Meta-analysis of the psychometric properties of the Pain Catastrophizing Scale and associations with participant characteristics.

Abstract

The aims of this study were to review the psychometric properties of the widely used Pain Catastrophizing Scale (PCS) using meta-analytic methods, and to investigate the relationship between PCS scores and participant characteristics. A systematic search from 1995 found 229 experimental, quasi-experimental and correlational studies that report PCS scores. Multivariate regression explored variables related to pain catastrophizing and participant demographics. Across studies, good internal reliability (á = 0.92, 95% CI 0.91 - 0.93) and test-retest reliability scores (Spearman ñ = 0.88, 95% CI 0.83 - 0.93) were found for PCS total scores but not for subscales. PCS scores were unrelated to age or gender, but strongly related to participants pain type, highest in those with generalized pain. Language of the PCS also affected PCS scores, with further research necessary to determine linguistic, cultural or methodological (e.g. sampling strategy) influences. Study type influenced PCS scores with non-randomized controlled trials reporting higher PCS scores than other study types, but results were confounded with pain diagnosis, as controlled trials were more likely than quasi-experimental studies to recruit clinical samples. The meta-analytic results provide insight into demographic influences on pain catastrophizing scores and highlight areas for further research. The advantages of systematic review and meta-analytic methods to achieve greater understanding and precision of psychometric properties - in this case, of the PCS - are applicable to other widely used outcome tools.