RESUMO
BACKGROUND: Studies have addressed the effect of multimodal pain rehabilitation (MMR), whereas criteria for selection are sparse. This study examines whether higher scores on musculoskeletal pain measures are associated with selection to MMR, and whether this differs across gender. METHOD: A clinical population of 262 male and 589 female patients was recruited consecutively during 3 years, 2007-2010. The patients were referred from primary care to a pain rehabilitation clinic in Northern Sweden for assessment and selection to MMR. Register-based data on self-reported pain were linked to patients' records where outcome (MMR or not) was stated. We modelled odds ratios for selection to MMR by higher scores on validated pain measures (pain severity, interference with daily life, pain sites and localized pain vs. varying pain location). Covariates were age, educational level and multiple pain measures. Anxiety and depression (Hospital, Anxiety and Depression Scale) and working status were used in sensitivity tests. RESULTS: Higher scores of self-reported pain were not associated with selection to MMR in multivariate models. Among women, higher scores on pain severity, pain sites and varying pain location (localized pain = reference) were negatively associated with selection to MMR. After adjustment for multiple pain measures, the negative odds ratio for varying location persisted (OR = 0.59, 95% CI = 0.39-0.89). CONCLUSION: Higher scores on self-reported pain did not guide selection to MMR and a negative trend was found among women. Studies of referral patterns and decision processes may contribute to a better understanding of the clinical practice that decides selection to MMR.
