An Anatomical Approach to the Problem

Gastrointestinal bleeding is a frequent problem and the localization of the site of hemorrhage may be a challenge. Many lines of approach have contributed information expressing the differentiation of the causes for bleeding. It was felt that the. reported anatomical investigations of the problem were not extensive enough and that an additional study of an autopsy series would be enlightening. Therefore, this work was undertaken. The importance of the gastrointestinal bleeding problem is illuminated by the observation that hemorrhage as a cause of death at autopsy at the Cook County Hospital has more than doubled since a previous review of 1938.62 LITERATURE Chalmers and co-workers,11 reporting from a similar type of charity institution, stated that in 24 of 101 necropsies, fatal gastrointestinal hemorrhage was not suspected clinically. The clinical accuracy for the diagnosis of the source of bleeding was 53%. This difficulty in ascertaining site of bleeding has been noted by others.* Crohn 18 states that 21 % of his cases went undiagnosed ; Schiff " found bleeding in 26.4% of his patients un¬ explained. The inaccuracies of diagnosis must be con¬ sidered with knowledge of the inherent limi¬ tations of x-ray.21 At our own institution, in only 54% of patients with a bleeding ulcer was a roentgenological diagnosis made. The ulcer crater may be filled with clot ; edema of the walls may be present ; bleeding ulcers tend to be shallow and heal rapidly. For these reasons, some gastroenterologists sub¬ mit their patients to study within the first few days of bleeding.f Further, gastroscopy is of limited aid except in gastric ulcer.r'Most esophagoscopists are reluctant to ex¬ amine their patients if varices are present. Therefore, the only certain means of identi¬ fying the cause of bleeding is anatomic dis¬ section.