Is There Such a Thing as an Unnecessary Diagnostic Test?
As I was driving home from work this morning after a night shift I heard a story about the health care debate on the radio news. The topic was one of the most poorly understood concepts in the debate: unnecessary diagnostic testing.
Many people have cited the elimination of unnecessary tests as a way to reduce health care costs. As a medical clinician I would like to know how an unnecessary test is defined.
It is not possible to know whether a diagnostic test is necessary without knowing the results. In other words, it is not possible to know whether a diagnostic test is necessary without running it. Only with hindsight can one determine whether the test was needed.
Consider the following example from veterinary medicine. Imagine your dog has suffered a two day period of lethargy and poor appetite. You authorize your veterinarian to run basic blood tests. The results are normal, but the ratio of blood sodium and potassium levels is on the low end of what is considered normal.
In cases such as this, the overwhelming majority of dogs do not have a serious condition. Transient gastrointestinal upset due to dietary indiscretion is most likely. An adverse reaction to a mild toxin also may be to blame. A self-limiting viral infection or dozens of other mild pathologies could be the cause of the issue. All of these problems generally resolve on their own.
But a much more serious problem can cause the symptoms and initial test results I have described: Addison’s disease. Addison’s disease is much less common than the milder problems I have described above. But if it is not detected and treated it is deadly.
The test for Addison’s disease is specialized and expensive. It is non-invasive and low risk. Without the test, it is not possible to know whether the syndrome is causing the symptoms. The treatment for Addison’s disease can be harmful to dogs who don’t have the syndrome.
Imagine that one out of every thousand dogs with lethargy, poor appetite, and a borderline sodium to potassium ratio has Addison’s disease. If a thousand dogs develop the symptoms I have described (and initial test results are as I have described), 999 of them will get better without treatment. One will die.
Those may sound like good odds. But what if your dog is the one?
If you think a one-in-a-thousand chance isn’t bad, I hope you don’t buy lottery tickets. The odds of hitting the mega jackpot make one-in-a-thousand look almost certain.
When I treat dogs like the imaginary one in this post, I always recommend the test for Addison’s disease. After I get the results, I usually learn that the test was not necessary. But every once in a while the test saves a dog’s life. In those instances the test was very necessary indeed.
You have questions. 
An “unnecessary” diagnostic test is very easy to define. It is one where the clinician has no reasonable expectation for obtaining informative results.
While I take your meaning, in that not every test will prove to be informative (positive result). However, tests should only be performed if they are expected help to establish a diagnosis or treatment plan. In this case, a negative test result can still be helping to establish a diagnosis or treatment. Therefore, even that test could be seen as “necessary.”
I think you would agree that there are such things as “unethical” diagnostic tests — tests whose only benefit is a financial one to the clinician. Similarly, there are “stupid” tests — tests that are conducted because the clinician fails to apply basic clinical logic. And finally, there are “cover your ass tests” — tests that are conducted only because they are traditionally included within the so called standard-of-care that protects the clinician from litigation, but are unlikely to provide information beyond those already obtained through experienced clinical observation.
All of these examples, and perhaps a few more, would fall under the label: “Unnecessary Diagnostic Test.” We would all benefit by eliminating them.
One concrete example I can think of is ordering a molecular HPV test at the same time as a pap smear. Roche, who makes money on the test, runs ads on TV telling women that their pap test is not enough, and they should have their clinician order an HPV test simultaneously. Normally the HPV test is a reflex order, only placed once an abnormal pap result is entered. What Roche doesn’t mention is that this test costs hundreds of dollars and is not likely to catch something a pap test missed. This is an unnecessary test. Obviously, pets don’t get pap smears, but pets do get other tests. Imagine bringing a dog in for a respiratory infection and having the clinician take blood, stool and skin cultures along with the nasal swab. Imagine also running the full breadth of micro on the cultures, from virology to parasitology. These would be unnecessary tests. Doctors do this *all the time* to be sure they don’t miss something, even when they can be 100% positive there is a horse in the room, not a zebra.
My Labrador was one of those 1 in a thousand who was diagnosed with Addison’s….. Ask me how I feel about that test :)
In the end it all comes down to trust of your Vet. When my Vet tells me a test should be done I trust that she needs more info. I know she doesn’t have a crystal ball.