An excerpt from A Return to Healing.

Patients often ask me to do whatever tests are necessary to make sure they don’t have cancer or any other really bad diseases. It’s called screening: Look for diseases before patients exhibit symptoms, then fix the problems before they can rear their ugly head. Many of my patients believe that they have had their lives saved this way, always by a caring and thorough doctor who, thank God, performed that “necessary and lifesaving” screening test just in the nick of time.

Most blood tests we order are screening tests. We send off a survey of chemical measurements, not because patients have any particular symptoms, but because it’s their yearly visit. As thorough Flexnerian doctors (named after Abraham Flexner, whose 1911 Flexner Report was backed by powerful industry leaders and transformed medicine from a hands-on healing profession into a research-centered, numbers-driven industry), this is what we do. We acquire data. In fact, the entire physical exam—whether it’s listening to your heart or your neck, feeling for lumps, or sticking our fingers up your rectum—is characterized as screening when you present without any particular symptoms.

In the world of cancer screening, there are three basic categories:

• Birds: Cancers that, even if we catch them early, are already too far gone. Cancers such as ovarian and pancreas are so aggressive that screening for them will not save a single life.
• Rabbits: Cancers that, like breast cancer, have a predilection to cause harm and may be more easily eradicated if found and treated early. All cancer screening should be directed against rabbits, which include things such as breast and colon, and maybe some lung cancers. As with breast cancer screening, and as we’ll see with lung cancer screening, even screening for rabbits provides limited benefit and substantial potential harm. Perhaps the most onerous flaw of rabbit screening is that Flexnerian doctors assure their patients that the test is scientifically valid and is an accurate gauge of underlying disease. That is Flexnerian logic; science is absolute, what the test shows must be right.
• Turtles: Cancers that progress so slowly, if at all, that finding and treating them early derives no benefit. These cancers just hang around minding their own business right up until a patient dies; if they are going to kill us, usually finding them early doesn’t prevent that unfortunate result, and treating them does not help people live longer. But, with screening, we can find turtles, tell a patient he or she has cancer, and then fix that cancer, touting the lifesaving feats that our precise testing can offer. When we find and fix a turtle, we have done nothing to help our patient; in fact, we often expose him to unnecessary harm. But people don’t want to hear that. They are convinced that their lives were saved by the very thorough and caring doctor who ordered the test and removed the cancer. They are now cancer survivors, labeled with a lethal disease by a medical system that thrives when we can convince more people that they are sick.

Prostate cancer is the poster child of a turtle, something we discussed and cited in chapter 4, but it is worth mentioning again in this context. About a quarter or more of men who die have prostate cancer sitting harmlessly in their bodies, having probably been there for years or decades, not bothering them at all. Millions of men have these turtles; if we test all of them, we’ll find them. It is estimated that about 99 percent of prostate cancers do not kill a person even 10 years after discovery, and treatment of prostate cancer does not change the prognosis. The same amount die whether treated or not, and the vast majority of people with prostate cancer do just fine without any intervention.

So, why even look for it?

It turns out that turtle hunting is a big business. Urologists in particular make a lot of money by finding and fixing prostate cancers. Not to mention that patients beset by cognitive biases believe that it’s always better to find and remove any cancer than to leave it alone.

Therein lies an unfortunate reality of screening. In a medical society built on the premise that if we measure and fix things, then we will be healthier and live longer, people have a hard time grasping the flaws of our current screening processes. Like with all tests and screening, a blood test to detect prostate cancer (prostate-specific antigen, or PSA) gives us a precise number that opens up a window to a horrible ailment that may be silently lurking in our bodies. Such scientific precision makes us think that one test can save our lives, and so it brings us down a very long and dangerous road about which we are totally blinded.

Mr. D is an elderly but active man, a spy in his former life, whose doctor ordered a PSA as part of a thorough exam and laboratory screen. It was high, and a biopsy showed cancer. He had hormonal treatment and radiation. His doctors said that they would monitor his treatment’s progress by following the PSA level; they could not assess treatment efficacy by asking if any of his symptoms resolved, since he had no symptoms. Nevertheless, the patient and his wife were thrilled that he would be cured. In the world of fix the number, not the patient, the treatment was a great success. His PSA dropped from 90 to below 1 in a matter of months. There was only one catch. Since starting his treatment, Mr. D was exhausted. He couldn’t walk because of leg weakness, and he could barely breathe. When I first met him after his treatment sent him to a nursing home, his wife still extolled the treatment that eradicated his prostate cancer, and she questioned my advice to stop it.

“Will we monitor the PSA if you stop the treatment?” she asked me.

“There’s no reason for that,” I tried to explain. “It’s likely to go up, but we don’t care. We’re not going to treat the number. You see what the treatment did to him.”

“Then how will we know if the cancer is spreading?”

“We won’t,” I told her. “But that’s OK. The cancer isn’t the problem. The treatment is. If he starts feeling poorly, then we can talk about what to do.”

Mr. D may have improved once we stopped the treatment, but the damage was done. He remained wheelchair-bound in the nursing home for the rest of his life. We never did check his PSA again. He had no cardinal signs of metastatic prostate cancer, no weight loss, no bone pain, no urinary issues. He was just weak and remained that way until he died in his sleep one night.

Medicare alone pays about $1.2 billion over three years for ineffective prostate cancer treatment. That’s small fries compared to many unnecessary medical interventions, but it does help prop up the salaries of urologists, whose society endorses such treatment. What makes prostate cancer similar to other Flexnerian screening tests is that the cost is incurred by measuring and fixing a number in a healthy patient without symptoms, potentially triggering harm and even more downstream cost.

Andy Lazris is an internal medicine physician. Alan Roth is a family physician. They are the authors of A Return to Healing.