For the better part of the past 20 years, I’ve been ignoring advice from doctors to submit to cancer screenings of various types. Most of my friends and family chalk this up to stubbornness or idiocy or some sort of death wish. I’ve simply been leaning all these years on the vague suspicion that these tests would have no discernible influence on my lifespan.
So it was with some satisfaction that I stumbled last week on a study presented at the annual meeting of the American Association for the Advancement of Science. A statistical analysis from the University of Virginia, the report suggests that cancer screenings don’t ensure a longer life and may result in unnecessary treatments that may lower your quality of life.
Doctors generally justify mammograms, pap smears, colonoscopies, PSA tests, and other screenings by citing statistics showing a percentage reduction in mortality that result from these preventive actions. The argument basically goes something like this: If 10 people out of 100,000 die from colon cancer in a year and colonoscopies reduce that number to six, then you’ve got a 40 percent reduction in mortality — and a strong argument for more tests. But study coauthor Karen Kafadar, a UVA statistics professor, argued that such calculations are misguided because they don’t consider how long a person would have lived if treatment began only after symptoms appeared.
The relevant question, she said, is this: “How much longer can a person whose case was screen-detected be expected to live, versus a case that was diagnosed only after clinical symptoms appeared?” And in the real world, somebody without symptoms who began treatment after a screening is more likely to be alive, say, five years later than someone who’s already exhibiting symptoms.
“People die anyway of various causes,” Kafadar said, “but most individuals likely are more interested in ‘How much longer will I live?’ Unfortunately, screening tests are not always accurate, but we like to believe they are.”
Kafadar’s findings echo results from a 2012 National Cancer Institute report showing how “lead-time bias” (measuring mortality risk before symptoms present themselves) and “length bias” (identifying less-aggressive cancers) contributes to the perception that these tests improve life expectancy.
Lead-time bias plays out like this:
A patient diagnosed with cancer based on symptoms at age 67 dies at 70. This translates to a five-year survival rate of 0 percent. But another patient who is diagnosed at age 60 after a cancer screening and dies at 70 indicates a five-year survival rate of 100 percent.
“This apparent dramatic increase in five-year survival is illusory,” explained Lisa Schwartz, MD, a professor of medicine at Dartmouth University. “In this example, the man does not live even a second longer.”
Length bias simply refers to the fact that screening tends to identify slower-growing, less-aggressive cancers. These may or may not ever become a problem, but most doctors will advise their patients that they need to be treated. This leads to overdiagnosis of harmless cancers, which inflates the five-year survival numbers.
Many doctors, unfortunately, don’t recognize that these biases exist and continue to preach the life-saving benefits of cancer screenings. Schwartz and her Dartmouth colleague Steven Woolshin, MD, surveyed more than 400 physicians and found about half of them believed that screenings saved lives if the tests produced more cancer diagnoses than among patients who were not screened.
“The majority of primary-care physicians did not know which screening statistics provided reliable evidence or whether screening works,” Schwartz said. “They were more likely to recommend a screening test supported by irrelevant evidence . . . than one supported by the relevant evidence: reduction in cancer mortality with screening.”
In fact, screenings don’t do much to lower mortality risks, Schwartz noted, because the chance that any specific cancer is going to kill us is much lower than we are led to believe. That may come as a surprise to most of us, given that more than a half million Americans succumb to some form of the disease each year, but a guy has less than a 2 percent chance of dying from colorectal cancer and less than a 3 percent chance of succumbing to prostate cancer, according to the American Cancer Society. The numbers are similarly low for women: Less than 3 percent will die of breast cancer and less than 1 percent from cervical cancer.
So why all the pressure to get tested? I think it has a lot to do with the way technology has altered the role of physicians.
I’ve been thinking lately about doctors and their reliance on technology after reading Technopoly, Neil Postman’s 1992 critique on America’s infatuation with technology. In it, Postman describes how physicians gradually lost their ability to effectively interact with patients. “Medicine is about the disease, not the patient,” he writes. “And, what the patient knows is untrustworthy; what the machine knows is reliable.”
As a result, doctors have seen their diagnostic skills — and their relationship to their patients — erode. Postman quotes from Stanley Joel Reiser’s Medicine and the Reign of Technology to explain further:
“. . . without realizing what has happened, the physician in the last two centuries has gradually relinquished his unsatisfactory attachment to subjective evidence — what the patient says — only to substitute a devotion to technological evidence — what the machine says. He has thus exchanged one partial view of disease for another. As the physician makes greater use of the technology of diagnosis, he perceives his patient more and more indirectly through a screen of machines and specialists; he also relinquishes control over more and more of the diagnostic process. These circumstances tend to estrange him from his patient and from his own judgment.”
The downside of this evolution is beginning to become clear even to those smarter and less stubborn than me. The percentage of men seeking PSA tests has dropped in recent years and the major public-health organizations have revised their recommendations for these and other screenings.
And at the AAAS meeting in which Kafadar presented her study, Constantine Gatsonis, chair of the Department of Biostatistics at Brown University, also cautioned against relying too heavily on these types of tests. Gatsonis, whose 2005 study on the efficacy of digital mammography is responsible for putting those machines in most hospitals, nevertheless noted that screening is not just about detection; it’s about the health of the patient.
“There is a growing concern of this notion of overdiagnosis,” he said. “Screening is finding small lesions that would not hurt you. Generally speaking with screening, especially as the modalities become more and more accurate and can see smaller and smaller things, the question is, is that good for you? It’s not a foregone conclusion.”
It is for this geezer. You can have your colonoscopies and PSA tests. I’ll trust my body, not some machine, to tell me how I’m doing.