Non-steroidal anti-inflammatory drugs and variceal bleeding: a case-control study

Abstract

Background/aim: The aim of this case-control study was to assess the risk of bleeding from esophageal varices associated with aspirin and non-steroidal anti-inflammatory drug consumption.

Methods: Between January 1992 and May 1994, patients admitted for bleeding from esophageal or gastric lesions related to portal hypertension were matched with a control patient of the same age and sex, who was free of gastrointestinal bleeding. A structured interview was conducted with the cases and controls to determine drug consumption during the 2 weeks preceding admission. Fifty-nine cases and 59 controls were recruited.

Results/conclusions: Use of aspirin was more prevalent among the cases than the controls (odds ratio 3.81; 95% confidence interval 1.36-11.64; p = 0.004). This difference remained significant in the subgroups of patients with a first episode of variceal bleeding (odds ratio 3.9; 95% confidence interval 1.2-13.9, p = 0.01), but was not significant in the subgroups of patients with a recurrent episode of variceal bleeding. The use of aspirin was associated with a high risk of a first episode of variceal bleeding, suggesting that patients with portal hypertension should avoid taking these drugs.