Real-world implementation of robot-assisted radical cystectomy at four Japanese institutions: perioperative outcomes, major complications, and mortality.

To describe the implementation of robot-assisted radical cystectomy (RARC) at four Japanese institutions, focusing on perioperative outcomes, major complications (Clavien-Dindo grade ≥3), and all-cause mortality within 90 days. We retrospectively analyzed 519 consecutive RARCs for bladder cancer (October 2014-March 2026). Each institution's series was divided chronologically into early (n = 261) and late (n = 258) halves to describe changes in practice (exploratory). Factors associated with major complications were assessed by logistic regression including institution. Survival among 90-day survivors (landmark analysis) was a secondary, exploratory outcome analyzed with Cox models stratified by institution. Within 90 days, 64 patients (12.3%) had a major complication (early 14.6% vs. late 10.1%; adjusted odds ratio 0.67 (95% CI 0.39-1.14)), and 13 (2.5%) died from any cause; six deaths (1.2%) were classified as complication-related. Patients who died were older and had poorer performance status. Pelvic lymph node dissection decreased from 95.4% to 83.3% (pNx, 4.6% to 16.7%), mainly at one institution, and pT3-4 disease decreased from 35.8% to 23.8%. Among 90-day survivors, the late phase was not significantly associated with overall survival (hazard ratio 0.70, 95% CI 0.45 to 1.08), cancer-specific survival (0.65), or recurrence-free survival (0.74), and major complications were not associated with survival. In this multicenter series, about one in eight patients had a major complication and 2.5% died within 90 days. Differences between the early and late groups in tumor characteristics, lymph node dissection, and follow-up preclude attributing any difference in outcomes to surgical experience.