Development and validation of a scoring system predicting failure of endoscopic epinephrine injection therapy in Taiwanese patients with bleeding peptic ulcers.

Background. Endoscopic epinephrine injection therapy (EIT) is always the first choice of treatment for bleeding peptic ulcers, although it fails in 15% to 30% of patients. The aim of this study was to develop a new scoring system to predict the failure in EIT, and to validate this scoring model prospectively. Methods. This study enrolled 125 patients who presented with the stigmata of hemorrhage of peptic ulcers and underwent EIT. Patients with coagulopathy were excluded from the study. Univariate analysis of the clinical and endoscopic parameters to predict failure in EIT was performed first. A multiple logistic regression was used to develop a scoring system. This scoring equation was further applied to 50 prospective patients with bleeding peptic ulcers to validate its predictive value. Results. EIT failed to arrest bleeding in 32 (25.6%) patients. Shock, blood transfusion of at least 500 ml, active bleeding, and ulcer size were effective in predicting failure in EIT as calculated by univariate analysis, while multivariate analysis showed shock, active bleeding and ulcer size to be the independent predictors. The scoring equation, defined as -3.14 + 1.29 (shock) + 0.99 (active bleeding) + 0.13 (ulcer size in mm), had a sensitivity of 81.8% and a specificity of 76.9% in predicting failure of EIT in the prospective cohort of 50 bleeding patients. Conclusions. The presence of shock, active bleeding (spurting or oozing hemorrhage) and ulcer size are risk factors of failure of EIT in Taiwanese patients with bleeding peptic ulcers. The scoring system based on these three parameters can predict failure in EIT, while alternative treatment should be considered in case patients fail EIT.