Procinéticos infusión antes de la endoscopia para la hemorragia digestiva alta aguda: un ensayo controlado, aleatorizado, doble ciego y controlado con placebo.

PURPOSE:

We investigated the relative benefit in the use of erythromycin/ metochlopramide to improve visualisation before endoscopyin in acute GI bleeding, assisted by subjective & objective criteria.

METHODS:

Forty-five patients with hematemesis were randomly assigned to receive placebo (intravenous saline), intravenous metochlopramide or intravenous erythromycin before endoscopy in a double-blind study. All the patients had gastric lavage done prior to the infusion of the study medicine/placebo. The study had 15 patients in each arm and randomization was done by the Department of Pharmacy.

RESULTS:

Characteristics of patients at admission were similar in the three groups. Nineteen patients (42%) had peptic ulcer disease & nine patients (20%) had esophageal varices. The gastric mucosa was entirely visualized by the endoscopist in 86% of patients in the erythromycin group, versus 66% in the metochlopramide and placebo groups (P > 0.05). The quality of examination of the upper gastrointestinal tract, assessed by using a four-point scoring system (from zero to 3 with zero as the worst and three as the best) was better in the erythromycin group (2.53/3) compared to metochlopramide (2/3) and placebo (2.2/3) with (P > 0.05). Clots were found in the stomach in 13% of the patients in the erythromycin group compared to 33% in the metochlopramide and the placebo group. Repeat endoscopy was needed in one patient in the erythromycin group and three patients in the placebo and metochlopramide group. The mean length of hospital stay was not different among the three groups. The mean amount of blood transfusion needed was 2.26 in the erythromycin group, 2.86 in the placebo group and 3.66 in the metochlopramide group. All patients tolerated erythromycin and metochlopramide with no adverse effects.

CONCLUSION:

The study did not detect a significant difference between placebo and either study medication There is a possibility of a type II error because of a relatively small sample size and high rate of success in the placebo arm. We estimated that, given the placebo response rate, about 100 patients needed to be enrolled in each arm to have a 90 percent chance of detecting a statistically significant (P value <0.05) difference in the medications. The study also suggests that a thorough lavage with normal saline may adequately cleanse the upper GI tract for good visualization in acute upper GI bleeding.