The perils of speculative medical testing: dealing with an answer to a question we didn’t ask

I’m a fan of the cartoonist who draws xkcd, which the author describes as “a webcomic of romance, sarcasm, math, and language.” This particular one well illustrates something that happens a lot in medicine, especially the ICU, because we order a lot of tests. We get back a test result that is nothing at all what we expected, one which gives an answer to a question we didn’t even ask.

Patients (and their families) not uncommonly want us to do a lot of tests, especially if what’s wrong is not obvious. Why not order a large panel of blood, urine, and x-ray tests? We might turn up something that will help. We might. But we also may turn up something that is unanticipated and totally unexpected. Once in a while this is useful and gives us the answer. But many other times we are puzzled by the result, and to resolve the puzzle we order even more tests, maybe ones that we never would have ordered otherwise. These new tests may turn up yet another unexpected result or two, which we pursue with yet more tests. You can see where this is going.

The best way to prevent this cascade of unintended consequences is to avoid scatter shot testing. We need to make sure there is some logical reason for the test, often from the patient’s story or examination, that makes sense. All medical tests have possible error built into them. It is often said if you order twenty tests you’ll get one that is a false positive, potentially leading to a wild goose chase.

If a doctor orders a test for you or your child, ask just what — quite specifically — the test is for. More care is not better care — it can also be worse care.


Leave a Reply

Your email address will not be published. Required fields are marked *