Medicare’s AI RPM Shift: Winners, Losers, and Your Investment Strategy

The digital health gold rush in AI-driven remote patient monitoring (RPM) is over. For venture capitalists and digital health growth equity investors, the era of broad market enthusiasm has slammed shut, replaced by a brutal focus on clinical utility and hard evidence. Medicare, the entity that controls the purse strings, is getting much tougher. This is creating a clear split in the market: companies that built their tech on strong, peer-reviewed clinical validation are locking in long-term revenue, while those that got by on marketing sizzle are facing a real threat of extinction.

The End of Easy Reimbursement: CMS Raises the Bar

For years, a flood of RPM solutions got by on a pretty loose reimbursement environment. That’s changed. The Centers for Medicare & Medicaid Services (CMS) has been steadily tightening the screws on its Physician Fee Schedule, especially for remote physiological monitoring (RPM) and remote therapeutic monitoring (RTM). This isn’t just bureaucratic paper-shuffling, it’s a hard pivot to value-based care, where you only get paid if you can prove your tech helps patients. The latest CMS Physician Fee Schedule final rules make this crystal clear. The CY 2026 Medicare Physician Fee Schedule (MPFS) final rule, which kicks in on January 1, 2026, brought major updates. The main Category I CPT codes for RPM (like 99453, 99454, and 99457) are still there, but CMS added new codes for shorter monitoring periods, finally acknowledging that clinical value can come from quick, frequent check-ins. For RPM, we now have CPT code 99445 for device supply over just 2-15 days and 99470 for 10-19 minutes of treatment management. RTM got similar codes, like 98985 for device supply and 98979 for the first 10 minutes of management. These changes effectively kill the old “16-day hurdle” and the “20-minute care management requirement.” But with this new flexibility comes a much heavier burden of proof. The focus on documentation, data interpretation, and active clinical intervention is intense now. Just collecting data gets you nothing. Your AI has to produce insights that a clinician can actually use to manage a patient better. The American Medical Association (AMA) and its CPT Editorial Panel are also a huge part of this, demanding a mountain of clinical evidence before they’ll even consider granting new Category I codes. Payers are tired of the hype, and as the market matures, they want proof. For investors, this means you have to scrutinize a company’s clinical evidence portfolio as intensely as you do its P&L statement.

Clinical Validation: The Ultimate Moat for Cardiac AI

In this new world, the companies that invested early in serious clinical utility trials are the ones pulling away from the pack. Just look at the difference between a player like iRhythm Technologies and a company like Biofourmis. They both use algorithms for remote monitoring, but their very different investments in clinical evidence create a huge gap in their long-term reimbursement security. iRhythm’s Zio XT patch is a perfect case study in how to build a “data moat” backed by unassailable clinical validation. Their AI cardiac monitoring platform has processed millions of labeled ECGs, creating a dataset that’s almost impossible for a new company to match. But here’s the key: iRhythm has relentlessly funded and published peer-reviewed clinical trials proving its platform’s value iRhythm clinical trial data. The AMALFI randomized clinical trial results from August 2025 showed their Zio LTCM service was better at diagnosing atrial fibrillation, and the AVALON real-world evidence study, also published in August 2025, confirmed it. Their clinical library is now over 135 original scientific manuscripts deep. Because they invested in evidence, they’re in a fantastic position for continued reimbursement and market dominance. Publishing research that shows superior diagnostic yield is exactly what CMS wants to see. On the other side, you have companies that prioritized a fast launch over building a foundation of clinical trials. They often get by with a slick pitch deck and claims of AI magic but have no peer-reviewed data to back it up, and they’re now looking at a revenue cliff. An FDA 510(k) clearance just means a device is similar to something else on the market. It says nothing about whether it actually improves outcomes or saves money in the real world, which is what payers now demand. For AI-driven SaMD, particularly in a high-stakes field like cardiac diagnostics, reimbursement now requires much more than just a regulatory nod. So what should investors be asking? Where is the real-world evidence (RWE) that shows fewer hospitalizations or lower healthcare costs? If a company can’t answer that, its path to scalable revenue is a fantasy.

Why Clinical Evidence is the Ultimate Moat in Regulated Healthcare Markets

For VCs and digital health growth equity investors, the bottom line is simple: clinical evidence isn’t a “nice-to-have” for AI-driven RPM. It’s the only competitive moat that will last. In a regulated market like healthcare where Medicare calls the shots, a deep portfolio of clinical evidence is what creates financial stability and defensibility. The companies that are building something durable are the ones that followed GMLP (Good Machine Learning Practice) from day one, making sure their AI models were built, tested, and watched for drift with scientific rigor. Having a QMS, an ISO 13485 certification, and solid HIPAA / HITRUST / SOC 2 compliance are just the ticket to entry. The real test is proving your AI solution actually helps patients and is cost-effective. That means investors need to dig into a company’s spending on prospective clinical trials, RWE generation, and health economic outcomes research. Are they just hoarding data, or are they producing real insights that doctors will trust and payers will fund? There’s a big difference between a simple Clinical Decision Support tool and a true Diagnostic AI. The latter has a much higher regulatory and evidence bar to clear, but once you do, your position is far more protected. Medicare’s new standards are cleaning house, getting rid of the “zombie companies” that raised a seed round and got an FDA clearance but don’t have the clinical chops to land big enterprise contracts or secure reimbursement. As an investor, you have to find the companies with a clear shot at Category I CPT codes that is supported by a pile of peer-reviewed studies. That is a non-negotiable part of due diligence now. Aligning with what payers demand is what will get you the exit multiples and create long-term value in this space.

Methodology and Source Note

We put this analysis together by digging into recent CMS Physician Fee Schedule final rules and comparing the public clinical evidence portfolios of key RPM companies. The strategic advice for investors comes from our ongoing work tracking the economics of health AI. We’ve referenced the Centers for Medicare & Medicaid Services (CMS), the American Medical Association (AMA), Biofourmis, and iRhythm Technologies. All facts about regulations and company data are based on publicly available information, including government filings and peer-reviewed scientific papers like those detailing Biofourmis clinical trial publications. The goal here is to offer a clear-eyed analysis of the market and reimbursement realities to help guide investment decisions.