BC-DAISY: A Breast Cancer Decision Aid System

Decisions about breast cancer are complex and preference-based. Existing decision aids that help identify women eligible for tamoxifen based on their breast cancer risk, such as the Breast Cancer Risk Assessment Tool (BCRAT), were not designed to identify women's risks for side-effects, her preferences for outcomes affected by treatment, nor the net balance of benefits and risks. It is possible to predict the risks of side effects of treatments according to patient characteristics, using patient-specific models.Our goal is to develop a decision aid, TXplore, that rates the overall benefit:risk profiles of various breast cancer prevention strategies according to a woman's risks for breast cancer, side-effects, and preferences. Each user will receive a customized report card grading each available prevention option, using preference-weighted risk:benefit grades; users can also explore personalized feedback from the program. Our hypotheses are that this tool can improve the implementation of appropriate prevention strategies, promote risk reduction behavior, and improve PCP's ability to identify and counsel high risk women.We propose building on modeling techniques that link the benefits and risks of preventive strategies to patient characteristics and preferences. We now seek to conduct focus groups and usability tests on end-users to optimize the design of the prototype TXplore.

The Specific Aims are:

To optimize the design of TXplore by focus groups conducted among diverse high-risk women and PCPs to a) identify the domains driving decisions about prevention and explore the framing of questions comparing the relative importance of one domain to another, synthesizing this information as preference trade-off questions; and b) explore the framing of risk:benefit grades and personalized feedback.
To assess the comprehensibility of preference trade-off questions by testing each question on a diverse sample of 50 high-risk women, revising and retesting.
To assess the usability of TXplore for patients and PCPsThe successful completion of this project will produce a novel tool for counseling women about their individual benefits and risks of breast cancer prevention strategies. It will be the first tool of its kind to integrate individual patient risks and preferences into decision support.