Using Illness Perceptions to Cluster Chronic Pain Patients Results From a Trial on the Chronic Pain Self-Management Program

Objectives: The aims of our study were (1) to identify possible subgroups of individuals with chronic pain based on their illness perceptions (IPs); (2) to examine whether these subgroups differed in health status and health expenditure; and (3) to examine whether the subgroups differed in their response to participation in a lay-led Chronic Pain Self-Management Program (CPSMP). Materials and Methods: Four hundred twenty-four participants in a randomized controlled trial on the CPSMP completed a questionnaire on their perceptions of their chronic pain condition at baseline. In addition, they completed a range of health status measures at baseline and 3 months after end of participation in the CPSMP. Health care expenditure was obtained from Danish health registers. We performed cluster analyses to identify possible subgroups based on the participants' perceptions of their chronic pain condition. Results: Cluster analysis of IPs resulted in 3 meaningful clusters, classified as distressed, certain cause, distressed, uncertain cause, and nondistressed, certain cause, respectively. The 2 distressed groups had significantly higher scores on pain catastrophizing, illness worry, and emotional distress than did the nondistressed. Moderator analyses showed, that the distressed, certain cause had significant positive effect of participating in the CPSMP on pain catastrophizing and emotional distress, whereas the distressed, uncertain cause had a decrease and the nondistressed, certain cause an increase in primary health expenditure. Discussion: Clusters based on IPs meaningfully distinguished individuals with chronic pain on health measures and outcome of participating in the CPSMP, and may prove useful in future studies to understand responses to treatment.