Video Preview
🎧 Part I Podcast free on Apple Podcasts and Spotify.
🎧 Part II Podcast episode for paid subscribers only. Also available on Apple Podcasts and Spotify.
To listen to paid episodes in Apple or Spotify, link your Substack subscription via the show settings on those platforms (instructions inside the Substack app under Subscriptions → Podcast).
Table of Contents
The problem all three pathways address
MCIT, the one that died
TCET, the one that exists, barely
RAPID, the newest lane
The side by side
What this means for device companies and investors
FAQ
Abstract
FDA authorization does not mean Medicare pays. The gap between market authorization and durable Medicare coverage routinely runs years, and for breakthrough devices that gap is the single biggest killer of commercial models.
MCIT (finalized January 2021, repealed November 2021) promised four years of automatic national coverage the day FDA authorized a breakthrough device. It never took effect.
TCET (final notice August 2024) is the current live pathway: capped at roughly five devices a year, nomination roughly 18–24 months before your anticipated FDA decision, coverage with evidence development on the back end. Exactly one device has publicly gone through the front door.
RAPID (announced April 2026, proposed procedural notice published August 2026) is the new joint CMS and FDA play: a proposed NCD posted the same day FDA authorizes, final coverage roughly 60 days later for Class II and 90 for Class III. Eligibility is narrow, IVDs are out, and TCET intake is paused while this stands up.
Comment period on the RAPID notice closes October 13, 2026. Nothing about coding or payment is addressed, and CMS has now moved from proposing to **finalizing** repeal of the IPPS and OPPS alternative breakthrough pathways on the payment side, starting with FY 2028 applications. Coverage is speeding up while payment is tightening. Plan accordingly.
The problem all three pathways address
Start with the number that makes device CFOs drink. FDA has handed out more than 1,284 breakthrough device designations as of March 31, 2026, and the designation comes with priority review, sprint discussions, extra FDA attention, the works. What it does not come with is a dollar. Medicare, which covers the population most breakthrough devices are actually built for, treats FDA authorization as the starting gun for a completely separate multi year process, not the finish line. A company can clear a PMA, pop champagne, issue the press release, and then discover that the largest payer in the country has no obligation to reimburse a single unit.
The mechanics of the gap are worth spelling out because the three pathways in this piece are all attempts to compress different parts of it. Coverage, coding, and payment are three separate machines. Coverage means Medicare agrees the thing is reasonable and necessary. That happens either nationally through a National Coverage Determination, which historically takes nine to twelve months once CMS actually opens one and CMS opens very few, or locally through Medicare Administrative Contractors making their own calls across a patchwork of jurisdictions. Most devices never get an NCD at all. They live in MAC purgatory, where coverage varies by region, claims get adjudicated case by case, and a sales team spends two years explaining to hospitals why the same procedure pays in Florida and denies in Ohio. Coding runs on its own annual and quarterly cycles through CPT and HCPCS. Payment gets set through the fee schedules and, for new inpatient tech, through mechanisms like the new technology add on payment. Each machine has its own clock, its own staff, and its own indifference to your burn rate.
Stack the clocks and the practical result is that a breakthrough device can sit in commercial limbo for four, five, sometimes seven years after authorization before Medicare coverage looks anything like settled. Venture math does not survive that. A Series C medtech company raising on the promise of Medicare volume cannot tell LPs the reimbursement plan is “wait for a MAC medical director to retire.” So for the better part of a decade, industry, Congress, and two administrations of very different flavors have taken swings at building an express lane. Three of those swings have names: MCIT, TCET, and RAPID. One is dead, one is on life support, and one just got published in the Federal Register with fresh paint and a 60 day comment clock. The comparison matters because diligence decks keep confusing them, and because which lane a company can actually use depends heavily on where it sits in its regulatory lifecycle.
MCIT, the one that died
Medicare Coverage of Innovative Technology was the maximalist version. Born out of a 2019 executive order pushing Medicare to keep pace with innovation, proposed in September 2020, and finalized in the last week of the first Trump administration in January 2021, MCIT said the quiet part loudly: any FDA designated breakthrough device would get automatic national Medicare coverage starting the day of market authorization and running four full years, with eligibility for devices authorized in the prior two years. No nomination process. No evidence plan. No CMS review of whether the pivotal trial enrolled anyone over 65. FDA says yes, Medicare pays, see you in four years.
