Robotic versus open pediatric ureteral reimplantation: Costs and complications from a nationwide sample

Abstract

We sought to compare complications and direct costs for open ureteral reimplantation (OUR) versus robot-assisted laparoscopic ureteral reimplantation (RALUR) in a sample of hospitals performing both procedures. Anecdotal reports suggest that use of RALUR is increasing, but little is known of the outcomes and costs nationwide. The aim was to determine the costs and 90-day complications (of any Clavien grade) in a nationwide cohort of pediatric patients undergoing OUR or RALUR. Using the Premier Hospital Database we identified pediatric patients (age<21 years) who underwent ureteral reimplantation from 2003 to 2013. We compared 90-day complication rates and cost data for RALUR versus OUR using descriptive statistics and hierarchical models. We identified 17 hospitals in which both RALUR and OURs were performed, resulting in a cohort of 1494 OUR and 108 RALUR cases. The median operative time was 232min for RALUR vs. 180min for OUR (p=0.0041). Incidence of any 90-day complications was higher in the RALUR group: 13.0% of RALUR vs. 4.5% of OUR (OR=3.17, 95% CI: 1.46-6.91, p=0.0037). The difference remained significant in a multivariate model accounting for clustering among hospitals and surgeons (OR, 3.14; 95% CI, 1.46-6.75; p=0.0033) (Figure). The median hospital cost for OUR was $7273 versus $9128 for RALUR (p=0.0499), and the difference persisted in multivariate analysis (p=0.0043). Fifty-one percent (55/108) of the RALUR cases occurred in 2012-2013. We present the first nationwide sample comparing RALUR and OUR in the pediatric population. There is currently wide variation in the probability of complication reported in the literature. Some variability may be due to differential uptake and experience among centers as they integrate a new procedure into their practice, while some may be due to reporting bias. A strength of the current study is that cost and 90-day postoperative complication data are collected at participating hospitals irrespective of outcomes, providing some immunity from the reporting bias to which individual center surgical series' may be susceptible. Compared with OUR, RALUR was associated with a significantly higher rate of complications as well as higher direct costs even when adjusted for demographic and regional factors. These findings suggest that RALUR should be implemented with caution, particularly at sites with limited robotic experience, and that outcomes for these procedures should be carefully and systematically tracked.