RESUMO
INTRODUCTION: Identifying patients at risk after cystectomy for cancer is essential. The POSPOM score is a non-specific urological surgery score for estimating postoperative hospital mortality. This study sought to validate the POSPOM score for predicting postoperative morbidity and mortality after cystectomy. METHODS: The study retrospectively included all patients undergoing cystectomy for muscle-invasive or locally advanced bladder cancer between 2010 and 2019 in one center. The primary objective was validation of the POSPOM score for calculating severe postoperative morbidity [Clavien-Dindo (CDC)≥3] and 90-day mortality after cystectomy. Secondary objectives were comparison to other predictive scores [Charlson (CCI), ASA]. RESULTS: At 90days, out of 167 patients, 26% (n=44) had a CDC≥3 complication and 8.4% (n=14) had died. POSPOM correlated with the risk of death at 90days (P<0.001) and postoperative transfusion (P<0.01). Patients with CDC≥3 complications had higher CCI and POSPOM (median 6.5 vs. 5, P<0.01 and 6.49% vs. 5.58%, P=0.029, respectively). Patients who died postoperatively had higher CCI and POSPOM (median 8 vs. 6, P<0.001 and 23.9% vs. 5.58%, P<0.001, respectively). The prognostic value of the POSPOM score for predicting mortality appears better [AUC=0.886 (0.798-0.973)] compared with CCI [AUC=0.812 (0.710-0.915)] and ASA [AUC=0.739 (0.630-0.849)], but not for predicting morbidity. CONCLUSION: This study confirms the robustness of the POSPOM score for estimating mortality and its limitations for predicting postoperative morbidity.