Device companies loved it for obvious reasons. It converted a coverage question into a regulatory one, and regulatory questions have knowable timelines. The problem, and the reason it never survived the change in administration, is that FDA and CMS answer genuinely different questions. FDA asks whether a device is safe and effective for its intended use, often based on trials that enroll few or zero Medicare beneficiaries. CMS asks whether the item is reasonable and necessary for the Medicare population specifically, a population that skews older, sicker, and more comorbid than the average pivotal trial cohort. MCIT bolted the second answer to the first and threw away the key. The new administration delayed the effective date twice during 2021, took comments, and repealed the rule outright in November 2021 before a single device ever received MCIT coverage. The stated rationale was straightforward: automatic coverage with no Medicare specific evidence requirement, and no practical mechanism for CMS to pull coverage on a device showing safety signals short of full FDA action, was a bridge too far. The rule died a paper death, undefeated and untested, which is why it still gets invoked at conferences like a martyred prophet.
What survived is more interesting than the corpse. The repeal notice conceded the underlying problem was real and promised a replacement that kept the speed while adding evidence guardrails. Every design choice in TCET, and honestly most of RAPID, reads as a direct response to a specific MCIT criticism. Automatic coverage became nominated coverage. No evidence requirements became evidence development plans. Unlimited eligibility became annual caps. MCIT is the ghost at every subsequent rulemaking, and CMS staff will tell you as much off the record with varying levels of eye rolling.
TCET, the one that exists, barely
Transitional Coverage for Emerging Technologies is what the repeal promise eventually produced, first as a notice with comment in June 2023 and then as a final procedural notice effective August 2024. The design is careful to the point of parody. A manufacturer self nominates roughly 18 to 24 months before its anticipated FDA decision date. CMS reviews nominations quarterly. Eligible devices must hold breakthrough designation, fall within a Medicare benefit category, not already be the subject of an existing NCD, and not be otherwise excluded from coverage by statute or regulation. In vitro diagnostics were effectively carved out at the start, with CMS punting nearly all diagnostic coverage back to the MACs on the theory that lab tests are their own specialized universe. That carve out alone removed a huge share of the breakthrough portfolio from contention, since diagnostics and algorithm driven tests are heavily represented in the designation pool.
For devices that make it in, the machinery is real but heavy. CMS and AHRQ work with the manufacturer on an evidence preview, essentially a systematic literature review that surfaces the gaps between the FDA evidence package and what Medicare wants to know. Then comes an evidence development plan that has to be approved before things move. The payoff is a commitment that CMS will aim to finalize an NCD within six months of FDA market authorization, with coverage typically granted under coverage with evidence development, meaning the company keeps generating data during a transition window generally understood to run about five years, after which CMS revisits and decides whether coverage becomes permanent, gets modified, or goes away. Compare that to the historical pattern where an NCD, if you could even get one opened, ate nine to twelve months after years of waiting in line, and TCET looks like genuine progress.
Now the uptake numbers, which are where the comedy lives. CMS said it anticipated accepting up to five TCET candidates per year, a cap it justified on staffing grounds, and which industry immediately pointed out was a rounding error against a designation pool adding well over a hundred new breakthroughs annually. Two years in, the publicly visible throughput is one device. EBR Systems’ WiSE leadless left ventricular pacing system became the first technology to enter the program, and CMS initiated the national coverage determination review in June 2026 with a projected final NCD in early 2027. That is the entire visible pipeline. One cardiac device, moving at what counts as blistering speed for an NCD and glacial speed for anything else in American commerce. Five slots a year was always going to be a lottery, and it turns out most manufacturers looked at the 18–24 month pre nomination requirement, the AHRQ literature review, the CED obligations, and the lottery odds, and quietly decided the MAC by MAC grind was less painful. Which brings the story to April 2026, when CMS essentially agreed with them.
