Enhancing Mammography Programs for Outreach, Wellness, Education, and Resources (EMPOWER) in Underserved Populations Study

Breast cancer screening inequities are drivers of disparities for rural and Latina women. Chatbots are increasingly popular in various healthcare contexts and can be easily accessed through smartphones, tablets, laptops, and desktops. Chatbots have many advantages for patient messaging, including providing scripted education and motivational information interactively, chunking information into digestible segments, and allowing for choice in the amount of information received at any one time. Thus, with chatbots this study can tailor the interaction based on individual patient factors. Chatbots are accessible to the vast majority of U.S. adults. While chatbots have been used successfully in some clinical contexts, there is a lack of studies that investigated the use of chatbots as part of a mobile screening program to increase adherence to follow-up recommendations about either screening or diagnostic care. Study staff will approach women at the time of or soon after a patient's routine breast cancer screening and invite the patient to participate in the study. Participants will complete baseline surveys at time of enrollment. This survey includes demographics and preferred contact method (e.g., text, email) that will be used to initiate the chatbot communication which can be completed via phone or website. Usual clinical procedures will be used to interpret participants' routine screening mammograms, including the use of the breast imaging-reporting and data system (BI-RADS), typically within 1 week of imaging. Based on their mammography results, women will be placed into two cohorts using their BI-RADS category. All participants will complete follow-up questionnaires, available in English and Spanish. For the follow-up surveys, the research team will contact participants via their preferred method (text or email) and send a link to the follow-up survey. If unsuccessful, the research team will then contact participants by phone to complete the survey via text or call. Once mammogram results are available, patients will be randomized 1:1 to either usual care or usual care with chatbot, with randomization in permuted blocks of size and stratified by age (\<55 vs. ≥55 years, a proxy for menopause and indicator of risk); rural/frontier vs. urban; language preference (English vs. Spanish); and cohort (normal vs. abnormal result). This study will randomly invite participants from four groups for focus group discussions (FGDs) based on their mammogram results and adherence to screening recommendations. This study aim to conduct 8 FGDs (2 per group, one in Spanish and one in English) with 8-10 participants each.