LBOVI-A-4Short-term gastrointestinal and cardiovascular safety of selective COX-2 inhibitors and non-selective nsaids: An instrumental variable analysis

Abstract

BACKGROUND A direct comparison of the gastrointestinal (GI) and cardiac safety of NSAIDs in elderly patients is not possible based on randomized trials since different trials employed non-overlapping patient populations and comparison exposures. We sought to estimate the risk reduction of GI complications and risk increase in acute myocardial infarction (MI) by celecoxib and rofecoxib compared with several non-selective NSAIDs addressing confounding by using instrumental variable methods. METHODS We identified 49,731 Medicare beneficiaries who initiated non-selective NSAIDs or selective COX2 inhibitors between 1/1/99 and 7/31/03. Risk increase of GI complications and MI in 180 days after initiation of NSAID therapy (rofecoxib, diclofenac, ibuprofen, and naproxen compared with celecoxib) was assessed using instrumental variable analysis adjusting un/measured confounders. RESULTS Compared with celecoxib, a tendency for increased risk of GI complications was observed for ibuprofen (risk difference = 0.93 events per 100 patients; −1.88; 3.73) and diclofenac (RD=5.20; −1.07;11.5). We found a clear increase in the risk of MI for rofecoxib (RD=2.17; 0.12;4.21) and diclofenac (RD=7.63; 0.24;15.0), while naproxen had the lowest risk (RD = −1.59; −4.62;1.44). CONCLUSIONS Diclofenac appears to have the least favorable safety profile among NSAIDs, while naproxen has the lowest risk of MI with no meaningful increase in risk for GI complications, making it a viable alternative to celecoxib. Clinical Pharmacology & Therapeutics (2005) 79, P83–P83; doi: 10.1016/j.clpt.2005.12.296