Accuracy of esophagoscopy performed by a non-physician endoscopist with a 4-mm diameter battery-powered endoscope

Background: A cost-effective technique is needed for screening of a broad population at risk for esophageal cancer. A solution would be to have non-physician endoscopists perform esophagoscopy with small-caliber battery-powered endoscopes. Methods: In a prospective blinded study, the diagnostic accuracy of sedated esophagoscopy performed by a trained nurse practitioner with a battery-powered 4-mm diameter endoscope was compared with that for a sedated standard video-endoscopy performed by a gastroenterologist. Patients were recruited to undergo peroral esophagoscopy by the nurse practitioner followed by sedated standard endoscopy by the supervising gastroenterologist, each blinded to the findings of the other. Major esophageal findings of nurse practitioner and gastroenterologist were compared. Results: Findings in 40 patients were analyzed. In 4 patients both endoscopists could not assess the presence or absence of columnar-lined esophagus because of severe erosive esophagitis (n = 3) or severe candida-esophagitis (n = 1). By using sedated standard endoscopy as the standard, on a per finding basis, esophagoscopy by the nurse practitioner had a sensitivity for columnar-lined esophagus of 89%: 95% CI [75%, 97%] and specificity of 96%: 95% CI [84%, 99%]. The missed columnar epithelium was a 3 × 3-mm island. For all lesions, the sensitivity of endoscopy performed by the nurse practitioner with the battery-powered endoscope was 75%: 95% CI [67%, 82%] and specificity 98%: 95% CI [96%, 99%]. The nurse practitioner missed all of 4 rings (3 considered clinically irrelevant). Conclusion: Esophagoscopy with a battery-powered 4-mm diameter endoscope by a non-physician endoscopist is feasible and accurate in detecting esophageal pathologies. It may be an efficient screening method for the detection of columnar-lined esophagus. There was a distinct underestimate of the presence of esophageal rings. (Gastrointest Endosc 2003;57:305-10.)