Abstract

Background and study aims: Management of hemobilia is often challenging. Recently, endoscopic hemostasis with a self-expandable metal stent (SEMS) has shown promising efficacy for controlling bleeding at the endoscopic sphincterotomy site. This study aimed to assess efficacy and feasibility of endoscopic hemostasis as bridge therapy for hemobilia. Patients and methods: Patients with hemobilia between 2008 and 2023 were retrospectively reviewed. We compared efficacy of hemostasis between the initial endoscopic hemostasis group (ENDO group) and the initial angiographic embolization group (EMBO group). The primary outcome was initial hemostasis success rate and the secondary outcomes were delayed bleeding rate, subsequent embolization rate, 28-day mortality, transfusion amount, time to first hemostasis, total hemobilia time, and incidence of hypovolemic shock. Results: A total of 26 patients with hemobilia were included in this study and 17 patients (65.4%) were identified as the ENDO group and nine patients (34.6%) were classified as the EMBO group. The success rate of initial hemostasis was 88.2% (15/17) in the ENDO group and 100% (9/9) in the EMBO group (P = 0.529). The rate of delayed bleeding in the ENDO group was 17.6% (3/17) and 0.0% (0/9) in the EMBO group (P = 0.529). Total hemobilia time was shorter in the ENDO group than in the EMBO group (mean: 281.5 ± 1022.4 minutes vs. 5002.8 ± 7982.6 minutes; P < 0.001) Stent insertion depth was associated with successful hemostasis without delayed bleeding. (<i>P</i> = 0.015). <b>Conclusions:</b> Endoscopic hemostasis using SEMS for hemobilia appeared to be a feasible bridge therapy.