INTRODUCTION: The appropriate use of benzodiazepines for sedation during prolonged therapeutic endoscopy has not yet been established. This retrospective observational study compared the safety and efficacy of diazepam (DZP) and midazolam (MDZ) under moderate sedation during gastric endoscopic submucosal dissection (ESD). METHODS: We studied 554 patients who underwent gastric ESD under sedation with DZP or MDZ combined with pentazocine. Sedation depth was assessed and recorded using the Richmond Agitation-Sedation Scale (RASS). According to the American Society of Anesthesiologists definition of sedation levels, RASS scores of-4 to-2 points indicated moderate sedation, whereas a score of-5 points indicated deep sedation. Sedation levels, respiratory and circulatory dynamics during the procedure, and the incidence of ESD-related pneumonia were compared. RESULTS: Of these, 273 and 281 patients received DZP and MDZ, respectively. No significant differences were observed in the occurrence of deep sedation (DZP:MDZ = 12.1%:15.4%) or in the proportion of patients who maintained moderate intraoperative sedation (76.2%:80.4%). Respiratory parameters showed no significant differences
however, blood pressure reduction was more common in the MDZ group (4.8%:11.0%, p=0.007). Multivariate analysis identified MDZ as a significant factor associated with blood pressure reduction. The incidence of ESD-related pneumonia did not differ between the two groups. CONCLUSIONS: DZP and MDZ were similarly effective in maintaining adequate sedation levels during gastric ESD. Respiratory depression did not differ between the groups
however, circulatory depression was more pronounced in the MDZ group.