The role of surgical experience in patient selection, surgical quality, and outcomes in robot-assisted radical cystectomy

Abstract

Background: Robot-assisted radical cystectomy (RARC) remains one of the most complex urological procedures. Due to regionalization of bladder cancer care, there is likely an imbalance in experience among urologists performing RARC. We sought to describe changes in patient selection, surgical quality surrogates and rates of complications in relation to surgical experience.

Methods: We retrospectively reviewed 409 consecutive patients with bladder cancer who underwent RARC between 2006 and 2017 by a single surgeon. The cohort was divided into 4 quartiles (Q1-Q4) according to surgical experience, based on the chronologic order at which RARC was performed. Baseline, perioperative and pathologic characteristics of patients were compared among the 4 groups. 30-day and 90-day complications were assessed using the Clavien-Dindo system. The association between surgical experience (quartile) and complications was assessed using multivariable logistic regression analyses.

Results: Median age (interquartile range [IQR] from 70-73 years), body mass index (IQR from 25 to 27 kg/m²) and preoperative glomerular filtration rate (IQR from 59 to 65 ml/min) were similar among all quartiles (all P > 0.05). Patients in Q4 had higher rates of previous abdominopelvic surgery (46.1% vs. 30.4%, P = 0.031) and American Society of Anesthesiologists score of 3 to 4 (72.3% vs. 47.1%, P = 0.003) compared to patients in Q1. Patients who underwent RARC in Q4 compared to Q1, had less estimated blood loss (250 ml vs. 350 ml, P < 0.001), shorter operative time (346 vs. 360 minutes, P < 0.001), and higher lymph node yield (22 vs. 17 nodes, P < 0.001). The 30-day and 90-day complication rates were 53% and 62%, respectively. Thirty-day complication rates were similar among all 4 quartiles (P > 0.05), but higher among patients in Q4 compared to Q1 within 90 days (74% vs. 54%, P = 0.01). On multivariable analysis, patients in Q4 were more likely to experience any 90-day complication (OR 2.03, 95%Cl 1.11-3.70) compared to Q1.

Conclusion: Our results show that with surgical experience, more complex cases can be performed while continuing to improve surgical quality. Nonetheless, there appears to be a trade-off between the increase in complexity of cases performed with experience and accepting higher rates of complications.

Keywords: Bladder cancer; Complications; Cystectomy; Learning curve; Robotics.