Re: Appropriate use of diagnostic tests in medical practice

As a longtime teacher in our UBC Faculty of Medicine, I was pleased to see the article by Dr Wong—“Appropriate use of diagnostic tests in medical practice”—in the May 2026 issue of the BCMJ.[1] Dr Wong points out the value of the Choosing Wisely recommendations.[1,2] It may be helpful to know that UBC medical students in their surgery clerkship block are advised to access the Choosing Wisely resources and are taught what I unashamedly call “Blair’s four rules of testing.” They are:

• Do not order a test unless the results will truly change your management.
• Do not order a test unless you know what the result will mean for your patient.
• Consider the risk of the test and whether this risk is appropriate in the context of your patient’s present and future health.
• If you are requesting a test, make sure you ask for it to be done correctly.

Students can learn to mimic physicists and their thought experiments: namely, to muse on how a patient’s management may be affected if a test were to come back positive, negative, or inside or outside the normal range. For instance, if a requested C-reactive protein test comes back high or normal, should that truly change one’s management, or has it been a waste of health care funds? Every test has a monetary cost, and a planetary cost, too.[2]

Along the same lines, if one asks for a test, one should have a good idea—before requesting it—of that test’s reliability parameters, such as accuracy, specificity, and sensitivity, in that patient’s context and the pretest probabilities.

No test is without risk. Some, such as angiograms and invasive biopsies, have known palpable risks, but there are risks of unintended consequences even with some apparently benign investigations. For example, a blood test done for no good reason that shows an abnormal result can sometimes lead the clinician and the patient into a Ulysses syndrome[3] of needless further testing, anxiety, and real danger. I have witnessed tragic deaths in such needless pursuits and health insurance refusals based on one specious blood test with no evidence of disease.

Last, if testing is warranted, it should be done correctly. Some cautionary examples include serum cortisol testing done at an inappropriate time of day, CT scans done without appropriate contrast, flat-plate abdominal X-rays done when upright views are also needed, and so forth. If in doubt, ask. Testing should be regarded as a form of consultation, a gathering of evidence. Indeed, “ordering” a test is perhaps better said as “requesting” a test.

Dr Wong, in his excellent article, typifies these four rules with good examples from his field of infectious diseases. His message that a “test is useful only if the result can meaningfully influence decision making” is vital for our medical students to learn and carry into their future practices, no matter what field of medicine they pursue.
—Geoffrey Blair, MD, FRCSC
Clinical Professor Emeritus, UBC

This letter was submitted in response to “Appropriate use of diagnostic tests in medical practice.”

References

1.    Wong D. Appropriate use of diagnostic tests in medical practice. BCMJ 2026;68:141-144.

2.    Choosing Wisely Canada. Accessed 23 May 2026. https://choosingwiselycanada.org/.

3.    Rang M. The Ulysses syndrome. Can Med Assoc J 1972;106:122-123.