Fun Science Fact #22: Evidence-based medicine is like good table manners - often preached, but rarely practiced.
A few years back, I went to see my doctor for a long-overdue physical. In and among all the poking and prodding, which even now I would prefer not to discuss, he ordered a round of blood work. At that point I figured a needle in the arm was the least of my worries, so I dutifully complied. A few days later, I got a call from his office. My cholesterol was on the high side, they told me, and I needed to stop by soon to pick up a prescription for Lipitor.
I’d been with my doctor for a long time at that point, and for the most part we had a pretty good relationship, but this didn’t seem reasonable to me. I asked for a copy of the lab results, and as I expected, almost everything—triglycerides, total cholesterol, HDL cholesterol—was in the normal range. Only the LDL results were high, and not by very much. Did I really need medical intervention? Fortunately, the American Heart Association had just come out with recommendations for who should and should not receive statin therapy (you can find them here, if you’re interested: http://circ.ahajournals.org/content/early/2013/11/11/01.cir.0000437738.63853.7a). I read through them carefully, and was happy to see that I fell very clearly into the “do not treat” category.
Why did this make me happy? Statins are widely prescribed today, and are rightly credited with helping to reduce the death rate from both heart attack and stroke. Interestingly, several studies have shown that your odds of suffering a cardiac event are reduced by taking statins even if your cholesterol numbers are already in the normal range. This has led some cardiologists to half-jokingly suggest that statins should be in tap water, like fluoride. If these drugs really reduce the probability that you’ll fall victim to what is still America’s leading cause of death, why not?
The answer, of course, is that every drug has side effects. Those for statins are generally fairly mild as these things go, but they’re not non-existent. The most prominent is muscle pain and weakness, which you can address by stopping therapy. The more significant effect in terms of making treatment decisions, however, is an increase in the rate of type II diabetes in statin users. Diabetes is not as serious as a heart attack, of course, but it’s not something you’d like to pick up on whim either.
In order to make a rational treatment decision in a case like this, you need to know several things: (1) the probability that you will suffer a heart attack or stroke without statin therapy; (2) the probability that you will suffer a heart attack or stroke even with statin therapy; (3) the probability that statin therapy will cause you to develop diabetes; and (4) the relative weighting (i.e. the seriousness) of developing diabetes vs. suffering a heart attack or stroke.
Here’s an example of this sort of decision making, from an earlier visit with a different doctor. In this case, I was seeking treatment for a plantar wart on my thumb. I expected my dermatologist to pull out a dewar of liquid nitrogen and freeze it off, but instead she offered me a prescription for a pill. I was a bit wary of using a systemic therapy to treat a local problem, and asked her what the possible side effects were. There was only one, she told me: gynecomastia. Gynecomastia, in case you were wondering, is the development of female breasts on a male’s chest.
Using the rubric above, I estimated (1) that the probability of my wart not going away without the pill was actually very low—liquid nitrogen is painful, but effective; (2) that the probability of my wart not going away with the pill was probably equally low; (3) that the probability of the pill causing me to grow breasts was also low, but unfortunately non-zero; and (4) most critically, that growing breasts would be much, much worse than failing to get rid of a wart. Needless to say, I told her to break out the nitrogen.
In the case of statin therapy, the AHA has already performed these calculations, and has provided a simple method for deciding whether the benefits of statins outweigh the risks. When I presented this to my doctor, however, he responded with a story about a former colleague of his who had just recently died of a heart attack, and reiterated that I should definitely start statin therapy immediately. The fact that his colleague had virtually nothing in common with me from a demographic or medical history standpoint didn’t seem to register with him, any more than the AHA recommendations did. His recommendation was made based on emotion rather than data. I politely thanked him when he handed me the prescription. I never took it to be filled.
The upshot of this is that we have reached a point where the resources are available, if you know where to look for them, to make most medical decisions on the basis of scientific evidence rather than emotion, opinion, or personal experience. However, a regrettably high percentage of physicians still prefer to trust their intuition regarding treatment decisions rather than the available data. When it comes to taking advice on major medical decisions, therefore, it’s a good idea to trust, but verify. Your doctor probably has your best interests at heart, but you’d really be better off if he spent a bit more time engaging his brain.