Feasibility of an electronic patient-facing cancer family history tool in medically underserved populations

We developed an electronic patient-facing family history collection tool including B-RST 3.0, PREMM₅ risk assessments and "limited family knowledge/structure" information designed for primary care settings. We evaluated the tool's performance compared with genetic-counselor-collected information for clinical risk stratification in a population with barriers to access.

English- or Spanish-speaking patients aged 18 to 49 were invited to participate. Individuals with limited family knowledge or at high or moderate risk based on their responses in the tool were offered genetic testing and counseling. We assessed overall agreement of family history collected in the tool compared with family history collected by the genetic counselors using Krippendorff's alpha (K-alpha). Multivariable logistic regression was used to assess characteristics associated with inaccuracy.

Most people (94%, n = 1711) who interacted with the tool completed it. Those included in the agreement analysis (n = 604) had a median age of 36.3 years, 81.6% were female, and 44.4% were Non-Hispanic White. Both the B-RST 3.0 and PREMM₅ had moderate agreement: 69.9% (K-alpha = .40, 95% CI [0.32, 0.47]) and 83.9% (K-alpha = .52, 95% CI [0.43, 0.60]), respectively. Agreement was high (96%) for people with clinically significant risk for one of the hereditary cancer syndromes. For B-RST 3.0, the factors significantly associated with inaccuracy were study site, sex, and race/ethnicity. For PREMM₅, age, sex, and education were associated with inaccuracy. Barriers to access were not associated with inaccuracy.

Implementation of this tool could increase identification of individuals at risk for hereditary cancer syndromes, including those with barriers to health care access.