The risk of upper gastrointestinal bleeding in elderly users of aspirin and other non-steroidal anti-inflammatory drugs: the role of gastroprotective drugs.

BACKGROUND AND AIMS: Although the administration of gastroprotective drugs may reduce the risk of gastrointestinal (GI) bleeding due to intake of non-steroidal anti-inflammatory drugs (NSAIDs) and aspirin during chronic treatment, no consensus exists as to whether such co-therapy is effective in short-term prevention, particularly in old age. The aim of our study was to evaluate the risk of bleeding associated with acute and chronic NSAID or aspirin therapy in elderly subjects, and the influence of gastroprotective treatment on such a risk.

METHODS: The study included 467 elderly NSAID or aspirin users and 1784 non-users, who consecutively underwent upper GI endoscopy. The use of NSAIDs and/or aspirin as well as gastroprotective drugs (misoprostol, H2-blockers, proton pump inhibitors) was evaluated during a structured interview. Upper GI tract bleeding was diagnosed on the basis of symptoms and endoscopic signs of recent hemorrhage.

RESULTS: 54.2% of patients were acute and 45.8% chronic users of NSAIDs or aspirin. The risk of bleeding was higher in acute [odds ratio (OR) 4.14, 95% CI 2.97-5.78] than chronic users (OR 1.71, 95% CI 1.1-2.67). The risk of bleeding, adjusted for age, gender, Helicobacter (H) pylori infection, and gastroprotective drug use were 7.87 (CI 4.90-12.60) in acute users and 3.97 (95% CI 2.27-6.96) in chronic users of NSAIDs and/or aspirin. The risk of bleeding was significantly associated with acute but not chronic use of regular-dose aspirin (OR 5.53, 95% CI 2.29-13.3), diclofenac (OR 4.44, 95% CI 2.21-8.93), ketorolac (OR 4.81, 95% CI 2.13-10.9), naproxen (OR 14.9, 95% CI 4.23-52.4) or nimesulide (OR 4.06, 95% CI 1.2-13.8). Piroxicam increased the risk of bleeding in both acute (OR 5.36, 95% CI 1.94-14.8) and chronic therapy (OR 5.53, 95% CI 1.23-24.9). In acute users, concomitant therapy with proton pump inhibitors reduced the risk of bleeding compared with non-users (OR 1.05, 95% CI 0.19-5.65), whereas co-treatment with H2-blockers was associated with a significantly higher risk of bleeding than in non-users (OR 3.40, 95% CI 1.28-9.02). Chronic users of NSAIDs or aspirin co-treated with proton pump inhibitors had a lower risk of bleeding (OR 1.12, 95% CI 0.21-6.07) than those treated with misoprostol (OR 1.91, 95% CI 0.33-10.9) or H2 blockers (OR 2.26, 95% CI 0.81-6.36).

CONCLUSIONS: The risk of upper GI bleeding is significantly higher in elderly acute vs chronic users of NSAIDs or regular-dose aspirin. In acute NSAID or aspirin users, co-treatment with proton pump inhibitors, but not with H2-blockers, may reduce the risk of bleeding compared with non-users.