Inside the AMA CPT Process: Lessons for MedTech Founders

Most people outside the reimbursement and coding world never really see how CPT codes come to life.

The AMA CPT process sits at a strange and fascinating intersection of medicine, economics, policy, specialty politics, and language. A single word in a descriptor can influence physician adoption, payment pathways, specialty ownership, utilization tracking, and even how investors perceive a technology category.

At the May 2026 AMA CPT Editorial Panel meeting, I secured approval for 3 new CPT codes across very different technologies. What many people do not see is that a successful CPT application is rarely the result of a few months of work. In many cases, the process stretches across years, building specialty society support, refining clinical workflows, generating evidence, pressure-testing utilization assumptions, navigating stakeholder politics, and repeatedly revising language before an application is ever presented to the Panel.

And even then, much of the debate comes down to something deceptively simple: what, exactly, is the code intended to describe?

One of the central tensions within the CPT process is the ongoing push and pull between describing the technology itself versus describing the physician or qualified healthcare professional work being performed.

Historically, CPT was built around physician actions. The language of the code set reflects that. Descriptors are intentionally grounded in verbs: insertion, programming, stimulation delivery, interpretation, analysis, imaging guidance. The CPT Editorial Panel is generally cautious about descriptors that sound too tied to a proprietary technology or that read more like product positioning than clinical work.

But medicine is changing. Many of today’s innovations, particularly in AI, digital health, neuromodulation, robotics, software-enabled diagnostics, and combination products, derive much of their value from technology-specific functionality that is difficult to fully capture through generic physician action verbs alone.

That is where the debate starts.

If a descriptor leans too heavily into technology features, there is concern that the code becomes overly narrow, difficult to scale, or too closely associated with a single company or platform. But if the descriptor becomes too generic and action-oriented, the uniqueness of the innovation can disappear entirely, weakening future payment arguments and reducing the signaling value that the code itself can carry in the market.

This tension is one reason why CPT timelines can become so long. Reaching consensus on descriptor language often requires balancing the perspectives of specialty societies, practicing physicians, advisors, payers, regulators, manufacturers, and the Editorial Panel itself. In many cases, the descriptor evolves through multiple iterations over several meeting cycles before the final wording is accepted.

But there is also an important “so what?” question that companies often underestimate. Securing a CPT code does not automatically mean a technology will be covered. It does not guarantee payment. And it certainly does not guarantee adequate payment. That reality is difficult for many innovators and investors to appreciate early on because the CPT process itself can feel like the finish line. In practice, it is often just the beginning of the next phase of work.

In some cases, the very descriptor compromises required to get a code approved can later create downstream challenges during coverage and payment discussions. A descriptor that becomes too broad may weaken the argument for differentiated reimbursement. A descriptor that avoids explicit references to technological functionality may make it harder to articulate why the service warrants unique payment treatment at all.

This is why reimbursement strategy cannot start with coding alone. The strongest companies tend to approach coding, coverage, payment, evidence generation, physician adoption, and health economic positioning as an integrated strategy from the beginning. They recognize that CPT is not simply an administrative exercise. It is a language strategy, a policy strategy, and often a long-term market-shaping strategy.

The descriptor ultimately becomes part reimbursement infrastructure, part regulatory artifact, part market positioning framework, and part category definition. Once it enters the code set, it can shape how an entire technology category is discussed for years afterward.

This year’s May meeting felt like another reminder that the future of CPT will increasingly be shaped by technologies that blur the boundaries between “clinical work” and “technology functionality.” As AI and software-driven interventions continue evolving, I suspect these debates inside the CPT process are only going to become more important, particularly as stakeholders wrestle not only with how to describe these technologies, but also whether the U.S. healthcare system is truly prepared to value and pay for them.

Do you have U.S. commercialization questions?

Whether it's how the multiple payer landscape works in the U.S., how to validate device or procedure coding, timelines associated with product uptake - we're happy to answer them.

Schedule and introductory call: https://calendly.com/nicolecoustier/introductions