By Editorial TeamAI in healthcare has a ton of promise, but great tech often dies on the vine because there’s no clear way for hospitals to get paid for using it. If you’re a digital health VC or a regulatory advisor, you have to understand Medicare’s New Technology Add-on Payments (NTAP). It’s not some academic concept, it’s the thing that determines whether an enterprise digital health company actually makes money and gets any market share. This is a look at how specific CMS payment pathways get cardiac AI adopted in the real world, giving investors a better way to judge their bets.
The NTAP Mechanism: Bridging the Reimbursement Gap for Novel Cardiac AI
The Centers for Medicare and Medicaid Services (CMS) created the New Technology Add-on Payment (NTAP) program to give hospitals a reason to try new things. Basically, if a new technology offers a big clinical improvement but isn’t covered by existing Diagnosis-Related Group (DRG) payments and costs too much, NTAP provides a temporary extra payment. This makes it less scary for a hospital to use an expensive new tool that their budget office would otherwise reject. For a new cardiac AI tool, getting an NTAP designation is a huge win, acting as a financial bridge from its first day on the market to when it gets a permanent payment code. This is especially true for cardiac AI that’s considered Software as a Medical Device (SaMD), since these tools often provide brand-new diagnostic capabilities that have no existing CPT code. Without that NTAP money, hospitals would lose money adopting the technology, which means it would never get off the ground. The program is really a subsidy for the early adoption period, letting hospitals get comfortable with the tech and generate the evidence needed to justify a permanent payment path later.
HeartFlow’s NTAP Journey: A Blueprint for Cardiac AI Commercialization
HeartFlow is the perfect case study of an AI company using the NTAP pathway to get its tech into hospitals. The company developed a non-invasive way to do fractional flow reserve analysis from a CT scan (FFR-CT), using AI to build a 3D model of a patient’s arteries and simulate blood flow to spot blockages. It gives doctors the physiological data they need about coronary artery disease without having to perform an invasive procedure. How did they get hospitals to pay for it? They got an NTAP designation. CMS created these add-on payments for certain clinical AI technologies, and HeartFlow was a textbook example of who should benefit. The company got its first NTAP approval back in 2017, which went into effect on January 1, 2018. This gave hospitals an extra payment of $1,450.50 for the technical component on eligible inpatient cases where HeartFlow FFR-CT was used, which made trying out the new diagnostic a much easier financial decision. That temporary NTAP payment was the foot in the door. It let HeartFlow prove its clinical value and build a base of evidence, eventually allowing them to transition off NTAP and onto permanent Category I CPT codes. Today, the 2026 hospital outpatient payment rate for FFRCT (APC 5724) is set at $877. CMS NTAP payment rates for HeartFlow FFR-CT That NTAP window was everything for HeartFlow’s market access. Hospitals could bring in a new diagnostic that improved patient care, and they could actually cover most of the cost while doing it. This temporary payment gave the company time to prove itself in the real world and secure its long-term payment strategy. HeartFlow’s success shows exactly how a focused, vertical AI healthcare company can thread the needle of reimbursement by solving one specific, high-value problem and playing by the regulatory rules.
Assessing Reimbursement Pathways for Novel AI Diagnostics: A Venture Capital Framework
For any VC in digital health, a startup’s reimbursement strategy is what separates a science project from a viable business. HeartFlow’s story gives us a practical framework for judging new AI diagnostics, especially in cardiology.
1. Early-Stage Regulatory Strategy and Data Moat Development
A solid regulatory plan is just the beginning. For a cardiac AI company, that usually means a 510(k) clearance or maybe a De Novo classification for something truly new. But the reimbursement work starts way before that with building a data moat. A company that has a proprietary, hard-to-replicate dataset, think of iRhythm and its millions of labeled ECG recordings, that makes its AI model better is just a much stronger bet. These are the truly AI-native companies, the ones that built their entire product, data pipeline, and business model around AI from the start and are set up for the long run.
