Although endoscopic examination in patients with bleeding of the upper gastrointestinal tract has improved diagnostic accuracy, it has not been useful in predicting clinical outcome and has not been shown to improve the patients’ prognoses. This article describes a subgroup of patients with acute upper gastrointestinal bleeding whose clinical outcome can be predicted at the time of endoscopy. In 28 of 317 patients who underwent endoscopy, a “visible vessel” was seen in an ulcer presumed to be the bleeding site. All 28 were later recommended for operation because of recurrent (86 per cent) or uncontrolled (14 per cent) hemorrhage. In contrast, 75 per cent of the remaining 289 patients in whom vessels were not seen, whether or not bleeding from ulcers, had single bleeding episodes managed medically. Since patients with a “visible vessel” can be expected to have uncontrolled or recurrent hemorrhage, surgical treatment should be considered at the time of endoscopy if such a vessel is seen. (N Engl J Med 300:1411–1413, 1979) BEFORE endoscopy was available, bleeding of the upper gastrointestinal tract was managed in an empiric manner. Patients whose bleeding could not be controlled medically were subjected to operation either without a preoperative diagnosis or with only a suggested diagnosis based on radiologic findings. In 1969, Palmer1 introduced the use of fiberoptic endoscopy for bleeding of the upper gastrointestinal tract and his “vigorous” approach has gained wide acceptance. From comparisons between this approach and more traditional methods, it became apparent that the bleeding lesion could be identified earlier and more accurately with the endoscope.1 2 3 4 5 Despite this diagnostic accuracy, improvements in morbidity. © 1979, Massachusetts Medical Society. All rights reserved.
