Since the benefits of mammography screening are uncertain for women >75 years and there are important
harms, guidelines recommend that primary care providers (PCPs) engage women >75 years in shared
decision making (SDM). The challenge is in weighing the trade-off between potentially finding a breast cancer
early which may in turn help an older woman live longer and/or better, or instead result in her being harmed by
work-up and/or treatment of a breast cancer that never otherwise would have caused problems in her lifetime
(overdiagnosis). Ideally, screening decisions should consider an older woman's breast cancer risk, health, and
preferences. However, 56% of women >75 years are routinely screened, including many with short life
expectancy, and few are informed of the harms. In prior work we found that older women want to talk about
these decisions with their PCPs. However, PCPs feel ill-prepared and unsupported to engage older women in
SDM and request an easy to use and flexible web-based tool (i.e., a conversation aid) that could provide
personalized information on the benefits and harms of mammography screening based on each patient's own
breast cancer risk and health. Conversation aids (CAs) are increasingly used and are designed to be brief,
require minimal training, fit into clinical work flow, and facilitate SDM. They have been shown to improve SDM
and to increase satisfaction; however, despite the need, no CA exists to support SDM around mammography
screening between PCPs and older women. Therefore, we, an interdisciplinary team of PCPs, experts in SDM
and implementation science, modeling researchers from the NCI-funded Cancer Intervention and Surveillance
Network (CISNET), and other expert stakeholders, aim to develop and test an interactive web-based
conversation aid (CA) on mammography screening for PCPs to use with women >75 years, and involved
family members during a visit. In Aim 1, we will leverage two CISNET models to estimate the personalized
benefits and harms of mammography screening in women >75 years based on age, race, comorbidity, and
breast cancer risk, necessary for decision-making. Model generated outcomes will be used to populate the CA
which we will develop and revise in Aim 2 using an established iterative end-user centered approach that
includes observing use of the CA by 10 PCPs practicing in diverse settings during 30 clinical encounters. Once
we have developed a CA that supports SDM, we will pilot test the CA in a pretest-posttest trial including 40
PCPs from Boston-area community, academic, and safety-net practices, 100 women >75 years and 25 of their
family members, and 10 medical assistants. The goal of the pilot study is to show that the CA is feasible to use
during primary care visits in diverse settings and to obtain pilot data on the CA's effects to inform
implementation and a future pragmatic trial. We hypothesize that our novel CA will increase PCP self-efficacy
to engage older women in SDM and as a result older, women will make more informed, value-concordant
screening decisions thereby maximizing the benefits of screening while minimizing its harms.
