Predictors of long-term mortality in older people with hip fracture

To determine 1-year mortality and predisposing factors in older people who had surgery after a hip fracture.

Prospective cohort study.

Public acute hospital, trauma service.

Patients (N=281) aged ≥65 years who were admitted to the hospital with a hip fracture from January 2009 to January 2010, and followed up for 1 year thereafter.

Not applicable.

Cumulative survival probability up to 1 year from surgery was calculated by means of Kaplan-Meier charts, and Cox regression models were performed to analyze the factors associated with mortality. Data were collected from medical charts and by interviews. Health status was evaluated using the American Society of Anesthesiologists rating, prefracture functional level with the FIM, and cognitive status with the Pfeiffer score.

The 1-year mortality for the 281 patients who were followed up was 21% (95% confidence interval [CI], 16.1%-25.9%). A non-weight-bearing status was associated with increased mortality in unadjusted analyses (hazard ratio [HR]=1.99; 95% CI, 1.16-3.43), but 5 other factors were identified when entered into the multiple Cox regression model: age (HR=1.05; 95% CI, 1-1.09), male sex (HR=2.92; 95% CI, 1.58-5.39), low health status (HR=2.8; 95% CI, 1.29-6.09), low prefracture function (HR=.98; 95% CI, .97-.99), and change of residence (HR=3.21; 95% CI, 1.43-7.17).

The overall 1-year mortality rate was 21%. Change of residence is the only potentially modifiable risk factor, independent of the following other traditional risk factors that were found: age, sex, health status, and prefracture functional level. Furthermore, 2 to 4 weeks of non-weight-bearing status, which is considered modifiable, is also associated with increased mortality rates in unadjusted analyses.