RAPID, the newest lane
The Regulatory Alignment for Predictable and Immediate Device coverage pathway arrived in two acts. Act one was the joint CMS and FDA announcement on April 23, 2026, heavy on ambition and light on mechanics, with CMS leadership talking publicly about delivering Medicare coverage within 60 to 90 days of FDA approval and officials floating that roughly 40 devices currently qualify with maybe 20 more on the bubble. Act two was the proposed procedural notice, CMS-3487-NC, issued in early August 2026 and published in the Federal Register days later, with a 60 day comment period closing on October 13, 2026. The final notice, whenever it lands, is what makes the pathway operational.
Here is the core mechanic, and it is genuinely novel. Instead of starting the coverage conversation after FDA authorization, RAPID moves it before the pivotal trial even starts. CMS and FDA jointly engage with the manufacturer at the IDE pre submission stage, while the pivotal study can still be shaped, so that a single trial design answers both the FDA question and the Medicare question. The study must enroll Medicare beneficiaries, which sounds obvious and has historically been anything but. In exchange, CMS commits to posting a proposed NCD the same day FDA grants market authorization, running a 30 day public comment period, and finalizing national coverage as soon as 60 days after authorization for Class II devices and 90 days for Class III. RAPID NCDs also jump the queue, getting priority over everything else on the CMS wait list, which the non breakthrough world will have feelings about.
Eligibility is where the fine print bites. The pathway targets breakthrough devices in FDA’s Total Product Life Cycle Advisory Program, with presumptive Class II devices heading toward a De Novo required to be in TAP and Class III devices heading toward a PMA eligible regardless of TAP participation. The device has to be at the pre submission stage of its IDE. Devices already enrolled in an active pivotal IDE study are out, though CMS explicitly asked for comment on whether some on ramp should exist for them, so expect the comment docket to be full of manufacturers arguing their trial is special. IVDs are excluded outright, continuing the TCET tradition of telling diagnostics companies to go talk to their MAC. Software as a medical device is not categorically excluded, which is one of the more quietly consequential lines in the notice given where the breakthrough pipeline is heading. And in the same breath, CMS proposed pausing TCET intake for new candidates as of publication, while devices already inside TCET continue and the separate parallel review program survives. The polite reading is consolidation. The accurate reading is that CMS looked at a pathway with one enrollee and decided not to keep staffing the front desk.
Two things the notice conspicuously does not do. First, it says nothing about coding or payment, so a RAPID device could hold a shiny final NCD 60 days after authorization and still be waiting on a HCPCS code and a payment rate, which is like being handed a key to a car with no engine. Second, and running in the opposite direction, CMS has now moved beyond proposal and in the FY 2027 IPPS and OPPS rules is eliminating the alternative NTAP and device pass through pathways that let breakthrough devices collect add on payments without demonstrating substantial clinical improvement, beginning with FY 2028 applications. So the same agency is accelerating coverage while raising the bar on payment. Anyone modeling RAPID economics without modeling the NTAP change is doing half a diligence.
The side by side
Since a proper table would violate the formatting religion of this publication, here is the comparison in prose, structured the way a diligence memo would ask the questions. On eligibility, MCIT was the open bar: any breakthrough device, full stop, retroactive two years. TCET narrowed to breakthrough devices in a benefit category with no existing NCD, minus IVDs, and capped at about five accepted candidates annually. RAPID narrows differently: breakthrough devices caught at the IDE pre submission stage, with presumptive Class II devices required to be in TAP and heading for a De Novo and Class III devices heading for a PMA eligible with or without TAP, IVDs excluded, but with working capacity estimated at 40 to 60 devices, roughly a tenfold expansion over TCET’s cap. On the evidence ask, MCIT demanded nothing beyond FDA authorization, which is why it is dead. TCET demands an evidence preview, an approved evidence development plan, and ongoing CED data generation. RAPID demands trial design alignment up front, including Medicare beneficiary enrollment in the pivotal, effectively front loading the evidence work into the study a company was going to run anyway.
On timing, MCIT promised coverage on day zero of authorization, lasting four years. TCET aims for a final NCD within six months of authorization, with transitional CED coverage running around five years before a permanence decision. RAPID promises a proposed NCD on day zero and a final one at day 60 or 90 depending on device class, though the honest asterisk is that the clock only starts after a company has spent years inside TAP (for Class II) and its IDE study, so the first same day NCD under this program plausibly lands closer to the end of the decade than anyone’s pitch deck admits. On current status as of late August 2026: MCIT is repealed and never operated. TCET is final, operational, paused for new intake, and carrying exactly one public passenger toward a projected early 2027 NCD. RAPID is a proposed notice with comments due October 13, 2026 and no effective date until the final notice publishes. On what each does for payment: nothing, nothing, and nothing, respectively, which is the most consistent design feature across all three.
