An evaluation of three validated comorbidity indices to predict short-term postoperative outcomes after prosthetic urologic surgery.

Abstract

Commonly used comorbidity indices include the Charlson Comorbidity Index (CCI), Elixhauser/Van Walraven Index (VWI), and modified frailty index (mFI). This study evaluates whether these indices predict postoperative readmissions and complications after inflatable penile prosthesis (IPP) and artificial urinary sphincter (AUS) placement. We identified adult males who underwent IPP or AUS placement using the State Inpatient and State Ambulatory Surgery and Services Databases for Florida (2010-2015) and California (2010-2011). CCI, VWI, and mFI scores were calculated for each patient. We extracted 30-day emergency department services, 30-day readmissions, 90-day device complications (e.g., removal, replacement, or infection), and 90-day postoperative complications (excluding device complications). Receiver-operating characteristic curves were constructed and areas under the curve (AUC) were compared between the indices using the VWI as the reference model. We considered an AUC < 0.7 to represent poor predictive power. We identified 4242 IPP and 1190 AUS patients. All three indices had AUCs and 95% confidence intervals less than 0.70 for all outcomes following IPP and AUS placement making these indices poor predictors for postoperative outcomes. There were no significant differences in predicting 90-day postoperative complications between the VWI (AUC = 0.59, 95% CI [0.54-0.63]), CCI (AUC = 0.59, 95% CI [0.54-0.63], p = 0.99), and mFI (AUC = 0.60, 95% CI [0.55-0.66], p = 0.53) for IPPs and VWI (AUC = 0.54, 95% CI [0.47-0.61]), CCI (AUC = 0.50, 95% CI [0.43-0.57], p = 0.30), and mFI (AUC = 0.52, 95% CI [0.43-0.60], p = 0.56) for AUS placements. All three comorbidity indices were poor predictors of readmissions and complications following urologic prosthetic surgeries. A better comorbidity index is needed for risk-stratification of patients undergoing these surgeries.