Self-rated work ability predicts return to work after occupational rehabilitation.

The study used absence of benefit payments for one month as a proxy for full return to work, so it could not see part-time, graded or sustained work participation. It was retrospective and did not include diagnosis, workplace conditions, or specific health problems, so it could not test whether the association differed across those factors. Questionnaire data were missing for some variables and were not imputed, so different analyses used different participant totals. Participants were selected for complex work and health problems after simpler interventions had failed, so the results may not apply to all sick-listed workers. The programs included several health and work interventions together, so the study cannot say which component changed work ability or return to work.

Do not read the Work Ability Score change as proof that rehabilitation caused return to work. The study was observational, and participants received several health and work interventions together, so it cannot show which part of the program mattered.