Hip fractures in South Africa: mortality outcomes over 12months post-fracture.

Abstract

The outcomes following hip fractures are unknown in sub-Saharan Africa. This study aimed to quantify the mortality rate (MR) following hip fractures and to identify predictors of mortality over 1year. In this cohort study, demographic, clinical, and biochemical characteristics of consecutive patients with low trauma hip fractures, admitted to the five public sector hospitals in eThekwini (formerly Durban), were recorded. Cox regression analyses identified predictors of mortality at 30days and 1year. In the 200 hip fracture patients studied, the mean age was 74.3years (SD ± 8.8) and 72% were female. Hospital presentation was often delayed, only 15.5% presented on the day of fracture. At admission, 69.5% were anaemic, 42% had hyponatraemia, 34.5% raised creatinine, and 58.5% hypoalbuminaemia. All received skin traction before 173 (86.5%) underwent surgical fixation. Median time from admission to surgery was 19.0days (IQR 12.3-25.0). Median hospital stay was 9.0days (IQR 12.3-25.0). Mortality rates were 13% and 33.5% at 30 and 365days, respectively. Over 1year, African patients were more likely to die than Indian patients (40.9 versus 30%, HR 11.5 [95% CI 1.51, 2.57]; p = 0.012); delays to surgery predicted death (HR 1.02 [95% CI (1.00, 1.04)]; p = 0.022). In multivariate analyses, death at 1year was most strongly predicted by an elevated serum creatinine (HR 2.43, 95% CI (1.02, 5.76), p = 0.044]. Hip fractures are associated with high MRs, in part explained by insufficient surgical capacity, highlighting the need for national efforts to improve hip fracture service provision.