Abstract: INFO11-FR
Leveraging Resources from Two Research Cores and the Veterans Health Administration to Assess the Value of Artificial Intelligence (AI)-Enabled Risk Prediction for Diabetic Kidney Disease in US Veterans
Session Information
- Informational Posters - II
November 03, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Diabetic Kidney Disease
- No subcategory defined
Authors
- Kim, Kibum, University of Illinois Chicago, Chicago, Illinois, United States
- Sarker, Jyotirmoy, University of Illinois Chicago, Chicago, Illinois, United States
- Abdelaziz, Abdullah, University of Illinois Chicago, Chicago, Illinois, United States
- Crook, Jacob L., University of Utah, Salt Lake City, Utah, United States
- Nelson, Richard E., University of Utah, Salt Lake City, Utah, United States
- Lu, Chao-Chin, University of Utah, Salt Lake City, Utah, United States
- Nyman, Heather A., University of Utah, Salt Lake City, Utah, United States
- LaFleur, Joanne, University of Utah, Salt Lake City, Utah, United States
Description
An artificial intelligence-enabled prognostic testing platform to guide management for early-stage diabetic kidney disease (DKD) has recently been granted FDA De Novo Marketing Authorization. An efficient allocation of limited resources in adopting and promoting such advanced technologies necessitate value evidence for a specific target population. To achieve this, a collaborative research model was developed by investigators from the University of Utah Data-driven Collaborative of Informatics, Pharmacoepidemiology & Health Economics Researchers (DeCIPHER), University of Illinois at Chicago Center for Pharmacoepidemiology & Pharmacoeconomic Research (UIC CPR), and Salt Lake City Veterans Affairs (VA).
The aim of this collaboration was to assess the cost-effectiveness of treatment decisions informed by the prognostic risk levels compared to the standard of care (SoC) pathway for US veterans with stages G1-G3b of DKD. DeCIPHER, in collaboration with Salt Lake City VA, organized the data access and provided resources to conduct a historic cohort study. Investigators at the UIC CPR built an economic model that involes the performance of this platform in identifying high-risk patients and state transitions across multiple DKD stages. Five-year incremental cost-effectiveness ratio (ICER) was calculated as a measure of the value.
Using contemporary epidemiology and economic outcome data from a historic cohort study comprising 700,000 DKD patients, the research team was able to calculate the value of this platform. The 5-year ICER for risk-level-informed comprehensive care versus SoC was found to be $116,300 per quality-adjusted life year (QALY) gained. The platform is cost-effective compared to SoC in 67% and 85% of probabilistic-sensitivity simulations at willingness-to-pay thresholds of $150,000 and $200,000 per QALY gained, respectively. This exemplary research partnership that leveraged the unique resources of each institution quantified the value of the progression platform for the US veteran population.
Funding
- Renalytix AI