PurposeTo identify predictors of early bleeding recurrence (< 30 days) after transcatheter embolization for refractory hemorrhage from gastroduodenal ulcers.Materials & MethodsRetrospective study of 60 consecutive emergency embolization procedures in hemodynamically unstable patients (41 males, 19 females, mean age 69.4 ± 15 years) referred from 1999 to 2008 for selective angiography after failed endoscopic treatment. The embolic agents used were metallic coils, microspheres, gelatine sponge particles, and glue, sole or in combination. Predictors of early rebleeding were tested with univariate analysis and multivariate logistic regression model, respectively.ResultsThe technical success rate was 95%. The primary clinical success rate was 71.9% (41 of 57 patients). Secondary clinical success occurred in three additional patients (77.2%). No major complications related to catheterization occurred. The periprocedural mortality rate was 26.7% (16 of 60), mostly related to underlying conditions. Early recurrence of bleeding was associated with longer time to angiography (P=.0005), more units of packed red blood cells transfused prior to the procedure (P=.0009), number of co-morbidities ≥ 2 (P=.005), and the use of coils as the only embolic agent (P=.003). Two factors were non confounding predictors of embolization failure: time to angiography > 2 days (odds ratio=70.94; P<.05) and co-morbid diseases ≥ 2 (odds ratio=45.06; P<.05). The use of anticoagulant and/or anti-inflammatory medications before procedure tended to predict failure treatment (odds ratio=11.83; .05<P<.1).ConclusionAngiographic embolization for hemorrhage from gastroduodenal ulcers should be performed early in the course of bleeding and not with coils alone in otherwise critically ill patients. PurposeTo identify predictors of early bleeding recurrence (< 30 days) after transcatheter embolization for refractory hemorrhage from gastroduodenal ulcers. To identify predictors of early bleeding recurrence (< 30 days) after transcatheter embolization for refractory hemorrhage from gastroduodenal ulcers. Materials & MethodsRetrospective study of 60 consecutive emergency embolization procedures in hemodynamically unstable patients (41 males, 19 females, mean age 69.4 ± 15 years) referred from 1999 to 2008 for selective angiography after failed endoscopic treatment. The embolic agents used were metallic coils, microspheres, gelatine sponge particles, and glue, sole or in combination. Predictors of early rebleeding were tested with univariate analysis and multivariate logistic regression model, respectively. Retrospective study of 60 consecutive emergency embolization procedures in hemodynamically unstable patients (41 males, 19 females, mean age 69.4 ± 15 years) referred from 1999 to 2008 for selective angiography after failed endoscopic treatment. The embolic agents used were metallic coils, microspheres, gelatine sponge particles, and glue, sole or in combination. Predictors of early rebleeding were tested with univariate analysis and multivariate logistic regression model, respectively. ResultsThe technical success rate was 95%. The primary clinical success rate was 71.9% (41 of 57 patients). Secondary clinical success occurred in three additional patients (77.2%). No major complications related to catheterization occurred. The periprocedural mortality rate was 26.7% (16 of 60), mostly related to underlying conditions. Early recurrence of bleeding was associated with longer time to angiography (P=.0005), more units of packed red blood cells transfused prior to the procedure (P=.0009), number of co-morbidities ≥ 2 (P=.005), and the use of coils as the only embolic agent (P=.003). Two factors were non confounding predictors of embolization failure: time to angiography > 2 days (odds ratio=70.94; P<.05) and co-morbid diseases ≥ 2 (odds ratio=45.06; P<.05). The use of anticoagulant and/or anti-inflammatory medications before procedure tended to predict failure treatment (odds ratio=11.83; .05<P<.1). The technical success rate was 95%. The primary clinical success rate was 71.9% (41 of 57 patients). Secondary clinical success occurred in three additional patients (77.2%). No major complications related to catheterization occurred. The periprocedural mortality rate was 26.7% (16 of 60), mostly related to underlying conditions. Early recurrence of bleeding was associated with longer time to angiography (P=.0005), more units of packed red blood cells transfused prior to the procedure (P=.0009), number of co-morbidities ≥ 2 (P=.005), and the use of coils as the only embolic agent (P=.003). Two factors were non confounding predictors of embolization failure: time to angiography > 2 days (odds ratio=70.94; P<.05) and co-morbid diseases ≥ 2 (odds ratio=45.06; P<.05). The use of anticoagulant and/or anti-inflammatory medications before procedure tended to predict failure treatment (odds ratio=11.83; .05<P<.1). ConclusionAngiographic embolization for hemorrhage from gastroduodenal ulcers should be performed early in the course of bleeding and not with coils alone in otherwise critically ill patients. Angiographic embolization for hemorrhage from gastroduodenal ulcers should be performed early in the course of bleeding and not with coils alone in otherwise critically ill patients.