Development of a nomogram predicting 90-day probability of severe complications in patients undergoing radical cystectomy

==inizio abstract==

INTRODUCTION AND OBJECTIVES: Despite the improvements in surgical technique and perioperative care, radical cystectomy (RC) is associated with significant incidence of perioperative complications also in contemporary series [1]. We developed a model predicting 30-day severe complication (grade ≥3) risk according to Clavien Classification System (CCS) to be used in the preoperative patients counseling.
METHODS: Data of 396 consecutive patients treated with RC and pelvic lymph node dissection at 17 institutions between April 2011 and March 2012 were prospectively collected.
Logistic regression analysis was used to predict 30-day severe complication risk, including the following variables: age, body mass index (BMI), American Society of Anaesthesiologists (ASA) score, hemoglobin levels (Hb), yearly cystectomy caseload and urinary diversion performed.
RESULTS: The overall incidence of 30-day severe complications was 15%.
Regression coefficients used to develop the nomogram were: age (OR 1.59, 95% CI 0.94-2.7), body mass index (OR 1.04, 95% CI 0.74-1.47), ASA score (OR 1.31, 95% CI 0.85-2.01), preoperative hemoglobin levels (OR 0.69, 95% CI 0.45-1.05), yearly cystectomy caseload (OR 0.73, 95% CI 0.54-0.99), and urinary diversion performed (ureterocutaneostomy: reference category; ileal conduit: OR 2.52, 95% CI 1.15-5.53; ileal neobladder: OR 6.04, 95% CI 2.21-16.5). The nomogram had a discrimination accuracy (c-index) of 0.68 and was well calibrated. The internal validation of the model with 200 bootstrap resamples demonstrated a discrimination accuracy of 0.64.
CONCLUSIONS: At the best of our knowledge we first developed a nomogram predicting 30-day severe complication risk according to CCS that may be a clinical tool to counsel patients in the preoperative setting about the potential impact of UD choice on perioperative outcomes. Larger cohorts and external validation are needed to confirm the clinical utility of the nomogram.

==fine abstract==