America’s breast-screening infrastructure lags science

As prevention research advances and cancer trends shift, universal age-based screening for breast cancer could one day be a protocol of the past.

The science of risk-based breast cancer screening is emerging, and AI’s integration makes scale promising. However, if health systems don’t create the infrastructure to explain and act on risk, we might be leaving out the cohort for whom the tests are most beneficial for to prevent, detect and treat breast cancers that could become the most severe.

The missed cohort 

Young women are getting breast-cancer diagnoses at higher rates. In the United States, breast-cancer incidence in women under 50 has risen by 1.4% each year, compared with 1% for the general population, according to the American Cancer Society.^[1] While US guidance has largely converged on recommending women get mammograms every year starting at age 40, medical societies and task-forces have wavered in their precise screening recommendations over the last several decades. The lack of clarity is merging with a growing at-risk cohort of younger women.^[2] 

“We have not changed the way we approach screening since 1980…before we even knew what an oestrogen receptor was,” says Dr Laura Esserman, a breast surgeon and the director of the University of California, San Francisco (UCSF) Breast Care Centre.

The at-risk screening momentum 

The WISDOM trial, a long-term study published in the Journal of the American Medical Association in December 2025, concluded that risk-based screening was just as effective as age-based screening.^[3] The study, which included more than 46,000 women from 2016 to 2023, found that risk-based screening did not increase the rate of late-stage breast cancers detected compared with universal screening. What’s more, nearly 90% of participants opted for risk-based screening when given the option, underscoring that the protocol is not only viable but preferred.

“If screening every year at 40 was so great, we wouldn't have 42,000 deaths a year,” says Dr Esserman, the lead investigator of the WISDOM trial. While a woman at high risk could already be developing later-stage breast cancer in her 40s and have required a more regimented screening protocol at an earlier age, a same-aged low-risk individual may not need screening for much longer. The idea is that you’re catching more severe cases earlier without overtesting and overburdening the system.  

Still, because nearly 80% of the original WISDOM trial participants were white, risk assessments must be drawn from broader, more diverse data sets. Black women face a higher death rate from breast cancer than white women.^[4]

The meaning of risk 

Risk depends not merely on family history, but on a mix of ancestry, biological traits and lifestyle. Research from the American Cancer Society highlighted this with the results that over 85% of breast cancers occurred in women who did not have a family history of breast cancer.^[5] For example, if you have a pathogenic variant, your lifetime risk for developing breast cancer is between 60% to 80%.^[6] The idea is also that if you know you are at a higher-risk, you may be more prone to do something about it. Modifiable risk factors include limiting smoking and maintaining a healthy body weight.^[7]

Last year the FDA approved a model to predict five-year breast-cancer risk with AI-assisted mammography.^[8] Alpa Patel, senior vice-president of population science at the American Cancer Society, says the organisation is building cancer-prevention studies that aim to identify someone's risk on a scale, which would be especially important for understanding the traits of the under-40 prototype getting a breast cancer diagnosis. “We’re getting very close on the science,” she says.  

Yet before promoting risk-based screening, policymakers must clearly define risk and make a coverage pathway for administering risk scores, as well as establishing treatment plans for targeted cancer care.

The infrastructure blindspot 

Infrastructure is necessary. While more data can be powerful, it’s essentially meaningless without the proper escalation and education channels. A similar phenomenon is playing out in the consumer-health space. Many consumer-health companies are capitalising on a growing interest for continuous biomarker monitoring, often in the name of prevention—similar to the reason to screen for cancers. The concern is that without actionable insights, you risk layering on tests that won’t produce long-term health outcomes.

If a risk-based score doesn’t have a proper care pathway, the innovation won’t move the needle. Ultimately, you can’t have truly equitable innovation without thinking about who won’t benefit from it, according to Shonta Chambers, the chief equity and community-engagement officer of the Patient Advocate Foundation. People often confuse availability with accessibility, but just because something is widely available does not mean it is accessible.

Education on risk-based screening must start at the community level. Health systems must also build new frameworks to define risk and ease the anxiety that comes with learning of future health complications. Take the growing popularity of full-body MRIs. While the most published findings occur in those who found a life-saving needle in the haystack, the concern is that when the bulk of findings are labelled incidental, but still within the range of normal, undue panic and further testing followed that doesn’t lead to improved outcomes.^[9] One example of improving this infrastructure and mitigating the psychological impact of knowing risk for breast cancer is investing in more genetic counsellors. Alongside education, another part of the infrastructure is advocating for clear care pathways and coverage.^[10]

The stakes are high—not just morally but economically. Catching higher-risk people earlier can greatly ease the burden on America’s healthcare system. A paper published in 2017 in Breast Cancer Research and Treatment found that for privately insured women under 45, insurers paid about $60,000 more a year after diagnosis for breast cancer at Stage III or Stage IV than Stage I.^[11]

It is easy to move fast when AI’s potential makes scalable solutions seem in sight. However, without ensuring education accompanies risk-based tests, and that the proper escalation-pathway framework is in place and the assessments are built on robust data, we risk exacerbating inequities.

Alexa Mikhail will continue this conversation on women’s health at HLTH USA 2026 in Las Vegas, November 15–18.

Views or opinions expressed are those of the author and any individuals cited, and do not necessarily reflect those of Economist Enterprise or any other member of The Economist Group.