Skip to:

COX-2 inhibitors were prescribed for millions who didn't need them, study finds

Millions of Americans who were prescribed Vioxx, Celebrex and other COX-2 inhibitors from 1999 to 2002 didn't need the medications and could have safely taken older, cheaper painkillers such as ibuprofen, according to a study co-authored by CHP/PCOR fellow Randall S. Stafford, published in the Jan. 25 issue of the Archives of Internal Medicine.

When Vioxx began being sold in 1999, it was touted for relieving pain without causing the gastritis and ulcers that some people developed from taking ibuprofen, naproxen and other painkillers known as non-selective non-steroidal anti-inflammatory drugs, or NSAIDs.

But over the next four years, it turned out that Vioxx was adopted way beyond that market niche: millions of people who had little risk of gastrointestinal bleeding ended up getting prescriptions for Vioxx, Celebrex and other medicines in their class, known as COX-2 inhibitors.

The overuse of these drugs is documented in a study published in the Jan. 25 issue of the Archives of Internal Medicine, co-authored by CHP/PCOR fellow Randall S. Stafford -- associate professor of medicine at the Stanford University School of Medicine -- along with University of Chicago researchers G. Caleb Alexander, MD, and Carolanne Dai. The researchers' findings reveal that 63 percent of the growth in COX-2 use from 1999 through 2002 occurred in patients with minimal risk of suffering gastrointestinal bleeding from NSAIDs.

Stafford said the problems associated with COX-2 inhibitors should serve as a cautionary tale about the growing trend of turning custom-fit medications into one-size-fits-all remedies. The researchers attribute the overuse of the drugs to several non-clinical factors that have spurred sales of other drugs as well - including heavy marketing and the tendency of patients and physicians to assume newer medicines are better.

"This phenomenon is not limited to COX-2 inhibitors," Stafford said, noting that it also happened with drugs for hypertension, diabetes and some infections. "There are a number of instances where use has expanded beyond the narrow clinical situations in which the drugs are most effective and cost-effective."

The medical profession has a term for expanding the use of a drug beyond its intended target population: "therapeutic creep." In the case of COX-2s, not only does this expansion mean that millions of people paid more for a drug without reaping a substantial medical benefit, but it now appears they unnecessarily exposed themselves to the risk of heart problems, Stafford said.

COX-2 inhibitors have been at the center of a controversy since last September when Vioxx was pulled from the market after clinical trials showed that it posed significant risks of heart attacks. Similar findings were also recently reported for Celebrex, although it is still available.

What makes the widespread use of COX-2s so troubling, Stafford said, is that they aren't any more effective at controlling pain than NSAIDs. While they were seen as a welcome alternative for those at greatest risk of gastrointestinal side effects from NSAIDs, they are more expensive. A daily dose of Vioxx at $2.64 costs about six times as much as a daily dose of ibuprofen at 42 cents, according to a 2002 compendium of drug prices.

For the study, Stafford and his colleagues looked at national databases that tracked patient visits to their physicians between 1999 and 2002 and the types of medications being either continued or prescribed. Additionally, they used a tool developed at Stanford in the 1990s to categorize patients according to their risk of gastrointestinal bleeding from NSAIDs.

According to the study, 73 percent of the patients had either a very low or low risk of GI bleeding from NSAIDs, meaning that there was no pressing medical reason for them to switch to COX-2 inhibitors. However, this was the group where the greatest growth in COX-2 use occurred. The study showed the number of physician visits associated with COX-2 use in these patients increased from approximately 9.5 million in 1999 to 20.9 million in 2002, accounting for 63 percent of the growth in the drugs' use during that time.

Patients most likely to benefit from COX-2s-those with either moderate or high risk of GI bleeding-accounted for the remaining 37 percent increase.

The study results were covered by several newspapers, TV news broadcasts and news Web sites around the country, including USA Today, the Los Angeles Times, the Chicago Sun-Times, the Seattle Times, the Palm Beach Post, Reuters, Health Day News, American Medical News, WebMD,, CBS MarketWatch,, CNN, and several TV stations including those in Los Angeles, Chicago, Hartford, Conn., and Memphis, Tenn.