Factors associated with mortality following hip fracture in Japan.

Various factors have been reported to increase the risk of death following hip fracture. However, our review of the literature indicates that previous studies were generally performed based on a rough classification of comorbidities. In this study, comorbidities were classified in detail, and the risk of death following hip fracture was investigated. Four hundred and eighty patients with hip fracture were enrolled. The patients' comorbidities and walking ability before injury were investigated using their own or their family's reports or their medical history, and the residences where the subjects were taken after discharge were recorded. Subsequently, the patients or their family were interviewed about whether they were alive or dead on January 1, 2002, by mail or telephone. A survival curve was drawn based on the Kaplan-Meier method. Cox proportional hazards regression models were used to determine the risk factors for death. An expected mortality rate for the Japanese population from 1991 to 2002 was obtained from a life table published by the Ministry of Health, Labour, and Welfare and compared to our observed mortality. The 1-year survival rate following hip fracture was 88.5%, which was a little lower than the expected survival rate. In subsequent years, the survival rate was lower than the expected survival rate. Being male and/or having a trochanteric fracture were risk factors for death. Patients who walked with a walker or other support or were nonambulatory before injury had an increased risk of death. Among the comorbidities, dementia, diabetes mellitus, and a history of gastrectomy or colonectomy were risk factors for death. Among the complications, pneumonia during hospitalization was a risk factor for death.