Predicting Subsequent Overuse Knee Injury Among Military Cadets and Midshipmen

This study presents a novel, internally validated prediction model for overuse knee injuries in a high-risk trainee population with prior knee injury. Subgrouping by prior injury status performed better than applying the model to the entire cohort, highlighting the potential efficiency of anatomically specific injury history as a first-level filter for development of injury prediction models. Although this specific study's model performance was moderate, the decision curve analysis supports its potential clinical utility for guiding targeted prevention efforts in military trainees.