U.S. trends in prescription nonsteroidal anti-inflammatory drug use among patients with cardiovascular disease, 1988-2016.

Abstract

Nonsteroidal antiinflammatory drugs (NSAIDs) have been associated with increased risk of adverse cardiovascular events prompting labeling revisions cautioning their use among patients with cardiovascular disease (CVD). However, little is known regarding long-term trends in real-world prescribing of NSAIDs within the CVD population. We aimed to characterize the use of prescription NSAIDs among U.S. adults with CVD from 1988 to 2016. We used the National Health and Nutrition Examination Survey cross-sectional data from 1988-1994 and 19992016 to identify participants aged greater than or equal to 18years with hypertension (defined by self-report, mean blood pressure ≥140/90mm Hg, or antihypertensive medication use), or aged greater than or equal to 20years with self-reported congestive heart failure (CHF), coronary heart disease (CHD), angina, myocardial infarction (MI), or stroke. Prevalence of prescription NSAID use was analyzed in 6-year examination periods. Weighted logistic regression was performed to test time trends in prescription NSAID use. Overall, prescription NSAID use declined among all CVD populations. The highest prevalence of overall prescription NSAID use was observed during the 1999-2004 examination years, thereafter declining through the 2005-2010 and 2011-2016 examination years: in patients with hypertension (13.9% [1999-2004] to 8.5% [2011-2016]), CHF (14.6%-8.5%), CHD (16.3%-7.4%), angina (17.6%-8.5%), MI (16.1%-9.0%), and stroke (15.7%-7.9%). Decreased use of COX-2-selective inhibitors was observed during the same period; whereas, nonselective NSAID use remained relatively stable. Trends in prescription NSAID use were reflective of the general adult population. Prescription NSAID use among patients with CVD appears to have declined from 1988 to 2016, primarily due to reduced COX-2-selective inhibitor use. Nonetheless, the prevalence of prescription NSAIDs has persisted among a subset of high-risk CVD populations.