External validation of the U-HIP prediction model for in-hospital mortality in geriatric hip fracture patients

•The objective was to externally validate the U-HIP prediction model for in-hospital mortality in geriatric hip fracture patients.•In this large international cohort, the model showed a good discrimination and fair calibration.•This model is freely available online and can be used to predict the risk of mortality, identify high-risk patients, and aid clinical decision making. Identification of high-risk hip fracture patients in an early stage is vital for guiding surgical management and shared decision making. To objective of this study was to perform an external international validation study of the U-HIP prediction model for in-hospital mortality in geriatric patients with a hip fracture undergoing surgery. In this retrospective cohort study, data were used from The American College of Surgeons National Surgical Quality Improvement Program. Patients aged 70 years or above undergoing hip fracture surgery were included. The discrimination (c-statistic) and calibration of the model were investigated. A total of 25,502 patients were included, of whom 618 (2.4%) died. The mean predicted probability of in-hospital mortality was 3.9% (range 0%-55%). The c-statistic of the model was 0.74 (95% CI 0.72–0.76), which was comparable to the c-statistic of 0.78 (95% CI 0.71–0.85) that was found in the development cohort. The calibration plot indicated that the model was slightly overfitted, with a calibration-in-the-large of 0.015 and a calibration slope of 0.780. Within the subgroup of patients aged between 70 and 85, however, the c-statistic was 0.78 (95% CI 0.75–0.81), with good calibration (calibration slope 0.934). The U-HIP model for in-hospital mortality in geriatric hip fractures was externally validated in a large international cohort, and showed a good discrimination and fair calibration. This model is freely available online and can be used to predict the risk of mortality, identify high-risk patients and aid clinical decision making.