New reports often reports the relative risk or benefits between two groups (say a control group versus treatment group). This number is usually deceptively higher than the absolute risk or benefit.
What is the difference between relative risk versus absolute risks?
KnowBreastCancer.org has a good explanation. Suppose that 2 person out of 100 in a control non-treated group ended up with cancer. And supposed that only 1 person out of 100 in a treatment group ended up with cancer. Hence the relative risk would be a 50% decrease in breast cancer in the treatment group. This is arrived at by taking the change and dividing by the initial value: (2 – 1) / 2 = 0.5 = 50% This is the number often reported. Sounds pretty good right?
But consider the absolute risk. Since the risk of cancer is 2% in the non-treated group, and the risk of cancer in the treated group is 1%, the treatment only reduced the risk from 2% to 1%. This is only a 1% reduction in absolute risk. Now it doesn’t sound that great. This number is arrived at by taking the change in percentage divided by the overall percentage: (2-1)/100 = 0.01 = 1%.
Tom Naughton is a comedian that did a bit at the Ancestral Health Symposium called “Science for Smart People” in which among the other topics talked about relative risk versus absolute risk…