A groundbreaking study has demonstrated that artificial intelligence-enabled chatbots can effectively match traditional approaches in facilitating cancer genetic testing services, potentially expanding access to vital genetic counseling resources.
The BRIDGE trial (NCT03985852), conducted from August 2020 to August 2023, found that a chatbot-based approach achieved equivalent outcomes to standard-of-care (SOC) methods in helping eligible individuals complete genetic testing and pretest services. The findings were published in JAMA Network Open.
Key Trial Findings
The study, which included over 3,000 participants, showed that the chatbot group (n=1554) and SOC group (n=1519) achieved similar completion rates for genetic testing (12.3% vs 13.6%) and pretest genetic services (25.7% vs 23.8%). The estimated differences were -1.3% and 2.0% respectively, meeting statistical criteria for equivalence.
"There's greatly increasing demand for pretest cancer genetic services for unaffected patients," explained Dr. Kimberly Kaphingst, lead author and director of cancer communication research at Huntsman Cancer Institute. "The rationale was to test whether we could use a chatbot as part of a genetic counseling workflow to supplement the role of certified genetic counselors."
Innovative Technology Implementation
The chatbot utilized in the study was a rules-based system enhanced with natural language processing capabilities. It delivered educational content through text, images, and video, replicating key information typically provided during standard pretest genetic counseling appointments.
Participants accessed the chatbot through a patient portal link and could progress through scripted information at their own pace. The system allowed for both predetermined topic exploration and open-ended questions, with complex queries being forwarded to the genetic counseling team.
Patient Population and Methodology
The trial enrolled patients aged 25 to 60 years who met specific criteria:
- No prior cancer diagnoses (except nonmelanoma skin cancer)
- No previous genetic counseling or testing
- Eligible for testing based on family history
- English or Spanish speaking
Patient identification was facilitated by GARDE, an open-source platform that analyzes family cancer history from electronic health records. The study population was well-balanced between arms, with mean patient ages of 43.5 and 44.1 years in the chatbot and SOC groups respectively.
Impact on Healthcare Delivery
Whitney F. Espinel, MS, CGC, from the Breast Cancer Risk Clinic at Huntsman Cancer Institute, highlighted the potential impact: "Having these types of tools to help supplement the genetic counseling process is going to be extremely helpful. We can utilize chatbots during pretest and cut out some of that need for the genetic counselor up front."
Future Directions
The success of the BRIDGE trial has prompted researchers to explore expanded applications of the technology. The Huntsman Cancer Institute has received additional funding from the National Cancer Institute to study the chatbot's effectiveness with Spanish-speaking women.
The research team is particularly focused on improving access for underserved populations, including:
- Spanish-speaking patients
- Rural and frontier populations
- Communities with limited broadband access
This innovative approach to genetic counseling service delivery represents a significant step forward in making genetic testing more accessible while maintaining high-quality care standards.
