Artificial intelligence in Medicare: utilization, spending, and access to AI-enabled clinical software.

Area of Science:

• Cardiovascular Medicine
• Health Informatics
• Medical Economics

Background:

• In 2018, Medicare introduced reimbursement for CT fractional flow reserve (FFRCT), the first AI-enabled clinical software for diagnosing coronary artery disease.
• Understanding the adoption patterns of such technologies is crucial for assessing their impact on healthcare delivery and patient outcomes.

Purpose of the Study:

• To quantify Medicare utilization and spending on FFRCT from 2018 to 2022.
• To characterize the hospitals, clinicians, and patients adopting FFRCT technology.

Main Methods:

• Analysis of 100% Medicare fee-for-service claims data.
• Comparison of FFRCT-adopting vs. non-adopting entities (hospitals, clinicians) and patients.
• Assessment of annual trends in FFRCT utilization and spending.

Main Results:

• FFRCT billing volume increased over 11-fold between 2018 and 2022.
• Adopting hospitals were larger, part of health systems, nonprofit, and profitable.
• Adopting clinicians were more likely to be cardiac specialists in larger practices.
• FFRCT patients were more likely to be male, White, and less likely to have Medicaid or disability benefits.

Conclusions:

• The initial 5-year diffusion of FFRCT was concentrated among well-resourced healthcare providers.
• Monitoring the spread of AI-enabled clinical software is essential to ensure equitable access to these diagnostic tools.