As seen in this Metafilter post asking for what to spend $3,000 of a Flexible Savings Account on come a couple of responses that illustrate your favorite and mine… statistics! (I’m kidding. But if you need evidence that it’s difficult for everyday people to make informed health care choices, look no further than that post.)

Here’s the problem with medicine: our tests are not perfect. They mess up sometimes. Granted, we try to choose the best tests so that they (or we) mess up less often, but it’s inevitable in some cases. So a test that looks for colon cancer may every so often tell someone they’re normal when they actually have colon cancer. Or a test to decide if someone has HIV may tell someone they have it when they really don’t. There are more benign examples, but these illustrate why us medical folk are so obsessed with data.

And here’s where doctors come in: they have to interpret the tests. I’m simplifying it, but doctors basically take their current suspicion that you have a disease, and then use a test to make them more likely to believe you have a disease, or less likely that you have a disease. Take me: I’m a pretty healthy 25 year-old. If I come to the doctor and say I’m having a heart attack, he’s probably not going to believe me. He’ll do a test, maybe just to make sure (an EKG?), but even if it’s positive , he probably still won’t believe me. Sure it’s possible, but the odds that a healthy 25 year-old is having a heart attack are so low that it’s much more likely that it’s some other disease I have. (If I were a 60 year-old diabetic obese man with high cholesterol and smoked 3 packs a day for 40 years, however, the doctor might be a bit more concerned.)

Now if that made sense (it’s hard to explain it without getting into numbers, which scare me), you’ll understand why getting a full-body CT or MRI scan is pretty pointless. Or even a “treadmill test,” as another poster recommends. If you’re totally asymptomatic and healthy, and don’t have any risk factors for diseases, even if there’s an abnormal little speck on an MRI, your doctor probably won’t be too concerned, because odds are, it’s probably nothing. There’s a term for these specks or masses–incidentalomas–because they’re found incidentally, and usually turn out to be nothing. A ton of us have them–in our pituitary, our thyroid, our adrenal glands–anywhere. The problem is that once you find one of these, often you want to figure out if it truly is benign (usually it’s nothing), and this causes a ton of anxiety, worrying, and health care spending as a result.

If you’re interested, these test qualities are known as the “positive predictive value” or the “negative predictive value.”

5 Comments »