I’ve just finished sitting through a wonderfully aptly named lecture: Probability and Statistics, in which, among other things, we learned (again) that the utility of various clinical tests depends at least as much and generally more on the patient and condition involved than on the specific test itself. From stress tests to mammograms to PSAs, the relationships of true and false positive and negatives, positive and negative predictive values all hinge on the prevalence of disease; or how likely is it that a given condition is present before you even do the test. Lots of times when you crunch those numbers, the best answer is not to do the test.

Perhaps not unsurprisingly, that tends not to go over real well with patients.

I think we do too many mammograms. Many people agree with me. I spend lots of time talking with my patients about why I don’t think they need a mammogram. Many women understand, but many others don’t. So I order it.

I try not to do routine PSAs. I preface the blood draw with a discussion about how he’s more likely to die with rather than of prostate cancer, and about the risks of incontinence and impotence as complications from its treatment. It’s often enough to talk them out of it, though not always.

I try so hard not to do stress tests unless I’m really worried about heart disease. Note that it has to be me who’s worried about it. Just because the patient is worried doesn’t impact the decision all that much. Far too often, the clinically unnecessary stress test only sets us on the treadmill (sorry about that) of sequential follow-up testing up to and including cardiac catheterization, with risks for all of those attendant complications (including death).

Unfortunately, there are still people who don’t understand what I’m trying to tell them, and sometimes walk away appalled that I “don’t think I need to know if I have cancer.” I know this because I’ve received scathing reviews on several online doctor rating sites saying precisely that.

Why?

How can knowledge be bad?

It’s a philosophical question that’s really hard to answer. Why wouldn’t you want to know about every possible little thing in your body that might hurt you? (cue the talk on incidentalomas) The answer, of course, is the phenomenon of overdiagnosis. Treating things that will never hurt you, like small prostate cancers in old men and and possibly DCIS (ductal carcinoma in situ, or so-called stage 0 breast cancer) exposes you to all the risks of treatment with none of the benefits, but it doesn’t feel like it. And that’s the problem.

I think the hidden agenda here is not dying. If you find or prevent every possible thing that can kill you, be it a subclinical cancer or atherosclerotic arteries, then you won’t die. Obviously it’s not true. But dying is one of those things to which we pay lip service: “Of course I know I’m going to die, but I really don’t want to.”

It doesn’t work that way, and I’m sorry. I still won’t recommend unnecessary screening tests, even at the cost of my online reputation. As for my patients, knowledge isn’t always more blissful than ignorance.

Lucy Hornstein is a family physician who blogs at Musings of a Dinosaur, and is the author of Declarations of a Dinosaur: 10 Laws I’ve Learned as a Family Doctor.