“Colonoscopy is the dominant approach to colorectal cancer screening in the United States. Among people who are screened, two thirds get a colonoscopy. This fact is easy to miss in Halpern and colleagues’ 2021 data, which included 9.2 million people reporting colonoscopy and 9.8 million reporting fecal immunochemical testing (FIT). That ambiguity is explained by the distinct screening intervals (every 10 years for colonoscopy and annually for FIT), whereas there is no ambiguity about the difference in the resources required each year: $24 billion for colonoscopy versus $0.6 billion for FIT.
Colonoscopy is clearly overused in the United States. It is unclear whether patients prefer it to stool testing. In historically underserved populations, the preference is for noninvasive stool testing over colonoscopy. It is also not clear whether colonoscopy is any more effective than stool testing. That equipoise is the foundation for 3 randomized trials of colonoscopy versus FIT. The U.S. trial, a Department of Veterans Affairs Cooperative Study of 50 000 people, was initiated in 2012 and is expected to be completed in 2028. Given that the data monitoring committee has access to outcomes data every 6 months and a stopping rule has not been triggered, I suspect the trial will find no clinically important difference. [..]
Halpern and colleagues estimated the cost of screening only for the population defined as eligible by the U.S. Preventive Services Task Force [$43 billion in 2021]. Yet perusal of the Behavioral Risk Factor Surveillance System and 2020 U.S. Census data suggests that there are millions of men younger than 55 years and millions older than 70 years who report being screened for prostate cancer. Screening of ineligible patients is particularly common among elderly people. Moss and colleagues estimated that more than half of community-dwelling adults older than 75 years were screened for colorectal cancer and almost three quarters of women older than 74 years were screened for breast cancer. Cervical cancer screening in women older than 65 years was also common, including in 37% of older women without a cervix. [..]
A full accounting of the cost of cancer screening must include subsequent unneeded treatment. Although the treatment of patients who are not destined to develop symptoms or die from their cancer is an unintended side effect of screening, the attendant costs are nonetheless attributable to screening. Overdiagnosis and overtreatment are most relevant to breast, prostate, thyroid, and melanoma skin cancer screening. Overdiagnosis is rare in colorectal cancer screening, although the increased rates of colectomy for nonmalignant polyps suggest overtreatment of precursor lesions. Given the enthusiasm to expand screening beyond heavy smokers, overdiagnosis and overtreatment may become relevant in lung cancer.
What is the value of all this? Promoters assert that cancer screening “saves lives.” Yet its effect is so small that randomized trials must enroll tens of thousands of participants to reliably detect a change in cancer-specific mortality—not all-cause mortality. General population screening is a long run for a short slide: As few as 1 person per 1000 who are screened over 10 years will benefit.
The conventional wisdom is that nothing bad happens to the other 999 who are screened. Yet false alarms are common—followed by more appointments, more follow-up testing, and more procedures. These require people to take more time off from work and pay more out-of-pocket costs. Some will experience complications. Then there are the few who are turned into patients and subjected to unneeded treatment.
Exaggeration of benefits and minimization of harms cause real problems. People see no reason not to be screened, even when their risk for dying from cancer is extremely low (young people) or their competing risks are extremely high (elderly people). Many still desire screening even if told that it does not reduce mortality and has harms. Cancer screening has become a public health imperative and a metric of health care quality.
Kudos to Halpern and colleagues for starting the accounting process on the cost of cancer screening in the United States. Other estimates will undoubtedly follow. But whether the number is $43 billion or $100 billion, questions remain about the value of the expenditure. Resources devoted to cancer screening would be better directed toward ensuring widespread access to effective cancer treatment. And addressing the social determinants of cancer risk—smoking, obesity, poverty, and unhealthy living conditions—would reduce death from multiple causes, not just cancer.”
Full editorial, HG Welch, Annals of Internal Medicine, 2024.8.6