Effect of surgical treatment on mortality after acetabular fracture in the elderly: A multicenter study of 454 patients

Objectives: Controversy exists regarding the effect of operative treatment on mortality after acetabular fracture in elderly patients. Our hypothesis was that operative treatment would confer a mortality benefit compared with nonoperative treatment even after adjusting for comorbidities associated with death. Design: Retrospective study. Setting: Three University Level I Trauma Centers. Patients/Participants: All patients aged 60 years and older with acetabular fractures treated from 2002 to 2009 were included in the study. Four hundred fifty-four patients were identified with an average age of 74 years. Sixty-seven percent of the study group was male and 33% female. Intervention: One of 4 treatments: nonoperative management with early mobilization, percutaneous reduction and fixation, open reduction and internal fixation, acute total hip arthroplasty. Main Outcome Measurements: Kaplan-Meier survival curves were created, and Cox proportional hazards models were used to calculate unadjusted and adjusted hazard ratios (HRs) for covariates of interest. Results: In contrast to previous smaller studies, the overall mortality was relatively low at 16% at 1 year [95% confidence interval (CI), 13-19]. Unadjusted survivorship curves suggested higher 1-year mortality rates for nonoperatively treated patients (21% vs. 13%, P < 0.001); however, nonoperative treatment was associated with other risk factors for higher mortality. By accounting for these patient risk factors, our final multivariate model of survival demonstrated no significant difference in hazard of death for nonoperative treatment (0.92, P = 0.6) nor for any of the 3 operative treatment subgroups (P range, 0.4-0.8). As expected, we did find a significantly increased hazard for factors such as the Charlson comorbidity index [HR, 1.25 per point (95% CI, 1.16-1.34)] and age [HR, 1.08 per year of age more than 70 years (95% CI, 1.05-1.11)]. In addition, associated fracture patterns (compared with elementary patterns) significantly increased the hazard of death with a ratio of 1.51 (95% CI, 1.10-2.06). Conclusions: The operative treatment of acetabular fractures does not increase or decrease mortality, once comorbidities are taken into account. The reasons for this are unknown. Regardless of the causes, the decision for operative versus nonoperative treatment of geriatric acetabular fractures should not be justified based on the concern for increased or decreased mortality alone.