Primary objective: To test the usefulness of a method to improve the measurement of prevalent mild traumatic brain injury (MTBI) among injured workers with a workers compensation claim.
Methods: Database codes were selected to identify MTBI cases in the Ontario workers compensation lost-time claims database. A random sample of 210 claims was selected, classified as MTBI or not, and used to calculate proportions with MTBI among code groups. The annual prevalence of MTBI in 1997 and 1998 was calculated by weighting the numerators with the appropriate proportions of MTBI within each code group.
Results: Four code groups were created: the head region, cranial region, concussion code group and the brain region. The proportion of MTBI in each group was 29%, 19%, 92% and 32%, respectively. The 1997 prevalence depended on the codes used, from 39/10,000 (95% confidence interval (CI): 35-44) for a weighted version of the 'concussion' code to 58/10,000 (95% CI: 50-65) for inclusion of all identified MTBI codes.
Conclusions: Restricting the enumeration of MTBI to specific 'concussion' codes can lead to under-estimation of the prevalence of MTBI in epidemiological studies using workers compensation data. Approximately six out of every 1000 lost-time claims are associated with MTBI. Given lost-time estimates of disability under-estimate the prevalence of this mild injury, MTBI, is an important workplace injury.