Using the capture-recapture method to ascertain the incidence of musculoskeletal injuries among elite military personnel.

Introduction: Injury epidemiology research traditionally uses a single musculoskeletal injury (MSI) data source, such as injury self-report (ISR) or medical record-review (MRR). MSI data obtained from ISR may be incomplete due to recall issues, and MRR data may be incomplete if medical attention was not sought, which is especially problematic in military populations. The objective of this analysis was to determine the ascertainment-corrected incidence of MSI among Special Forces (SF) military personnel using a novel statistical technique, capture-recapture (CRC).

Methods: This descriptive cross-sectional study used two sources of MSI data (ISR and MRR) collected at SF military installations. The CRC analysis, using the Chapman modification of the Lincoln-Petersen estimator, was conducted to assess the completeness of MSI data and to estimate the 1-year ascertainment-corrected cumulative MSI incidence.

Results: MSI data were available for 335 SF personnel (age: 28.9±6.2 years). At the time the study was conducted, only men could qualify as SF personnel. The percentage of participants identified as injured were 31.3% in the ISR, 26.6% in the MRR and 14.0% in both sources. The CRC estimate of the number of injured participants was 197.8 (95% CI 169.7, 225.8), with a CRC estimated 1 year cumulative incidence of 59.0% (95% CI 50.7%, 67.4%). Overall ascertainment was good (74.3%), but was relatively lower for spine MSI (49.0%) as compared with lower (73.7%) or upper (72.7%) extremity MSI. Ascertainment was highest for fractures (79.5%) versus strain (60.0%) or sprain (68.3%).

Conclusions: This was the first study to apply the CRC technique to estimate MSIs among SF personnel, who are at a very high risk of sustaining MSIs. Completeness of the MSI data varied by the MSI anatomic location and type. Further research into the application of the CRC technique in military populations is warranted, to provide an accurate estimate of the MSI burden.