OBJECTIVETo compare the clinical efficacy with meta-analysis between robot-assisted gastrectomy(RAG) and laparoscopy-assisted gastrectomy (LAG) for gastric cancer.

METHODSA literature search was performed in PubMed, Embase, Cochrane Library, Web of Science, Wanfang Data, CNKI(Chinese National Knowledge Infrastructure), and CBM(China Biological Medicine) databases for clinical researches published before July 2015 that compared RAG with LAG. Operative time, estimated blood loss, harvested lymph nodes, proximal margin, distal margin, hospital stay, conversion and complications were compared using weighted mean differences(WMD) and odds ratios (OR).

RESULTSSixteen studies were included in the analysis, comprising 5 764 patients(1 593 RAGs, 4 171 LAGs). RAG was associated with longer operative time (WMD=49.68, 95% CI: 35.54 to 63.82, P=0.000), less estimated blood loss (WMD=-26.10, 95% CI: -42.90 to -9.31, P=0.002), and shorter hospital stay(WMD=-0.72, 95% CI: -1.35 to -0.09, P=0.024). Conversion, mortality, complications, proximal margin, distal margin and harvested lymph nodes of RAG were similar to LAG. In meta-analysis results of distal gastrectomy and early-stage gastric cancer, RAG had more harvested lymph nodes (distal gastrectomy: WMD=2.23, 95% CI: 0.33 to 4.13, P=0.021; early-stage gastric cancer: WMD=2.58, 95% CI: 1.05 to 4.12, P=0.001).

CONCLUSIONSRAG can be performed safely with less estimated blood loss and more harvested lymph nodes as compared to LAG. Radical resection can be achieved by RAG.