What this means for device companies and investors
For operators, the strategic question is brutally stage dependent. A company that already holds market authorization gets essentially nothing from RAPID as proposed, and with TCET intake paused, the realistic play remains the old one: MAC engagement, local coverage, registry data, and lobbying for an NCD the traditional way. A company mid pivotal is in the awkward middle, currently ineligible for RAPID and invited to say so loudly in the comment docket before October 13. The real winners are earlier stage: a breakthrough designated company that has not yet filed its IDE pre submission can now architect one trial to serve two masters, which changes protocol design decisions being made this quarter. Enroll Medicare beneficiaries in the pivotal, get into TAP if you are a presumptive Class II De Novo, engage both agencies early, and the reward is a coverage timeline that compresses from years to weeks post authorization. That is not a marginal improvement. That is a different company.
For investors, the underwriting implications are worth being precise about. In a diligence model, the pathway a target can actually access should now be a checklist item with four boxes: breakthrough designation held, TAP participation status (for Class II), IDE stage, and whether the pivotal enrolls Medicare beneficiaries. A pre IDE breakthrough company that checks all four can credibly model national Medicare coverage inside a quarter of authorization, which pulls revenue inflection forward and materially changes the amount of capital needed to bridge from approval to commercial scale. That is the bull case, and it is real. The bear case deserves equal airtime. This is a proposed notice, not a final one, and procedural notices have died before, MCIT being the canonical example of a finalized rule that still never operated. The pathway addresses coverage only, so the coding and payment clocks still run at their old speeds, and the now finalized NTAP and pass through alternative pathway repeal means the inpatient payment sweetener that breakthrough devices enjoyed is going away for FY 2028 applications and beyond, with limited grandfathering. Capacity claims of 40 to 60 devices assume CMS staffing that has not been demonstrated at anything close to that scale, given the same coverage shop that could only process five TCET candidates a year now proposes to run ten times the volume with same day deliverables. And history offers a sobering comp: parallel review, the last big structural alignment between FDA and CMS, existed for over a decade and moved a tiny handful of products, because alignment on paper does not equal throughput in practice.
The portfolio level read: expect a premium on breakthrough designated, pre pivotal assets with Medicare heavy indications, expect diagnostics companies to stay exactly as stuck as they were, and expect at least one 2027 vintage medtech deal to blow up because someone modeled RAPID timelines for a device that was already three years into its IDE. Also watch the comment docket. The fights over active IDE eligibility, SaMD treatment, and coding linkage will tell you what the final notice looks like before CMS does.
FAQ
Is MCIT still active? No. It was finalized in January 2021, delayed twice, and repealed in November 2021 without ever taking effect. No device ever received coverage under it. If a pitch deck references MCIT coverage as a go to market plan, that deck was written by someone who stopped reading the Federal Register five years ago.
How long does TCET coverage last? Coverage under TCET is granted through coverage with evidence development, with the transitional period tied to the device’s approved evidence development plan, generally understood to run about five years. After that, CMS reviews the accumulated evidence and decides whether coverage becomes permanent, gets modified, or ends. Note that new TCET intake is paused under the RAPID proposed notice, though devices already in the program continue.
What replaced MCIT? Formally, TCET replaced it in 2024 with evidence requirements and a five per year cap. Practically, RAPID is the spiritual successor, restoring the day zero coverage ambition while fixing the original sin by moving evidence alignment before the pivotal trial instead of skipping it entirely.
Does RAPID cover diagnostics? No. IVDs are excluded from the proposed pathway, as they effectively were from TCET, with coverage decisions remaining with the MACs. Software as a medical device is not categorically excluded, which matters more each year.
Does RAPID solve payment? Not even a little. The proposed notice addresses coverage only, coding and payment run on their existing cycles, and CMS has now finalized the repeal of the NTAP and OPPS alternative pathways for breakthrough devices beginning with FY 2028 applications. Fast coverage plus slow coding plus tighter add on payments is the actual 2027–2028 operating environment. Model all three
.