Common comorbidity indices fail to predict short-term postoperative outcomes following male urethroplasty.

ABSTRACT

PURPOSE: To evaluate whether commonly used comorbidity indices [Charlson Comorbidity Index (CCI), Van Walraven Index (VWI), and modified frailty index (mFI)] predict postoperative readmissions and complications after urethroplasty. METHODS: Patients undergoing urethroplasty for urethral stricture from the State Inpatient Database and State Ambulatory Surgery and Services Database for Florida (2010-2015) and California (2010-2011) were identified. We calculated CCI, VWI, and mFI scores for each patient. We extracted the following adverse outcomes: 30 day ER services, 30 day inpatient readmissions, and 90 day Clavien-Dindo III-V complications. We constructed receiver operating characteristic (ROC) curves and compared area under the curve (AUC), using the VWI as the reference index. RESULTS: We identified 908 urethroplasty patients. Among these patients, 11.5% (n = 104) of patients had a complication, with 4.8% (n = 44) specifically having 30-day ER services, 6.2% (n = 56) having 30 day readmissions, and 9.0% (n = 82) having 90-day Clavien-Dindo III-V complications. ROC curves demonstrated poor predictive performances for all four indices as no index achieved an AUC > 0.70. The indices performed similarly poorly but the mFI-DX was particularly poor at predicting 90 day Clavien-Dindo III-V complications (AUC = 0.49; 95% CI 0.43-0.55; p < 0.01). CONCLUSIONS: The CCI, VWI, and mFI have poor ability to identify patients who had an adverse event after urethroplasty. Our results support the need for a urology-specific comorbidity index to better identify at-risk patients undergoing urethroplasty.