A case study: nonsteroidal antiinflammatory drugs and gastrointestinal bleeding.

This paper describes a postmarketing drug surveillance study which examined the association of upper gastrointestinal (GI) bleeding and exposure to nonsteroidal antiinflammatory drugs (NSAID). The rate of upper GI bleeding was compared in 47,136 patients exposed to a NSAID and 44,634 unexposed patients. The risk of bleeding was 1.5 times higher in patients exposed to a NSAID (95% confidence interval, 1.2-2.0). Dose (p less than 0.01) and duration (p less than 0.001) response relationships were identified. The rate of upper GI bleeding was then compared among 88,044 patients exposed to only one NSAID. Sulindac was associated with the highest rate of upper GI bleeding. Control for multiple potential confounding variables did not change the results. These results suggest that NSAID are associated with upper GI bleeding, and that sulindac may lead to upper GI bleeding more frequently than other NSAID.