A systematic review of cross-cultural validation of the pain catastrophizing scale

Abstract

Background: Although it has been suggested that the different cultural and social environments between countries contribute to variations in pain catastrophizing (PC), an international comparison of PC in patients with chronic pain has not yet been reported. Prior to undertaking this comparison, a cross-cultural assessment of the pain catastrophizing scale (PCS) was undertaken to explore the different factor structures among each translated version of the PCS.

Methods: The protocol for this systematic review was prospectively registered on International Prospective Register of Systematic Reviews 2018 (CRD 42018086719). Electronic searches were conducted in the following databases: Ovid/Embase, Ovid/MEDLINE and Ovid/PsycINFO, and then 19 articles (16 language versions) were included in this review. Based on the COSMIN check list, we investigated language translation followed by five domains of cross-cultural validation: structural validity, internal consistency, test-retest reliability and hypotheses testing for construct validity in each study.

Results: We found that (a) there were inconsistent structural models among each translated version, leading to variant subdomain structures for rumination, magnification and helplessness; (b) all languages versions showed sufficient internal consistency when assessing whole items and (c) the correlation coefficients between pain intensities and total scores of the PCS among each sample of chronic pain varied across the studies.

Conclusions: These results indicate that the total score of the PCS could be compared across each translated version, however, caution is warranted when each subdomain of the PCS is compared between countries.

Significance: Although the pain catastrophizing scale has been translated into approximately 20 languages, methodological quality during their translation process has not been systematically assessed. We found that all languages versions showed sufficient internal consistency when assessing whole items, however, there were inconsistent structural models among each translated version, leading to variant subdomain structures for rumination, magnification and helplessness.