BACKGROUND: At radical cystectomy (RC), continent urinary diversion (CUD) provides functional outcomes that most closely approximate that of a native bladder. We tested the hypothesis that patients treated at high RC caseload hospitals and/or by high RC caseload surgeons have higher CUD rates. METHODS: We identified 9,493 bladder cancer patients treated with RC between 1998 and 2007, within the Nationwide Inpatient Sample. Univariable and multivariable analyses tested the relationship between hospital and surgical caseload at RC, and CUD rate. Generalized estimating equations models were used to adjust for clustering among hospitals and surgeons. RESULTS: Only 8% of patients received a CUD at RC. The CUD rate was 5 vs. 7 vs. 13% for low versus intermediate versus high annual hospital caseload (AHC) tertiles (P < 0.001). The CUD rate was 6 vs. 10 vs. 16% for low versus intermediate versus high annual surgical caseload (ASC) tertiles (P < 0.001). In multivariable analyses, and after adjusting for clustering, ASC emerged as independent predictors of CUD rate (P < 0.001), while AHC failed to achieve the independent predictor status for the same end point (P ≥ 0.1). CONCLUSIONS: Our findings indicate that CUD is performed in a minority (8%) of RC patients. Surgical caseload represents an important determinant of CUD rate, while hospital caseload failed to achieve independent predictor status. Efforts should be made to optimize CUD rate a RC.
BACKGROUND: At radical cystectomy (RC), continent urinary diversion (CUD) provides functional outcomes that most closely approximate that of a native bladder. We tested the hypothesis that patients treated at high RC caseload hospitals and/or by high RC caseload surgeons have higher CUD rates. METHODS: We identified 9,493 bladder cancerpatients treated with RC between 1998 and 2007, within the Nationwide Inpatient Sample. Univariable and multivariable analyses tested the relationship between hospital and surgical caseload at RC, and CUD rate. Generalized estimating equations models were used to adjust for clustering among hospitals and surgeons. RESULTS: Only 8% of patients received a CUD at RC. The CUD rate was 5 vs. 7 vs. 13% for low versus intermediate versus high annual hospital caseload (AHC) tertiles (P < 0.001). The CUD rate was 6 vs. 10 vs. 16% for low versus intermediate versus high annual surgical caseload (ASC) tertiles (P < 0.001). In multivariable analyses, and after adjusting for clustering, ASC emerged as independent predictors of CUD rate (P < 0.001), while AHC failed to achieve the independent predictor status for the same end point (P ≥ 0.1). CONCLUSIONS: Our findings indicate that CUD is performed in a minority (8%) of RC patients. Surgical caseload represents an important determinant of CUD rate, while hospital caseload failed to achieve independent predictor status. Efforts should be made to optimize CUD rate a RC.