External Validation of an Injury Risk Prediction Algorithm in US Army Paratroopers

BACKGROUND: Musculoskeletal injuries (MSKIs) pose a major threat to military readiness, affecting nearly 40% of personnel annually, driving over 5 million medical visits and 24 million limited duty days. Validated methods for identifying and preventing injuries are a priority for the Department of Defense. PURPOSE: To externally validate a previously developed prediction model for identifying MSKI risk in a U.S. Army airborne infantry brigade. STUDY DESIGN: Prospective cohort; model validation study. METHODS: U.S. Army paratroopers newly assigned to the 173rd Airborne Brigade (Nov 2022-Aug 2023) underwent baseline screening with a published risk prediction algorithm. The primary outcome was a duty-limiting MSKI within one year, identified through the Army's e-Profile database from the Medical Operations Data System. Model performance was assessed for discrimination, calibration, calibration-in-the-large, overall fit, and decision curve analysis, with internal validation using 2,000 bootstrap interactions. Performance was compared with the original model. RESULTS: 308 paratroopers were prospectively followed for one year (93,636 military exposure days), during which 30.7% sustained a duty-limiting musculoskeletal injury. The model demonstrated poor discrimination (AUC 0.59, 95% CI 0.52-0.66) and only marginal net benefit over a treat-all strategy. CONCLUSIONS: The model failed to discriminate injury risk (AUC 0.59) and did not provide reliable calibration or meaningful net benefit. This model does not justify implementation for injury risk screening in this specific setting.