"A new cancer drug reduces incidence of the disease by 50%." Sounds great, right? But a closer look reveals that the drug reduced cancer from just 2 people in 1,000 to 1. Fifty percent, sure, but nothing to call home about.
This is the distinction between relative risk and absolute risk that a new Twitter account is drawing attention to. Much like its viral counterpart, @justsaysinmice, RelativelyRisky points out the fine print in scientific studies beyond the attention-grabbing headlines that the research sometimes inspires. Relative risk is a comparison — how much more risk of a bad outcome one group bears compared to another — while absolute risk is just this measure for one group.
RelativelyRisky is run by an epidemiologist and PhD student who goes by the name Gid M-K online. In a blog post on Medium, he explained that the reporting of relative risk instead of actual risk leads to different interpretations of the same results. Giant relative risks can make it difficult for a person to understand what the risk of something happening to them is; however, it is a useful tool for science, since the absolute risk of any given outcome can vary substantially based on factors like age while relative risk stays relatively steady.
Communicators of science should be more conscious of how they present risk percentages. This doesn’t mean getting rid of relative risk entirely, but reporting should at the very least be include both measures of risk to give readers a fuller sense of what the data mean. There are even studies showing that this approach helps. So next time you see what you think might be a sensationalized headline, click over to RelativelyRisky to see what the actual risk to you might be!