Sphincterotomy - Induced Hemorrhage: Prevalence, Risk Factors and Endoscopic Hemostasis / 대한소화기내시경학회지

ABSTRACT

BACKGROUND/AIMS: Endoscopic biliary sphincterotomy (EST)-induced hemorrhage occurs in approximately 0.5-12% of procedures. We prospectively investigated the risk factors of EST-induced hemorrhage and evaluated its safety as well as the effectiveness of endoscopic hemostasis. METHODS: One thousand three hundred and four patients, who underwent EST between July 1996 and June 1998, were enrolled. As a hemostatic treatment, epinephrine spray was initially used. If bleeding persisted, epinephrine injection was performed consecutively. In patients with exposed vessels, epinephrine injection followed by alcohol injection was given. RESULTS: EST-induced hemorrhage occurred in 136 (10.4%) patients. Types of sphincterotome (needle-knife sphincterotome, p=0.0079) and cutting speed (so-called, zipper cut, p=0.03) were revealed as significant variables for the occurrence of bleeding. Once bleeding occurred, patients with an associated ampullary lesion (impacted stone or cancer) or with coagulopathy were more likely to bleed profusely. Initial hemostasis was achieved in all patients. However, rebleeding occurred in eight patients who were initially classified in the moderate or severe bleeding group. Finally, EST-induced hemorrhage was successfully controlled in all patients after 1-3 treatment sessions (mean 1.1 sessions). The difference in the incidence of complications between the groups with and without endoscopic hemostasis was not statistically significant. CONCLUSIONS: The use of needle-knife sphincterotome and cutting speed were independent risk factors for bleeding occurrence. Once bleeding occurred, its severity was affected by the associated ampullary lesion (impacted stone or cancer) or coagulopathy. Endoscopic hemostasis with epinephrine and/or alcohol was effective and safe in EST-induced hemorrhage.