Responsiveness of five condition-specific and generic outcome assessment instruments for chronic pain

BACKGROUND: Changes of health and quality-of-life in chronic conditions are mostly small and require specific and sensitive instruments. The aim of this study was to determine and compare responsiveness, i.e. the sensitivity to change of five outcome instruments for effect measurement in chronic pain. METHODS: In a prospective cohort study, 273 chronic pain patients were assessed on the Numeric Rating Scale (NRS) for pain, the Short Form 36 (SF-36), the Multidimensional Pain Inventory (MPI), the Hospital Anxiety and Depression Scale (HADS), and the Coping Strategies Questionnaire (CSQ). Responsiveness was quantified by effect size (ES) and standardized response mean (SRM) before and after a four week in-patient interdisciplinary pain program and compared by the modified Jacknife test. RESULTS: The MPI measured pain more responsively than the SF-36 (ES: 0.85 vs 0.72, p = 0.053; SRM: 0.72 vs 0.60, p = 0.027) and the pain NRS (ES: 0.85 vs 0.62, p < 0.001; SRM: 0.72 vs 0.57, p = 0.001). Similar results were found for the dimensions of role and social interference with pain. Comparison in function was limited due to divergent constructs. The responsiveness of the MPI and the SF-36 was equal for affective health but both were better than the HADS (e.g. MPI vs HADS depression: ES: 0.61 vs 0.43, p = 0.001; SF-36 vs HADS depression: ES: 0.54 vs 0.43, p = 0.004). In the "ability to control pain" coping dimension, the MPI was more responsive than the CSQ (ES: 0.46 vs 0.30, p = 0.011). CONCLUSION: The MPI was most responsive in all comparable domains followed by the SF-36. The pain-specific MPI and the generic SF-36 can be recommended for comprehensive and specific bio-psycho-social effect measurement of health and quality-of-life in chronic pain.