2. NTAP Eligibility and Application Readiness
A startup whose AI offers a genuine clinical leap forward for an unmet need has to be thinking about NTAP eligibility from day one. That means they need to have read the CMS NTAP guidelines and studied past decisions. A company will be in a much stronger position if it can clearly show why its tech isn’t covered by existing DRGs and how it meets the cost threshold. Getting a Breakthrough Device Designation can also put them on a faster track for NTAP, as we saw with an AI tool for detecting cardiac amyloidosis. As an investor, you have to dig into a company’s NTAP readiness, from the quality of its clinical data to whether its team actually understands the application process.
3. Transition to Permanent Reimbursement: CPT Codes and Real-World Evidence
NTAP is a temporary fix, usually for just two or three years. The end game is always permanent reimbursement through an established CPT code. To make that jump, a company needs a mountain of real-world evidence (RWE) that proves its tech is clinically effective and cost-effective across different kinds of patients. This RWE, pulled from EHRs, patient registries, and claims data, is what backs up the clinical trial data and makes a convincing case to both the FDA and payers. Companies that are smart enough to collect and analyze RWE from the very beginning will have a much smoother path to getting a Category I CPT code. Just look at Anumana, the first ECG-AI to secure CPT codes. That achievement alone creates a massive reimbursement moat.
4. Understanding the Ecosystem: QMS, GMLP, and Patent Thickets
Reimbursement isn’t the whole story. Investors also have to look at a company’s operational maturity. Having a Quality Management System (QMS) certified to ISO 13485 is a clear sign that a medical device company is mature and compliant. For AI specifically, following Good Machine Learning Practice (GMLP) principles from the FDA, Health Canada, and MHRA is non-negotiable for ensuring an adaptive AI model remains safe and effective, especially for preventing algorithmic drift over time. You also have to know the competitive IP situation. HeartFlow, for example, surrounded its CT-FFR technology with a dense patent thicket, making it very difficult for anyone else to compete. ISO 13485 standard for medical devices
Conclusion
For vertical AI companies in healthcare, especially in cardiac diagnostics, figuring out the reimbursement puzzle is just as hard and just as important as building the tech itself. The Medicare New Technology Add-on Payment program offers a short-term but essential path for new AI to get a foothold in the market. HeartFlow’s experience with FFR-CT provides a clear playbook for how a specialized AI company can get over the initial payment hurdles. For VCs, the job is to find the companies that understand this game, the ones with a clear-eyed strategy for NTAP, a plan for building the evidence for permanent reimbursement, and a solid regulatory foundation. Those are the companies that will define the next generation of healthcare. CMS NTAP program guidelines overview
Frequently Asked Questions
What is the primary purpose of the Medicare New Technology Add-on Payment (NTAP) program?
The NTAP program was established by the Centers for Medicare and Medicaid Services (CMS) to incentivize hospitals to adopt new technologies that represent a substantial clinical improvement. It provides a temporary, additional payment to mitigate the financial disincentive for hospitals to utilize innovative, often more expensive, technologies not adequately paid under existing Diagnosis-Related Group (DRG) payments.
Why is NTAP particularly relevant for novel cardiac AI solutions?
NTAP is particularly relevant for cardiac AI because many of these solutions, especially Software as a Medical Device (SaMD), introduce entirely new diagnostic or prognostic capabilities not covered by existing CPT codes. Without NTAP, hospitals might face significant financial losses when adopting such technologies, making widespread commercial adoption nearly impossible. The program essentially subsidizes the initial adoption phase, allowing hospitals to gain experience with the technology.
How did HeartFlow leverage NTAP to achieve commercial success?
HeartFlow successfully utilized NTAP to drive hospital adoption of its fractional flow reserve technology, receiving its first NTAP approval in 2017. This designation provided hospitals with an additional payment for using HeartFlow FFR-CT, significantly reducing the financial burden associated with adopting this novel diagnostic tool. This temporary payment mechanism created a window of opportunity for HeartFlow to establish its clinical utility, build a robust evidence base, and eventually transition towards more stable, permanent reimbursement mechanisms.
What key considerations should venture capital partners assess regarding reimbursement for novel AI diagnostics?
Venture capital partners should assess a startup’s early-stage regulatory strategy, including securing 510(k) clearance or De Novo classification, and the development of a strong data moat. Additionally, they should evaluate the company’s NTAP eligibility and application readiness, proactively assessing how the AI solution represents a substantial clinical improvement and addresses an unmet need.
