Esophageal capsule endoscopy in patients with suspected esophageal disease:: double blinded comparison with esophagogastroduodenoscopy and assessment of interobserver variability

Background and study aims: The study aim was to compare the diagnostic yield of capsule endoscopy and esophagogastroduodenoscopy (EGD) in patients with suspected esophageal disease. Secondary aims were to assess interobserver variability of capsule endoscopy readings and safety. Patients and methods: In total, 98 patients (53 men, 53 13 years) with an indication for EGD were included. The patient population was artificially enriched to include two thirds of patients with abnormal esophageal findings at EGD, which was followed by capsule endoscopy. Capsule recordings were blindly read by three endoscopists, one from the center that recruited the patient and two from the other center. Study outcomes were the findings described on EGD and capsule endoscopy, agreement between EGD and capsule endoscopy for findings, quality of the capsule recording, and interobserver agreement for capsule endoscopy quality and findings. Results: EGD was normal in 34 patients and showed esophageal findings in 62 (esophagitis 28, hiatus hernia 21, varices 21, Barrett's esophagus 11, others 7). Average esophageal transit time of the capsule was 361 393 seconds. Capsule endoscopy was normal in 36 patients but detected esophagitis in 23, hiatus hernia in 0, varices in 23, Barrett's esophagus in 18, and others in 4. The positive predictive value of capsule endoscopy was 80.0% and the negative predictive value was 61.1%. Overall agreement per patient was moderate between EGD and capsule endoscopy for the per-patient (kappa = 0.42) and per-findings (kappa = 0.40) analyses. Interobserver agreement between capsule endoscopy readings was moderate for findings (kappa = 0.39) and quality assessment (kappa = 0.24). No adverse event was observed after either EGD or capsule endoscopy. Conclusion: In this study, despite artificial prevalence enrichment, capsule endoscopy showed a moderate sensitivity and specificity in the detection of esophageal diseases.