Run American women through the Tyrer-Cuzick model and 5.6% exceed the 20% lifetime breast cancer risk threshold; run the same women through other commonly used models, and the figure falls to between 0.4% and 0.9%. The discrepancy sits at the center of a new JAMA Internal Medicine Viewpoint questioning whether risk-based screening is moving faster than the systems designed to support it.

Aliberti and colleagues argue that risk-based screening — newly endorsed by updated American College of Radiology recommendations and supported by WISDOM trial findings showing no increase in advanced-stage cancer detection compared with annual screening — has outpaced the infrastructure meant to deliver it. In their case vignette, a mammography report auto-calculated a Tyrer-Cuzick lifetime risk of 25% from structured EHR fields, some of them blank, and recommended annual MRI; the primary care physician, working from the patient's actual family history, recalculated 19.2% — just under the threshold that triggered the recommendation. Models disagree, thresholds remain unsettled (20% versus 25%, or five-year cutoffs as low as 1.67%), and reports now reach patients before any clinician can vet them. Primary care physicians are left to validate or walk back a number, often without the time or context to do either well.

Here is the part worth sitting with: the recommendation arrives as fact. A score assembled partly from incomplete data lands in front of the patient beside an MRI recommendation citing professional guidelines, while the clinical reasoning, by design, comes afterward.

The evidence underneath is less settled than the page suggests. There is no consensus on which model or threshold defines high risk, data comparing mortality between annual and biennial screening intervals remain lacking, and prevention medications reduce incidence but have not been shown to reduce mortality. Even the definition of "high risk" ranges from lifetime-risk thresholds of 20% to 25% to five-year thresholds as low as 1.67%.

The clinical move is small but consequential: confirm the inputs before acting on a radiology-generated score, since a recalculation can move a patient across the line. Institutionally, the authors call for interdisciplinary working groups to settle the model, threshold, and messaging up front. The patient's decision, they write, "underscores the essential role of clinical context, rather than automated outputs."

One author reported grants from National Institutes of Health/National Institute on Aging and royalties from UpToDate. No other disclosures were reported.

Source: JAMA Internal Medicine