AIM:To investigate risk factors and adverse events related to high-dose diazepam administration during endoscopic submucosal dissection for gastric neoplasias.METHODS:Between February 2002 and December 2009,a total of 286 patients with gastric epithelial neoplasia underwent endoscopic submucosal dissection in our hospital.To achieve moderate sedation,5-7.5 mg of diazepam was administered intravenously by non-anesthesiologists.Intermittent additional administration of 2.5-5 mg diazepam was performed if uncontrollable body movement of the patient was observed.All patients were classified into groups based on the required diazepam dose:low-dose (≤ 17.5 mg,n=252) and high-dose (> 17.5 mg,n=79).RESULTS:Differences between the low-and highdose diazepam groups were observed in lifetime alcohol consumption (0.30 ± 0.48 vs 0.44 ± 0.52 tons,P=0.032),body weight (58.4 ± 10.3 vs 62.0 ± 9.9 kg,P=0.006),tumor size (15 ± 10 vs 23 ± 18 mm,P < 0.001),lesion location (P < 0.001) and the presence of ulcerative findings (14/238 vs 18/61,P < 0.001).Multivariate analysis identified all five variables as independently related to required diazepam dosage.In terms of adverse reactions to diazepam administration,paradoxical excitement was significantly more frequent in the high-dose diazepam group (P < 0.001).CONCLUSION:Intermittent administration of diazepam enabled safe completion of gastric endoscopic submucosal dissection except in patients who were alcohol abusers or obese,or who showed complicated lesions. To investigate risk factors and adverse events related to high-dose diazepam administration during endoscopic submucosal dissection for gastric neoplasias. METHODS: Between February 2002 and December 2009, a total of 286 patients with gastric epithelial neoplasia underwent endoscopic submucosal dissection in our hospital. To achieve moderate sedation, 5-7.5 mg of diazepam was administered intravenously by non-anesthesiologists.Intermittent additional administration of 2.5-5 mg diazepam was performed if uncontrollable body movement of the patient was observed. All patients were classified into groups based on the required RESULTS: Differences between the low-and high-density diazepam groups were observed in lifetime alcohol consumption (0.30 ± 0.48 (n = 25) and high-dose vs 0.44 ± 0.52 tons, P = 0.032), body weight (58.4 ± 10.3 vs 62.0 ± 9.9 kg, P = 0.006), tumor size (15 ± 10 vs 23 ± 18 mm, P <0.001) 0.001) and the presence of ulcerative findings (14/238 vs 18/61, P <0.001) .Multivariate analysis identified all five variables as independently related to required diazepam dosage. In terms of adverse reactions to diazepam administration, paradoxical excitement was significantly more frequent in the high- dose diazepam group (P <0.001). CONCLUSION: Intermittent administration of diazepam enabled safe completion of gastric endoscopic submucosal dissection except in patients who were alcohol abusers or obese, or who showed complicated lesions.