[Non-variceal upper digestive hemorrhage: evaluation of endoscopic sclerosing treatment]

We appraise the efficacy of endoscopic injection treatment in 112 patients with gastroduodenal bleeding lesion versus 78 control patients (historical group), employing adrenaline (29 patients), absolute ethanol (52 patients) o adrenaline and ethanol (31 patients) as sclerosants. We observe in the treated group a diminution in rebleeding, lower transfusional blood needs and shorter hospital staying, with high stadistic significance versus the control group (p less than 0,001). Among treated patients the larger relapsing index (29.6%) was in patients treated with adrenaline injection only, against the group treated with ethanol or adrenaline and ethanol (10.2%) (p less than 0.05). The demand of surgical treatment was not significant between both groups, except in patients with actively jet bleeding, because urgent treatment was necessary in 37.5% of endoscopically treated patients versus 88.8% in not endoscopically treated. The mortality was similar in both groups.