn from 19 practices (see Table 1). Inclusion criteria were that the patient was over the age of 18 and reporting pain for 3 months or more; exclusion criteria were inadequate literacy to read and comprehend the research questionnaire, visibly affected by drugs or alcohol or actively psychotic. The mean age of the participants was 43 years (standard deviation ðSDÞ ¼ 13:83, range ¼ 19 – 77). Fiftytwo participants (57.8%) were male. The majority of the participants (60%) were married, 22% were single, 8% were divorced and 8% were in de-facto relationships. The majority of the participants (42%) had 3 years of secondary education. Seventeen percent of the participants had completed a course at college, and 8% had university degree. Forty-one percent of the participants were unemployed, 28% percent were blue-collar workers, 9% were self-employed, 4% of the participants were engaged in managerial positions, 11% held professional positions, 6% were retired and 6% of the participants were students. Eightfive percent of the sample was of Australian or New Zealander descent, the remaining 15% was identified as British or European. Pain site varied, though 28% of the sample presented with low back pain. Cause of pain Table 1 Proportions of participants from each practice type Practice type
Fee/non-fee Participants Percentage of sample
Physiotherapist Chiropractor Medical centre Bulk-billing centre Orthopedic clinic Hand clinic Fracture clinic Public physiotherapist Rehabilitation unit
Fee Fee Fee Non-fee Non-fee Non-fee Non-fee Non-fee Non-fee
Total
10 12 7 8 17 1 9 6 20
11.1 13.3 7.8 8.9 18.9 1.1 10.0 6.7 22.2
90
100.00
286
S. Habib et al. / Pain 104 (2003) 283–290
included disc pathology (28%), neuropathic pain (18%), osteoarthritis (17%), rheumatoid arthritis (11%), fibromyalgia (6%), ligament damage (9%) and carpal tunnel syndrome (2%). The remaining 8% of the participants had no clear diagnosis. Pain duration ranged from 4 months to 30 years (M ¼ 6:6 years, SD ¼ 13:85). Average pain severity was measured on a numerical rating scale where 0 referred to no pain and 100, the worst pain ever experienced. The mean average pain rating was 53 (SD ¼ 21:81, range ¼ 10 – 100). Seven percent of the participants reported that they were currently receiving compensation, 9% were in the process of applying for compensation or disability payment and 12% were currently litigating in relation to their pain. The participants did not receive any reward or incentive for completing the questionnaire. Non-respondents: Within the practices sampled, nine individuals declined to participate in the questionnaire survey. The majority of the individual refusals came from fee-paying practices and were male. The mean age was 54 years (SD ¼ 12:5, range ¼ 28 – 76). Non-respondents had an average of 2 years of secondary education. Three of the refusing individuals were retired, two were unemployed, one was self-employed, two were blue-collar workers and one was engaged in home duties. Five of the nonrespondents reported a pain duration of greater than 5 years. The remaining four non-respondents reported a pain duration of more than 2 years but less than 5 years. The reasons for non-participation were pain too great, not enough time and not interested. Non-respondents tended to be older and had lower levels of education than respondents; however, the differences between respondents and nonrespondents were not statistically significant.
the practice or clinic, prior to consultations with medical practitioners, physiotherapists and chiropractors.
2.3. Questionnaire
3.3. Differences between individuals in each stage of change
The questionnaire was designed to elicit demographic and medical information, pain intensity, duration and coping strategies and compensation and litigation status. Stage of readiness to adopt a self-management approach to chronic pain was assessed using the PSOCQ (Kerns et al., 1997). Two additional measures of affect were used; however, these findings are reported elsewhere (Habib, 2002). The final page on the questionnaire provided brief information regarding pain-management workshops based on a self-management approach that might be offered as a later part of the research. Participants were informed that if they wished to provide their name and contact phone number they would be contacted in due course with further information.
