Liars Figuring (April 13, 2021)

"Figures don't lie, but liars sure can figure" is a wonderfully penetrating quote that is generally attributed to Mark Twain. (There ARE party poopers out there who insist the expression is much older than that. Pooh to them.) It is so simple to twist or shade a number to make it say something different from what it really does, if you are inclined to do so. And that reprehensible activity is nowhere more common than in the health biz.

Many of us are spooked by numbers to begin with, and therefore all too easily intimidated by people who seem to be on good terms with them. What that means that all an authoritative-sounding person has to do is toss an impressive figure out there, and we snap it up like hungry carp and let it guide our thinking. That is a bad path to follow, especially when it comes to matters of health and disease.

In my somewhat informed opinion, I would say that the most fudged figure that I run into in the medical/wellness field is that of "relative risk". It may not be the absolute lyingest but it is pretty darned close, and it IS the one that frosts me the most.

"Relative risk" is -- take a deep breath here -- a ratio of the probability of an event occurring in the exposed group versus the probability of the event occurring in the non-exposed group. Quick, an example! Maybe the occurrence of lung cancer (the "event") in smokers (the "exposed group") versus nonsmokers (the "non-exposed group").

If 1 out of 100 nonsmokers develops lung cancer, but 20 out of 100 smokers develop lung cancer, the "relative risk" is 20; meaning, smokers are 20 times more likely to develop lung cancer than nonsmokers. (Those figures aren't too far off the actual ones, by the way.)

Being twenty times more likely to get a fatal disease because of something you're doing IS scary. But it's not the relative risk alone that makes it scary. It's the ABSOLUTE risk -- the likelihood of actually getting lung cancer -- that makes the relative risk genuinely frightening. About 250,000 people per year in the U.S. are diagnosed with lung cancer. That means that every year about 12,000 nonsmokers and 238,000 smokers hit the lung cancer jackpot. Anyone can see easily how quitting smoking, or never starting in the first place, might really pay off for someone.

But, suppose. Suppose that only a hundred people (rather than 250,000) a year actually developed lung cancer? Since the relative risk is still 20:1, that would mean that only about 95 smokers and 5 nonsmokers per year would get lung cancer. Although you've got the same RELATIVE risk, the ABSOLUTE risk (of actually winding up with lung cancer) is so much smaller that most smokers would probably decide that it wasn't worth the trouble of quitting.

So. Whenever you hear that your chances of something are greater (or less) if you do or don't do something, even if it sounds really dramatic, that's not enough to go on. You have to know more than the odds of winning or losing. You have to know what the actual payoff would be as well.

Scaring people into committing to expensive and/or risky treatments with marginal or no evidence of benefit is a specialty of the medical profession, and is one of the reasons medical error is the third most common cause of death in the U.S. There have been multiple inquiries into the prevalence of useless or even harmful medical interventions, and you can learn more about that here:

https://www.sciencealert.com/recent-study-finds-400-medical-devices-procedures-or-practices-that-are-ineffective

The flip side to scaring people into getting unsafe or unproven treatment is, of course, scaring them AWAY from things that might actually be helpful. Again, using risk deceptively is a really useful tool for someone who wants to do this. It's used all the time to discourage people from taking herbs and nutritional supplements. "Twelve people a YEAR die from taking nutritional supplements! (Of course they've mostly just choked to death when trying to swallow the pills, and it's the same number of people who are killed and eaten by armadillos, but STILL...)". This is leveraging the idea of risk for propaganda purposes, and shame on anyone who does it.

Here's what sent MY Merde Meter clanging this past week, and prompted this entire diatribe. An article discussing the various SARS-CoV-2 vaccines was saying that the Pfizer vaccine was twice as likely to produce a severe allergic reaction as the Moderna vaccine. A no-brainer when it comes to choosing one then, eh?

No, not really. Because if you finish the article (always finish the article), it turns out that to date the Pfizer vaccine has racked up about 4.7 severe allergic reactions PER MILLION DOSES versus Moderna's 2.5. Hardly anything to get your panties in a bunch about. Twice as likely (relative risk), but still next to nothing (absolute risk).

Most people are a little nervous about medical interventions these days, period. And that is wise, when you consider that medical error is the third leading cause of death in the United States (have I mentioned that recently)? But when you are looking at the risk of doing something, you must look not just at the relative risk of it, but also the absolute risk of it, and then -- finally -- the risk of not doing it at all. In the case of any vaccine specifically, then, you're looking at the risk of reaction to the vaccine versus the risk of getting the disease PLUS the seriousness of the disease. Not all that hard to figure out.

--dr. diane holmes
  Copyright © 2021