Patient satisfaction, comprehension, and regret/conflict of their genetic testing decision after receiving genetics education via a virtual genetic counseling assistant: A randomized pilot trial

As genetic testing and personalized medicine have grown there has been an increased demand for genetic counseling services. Currently, there are several barriers faced by both patients looking to receive these services and genetic counselors (GCs) looking to provide care, including lack of access, billing and licensure issues, and insurance coverage issues for patients. Despite barriers to genetic counseling, a 2015 study demonstrated that 35.0% of patients expressed a strong desire to learn more about genetic testing (GT) and their genomic risk (Jagsi et al.). Artificial intelligence (AI) chatbots have already been implemented in genetic counseling clinics to provide genetic education and testing to patients while addressing these barriers. This feasibility pilot trial implements a novel AI chatbot, or <b>Ca</b>ncer <b>R</b>isk <b>E</b>valuation (CARE) digital assistant (also called Alex), in an oncology setting to educate and inform participants of their risk for hereditary cancer and the utility of GT. The aim of this trial was to understand patients’ attitudes on decisional conflict and decisional regret regarding their GT choice. Additionally, we explored if receiving genetics education via the AI chatbot had any effect on participants’ understanding of the genetic information being conveyed. The trial involved randomizing participants (n = 26) to one of two groups and administering a baseline and one-month follow-up survey to evaluate participants’ attitudes on their GT decision. The Enhanced Usual Care (EUC) group (n = 14) received a letter informing them of their hereditary cancer predisposition risk and a referral for genetic counseling services, while the Relational Assistant (RA) group (n = 12) received genetics education through the CARE digital assistant. Results showed that RA participants exhibited higher levels of GT decision satisfaction and lower levels of decisional conflict and regret, whereas EUC participants displayed the opposite on every measure. There was largely no difference between the two groups on GT knowledge. These results suggest that using an easy-to-navigate chatbot for genetics education leads to high satisfaction and confidence in the GT decision-making process. Additionally, the high rate of GT uptake suggests it is a viable model to minimize barriers to both genetics education and testing.