The abstract is read at the publisher; the summary is JClub's.
Machine learning models using US Army periodic health assessment data predict 24-month suicide risk with AUROC 0.72 and meaningful elevation only in the top 5% of predicted PHAs.
- Why it matters: Suicide is the leading cause of death among active-duty US Army soldiers, and preventive skills training needs targeting to improve cost-effectiveness.
- What they did: Using 2015–2019 PHA administrative records, the team trained and tested models on 668,684 soldiers with a 24-month risk horizon, evaluating suicide via National Death Index outcomes.
- The result: In the test sample, suicide prevalence was meaningfully elevated only in the highest 5% of predicted risk (sensitivity 18.6%), enabling risk-stratified identification for intervention eligibility.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.