The review searched four databases for machine learning models predicting outcomes after hip fracture, identifying 40 studies published between 2010 and 2026. Death was the most studied outcome, while delirium, complications, resource use, rehabilitation and surgical failure were examined less consistently. Models generally discriminated moderately to strongly, but external validation was inconsistent, calibration rarely reported and uncertainty estimates incomplete.
Why it is interesting: Death after hip fracture is a well-worked prediction task, yet validation and calibration reporting restrict clinical use.
Cite this item
BibTeX
Free sign-up. A digest of the new papers twice a week, grouped by domain.