RESUMO
Purpose Chronic musculoskeletal pain can have a major impact on ability to work. The work ability score is a commonly used single-item question to assess work ability but has not been fully validated yet. The aim of the present study was to evaluate test-retest reliability, agreement, construct validity, and responsiveness of the work ability score among sick-listed workers with chronic musculoskeletal pain. Methods Data of sick-listed workers with chronic musculoskeletal pain was routinely collected at seven rehabilitation centres in the Netherlands. Assessments included a set of questionnaires, administered at admission and discharge from a fifteen-week vocational rehabilitation program. Test-retest reliability was determined with the intraclass correlation coefficient. For agreement, the standard error of measurement and smallest detectable changes were calculated. Construct validity was assessed by testing hypotheses regarding Spearman rank correlation coefficient. Area under the curve obtained from the receiver operating characteristic curve and minimal clinically important change were determined for the total sample and work ability score baseline tertile groups to assess responsiveness. Results In total, 34 workers were analyzed for reliability and agreement, 1291 workers for construct validity, and 590 responded to the responsiveness questionnaire. Reliability reached an intraclass correlation coefficient of 0.89; 95% CI 0.77-0.94, a standard error of measurement of 0.69 points, and the smallest detectable change of 1.92 points. For construct validity, six of the seven predefined hypotheses were not refuted. The area under the curve was 0.76 (95% CI 0.71-0.81) allowing for discrimination between stable and improved workers, with a minimal clinically important change of 2.0 points for the total sample. Conclusion The work ability score showed good measurement properties among sick-listed workers with chronic musculoskeletal pain.
