Postoperative complications and surgical outcomes of robotic versus laparoscopic pancreaticoduodenectomy: a meta-analysis of propensity-score-matched studies.

Abstract

Background: Robotic pancreaticoduodenectomy (RPD) is used more commonly, but high-level evidence is still scarce. This meta-analysis aimed to compare the short-term outcomes between RPD and laparoscopic pancreaticoduodenectomy (LPD) using data collected from propensity score-matched (PSM) studies.

Materials and methods: We searched PubMed, Cochrane Library, Embase, and Web of Science databases for PSM studies comparing RPD and LPD. Risk ratios (RRs) and mean differences (MDs) with 95% confidence intervals were calculated.

Results: Ten PSM studies were included, encompassing 8106 patients (RPD group: 3695 patients; LPD group: 4411 patients). Compared with LPD, RPD was associated with a lower conversion rate (RR, 0.56) and blood transfusion rate (RR, 0.49), as well as a higher number of harvested lymph nodes (MD, 2.15). There were no significant differences observed in 30-day readmission (RR, 1.02), 90-day mortality (RR, 1.01), overall morbidity (RR, 0.94), major complications (RR, 1.06), operative time (MD, -8.00 min), blood loss (MD, -19.37 mL), reoperation (RR, 0.95), bile leak (RR, 0.93), chylous leak (RR, 1.40), postoperative pancreatic fistula (RR, 1.06), delayed gastric emptying (RR, 0.92), wound infection (RR, 1.12), length of stay (MD, -0.32 days), and R0 resection (RR, 0.98) between the groups.

Conclusions: Although LPD and RPD had similar surgical outcomes, RPD had the perioperative advantage over LPD in decreasing conversion rates and blood transfusion rates and increasing the number of lymph nodes harvested. Further randomized controlled trials evaluating the potential advantages of RPD over LPD are warranted.