Abstract

We identified 226 proximal femoral fractures prospectively in people aged over 60 years during one year in Stockport (population approximately 289,000). Our objective was to examine the association between primary treatment and outcome 6 months after the fracture using multivariate regression models. Higher haemoglobin concentrations and mental test scores on admission to hospital were associated with lower fatality within 6 months, whereas a high blood calcium-phosphate product, greater age, and an active medical problem at the time of the fracture were all associated with increased fatality. An intracapsular fracture, greater age, and a greater degree of dependency before the fracture all reduced the likelihood of regaining previous walking ability by 6 months. Good walking ability before the fracture increased the likelihood of walking well by 6 months. After adjusting for the effects of these extraneous variables there was no statistically significant association between treatment and outcome.