da Vinci robotic versus laparoscopic surgery in rectal cancer: a meta-analysis of postsurgery complications

Ziyan Luo,Dongzhu Zeng,Yan Shi,Peiwu Yu

DOI: https://doi.org/10.1007/s00384-016-2516-7

IF: 2.8

2016-01-01

International Journal of Colorectal Disease

Abstract:Dear Editor: Laparoscopic surgery (LS) and daVinci robot-assisted surgery (RS) are both used in rectal cancer treatment. Considering the anatomical features of rectal cancer, RS is superior to LS in mid-low set rectal operations and in anus preservation. However, it is unclear which surgical approach has a lower prevalence of postoperative complications and better outcomes. We therefore conducted a meta-analysis of controlled trials to compare postoperative complications and outcomes of LS and RS for resection of rectal cancer. We conducted a worldwide literature search of publications, assembled documents, and related references using the MedLine (PubMed), Embase, Ovid, CNKI, and WanFang databases. The search focused on controlled trials published between 1979 and 2014 comparing the use of RS and LS in rectal cancer surgery. We used the following search terms: da Vinci surgical system, robot, robotic, laparoscopic surgery, total mesorectal excision (TME), and rectal cancer. Two authors independently developed the selection including retrieval and data checks. Disagreements were resolved through discussion and, if necessary, by involving a third independent researcher.We assessed the quality of methods following standards of randomized controlled tests (RCTs) from Cochrane Reviewers Handbook 5.2. Specific evaluation criteria included allocation sequence, allocation concealment, blinding tests of doctors and patients, blinding tests of statistics keepers, incomplete data bias, selective reporting results and other potential factors affecting validity. If the above seven indices were in accordance with quality standards, the quality of the research was deemed high; if not, bias might exist. The inclusion criteria used to screen patients were as follows: (1) definitively diagnosed with rectal cancer and (2) had undergone surgery.We divided patients into two groups: those who underwent RS and those who underwent LS. The exclusion criteria used were as follows: (1) did not exhibit pneumoperitoneum or (2) had a medical history of colorectal surgery. We also excluded studies on nonrectal cancer and TME for nonrectal cancer. The outcomes included operative morbidity, anastomotic leakage, incision infection, intestinal obstruction, abdominal abscess, cleaned lymph node count, and unscheduled reoperation. We performed the meta-analysis using Review Manager (RevMan), version 5.2 (Cochrane Collaboration, London, United Kingdom). Enumeration data were examined using relative risk (RR) and 95 % confidence interval (CI). Overall effects were calculated using the Z test. Statistical significance was considered at P=0.05. For measurement data with the same units, statistical analyses were performed using the weighted mean difference, and for measurement data with different units, analyses were performed using the standardized mean difference. The studies were tested for heterogeneity before being combined. If heterogeneity did not exist, the fixed-effect model was used in the analysis; else the source of clinical heterogeneity was explored. If no clear difference was found, we used the random-effects model. To assess the potential for publication bias, we visually inspected funnel plots. The search identified five studies published in English for inclusion in the meta-analysis. There were 746 patients: 409 were treated with LS and 337 with RS. All studies had complete data records and reported main indices of previous studies. None of the studies lay outside 95 % CIs and there was no * Peiwu Yu yupeiwu01@sina.com