One-way analyses of variance were performed to examine predicted differences among the classified stage of change groups in the four PSOCQ scales. To control for alpha inflation associated with multiple comparisons, a family-wise Bonferroni alpha was set at 0.0125 (0.05/4) and the data are presented in Table 5. The results demonstrate significant differences among the stages of change for each of the PSOCQ scales, with the results identifying four specific stages. However, participants classified as being in either the Contemplation, Action
2.4. Procedure Every second patient who attended each clinic and met participation criteria was requested to complete a questionnaire. The questionnaires were given to individuals in
3. Results 3.1. Internal consistency scores Using the scoring method outlined by Kerns et al. (1997), each study participant was classified into one of the four stages based on their highest PSOCQ scale score as illustrated in Table 2. The scale score means, SD and the internal consistency co-efficients of the PSOCQ scale scores are presented in Table 3. Internal consistency co-efficients were greater than 0.75 for each of the four scales and are consistent with those reported by Kerns et al. (1997) in the original scaledevelopment sample and Jensen et al. (2000) in a pain-clinic and fibromyalgia sample. 3.2. Associations among the PSOCQ scales Inter-relations of the scale scores for the current sample are presented in Table 4. As illustrated in Table 4, there was no significant association between Pre-contemplation and Contemplation. Pre-contemplation was significantly, negatively associated with Action and Maintenance. Significant positive associations were demonstrated between Contemplation and Action; however, there was no significant association between Contemplation and Maintenance. Significant positive associations were demonstrated between Action and Maintenance.
Table 2 Proportions of participants in each stage of change Stage
n
Percentage
Total sample (N ¼ 90) Pre-contemplation Contemplation Action Maintenance
25 16 18 29
27.8 18.9 20.0 33.3
S. Habib et al. / Pain 104 (2003) 283–290 Table 3 Means, SD and internal consistency scores for PSOCQ scales PSOCQ scale
Mean
SD
Cronbach’s alpha
Total sample (N ¼ 90) Pre-contemplation Contemplation Action Maintenance
3.06 3.23 3.25 3.44
0.80 0.62 0.87 0.81
0.76 0.79 0.82 0.86
or Maintenance stage also had mean scale scores of three or over for Contemplation, Action and Maintenance, indicating more agreement than disagreement with items on each of the other scales. 3.4. Associations between stages of change and demographic variables Pearson correlations were calculated in order to explore relationships between PSOCQ scale scores and demographic variables. In order to correct for the large number of planned comparisons, results were considered significant if they were at, or below, the significance level of 0.01. The analyses revealed no significant relationships for age, ethnicity, marital status or average pain rating and any of the PSOCQ scale scores. Pain duration was negatively associated with Pre-contemplation (r ¼ 20:270, P , 0:01) and positively associated with Action (r ¼ 0:258, P , 0:01) and Maintenance (r ¼ 0:233, P , 0:01). The t-tests revealed no significant differences for feepaying versus non-fee-paying, married versus non-married or compensating versus non-compensating. The male participants had significantly higher mean Pre-contemplation scores than the female participants (male Precontemplation score, M ¼ 3:23, female Pre-contemplation score, M ¼ 2:82, P ¼ 0:02), and this fact is consistent with the findings of Kerns et al. (1997) in the original scaledevelopment sample. 3.5. Associations between stages of change and current coping strategies The second aim of the study was to determine the utility of the pain stages of change model (Kerns et al., 1997) in relation to its usefulness for treatment planning in a nonTable 4 PSOCQ inter-scale correlation co-efficients PSOCQ scale
Contemplation
Non-pain-clinic sample (N ¼ 90) Pre-contemplation 0.09 Contemplation Action Maintenance * P , 0:05, * * P , 0:01, * * * P , 0:001:
Action
Maintenance
0.45*** 0.32**
20.46*** 20.07 0.68***
287
