Cascade conversations: Empowering cancer genetic testing through Cascade chatbots.

e22621 Background: Cascade genetic testing refers to the process of offering genetic testing to blood relatives of individuals with known pathogenic mutations. Identifying cancer-associated pathogenic mutations offers relatives the opportunity for targeted surveillance and preventive interventions, which can reduce cancer incidence, morbidity, and mortality. However, cascade genetic testing remains critically underutilized, and patients traditionally must shoulder the burden of facilitating relative education. We created a Cascade Chatbot to assist families with genetic testing via gene-specific educational modules and facilitated access to genetic testing services. This quality improvement initiative investigates the acceptability of Cascade Chatbot in a gynecologic oncology clinic. Methods: All patients with pathogenic mutations in the BRCA1, BRCA2, MLH1, MSH2, MSH6, PMS2, ATM, RAD51C, RAD51D, PALB2, CHEK2, and BRIP genes presenting for care at a gynecologic oncology clinic were offered a Cascade Chatbot. Interested patients were provided with the Cascade Chatbot available via hyperlink and QR code to share with family members providing education and resources for testing. Patients were contacted two weeks following the receipt of the Chatbot to determine if they had shared the Chatbot with relatives. The primary outcome was the proportion of patients that accepted the Cascade Chatbot when offered at their gynecologic oncology outpatient visit. Results: One hundred consecutive patients (median age 43 years; IQR = 35-51.5 years) were offered the Cascade Chatbot; 85 (85%) identified as non-Hispanic white, 8 (8%) as Hispanic and white, 5 (5%) as Asian/Indian/Pacific Islander, 2 (2%) as black, and 1 (1%) as other/unknown ethnicity. Among patients approached, 59 (59%) had eligible relatives (blood relatives at risk for carrying the familial pathogenic mutation that had not yet completed genetic testing). Among patients with eligible relatives, 58 (98.3%) accepted the Cascade Chatbot. Two weeks following administration of the Cascade Chatbot, 44 (75.9%) patients had shared the tool with at least one relative, 8 (13.8%) had not but intended to, 3 (5.1%) patients opted not to share, and 3 (5.1%) could not be reached for follow up. Conclusions: Ninety-eight percent of patients with a hereditary cancer syndrome and relatives eligible for genetic testing accepted a Chatbot tool to facilitate cascade genetic testing. At two weeks, 76% of patients had shared the tool with relatives. Our work suggests that Chatbots may be a feasible and widely accepted tool to initiate conversations about family cascade genetic testing for patients across clinical settings, alleviating the burden currently placed on patients. Ongoing efforts are focused on evaluating the downstream impact of chatbot-driven education and facilitation on genetic testing uptake.