Capture-recapture estimation of unreported work-related musculoskeletal disorders in Connecticut.

Background: Estimates of the extent of musculoskeletal disorders (MSD) are usually based upon workers' compensation reports, although recent reports indicate that there may be widespread under-reporting of MSD.

Methods: An estimate of the incidence of arm and hand work-related MSD was made using capture-recapture analysis of the overlap between state workers' compensation reports and physician reports in Connecticut for 1995. The resulting estimate was compared to a population-based survey of MSD.

Results: There was very small overlap between the two state injury reporting systems: 6.7% of 793 reported workers' compensation cases, or 8% of 661 physician's reports. The estimate for MSD not captured by either system was 13,285, resulting in 14,686 (95% CI: 9,733-18,453) total reported and non-reported cases. This compares to an estimate of 13,775 cases (95% CI: 8,800-18,800) based on a phone survey.

Conclusions: This analysis points to substantial under-reporting of MSD in Connecticut: estimates of unreported cases exceed those officially reported by a factor of 11:1. The findings have an important bearing on injury prevention programs and policy making.