Development of a new comprehensive preoperative risk score for predicting 1-year mortality in patients with hip fracture: the HULP-HF score. Comparison with 3 other risk prediction models.

Abstract

The aim of this study was to develop a new comprehensive preoperative risk score for predicting mortality during the first year after hip fracture (HF) and its comparison with 3 other risk prediction models. All patients admitted consecutively with a fragility HF during 1 year in a co-managed orthogeriatric unit at a university hospital were assessed and followed for 1 year. Factors independently associated with 1-year mortality were used to create the HULP-HF (Hospital Universitario La Paz - Hip Fracture) score. The predictive validity, discrimination and calibration of the HULP-HF score, the American Society of Anesthesiologists (ASA) scale, the abbreviated Charlson comorbidity index (a-CCI) and the Nottingham Hip Fracture score (NHFS) were compared. Discriminative performance was assessed using the area under the curve (AUC) and calibration by the Hosmer-Lemeshow goodness-of-fit-test. 509 patients were included. 1-year mortality was 23.2%. The 8 independent mortality risk factors included in the HULP-HF score were age >85 years, baseline functional and cognitive impairment, low body mass index, heart disease, low hand-grip strength, anaemia on admission, and secondary hyperparathyroidism associated with vitamin D deficiency. The AUC was 0.79 in the HULP-HF score, 0.66 in the NHFS, 0.61 in the abbreviated CCI and 0.59 in the ASA scale. The HULP-HF score, the NHFS and the abbreviated CCI all presented good levels of calibration (p > 0.05). The HULP-HF score has a predictive capacity for 1-year mortality in HF patients slightly superior to that of other previously existing scores.