Validation of the Preoperative Score to Predict Postoperative Mortality in Patients Undergoing Radical Cystectomy

Standardized prediction of perioperative mortality risk is of major clinical concern in the radical cystectomy setting. We validated the recently developed Preoperative Score to Predict Postoperative Mortality (POSPOM) in a sample of 1083 consecutive cystectomy patients treated between 1993 and 2014. POSPOM was calculated as originally described based on age and 13 further parameters; three parameters which were not available in our database were ignored. Thirty-day and 90-d mortality were 1.0% and 4.1%, respectively. The areas under the receiver operator characteristic curves were 0.86 for 30-d mortality and 0.78 for 90-d mortality. Below the median of 27 POSPOM risk points, 30-d mortality was 0% and 90-d mortality was 0.5%. Above this level, the corresponding figures were 1.7% and 6.5%, respectively. The 30-d (p<0.0001) and even the 90-d mortality rates (p=0.004) were lower than the POSPOM-predicted in-hospital mortality rate for this sample (5.8%). Nevertheless, with its good discriminative accuracy, POSPOM might standardize the prediction of postoperative mortality after radical cystectomy. The absolute mortality figures in a high volume academic center were, however, lower than predicted based on nationwide collected data. PATIENT SUMMARY: With a good discriminative accuracy, Preoperative Score to Predict Postoperative Mortality might standardize the prediction of postoperative mortality after radical cystectomy. The absolute mortality figures in a high volume academic center were, however, lower than predicted based on nationwide collected data.