Pathways’ Pick of the Week: Reimbursing AI

article image
ARTICLE SUMMARY:

Artificial intelligence and machine learning tools don’t fit very neatly into current coverage, coding, and payment frameworks. But CMS has been making efforts to figure out the way forward. Excerpted from Pathways’ Picks for 2023: 5 Medtech Policy Issues Watch.

In the coming year, one key area to watch to help gauge CMS’ approach to incorporating next-generation technologies into the Medicare system is digital health and artificial intelligence/machine learning (AI/ML). These tools don’t fit very neatly into the current coverage, coding, and payment frameworks. But the agency has been making efforts to figure out ways forward, particularly for AI/ML tools. In particular, in 2021 and 2022, CMS issued a range of “requests for information,” as part of its annual payment rules, seeking feedback on fundamental considerations for reimbursing AI—for instance, to what extent it replaces, not just supplements, physician work and time. The agency has also increasingly started making reimbursement policies for individual AI-based devices, but any sort of underlying philosophy or framework to address the technology class is still in the works. Industry advocates worry that CMS could end up undervaluing AI/ML tools as a cheaper alternative to physician work instead of rewarding improved efficiency and care. It’s worth watching for any broader pronouncements from CMS in 2023 as it processes all the input it has received in the past two years. Meanwhile, private payors in the US and government payors worldwide are wrestling with the same issues.

As the earliest foundations of reimbursement frameworks for AI/ML devices start to be built, regulatory structures are becoming more fully formed. And it is likely there will be some significant developments on the regulatory side in 2023. Most significantly, the EU is advancing toward finalizing its sweeping AI Act by the end of the year. The legislation, which addresses practices and oversight for AI broadly, not just in healthcare, was proposed in 2021 and has since undergone multiple rounds of revision and debate. Medtech stakeholders remain worried that the new rules will add an overlay of duplicative requirements and assessments on top of the EU Medical Device and IVD Regulations. Case in point: TEAM-NB, the association of EU medical device notified bodies, recently circulated a position paper working through strategies for members that are already operating under the MDR/IVDR to also become designated to perform assessments under the AI regulation. On the flipside, MDR as it currently stands doesn’t necessarily respond to the particular challenges of AI/ML, such as adaptive algorithms, so there’s hope by some in industry that the focused AI regulation could help enhanced EU’s approach to the technology.

FDA, by many accounts, is more advanced in its approach to AI/ML-based systems compared to the EU, including its adoption of so-called predetermined change control plans (PCCPs), which are designed to accommodate adaptive algorithm systems. The agency is expected to issue a highly anticipated draft guidance document this year spelling out the application of PCCPs.

Of course, other countries are watching these efforts as they work toward their own AI regulatory approaches. And, in that context, both the EU AI Act and the FDA PCCP guidance will be very influential documents for the future of AI regulation globally.

  Trial MyStrategist.com and unlock 7-days of exclusive  subscriber-only access to the medical device industry's most trusted strategic publications: MedTech Strategist & Market Pathways. For more information on our demographics and current readership  click here.

×



Articles from David Filmore: