Report Absolute Risk!

Lisa Schwartz, Steve Woloshin and colleagues at the Veterans Administration hospital in White River Junction, Vermont are among my favorite researchers because they continually point out the flaws in standard medical reporting. Today's British Medical Journal carries their report on the failure of medical journals to note the absolute risk of a disease and an intervention when reporting the results of clinical trials.

What is absolute risk? Think of all the times you read a news account of a new drug that said it CUTS the risk of heart attacks IN HALF! Sounds pretty good, doesn't it? But what if you knew that only 1 in 10,000 people who didn't get the drug got a heart attack. That means the drug cut the risk from 1 in 10,000 to 1 in 20,000. Would you pay a thousand bucks a year for that level of protection?

Accurate reporting of such a study would not only report that the relative risk was cut in half, but that absolute risk was just one-one-hundredth of one percent (0.01%), and the drug cut the risk to five-one thousandths of a percent (0.005%).

Another way to report absolute risk is the number needed to treat (NNT) in order to reduce an incidence of a disease. So in this example, the number needed to treat to eliminate one heart attack is 20,000. This is an import concept for calculating the cost-effectiveness of treatments. Again, using this example, if the drug in question cost a thousand bucks a year, the health care system would have to spend $20 million to eliminate one heart attack.

Thinking about absolute risk allows clear thinking about the relative value of various health care interventions. So what did Schwartz and Woloshin find? They looked at 222 clinical trial reports in various journals and found that 68 percent failed to report the absolute risks in the abstract and half of those or nearly a third of the total did not report the absolute risks anywhere in the article.

Their conclusion: "Absolute risks are often not easily accessible in articles reporting ratio measures and sometimes are missing altogether—this lack of accessibility can easily exaggerate readers' perceptions of benefit or harm."

If that's what is happening in the journals, what is going on in the press? The answer is obvious. Most reporters never even ask, much less report the absolute risk numbers. It's an easy concept to master, and an easy question to ask the clinical investigators when reporting on the results of clinical trials. Keep that in mind the next time you read a news account of the latest "breakthrough" in medicine. Ask yourself, "Have they reported my absolute risk of suffering from this disease?"