[Digestive hemorrhage following gastroduodenal ulcers caused by non-steroidal anti-inflammatory agents].

The aim of this retrospective study was to describe the clinical and prognostic characteristics of bleeding ulcers following gastrotoxic agent use. From 1984 to 1988, 244 patients were admitted in our Intensive Care Unit for bleeding ulcers; 133 (54.5 percent) were using gastrotoxic agents. Patients not taking gastrotoxic agents were compared with patients taking non steroidal antiinflammatory drugs (n = 77) or aspirin (n = 35). Patients receiving non steroidal antiinflammatory drugs differed (P less than 0.05) from patients not taking gastrotoxic agents: they were older, had less history of ulcer, hemorrhage, and alcohol abuse. Patients taking aspirin had less severe hemorrhage and took anticoagulant therapy less frequently than patients not taking gastrotoxic agents. Patients taking aspirin were younger, had less severe hemorrhage and had more frequent history of alcohol abuse and ulcer disease than patients taking non steroidal antiinflammatory drugs. There was no difference with regard to endoscopic data and mortality (7 percent). Two thirds of hemorrhages following non steroidal antiinflammatory drugs occurred during the month following the onset or the modification of treatment. The distribution of non steroidal antiinflammatory drugs in the group of patients who bled after non steroidal antiinflammatory drugs differed from that of regional sales. These results suggest that upper gastrointestinal tract bleeding associated with the use of gastrotoxic agent are frequent, particularly in elderly patients, are not associated with poor prognosis, and occur soon after the beginning of treatment.