Don't just order tests
Ezra Klein writes:
. . . medicine is also used to comfort. Take a patient with heart palpitations. Odds are they're just benign skipped beats. A doctor may even know those odds. But when your heart jumps, it's scary. So the doctor runs the set of tests that distinguish them from deadly arrhythmias. These tests are, in Hanson's telling, wasted medicine, as they do nothing to improve biological function and are very costly. But the assurances they offer do much to improve quality of life, which is, along with extending the length of life, rather the point of medicine.
There will always be some level of "wasted" medicine that isn't, at the moment of prescription, sure to be waste (i.e, a diagnostic that could find a deadly disease, but doesn't), and some amount of medicine that's used to calm fearful patients. Neither of those show up on yearly physicals, but nor are they necessarily wasted dollars if your metric is improving patients' quality of life.
Funny that Ezra uses this example, we just used it today in medical school. A good doctor won't order tests merely for comfort. Unnecessary tests can in fact lead to more discomfort for a patient. In ordering tests, a physician must always remember the likelyhood of a false positive. The physician has to take into account the prevalence of the disease within the patient group to which the patient belongs.
For example, say this patient is a high school student. The prevalence of heart disease within this group is very low. For the purposes of this example, we'll say it's 1%. We'll also pretend that the tests Ezra refers to have a 90% chance of being positive if the patient has heart disease and a 95% chance of being negative if the patient does not. Statistics will tell the doc that, for any individual patient in this population, a positive test result in this case means that there is an 85% chance that the patient does not have heart disease. If the patient were an older gentleman (and the prevalence of heart disease in his group is 40%), there would be a 92.3% chance that a positive on the exact same test means that the patient has heart disease.
Its the doctor's job to know this. An unneccessary test could lead to an unneccessary proceedure. If you get a false positive, the next step might be a heart catheter, and that wouldn't do too much for the patient's comfort.
Here's a link to the original Hanson article.
alejandrogonzalez.typepad.com
Don't just order tests (part 3)
Megan McArdle writes the following:
Why does American medicine do so many blood tests, X-Rays, EKGs, and so forth? You can't blame it all on lawsuits; my doctor didn't test me for hyperthyroidism because she was afraid of the malpractice suit that would result from my losing too much weight and getting heart palpitations. Nor can you blame it on money; my doctor doesn't profit from giving me blood tests that all come back normal. And I don't think the lack of rational rationing can be the culprit either. To the extent that insurance companies have bad incentives, it should be to do too little, not too much. They should have incentives to ration this sort of thing, but they don't.
Lawsuits and testing were something we just finished discussing in medical school. Lawsuits can affect a doctor's decision to a test. Take this example I used earlier:
For example, say this patient is a high school student. The prevalence of heart disease within this group is very low. For the purposes of this example, we'll say it's 1%. We'll also pretend that the tests Ezra refers to have a 90% chance of being positive if the patient has heart disease and a 95% chance of being negative if the patient does not. Statistics will tell the doc that, for any individual patient in this population, a positive test result in this case means that there is an 85% chance that the patient does not have heart disease.
Pretend I order it. My epidemiology professor says I'm an idiot for doing so, but mom is really concerned and she saw the test on TV. The test comes back positive, but I know there's an 85% chance that it's a false positive. Doesn't matter. Now I've got a positive test result, and I've got to do something about it. What if the he actually does have coronary artery disease, something happens, and I end up in front of a jury with the plaintiff's attorney asking "Doctor, you had a positive test result and you did nothing?"
That "something" probably means ordering the gold standard test, a coronary angiogram. It's a low risk procedure, but my professor has seen one case where an experienced cardiologist perform this test and the patient died. Granted, that's exceedingly rare, but it's still something to think about.
alejandrogonzalez.typepad.com |