pain-clinic sample. Univariate statistics were used to examine the characteristics of the participants in each stage of change, particularly in relation to coping strategies, including medication use and willingness to participate in a self-management pain programme. A number of inconsistencies were discovered in relation to the predictions based on the model and reported behaviour. 3.5.1. Pre-contemplation According to the pain stages of change model (Kerns et al., 1997), individuals in the Pre-contemplation stage believe that their pain is purely a medical problem, thus, would not use self-management strategies and would not be interested in this type of approach. Contrary to the model, 44% of the participants in the Pre-contemplation stage indicated an interest in participating in a pain programme based on self-management strategies with no medical treatment involved. As it was not the intention of the researchers to examine the predictive ability of the PSOCQ, actual participation in treatment was not examined in this study; however, the large number of Pre-contemplators indicating an interest in this type of approach is thought to be inconsistent with the definition of Pre-contemplation provided by Kerns et al. (1997) in the scale development study. Although it could be argued that subsequent studies (e.g. Kerns and Rosenberg, 2000) have demonstrated that Pre-contemplators can be successfully engaged in treatment with a self-management focus (though they are less likely to complete treatment than Contemplators), this shows that either stages are less stable than originally thought to be, or the wording on the PSOCQ is not specific enough to accurately classify individuals into appropriate stages of readiness to adopt a self-management approach to pain. Further speculation about other possible reasons for Precontemplators expressing an interest in self-management programmes raised three possibilities. Firstly, it could be that completing the PSOCQ may have had a motivational effect, thus moving participant closer to considering a selfmanagement approach; secondly, individuals in these settings (i.e. where no pain clinic is available) may be willing to try any type of treatment; and thirdly, participants may have appreciated the attention from researchers and felt that participating in research validated their pain to others. 3.5.2. Contemplation The Contemplation stage is characterized by a consideration of the potential value of adopting a self-management approach, but a reluctance to relinquish pursuit of a medical solution (Kerns et al., 1997). Fifty-nine percent of the participants in the Contemplation stage indicated that they would like to participate in a pain-treatment programme based on self-management strategies with no medical treatment involved. 3.5.3. Action The Action stage is characterized by acceptance of a self-
288
S. Habib et al. / Pain 104 (2003) 283–290
Table 5 Mean scale scores for classified stage of change groups Classified stage Pre-contemplation (n ¼ 25) PSOCQ scale score mean (SD) Pre-contemplation 4.00 (0.40) Contemplation 3.07 (0.58) Action 2.52 (0.70) Maintenance 2.87 (0.64)
Contemplation (n ¼ 16)
Action (n ¼ 18)
Maintenance (n ¼ 29)
F-value
2.92 (0.74) 3.78 (0.77) 3.10 (1.0) 3.00 (0.76)
2.61 (0.47) 3.33 (0.41) 3.85 (0.51) 3.72 (0.57)
2.65 (0.56) 3.00 (0.50) 3.52 (0.66) 4.02 (0.47)
36.06* 7.92* 14.22* 21.30*
* P , 0:001:
management approach and active engagement in attempts to improve self-management skills (Kerns et al., 1997). In this study, 33% of the participants classified by the PSOCQ as being in the Action stage were using daily analgesics as the primary pain-coping strategy, and 11.1% were using daily narcotics to manage their pain. These participants were taking their medication on a pain-contingent basis, reported few active coping strategies and did not express an interest in participating in a pain-treatment programme based on self-management strategies. 3.5.4. Maintenance According to the pain stages of change model (Kerns et al., 1997), the Maintenance stage is characterized by beliefs reflecting an established self-management perspective and a desire to continue to acquire and maintain adaptive self-management strategies, which can include the appropriate time-contingent use of medication. In this study, however, 36.7% of the participants in the Maintenance stage reported using analgesic medication as a primary coping strategy on a daily, pain-contingent basis and 10% reported using daily narcotic medication on a pain-contingent basis to manage their pain. This dependence on medication and inappropriate use is inconsistent with a self-management approach. Also at odds with the model is the fact that 27% of the participants in the Maintenance stage indicated that they had further surgery planned. Sixteen percent of the participant in the Maintenance stage used marijuana to manage their pain (compared with 10% of individuals in the Pre-contemplation stage and 16.7% of individuals in the Contemplation stage).
4. Discussion The results of this study provide empirical support for the findings presented by Kerns et al. (1997) and Jensen et al. (2000) in terms of the internal consistency of the PSOCQ scales, with Cronbach’s alpha scores being greater than 0.75 for each of the scales in a sample of people with chronic pain in the community. Associations between the PSOCQ scale scores other than Pre-contemplation were generally consistent with the findings of Kerns et al. (1997) andJensen et al. (2000).
The results of this study support the assertion by Jensen et al. (2000) that individuals may present as agreeing with statements in a number of different stages for separate selfmanagement behaviours, so that, for example, someone may be in the Pre-contemplation stage for one behaviour (e.g. thinking that they will never be able to manage their pain without analgesic or opioid medication), the Contemplation stage concerning other behaviours (e.g. considering commencement of an exercise programme), the Action stage for other behaviours (e.g. in the process of learning relaxation strategies for pain) and the Maintenance stage for yet other behaviours (e.g. regular use of coping self-statements). As noted by Dijkstra et al. (2001), the items do not appear to be specific enough regarding the particular activity to which the statement refers and may, therefore, cause respondents to endorse items that do not relate to the type of self-management intended by the developers of the scale. For example, a number of participants in the Maintenance stage indicated that the ‘new way’ they had learned to manage their pain was the use of marijuana. In addition to the wording on the questionnaire items, the scoring method may also be contributing to the overlap in stages. The method of scoring in the PSOCQ has been discussed in a number of earlier studies (e.g. Keefe et al., 2000; Dijkstra et al., 2001). Whilst in the present study, the scoring method employed was the one suggested by Kerns et al. (1997) in the scale development study, it is possible that different ‘staging’ techniques would yield different results by providing a profile or pattern of responses rather than a score on an individual scale. However, given the lack of clarity regarding the self-management activities to which the items refer, it is unlikely that a different scoring method would significantly enhance the clarity or validity of the findings. In the original scale-development sample (Kerns et al., 1997), the authors noted an unexpectedly high correlation between Action and Maintenance and argued that greater discrimination may be found in a community-based sample, such as the one described in the present study. It is evident, however, that Action and Maintenance are also highly correlated in this sample and this may be due to failure of the PSOCQ Maintenance stage items to include a temporal dimension, thus distinguishing Action from Maintenance. Use of a time-frame within the items may help to elicit more
S. Habib et al. / Pain 104 (2003) 283–290
accurate responses and would increase the validity of the measure, as the temporal dimensions of the stages of change are theoretically important. 4.1. Study limitations The study may have been strengthened in a number of ways. Firstly, the relatively small sample size in this study and the small number of participants from each recruitment site limit the generalizability of these findings. A larger sample may also have provided stronger support for the stability of the correlations and the reliability of the findings. Secondly, the research participants were heterogeneous in terms of diagnosis and included patients with osteoarthritis, rheumatoid arthritis and fibromyalgia. These patients could be different in that they may only be motivated to self-manage when the symptoms are present (Turk and Rudy, 1990). Given that the participants were seeking treatment and relief of symptoms at the time of the study, it is likely that they were in an acute phase during this period. However, further research may be strengthened by excluding individuals who have chronic conditions that are characterized by fluctuations between remission and recurrent acute phases or designing specific studies to examine these sub-groups, such as the study of Keefe et al. (2000). Finally, the limited number of criterion measures restricts the investigation of the psychometric properties of the PSOCQ in this non-pain-clinic sample. Obviously, further studies are required to validate these preliminary findings with the PSOCQ in a non-pain-clinic population. Despite these limitations, the following conclusions can be drawn. Whilst the stages of change model may be a useful conceptual framework for understanding some types of behaviour change, the theoretical basis of the construct of readiness to adopt a self-management approach to pain needs to be further developed. This will require a more unified understanding and definition of what a self-management approach means to individuals who have chronic pain; as clearly, our current conceptualization of this type of approach is somewhat different from that of our patients. Future research efforts should also be directed towards the development of a valid means of assessing readiness to change in this population. A useful measure would discriminate clearly between specific self-management activities with the inclusion of a temporal dimension to the items, and would assess both belief and behaviour in relation to self-management activities, thereby enabling us to develop more appropriate and effective treatment strategies that are better suited to the motivational needs of our patients.
