Evaluating safety and quality of robotic-assisted gastric cancer surgery: meta-analysis and meta-regression

Robotic-assisted surgery is expanding globally. The UK's National Institute for Health and Care Excellence recently cautioned due to a paucity of high-quality evidence. To address this, a systematic review, meta-analysis, and meta-regression were undertaken to evaluate the quality and safety of robotic-assisted gastrectomy (RAG) versus conventional approaches for gastric cancer.

Systematic searches were conducted on MEDLINE, Embase, Web of Science, and Scopus (2 May 2025) for studies comparing RAG to open or laparoscopic gastrectomy up to 30 April 2025. Primary outcomes were Clavien-Dindo grade ≥ II complications (CD ≥ II; safety) and margin-positive resections (quality). Risk of bias was assessed using the Risk Of Bias In Non-randomized Studies of Interventions and Cochrane Risk of Bias v2.0 tools. Heterogeneity and evidence certainty were evaluated using meta-regression and GRADE assessment.

In all, 90 studies (65 296 patients) were included; only three studies were randomized clinical trials and 72 were from East Asia. In 44 studies (12 102 patients) RAG was associated with significantly lower CD ≥ II complications (odds ratio (OR) 0.74; 95% confidence interval (c.i.) 0.64 to 0.86); heterogeneity was low (I2 = 21.4%). Seven studies had a low risk of bias. From 35 studies on margin status, RAG had fewer R1 resections (OR 0.74; 95% c.i. 0.51 to 1.07); heterogeneity was moderate (I2 = 34.0%). Adoption year, industry funding, extent of resection, and tumour stage were identified as sources of heterogeneity. Three studies were at low risk of bias. Certainty was very low for both outcomes.

Although there may be potential benefits of robotic-assisted surgery, cautious adoption is warranted given the current uncertainty. Safe adoption requires standardized training, competency benchmarks, and limiting industry involvement. High-quality evaluation through randomized trials and parallel health economics is urgently needed to inform future policy and practice.