Determinants of mortality after hip fracture surgery in Sweden: a registry-based retrospective cohort study

Abstract

Surgery for hip fractures is associated with high mortality and morbidity. The causes of poor outcome are not fully understood and may be related to other factors than the surgery itself. The relative contributions of patient, surgical, anaesthetic and structural factors have seldom been studied together. This study, a retrospective registry-based cohort study of 14 932 patients undergoing hip fracture surgery in Sweden from 1st of January 2014 to 31st of December 2016, aimed to identify important predictors of mortality post-surgery. The independent predictive power of our included variables was examined using Cox proportional hazards modeling with all-cause mortality at longest follow-up as the outcome. Twelve independent variables were considered as interrelated ‘exposures’ and their individual adjusted effect within a single model were evaluated. Kaplan-Meier curves were also generated. Crude mortality rates were 8.2% at 30 days (95% CI 7.7–8.6%) and 23.6% at 365 days (95% CI 22.9–24.2%). Of the 12 factors entered into the Cox regression analysis, age (aHR1.06, p < 0.001), male gender (aHR 1.45, p < 0.001), ASA-PS-class (ASA 1&2 reference; ASA 3 aHR 2.12; ASA 4 aHR 4.79; ASA 5 aHR 12.57 respectively, p < 0.001) and PACU-LOS (aHR 1.01, p < 0.001) were significantly associated with mortality at longest follow-up (up to 3 years). University hospital status was protective (aHR 0.83, p < 0.001) in the same model. Age, gender and ASA-PS-class were strong predictors of mortality after surgery for hip fractures in Sweden. University hospital status and length of stay in the postoperative care unit were also identified as modifiable risk factors after multivariable adjustment and require confirmation in future studies.