Risk stratification of upper GI bleeding with an esophageal capsule

Background: Analysis of upper GI bleeding (UGIB) presentations to our institutions suggests that many patients admitted for endoscopic investigation could be managed safely as outpatients. Objective: To learn whether an esophageal capsule could identify a low-risk group of patients with UGIB who could safely wait for elective EGD. Design: Diagnostic, nonrandomized, single-blind (investigator) study. Setting: Three tertiary-care referral centers. Patients: Eighty-three consecutive adult patients referred for management of UGIB. Intervention: A capsule endoscopy (CE) was performed before EGD for the investigation and management of UGIB. Main Outcome Measurements: Detection rates of UGIB source and identification of a low-risk group of patients who would have been suitable for outpatient EGD based on CE findings. Results: In total, 62 of 83 patients (75%) had a cause for bleeding identified. Findings were concordant across both modalities in 34 patients (55%). Twenty-one patients (38%) with positive EGD results had negative CE results, 7 of whom were due to lack of duodenal visualization alone. However, 7 of 28 patients (25%) with normal EGD results had positive CE results. The subgroup of patients with duodenal visualization on CE, 23 of 25 (92%), were concordant with EGD for low-risk lesions that would have been suitable for outpatient management. Limitations: Low duodenal visualization rates with CE and low concordance between EGD and CE. Conclusion: Although CE is not currently ready to be used as a triage tool, when duodenal visualization was achieved CE correlated well with EGD findings and identified 92% of patients who may have been managed as outpatients.