While doing research for a New York Times article on cancer screening in elderly patients, Liz Szabo came across some statistics that shocked her. Nearly 20% of women over 70 with severe cognitive impairment received screening mammograms in 2002. More than half of patients with a very high mortality risk still receive prostate cancer screenings.
And this doozy: Almost 60% of women who received a hysterectomy were screened for cervical cancer in 2010, even though their cervix had been removed.
It makes sense that we would stop doing cancer screening for those who are unlikely to benefit from them, but as Szabo highlights in the article (and in a lively tweetchat), unnecessary cancer screenings are widespread. This happens for several reasons – patients and their families expect routine screenings, doctors often don’t have time to talk through the risks and benefits of screening, and financial incentives and guidelines push clinicians to screen more.
And the potential risks of cancer screenings aren’t well-known. As the daughter of an elderly patient was quoted in the Times, “If the screening is not too invasive, why not? I want her to have the best quality of life possible.” However, screenings can have complications (which are more likely in elderly patients) and if something is found, patients may undergo invasive and painful procedures like biopsies or surgery.
In an editorial in the Annals of Internal Medicine, Dr. Gilbert Welch and Dr. Otis Brawley point out another negative consequence of excessive cancer screening that is rarely talked about – the impact on data. Cancers such as thyroid, prostate, kidney, and breast cancer are known as “scrutiny-dependent,” which means the number of cases diagnosed is dependent on how closely we look for them.
As imaging technology has gotten more sophisticated and early detection has been increasingly prioritized, we find more cancers. This makes it seem like the incidence of cancer is increasing, when in fact, we’re just finding more of them.
This “scrutiny effect” also leads to misinterpretation of risk factors. For example, people with family members who have prostate cancer are also more likely to get screened, which leads to cancer being found in more of those patients.
“Family history influences how hard we look for prostate cancer and therefore how much we find,” says Welch, quoted in STAT, “The risk factor becomes a self-fulfilling prophecy.”
As Shannon Begley writes in STAT, the same effect has happened with women being at “greater risk” of thyroid cancer. However, men and women have an equal mortality rate from thyroid cancer, suggesting that the gender disparity is because women are more likely than men to get checked for thyroid cancer, not that gender is actually a risk factor.
How do we separate the real risk factors from effects of excessive screening? Welch and Brawley recommend focusing on risk factors of cancer mortality rather than incidence, to get a better picture of cancer risk without just measuring those more likely to be screened.