Many people assume that the lack of pro-active testing for every possible problem in medicine is simply done as a cost savings measure. Patients will sometimes complain to me that their insurance won’t cover a periodic full-body MRI, or a routine stress test, or blood testing in the absence of any symptoms for this disease or that. And there is some validity to this critique: it’s certainly true that insurers like to spend as little money as possible on patients. But to be fair, there is another side to the argument, which is that – contrary to what many assume – testing for problems that aren’t apparent can sometimes in fact be harmful.
This is a difficult topic to grasp, so I’m going to break this post down into two parts. Today, I will outline some of the logic behind why tests can sometimes create more problems than they prevent. In next week’s post, I will give a few actual anecdotes from my career that demonstrate the point.
How could it possibly be that ordering tests to screen for health problems can be harmful?
There are a number of factors to consider, but the most important one is: what are the odds of turning up a “false positive” result to the test? And what harm might accrue to the patient as a result?
To take the emotion out of the discussion, let’s imagine a fictitious scenario in which cancer of the pinky was ravaging the United States. Let’s further suppose that doctors developed a screening test for this cancer, called the “Pinkygram.”
Now consider the following scenario: Angela gets a pinkygram, which shows a suspicious spot. This leads to a biopsy, which shows an early pinky cancer. This is removed, and she goes on to live a long healthy life. At the same time, her twin sister Barbara also gets a pinkygram, which shows a suspicious spot. This leads to a biopsy, which turns out to be benign, and she too goes on to live a long healthy life.
So far, giving our fictitious sisters their pinkygrams seems like a great idea. We did two tests, found two suspicious spots, and diagnosed one cancer. We saved one life (a huge positive) at the expense of putting Barbara through a scary, painful, and expensive medical procedure that didn’t result in adding any benefit to her health (a negative). But clearly this is a good trade-off. Had we not done the pinkygrams, Barbara would have been spared an unnecessary (though brief) ordeal, but Angela would have died.
But what if for every Angela we find, there are a hundred Barbaras? Now we’re faced with a bit of a tougher scenario. We are putting one-hundred women through an expensive, painful, risky procedure in order to save one life. On the whole however, this still seems like a good tradeoff. One-hundred women have been put through a bad ordeal, which is certainly unfortunate, but in the long-run they are going to do just fine. In exchange for that negative, we’ve saved Angela’s life. I think most people would take this.
However, now let’s suppose that one of those one-hundred women has a bad complication from their biopsy procedure. To make the example really clear, suppose that one of our Barbara’s has a severe allergic reaction to the anesthetic during her biopsy and dies on the table. So the “tally” now looks like this:
One life saved (Angela, who might have died of pinky cancer but now won’t)
One death created (A Barbara who would not have died of pinky cancer and might have gone on to live a long life, but is now dead from the procedure that was done to prove her absence of pinky cancer)
99 other women who have been put through a great deal of pain, stress, anxiety, and financial expense.
Now, on the whole, our pinkygrams have caused more harm than good.
This is precisely the sort of reasoning that leads from time to time to medical organizations putting out recommendations to do fewer mammograms, or to stop screening men for prostate cancer after the age of 70 or so.
To be fair, in real life the data is not so unambiguous as in our hypothetical pinky example above. It’s hard to do a full “cost/benefit” analysis of a test that is used on millions of people, because there are so many variable outcomes that may be over or undercounted. And cancer is understandably such an emotional disease for so many people, that a pure “cost/benefit” analysis can never give us a full picture of the situation (how do you account for the value of reassurance and peace of mind, for example?)
The right thing to do is not always clear cut, and the recommendation to do fewer mammograms, or less prostate cancer screening, or any other specific scenario may well be the wrong one. My point is not to take a position on any particular screening test, but rather to underscore that a reasonable person who is motivated only by a desire to help people be as healthy as possible can come down on either side of the question.
They say a picture is worth a thousand words, but in my profession, I find that an anecdote about an actual patient is worth an entire research paper full of data. Therefore in my next post, I am going to discuss a few patients from my own career who demonstrate the pitfalls of “overtesting.”
