[Descriptive analysis of 660 cases of upper digestive hemorrhage: relation of emergency endoscopy and mortality].

: A study was made of 660 patients admitted for upper gastrointestinal bleeding in which the endoscopic findings were analyzed, excluding patients with esophageal varices and cases in which the lesions responsible for bleeding were not appreciated (5.7% diagnostic failure). 71.7% were males and 38.3% females, average age was 59.9 +/- 17.5 years and average hematocrit on admission was 27.4 +/- 8.2. 58.3% were over 60 years-old. There was a history of ingestion of gastro-erosive drugs in 56.8%. Gastroscopy was performed within 24 hours of admission in 77.4%, the first clinical symptom of bleeding being melenas in 50%. Peptic ulcer was the most common cause of upper gastrointestinal bleeding. Eighty-one patients required surgery, 30 early and 51 delayed. The global mortality was 3.6% and surgical mortality, 13.5%. After relating the endoscopic groups of upper gastrointestinal bleeding and individual risk with mortality rates, emphasis is placed on the need for early endoscopy as a diagnostic tool of prognostic value.