Immediate endoscopic diagnosis of upper gastrointestinal bleeding. Its accuracy and value in relation to associated pathology.

Two hundred sixty-two patients with active upper gastrointestinal (GI) bleeding underwent panendoscopy between July 1970 and March 1973. There was 100% accuracy of endoscopic diagnosis as to the anatomical site of bleeding; the etiopathologic definition was 94.7% accurate. The series was divided into two groups, 116 with "liver disease" and 146 with "no liver disease." There were 107 patients with varices: 21 fell into no liver disease (small varices) and 86 into liver disease (39 small and 47 large varices). All had associated gastritis. Three endoscopic bleeding patterns were identified in the liver disease group. Only 27% of the patients in the liver disease group with varices (cirrhotics) had frank variceal hemorrhage, whereas 57% bled from hemorrhagic gastritis. The diagnostic unit provided early diagnosis, meaningful therapy, organized data gathering, and rough estimates of ultimate prognosis.