RESUMO
BACKGROUND: Roland-Morris Disability Questionnaire for general pain (RMDQ-g) is an instrument adapted to assess disability in patients with pain in any region of the body. OBJECTIVE: To perform the structural and criterion validity of the RMDQ-g in Brazilian patients with chronic pain. DESIGN: A cross-sectional study. METHODS: We included native speakers of Brazilian Portuguese, of both sexes, aged ≥18 years, with pain for at least 3 months in any region of the body. Participants eligible for the study responded to an online form containing personal and clinical data, and assessment instruments. We used the confirmatory factor analysis and considered the following fit indices: chi-square/degree of freedom (DF), comparative fit index (CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA). In the comparison between models, we considered the structure with the lowest values of the Akaike information criterion (AIC) and sample-size adjusted Bayesian information criterion (SABIC). We assessed criterion validity via Spearman's correlation coefficient (rho) to correlate the long and short versions. RESULTS: The study consisted of 297 participants with chronic pain. The main sites of pain were the lumbar region (40.7%), thoracic (21.5%), and neck (19.5%). Mean pain intensity was greater than 5 points. The 24-item long version and the 15-item short version had adequate fit indices (chi-square/DF ≤ 1.77, CFI ≥0.97, TLI ≥0.96, and RMSEA ≤0.05). However, when comparing structures, the short version was the most appropriate because it had the lowest values of AIC (2562.05) and SABIC (2577.72). Criterion validity was acceptable (rho = 0.94) and internal consistency as well (Cronbach's alpha = 0.87). CONCLUSION: The structural validity and criterion validity of the RMDQ-g with one domain and 15 items is the most appropriate version and should be considered in the clinical environment and in research for measuring disability in patients with chronic pain in any region of the body.
