Endoscopic control of upper gastrointestinal hemorrhage with a bipolar coagulation device.

It has been difficult to determine the real efficacy of endoscopic treatment for upper gastrointestinal tract bleeding sites for several reasons. First, since 80 per cent of an unselected group are expected to stop bleeding spontaneously, it is important to focus upon those individuals who continue to bleed instead of a group in whom bleeding would have stopped spontaneously in the majority. Second, it is difficult, if not impossible, to have comparable groups of patients with similar lesions and similar rates of bleeding who can be randomized into different treatment groups. This report describes the use of a bipolar endoscopic coagulation device in 28 patients with active massive upper gastrointestinal tract hemorrhage who represent 10 per cent of the patients with hemorrhage during a one year interval. Endoscopic treatment controlled bleeding initially in 23 of these patients. Another eight patients with recent hemorrhage who were at high risk for recurrent bleeding (visible vessels) had endoscopic coagulation without subsequent hemorrhage. Immediate operations were required in five of the 28 and delayed operations in another four. Mortality in the patients treated by endoscopic or surgical therapy was comparable (20 per cent), but no patient died of hemorrhage. The high mortality in this group of patients is explained by associated illnesses. B-C is as effective as other endoscopic treatments for nonvariceal sources of upper gastrointestinal tract hemorrhage. This modality is relatively cheap compared with other devices, is theoretically less complicated and has minimal risk to the individual patient. Because of these considerations, it is a technique which deserves wider application and may become the endoscopic treatment of choice for control of upper gastrointestinal tract hemorrhage. Patients with endoscopic control of upper gastrointestinal tract bleeding avoid perioperative morbidity, have a lower transfusion requirement and may have a shorter hospital stay than comparable individuals who require operative control of bleeding sites.