Abstract
OBJECTIVE: To investigate recent national trends in nonsteroidal antiinflammatory drug (NSAID) and acetaminophen use for osteoarthritis (OA). METHODS: Using data from the 1989-98 National Ambulatory Medical Care Survey, a representative sample of US office based physician visits, we assessed 4471 visits by patients 45 years or older with a diagnosis of OA. We examined cross sectional and longitudinal patterns of OA pharmacotherapy. The independent effects of patient and physician characteristics on NSAID and acetaminophen use were examined using multiple logistic regression analysis. RESULTS: Pharmacological treatment for OA (either NSAID, acetaminophen, or both) has steadily decreased from 49% of visits (1989-91) to 46% (1992-94) to 40% (1995-98) (p = 0.001). Reduced NSAID use over this time period (46% to 33%; p = 0.001) was partially offset by a modest increase in acetaminophen use (5% to 10%; p = 0.001). Among individual NSAID, ibuprofen (5.7% of OA visits), nabumetone (4.9%), naproxen (4.6%), and aspirin (4.4%) were the most frequently reported in 1995-98. For patient visits in 1995-98, 45 to 59-year-olds (38%) received NSAID more often than 60 to 74-year-olds (34%) or patients older than 75 (28%; p = 0.029). Other possible predictors of OA therapy included patient race and physician specialty. CONCLUSION: The decline in the use of NSAID from 1989 to 1998, especially among elderly patients, and the frequent selection of safer medications may reflect awareness of the literature citing the risks of nonsteroidals for OA. However, variations in prescribing patterns among different patient populations and the modest use of acetaminophen, despite evidence supporting its efficacy, suggest that better assimilation of the literature into medical practice is needed to optimize OA therapy.
