Robotic-Assisted Lobectomies in the National Cancer Database.

Abstract

Robotic-assisted thoracoscopic surgery (RobATS) lobectomy is becoming more common for the treatment of lung cancer. As with any relatively new technology, there is the assumption that greater experience leads to greater proficiency. The objective of this study was to analyze outcomes of patients undergoing RobATS lobectomy as hospitals gain experience, and to describe outcomes after conversion to open procedures. The National Cancer Database (NCDB) was used to analyze robotic lobectomies for lung cancer from 2010 to 2014. Individual hospitals were categorized by the year they began reporting robotic lobectomies to the NCDB. Primary outcomes were perioperative morbidity and mortality and rate of conversion to open lobectomy. There were 7,645 robotic lobectomies identified from 465 hospitals. The overall conversion rate was 9.2% (n = 702). A propensity-matched analysis showed no significant difference between experienced and inexperienced hospitals with respect to 30-day mortality (1.07% vs 2.03%, p = 0.092) or 90-day mortality (2.35% vs 3.63%, p = 0.104). Conversion to open was a predictor of 30-day mortality (odds ratio [OR] 2.54, CI 1.56 to 4.14) and 90-day mortality (OR 2.68, CI 1.83 to 3.91). Patients who underwent conversion had higher 90-day mortality compared with patients not undergoing conversion, in years of experience: 2 (p = 0.043), 3 (p = 0.002), and 4 (p = 0.003). Mortality after RobATS lobectomy at experienced hospitals is not significantly different than at inexperienced hospitals. Though conversion rates decrease with experience, patients who undergo conversion have higher mortality than those who do not, particularly in hospitals with more experience. This suggests that a deliberate effort to increase experience with and improve patient selection for RobATS lobectomies may ameliorate the conversions and their attendant sequelae.