AI-Based Risk Stratification Proves Cost-Effective for Diabetic Kidney Disease in Veterans

Key Takeaways:

• US veterans exhibit higher rates of diabetes, making them more likely to develop diabetic chronic kidney disease (DKD), which has faster disease progression and higher health care costs.
• An artificial intelligence (AI)-driven in vitro kidney disease risk assay (AIKD) is a potential cost-effective risk stratification method whose implementation could benefit the US Veterans Health Administration (VHA) health care system.
• AIKD garnered an additional $1317 per 0.0113 quality-adjusted life-year (QALY) gained when compared with a traditional risk stratification model, resulting in an estimated incremental cost-effectiveness ratio (ICER) of $116 349 per QALY gained.

The rise of diabetes, a known risk factor for chronic kidney disease (CKD), has led to an expected increase in DKD. DKD results in faster disease progression and higher health care costs than patients with non-diabetic CKD. Additionally, higher rates of DKD have been found in veteran populations due to the high prevalence of diabetes among US veterans.

Advancements in technology, particularly in AI, show promise in improving risk stratification. AIKD has recently emerged as a new method of assessing the risk of kidney disease progression.

A study aimed to compare the cost-effectiveness of AIKD in risk stratification against the procedural Kidney Disease: Improving Global Outcomes (KDIGO) guidelines. Since veterans are 34% more likely to have kidney disease, the study group included patients with DKD receiving care within the US VHA health system.

AIKD Demonstrates Slight Increases in Care Costs and QALY

Traditional KDIGO risk stratification costs $145 120 for 2.8164 QALYs while AIKD risk stratification has a higher cost of $146 437 for 2.8277 QALYs over a 5-year period. This results in an additional $1317 per 0.0113 QALY gained, an estimated ICER of $116 349 per QALY gained for AIKD.

A sensitivity analysis conducted by researchers found that the ICER was more impacted by the sensitivity of AIKD and KDIGO and their ability to correctly exclude patients with no risk of progressive decline in kidney function. Furthermore, the analysis showed AIKD as the cost-effective method of risk stratification in 69% of the simulations at a willingness-to-pay threshold of $150 000 per QALY gained.

Although AIKD resulted in higher care costs, those incremental costs are offset by the QALYs gained, and the model may be improved with future research and technology updates. The study demonstrated AIDK as a cost-effective strategy for kidney disease risk stratification.

According to the authors, “Once properly implemented, AIKD can assist providers in referring high-risk DKD patients to specialty care or in making treatment decisions for switching to renal protective medications.”

Reference

Sarker J, Abdelaziz AI, Crook J, et al. Cost-effectiveness analysis of artificial intelligence-driven risk stratification in patients with diabetic kidney disease in the US veterans population. Kidney Med. 2026;8(3):101261. doi:10.1016/j.xkme.2026.101261