The aim of this study was to systematically compare the efficacy and safety of remimazolam versus propofol for sedation in gastrointestinal endoscopic procedures by synthesizing evidence from randomized controlled trials. Specifically, the study sought to evaluate differences in induction and recovery profiles, procedural performance, patient and endoscopist satisfaction, and the incidence of adverse cardiopulmonary events.

A comprehensive search of PubMed, Web of Science, Scopus, Cochrane, and Embase was conducted from inception to January 2025. Randomized controlled trials comparing remimazolam and propofol in adult patients undergoing endoscopy were included. Primary outcomes included induction time, total sedation time, sedation success, procedure time, recovery parameters, and satisfaction. Safety outcomes included hypotension, bradycardia, hypoxemia, respiratory depression, postoperative nausea/vomiting, injection pain, and body movement. Random-effects models, subgroup analyses, sensitivity analyses, and meta-regression were performed.

From 2,304 records, 37 randomized controlled trials involving 8,533 patients met inclusion criteria. All studies were conducted in China or South Korea between 2020–2024. Risk of bias assessment showed 35 studies at low risk, one with some concerns, and one at high risk.

Induction time showed no overall difference between remimazolam and propofol (MD 0.13 s; 95% CI −0.05 to 0.31), though subgroup analysis showed significantly longer induction with remimazolam when MOAA/S = 1. Remimazolam was associated with shorter total sedation time (MD −1.84 min; 95% CI −3.07 to −0.61). Sedation success, procedure time, time to discharge, time to recovery, awake time, and endoscopist satisfaction showed no significant differences between drugs. Patient satisfaction slightly favored remimazolam (RR 1.06; 95% CI 1.00–1.11).

Safety outcomes consistently favored remimazolam, with significantly lower risks of hypotension (RR 0.42), bradycardia (RR 0.42), respiratory depression (RR 0.37), hypoxemia (RR 0.42), and injection pain (RR 0.11). No significant differences were found for postoperative nausea/vomiting or body movement. Subgroup analyses across gastroscopy, colonoscopy, and ERCP highlighted procedure-specific differences, with remimazolam demonstrating longer induction time in gastroscopy and colonoscopy, but comparable performance in ERCP. Meta-regression showed age significantly modified induction time (P = 0.0239) but did not influence other outcomes.

Remimazolam provides a markedly safer sedation profile than propofol for GI endoscopy, with significantly reduced cardiopulmonary adverse events and comparable efficacy across most clinical endpoints. Although induction may be slower, especially in certain procedures, remimazolam’s overall performance supports its integration into routine endoscopic sedation practice. Larger multicenter trials with standardized MOAA/S criteria and more diverse populations are needed to refine clinical recommendations.