A CT scanner. Not a good test for anxiety. Credit: Wikimedia Commons/1weezie23

Every primary care doctor has been faced with this situation. A patient reports vague symptoms and is very worried that they are a sign of a catastrophic illness. The symptoms aren’t even slightly suggestive of the disease the patient is worried about, but the patient’s neighbor’s brother-in-law was just diagnosed with the same disease, and so the patient is pretty sure that he has it too. The doctor is not at all suspicious that the patient has this disease. The doctor believes that the patient is simply anxious, and that his symptoms are either caused by his anxiety or are normal bodily sensations that are being magnified and given lots of attention because of the news about the neighbor’s brother-in-law.

What can the doctor do? One option is to order a test – a CT, a MRI, blood tests, whatever would rule out the specific disease the patient is worried about. The doctor is not ordering the test because he is actually curious about the results. He thinks the probability of an abnormal result is extremely low. He is ordering the test simply in the hopes that a normal result will reassure the patient, decrease the anxiety, and maybe even lead to the resolution of the symptoms by letting the patient focus on something else.

The temptation to order the test is pretty great (especially if the doctor owns the testing equipment). But will it work? Will the normal test result fix the problem?

A study published this week in JAMA Internal Medicine attempted to answer that question. Researchers compiled all previous published randomized trials that assessed diagnostic testing done for symptoms that were unlikely to represent serious illness. They found that on average the patients’ reported anxiety and symptom severity did not decrease after the result was normal.

So when the disease being investigated is very unlikely, ordering a test just to reassure a patient doesn’t actually reassure the patient.

It might be more effective to take the time to understand the cause of the anxiety. Perhaps the patient is actually very close to the neighbor’s brother-in-law and is himself devastated by the bad news and simply needs to express how sad he is for his friend. Or perhaps he has health anxiety (hypochondriasis) and has been to a dozen doctors in the last six months with different symptoms getting myriad normal tests. The former just needs some sympathetic listening. The latter needs cognitive behavioral therapy. Neither benefit from diagnostic testing.

Another reason to avoid testing for a disease that is very unlikely in a given patient has to do with math. I wrote last year that screening for most diseases is not helpful. One of the reasons is that no test is perfect. If the likelihood that the disease is present is extremely small, an abnormal test is more likely to be caused by an test error than by the disease being present. So testing patients that are almost certainly healthy raises the possibility of false positives due to test errors. That won’t reassure anyone and will likely lead to more tests to pursue the spurious abnormal result.

Doctors need to learn to say to patients “That doesn’t sound worrisome. Let’s just keep an eye on it.” without being dismissive. Patients need to learn that a system that pays more for testing than listening will deliver more testing than listening.

Learn more:

In many patients, diagnostic testing isn’t reassuring after all (LA Times)
‘Worried Well’ Often Ignore Negative Test Results: Study (US News)
Reassurance After Diagnostic Testing With a Low Pretest Probability of Serious Disease (JAMA Internal Medicine)
Doctor, Test Me for Everything (My post from last year explaining why some screening tests are harmful)
It’s Not All in Your Head: How Worrying about Your Health Could Be Making You Sick and What You Can Do about it (A very helpful and authoritative book for patients with health anxiety)