References Biller N, Arnstein P, Caudill M, Federman C, Guberman C. Predicting completion of a cognitive-behavioural pain management program by
289
initial measures of a chronic pain patient’s readiness for change. Clin J Pain 2000;16:352 –9. Carosella AM, Lackner JM, Feurstein M. Factors associated with early discharge from a multi-disciplinary work rehabilitation program for chronic low back pain. Pain 1994;57:69–76. Chapman CR, Sola AE, Bonica JJ. Illness behaviour and depression in pain center and private pain practice patients. Pain 1997;6:1–7. Compas BE, Haaga DAF, Keefe FJ, Leitenberg H, Wiliams DA. Sampling of empirically supported psychological treatments from health psychology: smoking, chronic pain, cancer and bulimia nervosa. J Consult Clin Psychol 1998;66:89– 112. Crook J, Rideout E, Browne G. The prevalence of pain complaints in a general population. Pain 1984;18:299– 314. Crook J, Weir R, Tunks E. An epidemiological follow-up survey of persistent pain sufferers in a group family practice and specialty pain clinic. Pain 1989;36:49–61. Dijkstra A, Vlaeyen J, Rinjen H, Nielsen W. Readiness to adopt the selfmanagement approach to cope with chronic pain in fibromyalgic patients. Pain 2001;90:37–45. Habib S. Chronic pain: assessing and enhancing readiness to adopt a selfmanagement approach. Doctoral dissertation, James Cook University, Queensland, Australia; 2002 (unpublished). Jensen MP. Enhancing motivation to change in pain treatment. In: Gatchel RJ, Turk DC, editors. Psychological approaches to pain management: a practitioners handbook. New York, NY: Guilford Press, 1996. p. 78– 111. Jensen MP, Nielson W, Romano J, Hill M, Turner J. Further evaluation of the Pain Stages of Change Questionnaire: is the transtheoretical model of change useful for patients with chronic pain? Pain 2000;86:255– 64. Keefe FJ, Lefebvre JC, Kerns RD, Rosenberg R, Beaupre P, Prochaska J, Prochaska JO, Caldwell DS. Understanding the adoption of arthritis self-management: stages of change profiles among arthritis patients. Pain 2000;87:303 –13. Keefe FJ, Salley AN, Lafebvre JC. Coping with pain: conceptual concerns and future directions. Pain 1992;51:131–4. Kerns RD, Bayer LA, Findley JC. Motivation and adherence in the management of chronic pain. In: Block AR, Kremer EF, Fernandez E, editors. Handbook of pain syndromes: biopsychosocial perspectives. Hillsdale, NJ: Lawrence Erlbaum, 1999. p. 99–121. Kerns RD, Rosenberg R. Predicting responses to self-management treatments for chronic pain: application of the pain stages of change model. Pain 2000;84:49–55. Kerns RD, Rosenberg R, Jamison RN, Caudill MA, Haythornwaite J. Readiness to adopt a self-management approach to chronic pain: the pain stages of change questionnaire (PSOCQ). Pain 1997;72:227–34. King SA, Snow BR. Factors for predicting premature termination from a multidisciplinary inpatient chronic pain program. Pain 1989;39:281 –7. McConnaughy EA, Prochaska JO, Velicer WF. Stages of change in psychotherapy: measurement and sample profiles. Psychother Theory Res Pract 1983;20:368 –75. Morley S, Eccleston C, Williams A. A systematic review and meta-analysis of randomized controlled trials of cognitive-behaviour therapy and behaviour therapy for chronic pain in adults, excluding headache. Pain 1999;80:1–13. Prochaska JO, DiClemente CC. Transtheoretical therapy: towards a more integrated model of change. Psychother Theory Res Pract 1982;19: 276– 88. Prochaska JO, DiClemente CC. Towards a comprehensive, transtheoretical model of change: states of change and addictive behaviours. In: Miller WR, Heather N, editors. Applied clinical psychology, 2nd ed. Treating addictive behaviours, New York, NY: Plenum Press, 1998. p. 3–24. Prochaska JO, Velicer WF, Rossi JS, Goldstein MG, Marcus BH, Rakowski W, Fiore C, Harlow LL, Redding CA, Rosenbloom D, Rossi SR. Stages of change and decisional balance for twelve problem behaviours. Health Psychol 1994;13:39–46. Richmond RL, Carmody TP. Dropout from treatment for chronic low back pain. Prof Psychol Res Pract 1999;30:51–5.
290
S. Habib et al. / Pain 104 (2003) 283–290
Turk DC. Customising treatment for chronic pain patients: who, what, and why. Clin J Pain 1990;6:255 –70. Turk DC, Rudy TE. Neglected factors in chronic pain treatment outcome studies – referral patterns, failure to enter treatment and attrition. Pain 1990;43:7–25. Turk DC, Rudy TE. Neglected topics in the treatment of chronic pain
patients – relapse, non-compliance, adherence enhancement. Pain 1991;44:5–28. Turk DC, Miechenbaum D, Genest M. Pain and behavioural medicine: a cognitive–behavioural perspective. New York, NY: Guilford Press, 1983.