AIM: To evaluate the safety of unsedated transnasal small-caliber esophagogastroduodenoscopy (EGD) for elderly and critically ill bedridden patients.METHODS: One prospective randomized comparative study and one crossover comparative study between transnasal small-caliber EGD and transoral conventional EGD was done (Study 1). For the comparative study,we enrolled 240 elderly patients aged > 65 years old.For the crossover analysis, we enrolled 30 bedridden patients with percutaneous endoscopic gastrostomy(PEG) (Study 2). We evaluated cardiopulmonary effects by measuring arterial oxygen saturation (SpO_2) and calculating the rate-pressure product (RPP) (pulse rate × systolic blood pressure/100) at baseline, 2 and 5 min after endoscopic intubation in Study 1. To assess the risk for endoscopy-related aspiration pneumonia during EGD, we also measured blood leukocyte counts and serum C-reactive protein (CRP) levels before and 3 d after EGD in Study 2.RESULTS: In Study 1, we observed significant decreases in SpO_2 during conventional transoral EGD,but not during transnasal small-caliber EGD (0.24% vs-0.24% after 2 min, and 0.18% vs -0.29% after 5 min,P = 0.034, P = 0.044). Significant differences of the RPP were not found between conventional transoral and transnasal small-caliber EGD. In Study 2, crossover analysis showed statistically significant increases of the RPP at 2 min after intubation and the end of endoscopy (26.8 and 34.6 vs 3.1 and 15.2, P = 0.044, P = 0.046),and decreases of SpO_2 (-0.8% vs -0.1%, P = 0.042) during EGD with transoral conventional in comparison with transnasal small-caliber endoscopy. Thus, for bedridden patients with PEG feeding, who were examined in the supine position, transoral conventional EGD more severely suppressed cardiopulmonary function than transnasal small-caliber EGD. There were also significant increases in the markers of inflammation,blood leukocyte counts and serum CRP values, in bedridden patients after transoral conventional EGD, but not after transnasal small-caliber EGD performed with the patient in the supine position. Leukocyte count increased from 6053 ± 1975/L to 6900 ± 3392/L ( P = 0.0008) and CRP values increased from 0.93 ± 0.24 to 2.49 ± 0.91 mg/dL ( P = 0.0005) at 3 d after transoral conventional EGD. Aspiration pneumonia, possibly caused by the endoscopic examination, was found subsequently in two of 30 patients after transoral conventional EGD.CONCLUSION: Transnasal small-caliber EGD is a safer method than transoral conventional EGD in critically ill,bedridden patients who are undergoing PEG feeding.