Grant Number:	1R01CA315493-01 Interpret this number
Primary Investigator:	Smith, Shawna
Organization:	University Of Michigan At Ann Arbor
Project Title:	Evaluating Strategies to Minimize Unnecessary Interventions in Older Women with Breast Cancer
Fiscal Year:	2026

Project Summary/Abstract Over 30% of women diagnosed with breast cancer are aged 70 or older. Nearly all of these women present with early-stage, hormone receptor-positive (HR+) disease, which has an excellent prognosis. As a result, the chance of an older woman with breast cancer dying from her disease is less than 1%. Despite this favorable outlook, many older women receive high-intensity treatments that are unlikely to provide significant benefits. While surgery and hormone therapy improve survival rates in older women with HR+ cancer, other common treatments in younger women—such as axillary staging with sentinel lymph node biopsy (SLNB) and adjuvant radiotherapy after lumpectomy—do not offer any survival advantage. Based on these data, national guidelines recommend omitting SLNB and adjuvant radiotherapy in most older women. Nonetheless, many continue to undergo these interventions at alarming rates (>80% and >50%, respectively), exposing a clear gap between evidence and practice. These unnecessary procedures can increase morbidity, trigger care cascades leading to other low-value treatments and raise healthcare costs by over 35%. Our previous research identified factors driving overuse at the patient, clinician, and facility levels, and we tested on a small scale the effectiveness of patient- and clinician-focused strategies to reduce this overuse. For example, although older women generally accept age-based guidelines, they often lack awareness of the option to avoid certain treatments and may be influenced by their innate medical maximizing-minimizing preferences toward more intervention. We demonstrated that an age-specific decision aid is acceptable to patients, and that messaging tailored to their values can lower their preferences for low-value treatments in hypothetical scenarios. At the same time, we found that clinicians overestimate the benefits, minimize the harms of these therapies, and lack the skills to discuss omitting treatments. To address these issues, we tested educational sessions, clinician scripts, and decision support tools based on patient frailty. Finally, while high-volume and academic centers show less overuse compared to low-volume and community centers, de-implementation remains inconsistent, with most facilities overusing one or both treatments. Overall, this work supports further testing of a comprehensive, multi-level, multi-component strategy to achieve complete and sustained de-implementation that is applicable across diverse hospital settings. In this study, we plan to 1) adapt this multi-level, multi-component approach for three sites, 2) evaluate its effectiveness at 18 Michigan sites using a pragmatic, stepped wedge cluster randomized trial, and 3) examine key implementation outcomes that could influence its effectiveness, generalizability, or sustainability.





None