Geriatric acetabular fractures are becoming increasingly common. Surgeons must balance the long-term benefits of surgery with the risk of postoperative complications. Several risk stratification models have been adapted to assist surgeons with this decision-making. We compared the accuracy of the Elixhauser Comorbidity Measure (ECM), the Charlson Comorbidity Index (CCI), and the Combined Comorbidity Score (CCS) for predicting adverse events and postoperative discharge destination after surgical treatment of geriatric patients with acetabular fractures.
A search of the National Inpatient Sample for patients over the age of 65 yr who had fixation of an acetabular fracture between 2002 and 2014 was undertaken. Logistic regression models of basic demographic variables and the ECM, CCI, or the CCS were used to predict inpatient mortality, complications, extended length of stay, and discharge disposition. The predictive discrimination of each model was evaluated using the C-statistic.
A total of 2,497 patients were identified. The model using demographic variables and the CCS outperformed the corresponding ECM and CCI models, with an area under the curve (AUC) of 0.829 for mortality (compared to 0.791 and 0.689, respectively), 0.791 for cardiac complications (compared to 0.694 and 0.704, respectively), 0.789 for renal complications (compared to 0.787 and 0.683, respectively), and 0.760 for pulmonary complications (compared to 0.750 and 0.662, respectively).
The CCS was the best predictive model for assessing postoperative complications, followed by the ECM then CCI. Our results may assist in preoperative decision-making for geriatric patients with acetabular fractures.
Level of Evidence: