A Plea for Better Cancer Screening Guidelines

A Sensible Medicine Guest Post

Screening is a key component of cancer prevention. The rationale is straightforward: early detection can lead to prompt treatment, potentially improving patient outcomes. Physicians rely on published cancer screening guidelines like those produced by the United States Preventive Services Task Force (USPSTF). While the recommendations from guideline panels are, to an extent, evidence-based, they are often insufficient for the needs of practicing physicians. By oversimplifying nuanced trade-offs regarding risks and benefits, cancer screening guidelines frequently end up getting in the way of informed decision-making.

Guideline panels prioritize population-level impact. As such, most guidelines encourage screening for nearly everyone within the target population. But such blanket recommendations are problematic for physicians who must apply them to individual patients. We know that some patients who undergo screening will still die of cancer despite early detection. Others will face unnecessary harm. Many will endure years of screening for a condition they never develop.

The potential harms of cancer screening are well-known and have been covered extensively on Sensible Medicine. Yet they remain underappreciated by physicians and patients. To a great extent, this is the result of how benefits and harms are characterized in professional guidelines. According to one study, over two-thirds of U.S. guidelines recommending cancer screening did not quantify harms and benefits at all, or reported them in a way that made comparison difficult. For instance, some guidelines present the cumulative benefits of years of screening (such as the reduction in cancer deaths over a decade) while citing the harms of only a single test (like the false positive rate).

Furthermore, cancer screening guidelines rarely disclose what level of benefit justifies screening despite the harms. This lack of transparency may explain variation in recommendations between guideline organizations, even after examining the same evidence. The inconsistency is confusing, and without clarity about how these judgments are made, clinicians are left guessing about which advice to follow.

To illustrate these shortcomings, consider the USPSTF’s final recommendation statement on breast cancer screening, which recommends that all women be screened every other year from ages 40 to 74. According to their estimates, this would prevent approximately 8.2 breast cancer deaths per 1000 women over a lifetime of screening. When applied to tens of millions of women in the US, this is by no means insignificant. But for the woman sitting in front of me in my exam room, she is far more likely to be one of the 992 whose lives will not be saved.

Yet the USPSTF’s messaging on breast cancer screening overstates its benefit. Meant for display in primary care offices, the infographic accompanying their recommendation heavily emphasizes the life-saving benefits of screening while offering limited meaningful context. Rather than simply stating that screening saves approximately eight lives per 1000 women, it cites the relative risk reduction of initiating screening at age 40, a figure that is easily misconstrued.

What is more troubling is that a discussion guide intended to support shared decision-making omits statistics essential to a basic, informed understanding (such as the absolute reduction in deaths) in favor of unquantified generalizations. The possibility of false results and the resultant excessive treatments is briefly mentioned, although without specific numbers and framed in a way that minimizes their significance. Even the clinician summary, which purports to provide clinicians with “key information about the recommendation,” lacks the requisite data^⁠ for informed consent.¹

When a guideline declares that everyone should complete a test, it creates the illusion of a “right” decision and a “wrong” one. This may nudge a physician towards paternalism, encouraging patients to screen even if it conflicts with patient desires or values. Complicating matters further, guideline recommendations form the basis for quality incentives that commonly influence physician compensation. Tying pay to test completion binds physicians in an unsuitable conflict of interest.

There is, however, a path forward. Prostate cancer screening guidelines deserve credit for abandoning universal screening in favor of shared decision-making, while also doing better at reporting harms. Yet, a recent study reported in the New York Times links changes in prostate screening recommendations with harm.

Specifically, the researchers connect an observed rise in advanced disease with the USPSTF’s decision to recommend against PSA screening in 2012. That stance broke from other major societies that endorsed shared decision-making. Assuming this correlation proves true, it further illustrates the danger of making sweeping recommendations–even when they are against screening. In the end, the mistake is the same: reducing a complicated, individual decision to an oversimplification.

Future guidelines should inform, rather than decide. At a minimum, harms and benefits must be quantified in absolute terms for easy comparison. Second, “one-size-fits-all” screening recommendations should be avoided. Guideline panels should instead leverage their expertise to create patient-friendly decision aids^⁠ that equip physicians for personalized shared decision-making.²

Beyond informing physicians, guidelines shape public understanding. The debate following President Biden’s prostate cancer diagnosis revealed that many Americans view cancer screening with false clarity. As Dr. Cifu noted in July, screening decisions at the individual level are inherently complex. We must reexamine how we communicate the science of cancer screening to the public. Balanced, transparent, and patient-centered guidelines can effectively combat cancer while also enabling patients to make informed choices. Amid rising mistrust in American healthcare, empowering patients isn’t just good medicine—it’s good science.

Bobby Scott, MD is a practicing family physician and associate professor at Wake Forest University School of Medicine. On his Substack, Statcast & Stethoscopes, he explores what baseball can teach us about making better decisions in medicine.

1. A clinician seeking the absolute risk reduction would have difficulty finding the information. It is buried in a table at the very end of the guideline: Table 2. Estimated Median Lifetime Benefits and Harms of Biennial Screening Mammography With Digital Breast Tomosynthesis for a Cohort of 1000 Women and a Cohort of 1000 Black Women by Starting Age of 40 vs 50 Years.

2. Ironically, the USPSTF’s decision aid in their 2018 guideline on prostate cancer screening is exemplary: Is Prostate Cancer Screening Right for You?