NSAID Usage: Impact of Safety Data and Product Withdrawals on Prescribing Trends

the most commonly used classes of medications, accounting for 3% of all prescriptions dispensed in the U.S. About 50 million Americans receive NSAID therapy for both chronic disorders, including rheumatoid arthritis and osteoarthritis, and acute conditions such as postoperative pain, headache, tendonitis or bursitis, and dysmenorrhea. All NSAIDs inhibit the COX enzymes responsible for prostaglandin (PG) synthesis, but the COX-2–selective agents, as compared with COX-1, preferentially inhibit COX-2. The COX-1 enzyme is constitutively expressed and primarily maintains homeostatic function, especially in the gastrointestinal (GI) tract, platelets, and kidneys. COX-2 is an inducible enzyme in inflammatory states, but it is also expressed in small amounts in certain tissues, including the kidney, brain, and bone, when no apparent inflammation is present. Therefore, it is generally accepted that the risk of GI toxicity is reduced when COX-2– selective NSAIDs are used. The introduction of the COX-2–selective NSAIDs—celecoxib (Celebrex, Pfizer), and rofecoxib (Vioxx, Merck)— to the U.S. market in 1999 led to an increase of almost 50% in the total number of NSAID prescriptions dispensed. The number of prescriptions increased from 62 million (in 1998) to 92 million (in 1999), with COX-2 drugs accounting for two thirds of this increase. A third COX-2–selective NSAID, valdecoxib (Bextra, Pharmacia/Pfizer), became available in 2001, but it had little effect on the total number of COX-2 drugs prescribed annually. Recent reports of adverse effects associated with the use of COX-2–selective and some nonselective NSAIDs, followed by the market withdrawal of rofecoxib (in late 2004) and valdecoxib (in 2005), have influenced clinicians to change their prescribing practices. Our goal in this article is to characterize NSAID usage before and after the withdrawal of rofecoxib and valdecoxib from the market.