Abstract

OBJECTIVES: To measure the incidence of back pain relapse (causing three consecutive days off work on medical advice) and of short sickness absence (less than three consecutive days), and to determine whether the incidence of such events was affected by overall pain and specific pain related to simple daily movements (functional capacity) assessed at discharge. METHODS: A cohort of workers with a first compensated episode of back pain was prospectively followed up from return to work after rehabilitative treatment. Follow up among 230 workers was carried out monthly by phone for a maximum of six months. Crude and adjusted rate ratios (RRs) along with 95% confidence intervals (95% CIs) were estimated with the Cox's proportional hazards model. RESULTS: Within six months of return to work, 29 workers (12.6%) had relapsed, and another 15 workers (6.5%) had a short sickness absence. 50% of relapses had occurred within 42 days of return to work whereas this figure was 28 days for short sickness absence. In a multivariate model that considered pain and clinical variables at discharge only a scale combining all pain variables (specific daily movements as well as the visual analog overall pain scale) contributed to relapse and short sickness absence as the outcome (RR (95% CI)) (1.53 (0.96-2.43)); the same was true in a model considering pain and workers' views on desired changes to work conditions (1.60; 1.08 to 2.36). CONCLUSIONS: Incidence of relapse or short sickness absence in the first six months after return to work was 19.1%. Of all measured prognostic variables (sociodemographic, clinical, workers' views, and pain), only overall pain and pain associated with carrying out simple daily movements were helpful in predicting relapse or short sickness absence.