Unsedated transnasal ultrathin esophagogastroduodenoscopy may provide better diagnostic performance in gastroesophageal reflux disease.

Transnasal ultrathin esophagogastroduodenoscopy (N-EGD) with less gagging reflexes under non-sedation is likely suitable for the diagnosis of gastroesophageal reflux disease (GERD), however, N-EGD might have drawbacks, including its low image resolution. Limited information is available regarding the diagnosability of N-EGD for GERD. We compared the utility and gagging reflexes of three different endoscopies, including N-EGD, ultrathin transoral EGD (UTO-EGD) and conventional oral EGD (CO-EGD), in the diagnosis of GERD. We performed screening endoscopy in 1580 patients (N-EGD n=727, UTO-EGD n=599, CO-EGD n=254) and compared the frequency distributions of the severity of reflux esophagitis, hiatus hernia, and Barrett's epithelium to estimate the diagnostic performance of each endoscopy. We also analyzed patients' tolerability of endoscopy by the subjective evaluation of gagging reflexes. In the diagnosis of reflux esophagitis and Barrett's epithelium, there was no significant difference in the frequency distributions of the severity of the diseases among three EGDs. However, the incidence of Barrett's epithelium was higher than that in the previous nationwide survey of GERD in Japan. The evaluated size of hiatus hernia was smaller in N-EGD than in two other peroral endoscopies. The size of hiatus hernia correlated significantly with severity of gagging reflexes that was also lowest when diagnosed with N-EGD. N-EGD had an equivalent performance in the diagnosis of reflux esophagitis and Barrett's epithelium compared with CO-EGD. Enlargement of hiatus hernia induced by gagging reflexes was minimal in N-EGD, resulting in its better performance in the diagnosis of Barrett's epithelium.