Search This Blog

Follow by Email

Saturday, June 19, 2010

The pseudoscience of medicine

The training that leads to becoming a physician is long, taxing and requires academic stamina and intelligence to complete. Nevertheless, most of what we eventually learn is practical: how to take care of patients in sickness and health. This is as it should be, since that is what we mostly do. Nevertheless, because we take many many hours of science related classes, most physicians consider themselves to be scientists. And that we, mostly, are not.

In my years of training I have learned how to construct a hypothesis, test it and use my data to make a conclusion. I know how to document my data, and I know how to perform simple statistical analyses.  I know how to interpret statistics I read in other peoples' work, for the most part. But because I am always looking for ways to use the science I read to help me in patient care, I often make inferences that are speculative and probably just plain wrong.  It works for me, though. I need to plug the science I read into the craft of medicine I practice in order for it to be meaningful to me, and sometimes my inferences might just be correct.

Take for instance standard clinical trials which look at the effect that a certain intervention, say a drug treatment for cancer, has on a group of people.  That clinical trial will show that the in the group getting the drug the cancer will go away for a certain percentage of the people getting it. This result will be compared to results for a placebo group or a group getting a different drug. If the group getting the drug has a higher level of response than the placebo or different drug group, the interpretation will be that the new drug works. This is where the speculation starts to be misleading. I will then tell the patient I see that this new drug works best, and the patient may then choose to take it rather that watching and waiting or taking the other drug.  But it isn't necessarily true for this patient that the drug works best, because patients are different, and withing the group that got the new drug, there are very likely patients who would have done better getting no drug or another drug. So I really can't, and shouldn't tell the patient that the drug works better. But just to make it as simple as possible, I do. And most doctors, until they sit down and think about it believe that this is true, that the drug that comes out on top in the clinical trials is the best drug, and they will proceed to use it preferentially.

I have been attempting to explain this sort of thing to my patients more often since I have been thinking about it, but I think it just makes them uneasy.  They want an answer from me: what choice is best.  Now that is not true of all patients. Some of my more thoughtful patients are glad to have many options open to them. It is more honest to discuss these things, but they are complicated and definitely not reassuring.  The use of estrogen is a frequent subject for these discussions.  Estrogen causes various harms in some people, including increasing the risk of breast cancer and vascular events when combined with progesterone, but it also saves people from breaking their hips and reduces the risk of colon cancer. It definitely helps relieve the sleeplessness and hot flashes of menopause as well.  So is it good for a woman or bad for her? I guess it depends on what she values.

Another thing that physicians do that makes us feel like scientists is we measure things.  We measure how much pressure it takes to stop the blood flowing in someone's arm. We call that the blood pressure.  We measure the number of blood cells in a cubic centimeter of blood. We measure weight, temperature, height and head circumference.  We count the number of times we feel the blood pulsing in someone's wrist per minute. We are reassured of a person's health based on these numbers. The numbers themselves may be misleading, as in the case of the blood pressure. The pressure it takes to stop the blood flow in the arm can go up if the arteries are particularly tough and springy. We don't necessarily know that this is a bad thing. The blood pressure can vary depending on recent exercise, time of day and emotional state. But that really isn't the most basic problem.  What I think is more basic is the fact that we have decided that the things we can measure, and routinely do measure, are the important things, and we mainly base our studies on these pieces of data that we have decided are important because we can quantify them.

I recognize that medicine has, at times, significantly improved the quality and quantity of peoples' lives, so disrespecting it based on its fallacies is unkind and unfair. I would really just like to see my medical profession lighten up and recognize that much of what we see as fact is not. This could nicely dovetail with the recent emphasis on what is called "shared decision making." We have come a long way from the paternalistic past of medicine, and have another long way to go.

No comments: