Labos: More tests don’t make us healthier
Precautionary exams and procedures can do more harm than good, Christopher Labos warns
“Do you really need that test?”
That’s the question raised by a new report, “Unnecessary testing in Canada,” which finds that 30 per cent of low-risk patients may be getting needless exams and treatments.
On the surface, the problem is obvious. Unnecessary testing wastes money. It leads to backlogs in the health-care system and creates waiting lists. For society as a whole, it is clearly detrimental.
But is asking for tests a good idea for a patient who wants to be cautious?
Intuitively, we believe that more tests are better for us, and increase our chances of finding some disease we may not know we had.
But the truth is more complicated.
First, the tests themselves are not benign. X-rays, CT scans and nuclear scans all expose patients to ionizing radiation. If done sparingly, the risk is minimal. However, there is mounting evidence that repetitive testing exposes patients to significant amounts of radiation, which increases their risk of developing cancer.
If the scans are medically necessary, that’s one thing. But patients who are getting yearly scans just as a checkup are potentially exposing themselves to non-trivial amounts of radiation.
Second, these scans often pick up non-significant findings. The term “incidentaloma” has been coined to describe masses that turn up on routine scans done for other reasons. These findings then provoke other tests, scans and biopsies to find out if they are dangerous, which they usually aren’t. And these extra tests and procedures carry some risks.
Consider a 2011 case report of a woman presenting to an emergency room with chest pains, worried she might be having a heart attack. Despite her work-up being normal and that she was at low risk for heart disease, she got a CT scan to look at her arteries. The scan suggested maybe mild disease, but was equivocal enough that she was sent for an angiogram.
The angiogram was normal, but caused a tear in her aorta, which led to emergency surgery, which caused her to have a heart attack and finally led to a heart transplant.
Remember, the CT scan that started all this was unnecessary; it was done simply for the purpose of being extra-thorough.
Unfortunately, if you do enough unnecessary tests, you might cause a complication.
But what about that rare person where some fancy test will find something important? Perhaps getting regular scans or MRIs will find an early cancer that could be treated? We naturally believe that an earlier diagnosis leads to better outcomes. But the truth is a little murkier.
Screening for diseases like cancer intuitively makes sense but in reality is not necessarily beneficial. Sometimes cancer tests pick up cancers earlier but don’t extend survival. Therefore you just end up living longer with a diagnosis without delaying death. This concept is called lead-time bias.
Sometimes tests preferentially pick up slow-growing, non-aggressive tumours that probably wouldn’t have been lethal if they had been left alone. This is called lengthtime bias. The best example of this: prostate cancer, where many cases are nonaggressive and can be safely observed without treatment.
In situations like cervical cancer and colon cancer, cancer screening is of proven benefit. But in many other situations the benefits are less clear.
For example, mammograms are not recommended for younger women because they have denser breast tissue, which makes mammograms less effective and prone to false positives. If a woman under 50 (with no high-risk features) has a positive mammogram, the probability that she actually has cancer is about four per cent. Therefore, routine mammograms are not very useful and not recommended before age 50.
Over-testing and unnecessary testing are bad for society as a whole. But it’s bad for you as an individual, too. So when you see your doctor, ask: “Do I really need that test?”