Bridging Technology and Pretest Genetic Services: Quantitative Study of Chatbot Interaction Patterns, User...

Area of Science:

• Genetics
• Health Informatics
• Digital Health

Background:

• Chatbots, or conversational agents, are increasingly utilized for genetic service delivery.
• Limited research exists on user interactions with healthcare chatbots in real-world settings.

Purpose of the Study:

• To examine user interaction patterns with a cancer genetics education chatbot.
• To investigate associations between user characteristics, chatbot interactions, and genetic testing decisions.

Main Methods:

• Analysis of data from a pragmatic trial comparing chatbot intervention to standard care for hereditary cancer genetic testing.
• Participants received educational content via chatbot, with options for supplementary prompts and open-ended questions.
• Logistic regression models assessed relationships between factors and genetic testing decisions.

Main Results:

• 83.5% of users completed the chatbot interaction, and 80.6% of completers were willing to pursue genetic testing.
• Selecting more prompts or asking open-ended questions was associated with a lower likelihood of pursuing genetic testing.
• Interaction patterns were generally not associated with clinical or sociodemographic factors, suggesting potential scalability.

Conclusions:

• Chatbots demonstrate effectiveness in engaging users and high acceptability for genetic education.
• Sociodemographic factors did not significantly influence interaction patterns, indicating broad applicability.
• Future chatbot designs should address high information needs to better support informed genetic decision-making.