De-implementation of Mammography Overuse in Primary Care Settings

De-implementation is recognized as a critical but understudied area within implementation science (IS). Research is needed to determine the optimal methods and approaches for identifying, selecting, and tailoring de-implementation strategies. De-implementation of routine cancer screening in older patients, such as mammography screening for breast cancer, offers excellent opportunities for both advancing the science of de- implementation and improving care delivery and health outcomes. While national guidelines do not support routine mammography in older patients and recommend consideration of morbidities, life expectancy and patients’ informed preferences, ~56% of patients ≥75 years report receiving mammography, including 50% of individuals with life expectancy <10 years. Our preliminary research identified multi-level challenges and facilitators to de-implementation of mammography overuse among older patients at the organizational (e.g. system alerts, patient reminder letters), provider/clinic (e.g., knowledge, clinic norms), and patient (e.g. habit, knowledge) levels. Informed by the Knowledge-to-Action Model, we propose a study for de-implementation of mammography overuse (i.e., reducing the frequency or cessation of mammography) across a large healthcare system serving in New York City. In Aim 1, we will identify a range of de-implementation strategies at the patient, provider, and organizational levels for reducing mammography overuse in patients ages ≥75 years. We will use a crowdsourcing method, successfully applied as an emerging participatory IS approach (innovation tournament) to generate rapid data collection from all relevant stakeholders (patients/family members, providers/administrators from the community and multiple healthcare systems) on factors that influence de-implementation. Combining this data with our rich qualitative preliminary data, and principles from Dual Process Theory, we will develop distinct de-implementation strategies for refinement in Aim 2. In Aim 2, we will prioritize and tailor de-implementation strategies at patient, provider/clinic, and organizational levels. We will recruit 12-15 experts to prioritize strategies based on feasibility and acceptability, and propose important attributes (e.g., duration, frequency, content) for each strategy, and employ discrete choice experiment to elicit patient preferences for modifiable attributes of each prioritized strategy. In Aim 3, we will evaluate the feasibility, acceptability, and use of the tailored de-implementation strategies in a pilot cluster randomized trial . Using a sequential mixed-methods design, we will assess use of strategies, de-adoption outcomes (e.g. reduction of mammography overuse), and theoretical mechanisms of strategies at the patient, provider, and organizational levels. Data will establish feasibility and provide preliminary data for effectiveness of strategies to be tested in a future Hybrid 2 trial, and lay the groundwork for advancing de-implementation frameworks and methodological approaches for selecting de-implementation strategies to reduce the use of low-value care.