Breast Cancer Screening—Age-Stratified Benefits, Harms, and the Case for Risk-Based Shared Decision Making

Breast cancer screening guidelines have increasingly relied on expert opinion rather than robust clinical trial data to justify supplemental MRI in high-risk, non-BRCA1 women. While MRI improves cancer detection rates in average-risk populations, this detection advantage has not translated to demonstrated mortality reduction in randomized controlled trials, raising the question of whether earlier identification without survival benefit represents clinical progress or diagnostic expansion. This article presents comprehensive age-stratified data demonstrating that mammography delivers modest but real RCT-proven mortality benefits. Across randomized trials and pooled analyses, mammography yields modest absolute reductions in breast cancer mortality, ranging from approximately 0.03% (NNS 3,300) in women aged 40–49 to approximately 0.33–0.50% (NNS 200–300) with biennial screening in women aged 60–69 years, based on pooled RCT analyses and CISNET modeling. Over approximately 10 years, screening harms vary substantially by interval: about half of women (~56%) undergoing annual digital mammography experience at least one false-positive callback requiring additional imaging, compared with about one-third (~38%) with biennial screening. Among screened women, approximately 1 in 9 (~11%) receiving annual screening and 1 in 14–15 (~7%) with biennial screening receive false-positive biopsy recommendations and most of these recommended biopsies are ultimately benign. Digital breast tomosynthesis (DBT) modestly lowers cumulative false-positive recall rates compared with digital mammography. Supplemental MRI for non-BRCA1 high-risk women (e.g., BRCA2 carriers, CHEK2/ATM/PALB2 mutations, or lifetime risk ≥20% per Tyrer-Cuzick model) lacks any randomized trial evidence of mortality reduction, generating substantial harms, including heightened ductal carcinoma in situ (DCIS) overdiagnosis. MRI increases recalls and benign workups, particularly in the initial screening round. In the DENSE RCT (women 50–75 with extremely dense breasts), the false-positive rate was ~8% per screening round (79.8/1000), with positive predictive value (PPV) 17% for recall and 26% for biopsy. In the DENSE randomized trial (women 50–75 with extremely dense breasts), invitation to supplemental MRI reduced invasive interval cancers over 2 years (RR 0.47) and reduced overall interval cancer rates; effects on advanced disease and mortality remain unknown. In contrast, BRCA1 carriers demonstrate exceptional benefit (NNS=9, 80% relative risk reduction). This analysis reveals a fundamental asymmetry: screening programs with number needed to screen (NNS) >300 are "strongly recommended" based on consensus, while pharmaceutical drugs with number needed to treat (NNT) of 50 would typically face scrutiny from FDA due to efficacy concerns. This discrepancy persists due to different regulatory philosophies, semantic shifts conflating detection with mortality benefit, population-level framing obscuring individual harms, and aligned institutional incentives. The 2025 WISDOM trial provides landmark evidence that risk-stratified screening is superior in efficiency to age-based universal screening, with biennial mammography achieving non-inferior late-stage cancer detection for average-risk women (62% of population) while reducing false-positive burden by 25–30%. The article proposes actionable reforms: explicit evidence-grade labeling (evidence level vs. consensus strength), absolute risk quantification, shared decision-making templates for all NNS >50 scenarios, DCIS de-escalation through observation trials with risk stratification by grade, and implementation pathways accounting for health equity barriers. An evidence-guided orientation reframes screening away from fear of undetected cancers toward transparent, values-aligned medicine serving women's health interests. NOTE (Timing clarification—COMET DCIS Trial):Earlier text in this article referenced “2026–2027” as an estimated window for longer-term COMET outcomes. COMET’s published results to date report short-term outcomes; longer-term outcomes (including 5-, 7-, and 10-year follow-up analyses) remain pending, and publication timing is not guaranteed to map to specific calendar years. This clarification does not change the article’s conclusions or benefit–harm framing.