Reliability and Validity of the Turkish Pain Behaviour Scale in Adults with Chronic Nonspecific Low Back Pain
Elif Esma Bayraktar, Irmak Çavuşoğlu, Yağmur İldeniz, Nuray AlacaBackground/Objectives: Pain behaviors observed during movement may complement self-reported and performance-based assessment in chronic nonspecific low back pain (NSLBP). The Pain Behaviour Scale (PaBS) quantifies clinician-observed pain behavior severity during functional tasks, but no Turkish version has been evaluated. This study translated and culturally adapted the PaBS into Turkish and examined its reliability, agreement, measurement error, and construct validity. Methods: This cross-cultural adaptation and psychometric validation study included 102 adults with chronic NSLBP. The PaBS was translated, back-translated, reviewed by an expert panel, and pilot tested. Participants completed clinical questionnaires and standardized physical performance tests. Two independent raters scored the PaBS at baseline; one rater repeated scoring after one week. Reliability was analyzed using ICCs. Measurement error, agreement, and construct validity were assessed using SEM, MDC95, percentage agreement, Cohen’s kappa, Bland–Altman analysis, and predefined correlation hypotheses. Results: Total-score interrater reliability was excellent (ICC = 0.95), and intrarater reliability was high (ICC = 0.96), although the latter should be interpreted cautiously because participant status changed between sessions. MDC95 values were 2.05 and 1.85. Individual behavior agreement ranged from 81.4% to 100.0%, but item-level findings should be supplementary. All construct-validity correlations were in the expected direction; however, most were stronger than predefined expectations, and only 1 of 13 hypotheses met both direction and magnitude. Conclusions: The Turkish PaBS appears reliable for assessing observed pain behavior severity during functional movement in adults with chronic NSLBP. Construct-validity findings should be considered preliminary because stronger-than-expected correlations may reflect construct overlap with disability, fear-avoidance, and physical performance.