Validation of a questionnaire for central nervous system aspects of joint pain: the CAP questionnaire

Abstract

Background and Aims Neuropathic-like pain, fatigue, cognitive difficulty, catastrophising, anxiety, sleep disturbance, depression, and widespread pain associate with a single factor in people with knee pain. We report the Central Aspects of Pain questionnaire (CAP) to characterise this across painful musculoskeletal conditions. Methods CAP was derived from the 8 item CAP-Knee questionnaire, and completed by participants with joint pain in the Investigating Musculoskeletal Health and Wellbeing survey. Subgroups had osteoarthritis, back pain or fibromyalgia. Acceptability was evaluated by feedback and data missingness. Correlation coefficients informed widespread pain scoring threshold in relation to the other items, and evaluated associations with pain. Factor analysis assessed CAP structure. Intraclass Correlation Coefficient (ICC) between paper and electronic administration assessed reliability. Friedman test assessed score stability over 4 years in people reporting knee osteoarthritis. Results Data were from 3579 participants (58% female, median age; 71 years), including subgroups with osteoarthritis (n = 1158), back pain (n = 1292) or fibromyalgia (n = 177). Across the 3 subgroups, ≥10/26 painful sites on the manikin scored widespread pain. Reliability was high (ICC= 0.89 (95% CI: 0.84–0.92)) and CAP scores fit to 1 and 2 factor model, with a total CAP score that was associated with pain severity and quality (r = 0.50–0.72). In people with knee pain, CAP scores were stable over 4 years at the group level, but displayed significant temporal heterogeneity within individual participants. Conclusions Central Aspects of Pain is reliably measured by the CAP questionnaire across a range of painful musculoskeletal conditions, and is a changeable state.