By Rebecca Burkholder, NCL Vice President for Health Policy
When we read a headline about a new study that shows a drug or procedure will reduce the risk of cancer by 10 percent, what does that really mean? It can definitely be challenging for consumers to interpret the evidence for themselves.
At the 10th Annual Cochrane Collaboration Meeting I attended last week in Colorado, how we look at evidence and interpret it was the subject of several sessions and workshops. For consumers to be able to interpret evidence, one of the basic concepts to understand is how study results are reported.
Steve Goodman, with Johns Hopkins Medical Institutions, explained at one session about the difference between relative and absolute risk and that most study results are reported in terms of relative risk. So why does that matter? There are two ways to report risk:
- Absolute risk: your own risk
- Relative risk: used to compare the risks in two different groups of people
Here is an example when talking about how a treatment may reduce the risk of a disease. Say women have a 4 in 100 risk of developing a certain cancer. Then, research shows that a new drug reduces the relative risk of getting the cancer by 50 percent. Sounds pretty good, right? It means that the “4” in 100 is reduced by 50 percent to”2” in 100. So, the absolute risk of getting the cancer only dropped by 2 percent (4 percent minus 2 percent equals 2 percent ).
Which headline sounds better?
“New drug reduces risk of cancer by 50 percent !”
“New drug results in 2 percent drop of cancer risk!”
Both of these are reporting on the same study results.
Absolute risk is generally better understood by the public, but relative risk is more often used to report results since is it usually a bigger number and looks a lot better than absolute risk.
Next time you are looking at a study on the Internet or hear about one on the news, make sure you are taking a critical look at how the results are reported, especially the risk that is reportedly being lowered. What does it really mean for you?