Ban All Relatives. Hope you like our new campaign and will join BAR in getting rid of them. No, we’re not planning to usurp uncles or mount an attack on aunties; we all love those relatives, and they make a mean Victoria sponge when we visit.
BAR is campaigning to ban the use of the relative risk ratio (collective sound of comatose people slumping into their armchairs). But hold! It’s the rocket fuel of the pharmaceutical industry, it drives the sale of new drugs—where much of Big Pharma’s profits are to be made—and gives the mass media exciting new things to talk about.
The relative risk ratio makes it seem that we’re all very busy being brilliant and pushing back the boundaries of disease, when actually we’re not, of course.
Let’s take as an example a news report you read: “New Drug X Reduces High Blood Pressure by 30 Percent.”
“Blimey!” you say, dropping your Twitter feed, “I’ll have some of that.” By the by, when we say “you,” we’re describing the average Everyman, and not the discerning WDDTY reader.
Eric Everyman rushes to his local surgery, Twitter feed in hand. He shows his doctor the impressive data on the new drug. The doctor confirms that indeed the drug is truly that effective at reducing high blood pressure, as he was assured of the fact by the drug company itself while on a vital, and all-expenses paid, conference in the Seychelles, fully sponsored by the manufacturer.
But it’s the illusion of the magician, whose sleight of hand is achieved with the use of relative risk. When people read that a drug has 30 percent effectiveness, they assume the drug improves a condition by, er, 30 percent. It doesn’t.
Let’s assume that Eric Everyman is 40 years old, doesn’t smoke, drinks only moderately and exercises by cycling on weekends. His risk of developing high blood pressure (hypertension) is around 10 percent. In other words, he is 10 percent more likely to develop hypertension than someone of his age who is fitter and, as a result, has a lower risk profile.
If Eric were to take New Drug X, we know his risk would drop 30 percent—but it’s 30 percent of his actual risk of 10 percent. In other words, the drug reduces his risk of hypertension by 3 percent—not the sort of headline rate that would have him rushing to his doctor. So, 30 percent is relative to someone’s actual risk, or “absolute risk,” as statisticians like to call it.
A recent study makes the point. Headlines that reported on the research exclaimed that women who were using a progestogen-only contraceptive, such as the Pill or a coil, were 25 percent more likely to develop breast cancer.
Again, it’s true, sort of—but that’s the relative risk a woman faces. A woman in her late 20s has an absolute risk of developing breast cancer of just 0.5 percent, and taking the pill increases that risk to 0.57 percent. By the time she reaches her late 30s, her risk of cancer is around 2 percent—still very low—and the Pill and coil raise that to 2.2 percent.1
Researchers from Oxford University monitored the health of around 30,000 women, some of whom were taking a progestogen-only contraceptive. The incidence of breast cancers in women using contraceptives for 15 years increases from 0.084 percent to 0.092 percent overall, the researchers concluded—but that didn’t make it into the headlines, the only thing many people ever read. Instead, reporters used the relative risk figure for that.
Everyone loves relative risk. It gives the media something to report on, it boosts drug company sales, and it gives academics a turn in the spotlight for a study that otherwise would be so underwhelming that it would be consigned to academic oblivion.
Mind you, drinking alcohol is risky. One alcoholic drink every day increases the risk of breast cancer by 10 percent. Worried? Not as much as the researchers would like you to be—they’re using relative risk again.
