The benefit of nonsurgical
therapy in the treatment of active nonvariceal upper gastrointestinal tract
hemorrhage is uncertain. I performed a prospective controlled trial of endoscopic multipolar
electrocoagulation for active upper
gastrointestinal hemorrhage. Patients were considered for entry if they had a bloody nasogastric aspirate,
melena, or
hematochezia, and any of the following: unstable vital signs, a requirement of greater than or equal to 2 units of blood per 12 hours, or a drop in hematocrit of greater than or equal to 6 percent in 12 hours. Forty-four patients were randomly assigned to receive multipolar
electrocoagulation or
sham multipolar
electrocoagulation if endoscopy revealed active
bleeding from an
ulcer (24 patients), a
Mallory-Weiss tear (17), or a
vascular malformation (3). The group receiving multipolar
electrocoagulation did significantly better in terms of hemostasis (90 percent vs. 13 percent, P less than 0.0001), mean (+/- SE) transfusion requirements (2.4 +/- 0.9 vs. 5.4 +/- 0.9 U; P = 0.002), mean number of hospital days (4.4 +/- 0.8 vs. 7.2 +/- 1.1, P = 0.02), and percentage needing emergency surgery or another intervention (14 vs. 57 percent, P = 0.01). Although mortality was lower in the group receiving multipolar
electrocoagulation (0 vs. 13 percent), this difference was not statistically significant. The mean cost of hospitalization for treated patients was less than half that for the controls ($ 3,420 +/- 750 vs. $ 7,550 +/- 1,480, P = 0.001). I conclude that multipolar
electrocoagulation markedly improves the hospital course in patients with major, nonvariceal upper
gastrointestinal hemorrhage.