Most tests are probabilistic and you have to think about the conditional probability given other indicators. Consider a urine culture for detecting UTIs. If you just get urine samples from a general population, some fraction will show positive (bacteria in the culture) falsely. It would be pointless to perform this test on the general population and to treat people that show positive . But if you perform this test conditioned on the fact that the patient is presenting other symptoms (burning sensation, fever etc.) then the same test very useful and the conditional false probability is low enough that it can be used to decide on treatment.

The same idea applies to most tests in medicine. That is why tests are mostly done conditionally.

But why can't they go the other way? Oh, you had bacteria in your urine, do you feel a burning or have a fever? You'd have to find a way to ask that doesn't incite a false response, but it still seems like it would be useful.

>Why can't a doctor see an anomaly and decide not to chase it down based on the same reasoning that they use to not scan for it

Mostly it is just how the medical industry works. A previous co-worker of mine had a wife that ran a small family practice. He was always griping about the money they spent on insurance for the place ($20k a year or quarterly, can't remember). I asked why it was so much, and case in point, he told of a story just in the prior year of a woman that came in for a knot on her breast. She was informed to come back in a month as sometimes hormones can cause them to come and go. Well they never saw her again but a year later they got a malpractice lawsuit from her since she now had breast cancer...

Some folks have raised plenty of practical/medicolegal reasons, but it's also worth noting that patients are often not comfortable being told that they have an "anomaly" that their doc is going to sit back and do nothing about. There's a lot to be said for the psychiatric implications of being aware of benign MRI findings.

Even if you (or anyone, for that matter) might be comfortable with that, I think you'd find that a lot of patients wouldn't be.

(1) Because we do not have the body of knowledge that tells us how to predict an anomaly is not worth pursuing. It would require a lot of major, expensive, long-term studies involving non- negligible chemical exposures in$ asymptomatic people in order to get a good idea of what the “healthy” anomaly looks like.

$ many of these MRIs require contrast material.

(2) Malpractice for a missed cancer in most states has an enormous lookback period and an automatic “doctor loses.” It doesn’t matter if the doc made a statistically appropriate call. So docs follow-up on anything but the most unambiguously benign lesions.

> statistically appropriate call.

but factually incorrect call is apparently OK, because they are treating "median person", and not the actual patient.

This has to do with money - and nothing else.

No, it has to do with avoiding harm to patients. I don’t make one red cent more or less from sending someone to an MRI or not. I do care about wasteful testing that is more likely to stress out a patient and send them down a rabbit-hole of follow-up tests with potential complications and costs (to them), if there’s not a reasonable chance of an outweighing clinical benefit for them.

And of course we make statistical calls! When you’re deciding -prospectively- whether a course of testing or treatment is in a patient’s best interest, you have to look at the stats on likely benefits and harms of various courses of action. I’d love to know what my patients have before hand, so I could avoid dealing with probabilities, but I haven’t been blessed with that particular power. I’d love for the data to exist to tell me how to interpret a result in my precise patient rather than a larger population she belongs to, but “solitary pulmonary nodules in 33-year old men who smoked twice in college and live half a mile from a freeway with two episodes of bronchitis in their teens” is a study that hasn’t been conducted yet. One day perhaps it will be, and if so, I’ll be grateful for it. But right now that data doesn’t exist, so we use what we have, and use our judgment to tailor it to the patient in front of us - imperfectly.

I recommend reading “overdiagnosed” by Gilbert Welch.

> judgment

> make statistical calls!

Really? Your fuzzy feels are not statistics.

Do you even at least try to know how large is the type 2 error in your guesswork? Is it better than flipping a coin?

We had a medical student who had her abdomen scanned as part of teaching. They found a cystic looking thing on her pancreas. She ended up getting it excised - it was benign and would've never caused problems, but she had multiple complications and 2 months in hospital as a result of the surgery.

I think Doctors (I am one) definitely err on the side of overtreating in those uncertain situations. Just like politicians, it feels more caring to act than to not act.

If you MRI 1000 people and leave the vague results in the hands of a bunch of physicians, they will tend to be interventional. This will lead to a spate of overtreatment.

Ways around this would be to consent patients for watchful waiting in the case of uncertain lesions. Even so, telling people they have weird looking lumps that you're not going to treat will cause serious anxiety.

   Why can't a doctor see an anomaly and decide not to chase it down based on the same reasoning that they use to not scan for it?

In the US at least, one answer is malpractice litigation.

I think the pointless trap is often the patient. I worry less than some but more than many, and while I usually defer to my doctor's advice to leave most of life's little lumps and bumps be, it's a challenge for me to accept the unknown. I imagine that many patients will demand answers once they know that any anomaly exists, no matter how likely the benign explanation is.

Yeah, just look at the number of people that demand antibiotics, even if a doctor thinks its viral, and the number of doctors that capitulate to those demands.

Yeah, emotions, money and politics aside, more information is surely better, even if not now, in the long run as processes improve.