Using the capture-recapture method to estimate the incidence of musculoskeletal injuries among U.S. Army soldiers

Musculoskeletal injury (MSI) data typically are obtained from medical chart-review (MCR) or injury self-reports (ISR). MSI incidence may be under-counted if only one source is utilized, as MCR will not capture MSI for which medical care was not sought, and ISR may be affected by issues with recall.

Objectives: The purposes of this study were to determine MSI incidence from two sources (MCR, ISR) and to estimate the incidence, after accounting for the under-counting in both sources, among a sample of U.S. Army soldiers.

Design: Descriptive cross-sectional study.

Methods: The estimated cumulative incidence during a one-year period was calculated from the two sources of MSI data using a novel statistical analysis (capture-recapture-CRC).

Results: MSI data were available for 287 soldiers (age: 27.5±6.3years (mean±standard deviation)). The one-year cumulative incidence of MSI was 17.8% (MCR), 19.5% (ISR), and 54.0% (CRC). CRC analysis showed that there was under-counting from both sources of data and the percent of CRC estimated MSI observed were 32.9% (MCR), 36.1% (ISR), and 57.4% (MCR and ISR combined). When analyzed by MSI type, percent of CRC estimated MSI counted from both sources was highest (75.0%) for fracture, followed by sprain (53.8%), strain (43.8%), and pain/spasm/ache (35.8%).

Conclusions: There was under-counting of MSI from both sources of data, and the under-counting varied by MSI type. There is a need for further investigation of the relative benefits of various sources of MSI data and the application of the capture-recapture analysis in military populations.