Monday, March 26, 2012
When It Comes To Reading About A New Medical Study, I Say Caveat Lector.
There are so many potential and intrinsic problems with medical study results that they could fill the pages of a book. In fact, I plan on writing such a manuscript in the not too distant future to help the lay person make some sense of the never-ending blizzard of medical studies whose results when reported by the media seem to confuse everyone, including doctors.
In the interim, here’s one of the key problems with the reporting of study results worth noodling on-the issue of relative risk versus absolute risk. For example, a recent headline trumpeted that depression plus diabetes doubles one’s risk for dementia. Wow, that sounds like a big deal to double the incidence of the fearsome dementia. However, while the study showed a 100% increase in risk, the actual rate for dementia only increased from 1% to 2%. In terms of relative risk, the incidence rate doubled. However, in relationship to absolute risk it merely increased 1%. That’s two very different sides of the same story.
Another example applies to Statins, the blockbuster cholesterol reducing drugs that decrease cardiac events 33%. However, that is the relative risk reduction. An absolute risk reduction analysis, however, tells a different story. Would it surprise you to learn that 100 people need to take a statin drug for one person to benefit? Statin studies show that cardiac events are decreased from 3 out of 100 to 2 out of 100 patients, which means that 100 people must take the drug and potentially suffer its side-effects, for one person to see a reduced risk for a heart attack or stroke.
A third example, as cited in a recent Wall Street Journal article titled When Risk Is a Red-Meat Issue by Carl Bialik speaks to the health risks associated with eating red meat. Although, I am not a fan of any significant meat consumption, I can appreciate the author identifying how the media’s handling of the study’s results was misleading as headlines focused on a relative death risk increase of 20% versus an absolute death risk rise of 0.2 percentage points. That’s a big difference.
I’ve read some 10,000 studies over the past year or so and it’s amazing to me how many ways others can interpret the same studies. That why people often are confused by new studies because to grab headlines the media often chooses the most contentious, surprising, or alarming aspect of the study, instead of the most actionable aspect. Unfortunately, the lay person never reads the actual study behind the headline and may not even get past the headline to read the study details. And that assumes that the article writer actually provides all the details of both the relative and absolute benefits.
As I said earlier, that are many deceptive aspects of reporting studies and until someone actually teaches people how to make sense of the avalanche of studies, the best advice really is the Latin phrase, Caveat Lector translated to “let the reader beware.”