Artificial Intelligence and Machine Learning for Outcome Prediction After Osteoporotic Hip Fracture: A Systematic Review and Meta-analysis of Prediction Model Performance.

purpose of reviewHip fracture is a major consequence of osteoporosis and is associated with mortality, complications, prolonged hospitalisation and loss of independence. This systematic review evaluated artificial intelligence (AI) and machine learning (ML) models for predicting clinical outcomes after hip fracture, focusing on model performance, validation, reporting quality and clinical readiness. RECENT

findingsThis PRISMA-compliant review was registered with PROSPERO (CRD420261414142). Searches of PubMed, Ovid MEDLINE, Embase and EBSCO CINAHL identified 40 eligible studies published between 2010 and 2026. Mortality prediction was the most frequently studied outcome, while delirium, medical complications, resource utilisation, rehabilitation outcomes and surgical failure were less consistently examined. AI/ML models generally showed moderate to strong discrimination, particularly for mortality prediction. Current AI/ML models show promise for outcome prediction after hip fracture, but inconsistent external validation, limited calibration reporting, incomplete uncertainty estimates and variable methodological quality restrict clinical implementation. Future research should prioritise transparent reporting, external validation and prospective clinical evaluation.