Chatbot more cost-effective for finding high-risk pathogenic variants in cancer genes
Click Here to Manage Email Alerts
Key takeaways:
- The chatbot identified three times as many patients with high-risk pathogenic variants as a traditional genetic counseling pathway.
- The chatbot also cost considerably less.
A cloud-based chatbot considerably improved identification of high-risk pathogenic variants in cancer genes, according to study results.
The chatbot identified three times as many patients with high-risk pathogenic variants and did so at a much lower cost, findings presented at Society of Gynecologic Oncology Annual Meeting on Women’s Cancer showed.
“Compared [with] current practice, utilization of a novel chatbot for genetic counseling and testing has the ability to improve access to care, identify more high-risk patients and decrease the cost of genetic testing,” researcher Kari L. Ring, MD, gynecologic oncologist at University of Virginia Health, told Healio.
Background and methods
An estimated 5% to 8% of the population carries germline pathogenic variants in a cancer gene. However, standard screening methods do not identify many of these patients, according to study background.
The need for new, effective strategies to ensure equitable access to high-risk care has increased given the rising number of actionable genes, the decreased cost of genetic testing and a shortage of genetic counselors, researchers wrote.
Ring and colleagues assessed whether a virtual genetic information assistant (Gia, Invitae Corporation) — a cloud-based chatbot developed to identify patients with high-risk pathogenic variants and help counsel them about genetic testing — could improve identification of high-risk patients.
Investigators also evaluated whether Gia could do this at a lower cost than standard approaches, including referral for genetic counseling.
Researchers used TreeAge software to build a decision tree to determine costs and pathogenic variant detection rates for Gia and genetic counseling.
Based on the estimated number of annual exams performed at a primary care site in rural Virginia over the course of a year, investigators added 2,600 patients to the model, with half in the Gia pathway and half in the genetic counseling pathway.
Researchers calculated probabilities at each decision point based on available evidence in the literature. They calculated costs based on Medicare reimbursement rates for genetic testing costs, relevant Current Procedural Terminology codes and Gia fees from Invitae Corporation.
Results and implications
Results showed 1,079 of 1,300 patients in the genetic counseling group would undergo screening for a pathogenic variant in a cancer gene, with 227 screening as high risk for a pathogenic variant, 20 presenting for genetic counseling, 19 completing genetic testing and one testing positive for a high-risk pathogenic variant.
Results showed 740 of 1,300 patients in the Gia group would complete pre-test screening, 200 would screen as high risk for a pathogenic variant, 58 would undergo genetic testing and three would test positive for a high-risk pathogenic variant.
Researchers estimated the total cost for exams, appointments and testing in the genetic counseling group — equal to the cost to identify one patient with a high-risk pathogenic variant — would be $462,190.
They estimated the cost for pre-test screening using Gia and genetic testing to be $17,425, with an estimated cost of $5,808 to identify one patient with a high-risk pathogenic variant.
“We are interested in improving equitable access to genetic testing and high-risk care for patients through utilization of novel approaches, such as the chatbot evaluated in our study,” Ring said. “Our primary aim was to evaluate whether we were able to identify more high-risk pathogenic variant carriers with utilization of the chatbot. Not only did we see that in the model, but also saw that we could identify more high-risk patients at a lower cost, which is key when discussing the upfront investment in technology into clinical practice.”
The decision tree calculation showed increased effectiveness of traditional genetic counseling — 0.67 compared with 0.46 for Gia — based largely on the higher number of patients who underwent pre-test screening. This led to an incremental cost-effectiveness ratio of $1,633 per correctly identified patient, according to investigators. However, Gia still identified more high-risk patients at a lower cost, they wrote.
“The acceptability of chatbots for counseling and testing remains a key question in different patient populations and among providers,” Ring told Healio. “Given broader indications for genetic testing and shortage of genetic counselors, especially in underserved patient populations, novel approaches to high-risk care like chatbots are critical to improve access to genetic testing and care.”
Collapse
Underkofler K, et al. Modeling and comparison of costs and high-risk variant detection between a genetic counseling chatbot and a traditional genetic counseling pathway. Presented at: Society of Gynecologic Oncology Annual Meeting on Women’s Cancer; March 25-28, 2023; Tampa, Florida.
Click Here to Manage Email Alerts