[Factors associated with failure of endoscopic therapy in gastric ulcer bleeding]

Endoscopic therapy is an effective technique in the control of bleeding due to peptic ulcer. However, bleeding persists or recurs in as many as 10-30% of patients. Gastric and duodenal ulcers present different clinical and endoscopic features and consequently the efficacy of endoscopic therapy and the factors associated with its failure should be studied separately.

To analyze the efficacy of endoscopic therapy in patients at high risk of persistent or recurrent bleeding due to gastric ulcer and to identify the factors associated with the failure of this technique.

We performed a retrospective study based on a clinical intervention protocol. Two hundred eight patients admitted for bleeding secondary to gastric ulcer with active bleeding or stigmas of recent bleeding who received endoscopic therapy between January 1992 and December 2001 were analyzed. Clinical, laboratory and endoscopic variables on admission, as well as the medical treatment and endoscopic procedure applied, were registered. Endoscopy was performed within 12 hours of admission. Patients were classified according to their response to endoscopic therapy: a) patients with limited bleeding, and b) patients with persistent or recurrent bleeding due to therapeutic failure. Intervention in patients with therapeutic failure was performed according to a previously established protocol. Variables that were statistically significant in the univariate analysis were included in a logistic regression model to identify those with an independent predictive value for failure of endoscopic therapy.

Definitive hemostasis was achieved after initial therapy in 181 patients (87%). The efficacy of a second procedure increased the percentage of hemostasis to 91% of the patients. In the logistic regression model, the only variables that were independently associated with initial therapeutic failure were: hemodynamic status on admission (p = 0.016; OR = 3.99), the need for transfusion of blood products prior to endoscopy (p = 0.025; OR = 3.48), upper localization of the gastric ulcer (p = 0.050; OR = 3.08) and unsatisfactory endoscopic therapy (p = 0.009; OR = 17.39).

These variables could contribute to the early identification of a subgroup of patients, which would enable us to increase medical-surgical surveillance and offer them other therapeutic alternatives.