Many doctors fail to understand the math of cancer prevention

All too often the most powerful illusions seduce us through truthful whisperings.  Let’s start with an obvious truth: Living a long and happy life after a cancer diagnosis is better than living a short miserable one.

Given a choice between receiving a diagnosis of metastatic cancer—an incurable life-ending-it’s-already-spread-to-your-brain neoplasm—versus the diagnosis of a localized, snip-it-out-and-it’s-done tumor: Who wouldn’t choose the latter?

And yet this simple truth causes doctors to embrace unproven screening tests, the result being millions of dollars of potentially wasteful medical care and an untold amount of unnecessary anxiety.

In a recent study published in the Annals of Internal Medicine, researchers surveyed primary care physicians and asked them a series of questions designed to determine what makes physicians embrace cancer screening tests.  In one portion of the survey, the researchers describe the hypothetical screening test as follows: “Screen detected cancers have better five year survival rates than cancers detected because of symptoms,” and then asked doctors whether this fact proved that the screening test “saves lives.”  A whopping 76% of physicians mistakenly concluded that the test was lifesaving.

I have written a couple recent posts about the prostate cancer screening controversy.  In these posts, I explain why the PSA test is so appealing—it catches cancers so early that they can be removed in their entirety; it means prostate cancer doctors can treat early illness rather than metastatic disease.

But when a screening test finds a tiny cancer—a tumor that on its own would have dozed indolently inside a person’s body without ever causing harm—that person has been harmed by the test, not helped.  That person will now undergo a painful and unnecessary treatment, living the rest of his life as a cancer “survivor” worried that the tumor will recur.

The only good thing about finding this early cancer is that this person has a great chance of surviving more than five years without dying from this cancer.  Of course, he would have survived five years without dying from the cancer anyway.  Five year survival statistics are not an appropriate way of determining whether a screening test works.

How strong is the illusion among doctors that five year survival rates are a good measure of whether a screening test is effective?  In one portion of their survey, the researchers described a screening test that increased five year survival rates.  Almost 70% of doctors said they would recommend that test to their patients, even though this five year statistic may simply reflect the identification of indolent, harmless cancers!

Worse yet, in another portion of their survey, the researchers presented doctors with information on a screening test that reduced the number of people dying from the cancer in question.  They explained that people who didn’t receive the test were more likely to die of the cancer than people who did.  This type of mortality reduction is the gold standard for determining whether a screening test does what it’s supposed to do—namely, identify an otherwise life-threatening illness while it is still curable and thereby save lives.  And yet, only 20% of doctors said they would definitely recommend this test to their patients.  I guess they were unimpressed, because they hadn’t learned whether the screening test was associated with higher five year survival rates!

Seduced by a simple but misleading truth, most physicians embrace cancer screening tests without comprehending what makes such tests harmful or beneficial.  When it comes to understanding the math of cancer prevention, too many physicians are no smarter than fifth graders.

Peter Ubel is a physician and behavioral scientist who blogs at his self-titled site, Peter Ubel and can be reached on Twitter @PeterUbel.  He is the author of Critical Decisions: How You and Your Doctor Can Make the Right Medical Choices Together.

View 3 Comments >