Recent headlines heralded unexpected benefits of a cholesterol-lowering drug: “Crestor Cuts Risk of Clots by 43 Percent.”
Crestor (rosuvastatin) made news because deep vein thrombosis can cause blood clots to lodge in the lungs. This is a potentially deadly complication that kills about 100,000 Americans annually.
Cutting the risk by almost 50 percent seems like an extraordinary bonus. But such numbers can be misleading. The analysis that led to such enthusiastic headlines refers to something called “relative risk.” Most people don’t have a clue what it actually means.
Of the 17,802 healthy men and women who participated in this study (the JUPITER trial), 94 had a blood clot during the multi-year study. There were 60 among those taking a placebo and 34 among those on Crestor.
Think about those numbers for just a moment. People taking Crestor suffered blood clots at a rate of 1.8 per 1000 people per year (New England Journal of Medicine, online March 29, 2009). Those on placebo had blood clots at a rate of 3.2 per thousand people year. That is called “absolute risk.” The difference between them reflects a 43 percent reduction—the relative risk.
What this means in real life is that if you are healthy, your risk of suffering a blood clot is very low. A medication to reduce that low risk may not warrant the hoopla.
Drug companies are good at taking advantage of relative risk numbers in their marketing efforts. Ads for Lipitor proclaimed: “In patients with multiple risk factors for heart disease, LIPITOR REDUCES RISK OF HEART ATTACK BY 36%*…”
On the surface, this seemed fabulous. Reducing the risk of a heart attack by a third is something a lot of people would like to do. They might interpret this to mean that out of 100 people, Lipitor would protect 36 from having a heart attack. Those would indeed be impressive odds.
But the asterisk told a different story. In fine print the ad went on to explain: “That means in a large clinical study, 3 percent of patients taking a sugar pill or placebo had a heart attack compared to 2 percent of patients taking Lipitor.”
What this means is that if 100 people took the drug and 100 people took the placebo, there would be 2 heart attacks among Lipitor users and 3 among placebo takers—1 fewer heart attack over the course of the study (which actually had many more people and ran longer than a single year). In other words, 99 people out of 100 who took the drug did not get any obvious heart attack protection. All of a sudden the odds don’t seem as appealing as a 36 percent relative reduction in risk.
Think of it another way. Imagine that you had to go in for gallbladder surgery and the surgeon told you that your likelihood of getting benefit was 1 out of 100. Would you be willing to sign a consent form, pay for the procedure and accept the risk of major surgery for such a modest chance of benefit? Most people would not.
Fortunately, surgery is usually far more successful than that. Pharmaceutical benefits, on the other hand, may be harder to assess. Before starting on any treatment ask your doctor about absolute risk reduction. Don’t rely on commercials bragging about relative risk.