NeurAxis Q2: Two Payers, One Number, and the Unit Economics

NRXS | Multibagger | Added 8/17/24 @ $3.18

NeurAxis reported its second quarter this morning: revenue of $1.928 million, up 116% year over year, the eighth consecutive quarter of double-digit growth and the largest quarter in company history. Gross margin reached 85.9%, up 230 basis points. Cash stood at $8.3 million with quarterly burn of roughly $1.0 million, improved from $1.5 million a year ago, and the at-the-market facility has been untouched since May.

Good numbers. But the quarter's real content was in three things the call clarified: how close the next coverage wins are, why coverage converts to revenue the way it does, and what the procedure actually pays the people who perform it. Taken together, they describe a business the market is still mispricing.

The Two Payers

Carrico was as direct as he gets about the pipeline. NeurAxis made significant gains in the quarter with two large commercial payers — two of the largest remaining holdouts without existing medical policy coverage — and management is cautiously optimistic that coverage lands in the second half of 2026 or early 2027.

When Craig-Hallum's analyst pressed on the source of his confidence, the answer was specific:

“My confidence comes from the fact that we've had direct conversations. I wouldn't be confident if we didn't have direct conversations with a payer who made comments or alluded to the fact that they believe this should — they also believe this should be a covered service. And these are direct firsthand conversations.”

Later in the call, answering our question on profitability, he went a step further: “I know dates and I understand when things are happening, but I'm not going to talk about that today.”

He knows dates. That is not language a CEO uses about discussions in an exploratory phase. This management team has a consistent pattern of saying less than it knows and delivering against what it implied — the December Anthem-scale win followed the same cadence of understated telegraphing. We take the two-payer signal seriously.

Carrico also reframed what the holdup has been all along: “The challenge has often been access to the right decision-makers rather than fundamental opposition to the therapy.” Once NeurAxis gets in front of the right medical director, the clinical case — the only FDA-cleared therapy in the category, guideline inclusion, a Category I code, an alternative to off-label drugs carrying black box warnings — wins the argument. The bottleneck has been getting the meeting. They now have the meetings.

The remaining large payers run a minimum of roughly 20 million lives each. Two of them landing means 40 million-plus new covered lives against the current 100 million-plus base. When fellow MCO contributor Lindsay Leeds asked whether 200 million total covered lives by the end of 2027 was a reasonable expectation, Carrico said he would be “highly disappointed” if they weren't over that mark — “and then some.” That is the closest thing to guidance this company gives.

The Seventy Percent Number

The most valuable single disclosure on the call came in response to Lindsay's follow-up question: what share of its patient population does a hospital need covered before it commits to a full IB-Stim program?

Carrico's answer was one word: 70%.

Sit with that number, because it explains the entire shape of this business — including the parts that have frustrated shareholders. Coverage does not convert to revenue proportionally. It converts as a threshold. A hospital at 60% coverage behaves almost identically to a hospital at 20%: a champion physician treats some patients, but the institution does not build a program, does not allocate dedicated clinic time, does not work the referral pathways. Cross 70% and the switch flips.

This is why a hundred million covered lives produces two-million-dollar quarters. It is not that coverage doesn't matter — it is that most hospitals are still sitting under the line, and partial coverage buys almost nothing.

And the quarter's own KPIs prove the mechanism. NeurAxis introduced a formal KPI framework this quarter — a disclosure upgrade worth applauding in its own right — and the numbers, year-to-date against 2025:

• Revenue: $3.5M in 2026 versus $1.8M in 2025 — up 98%.

• IB-Stim ASP: $1,003 in 2026 versus $772 in 2025 — up 30%.

• Internal prior auth approval rate: 31% in 2026 versus 12% in 2025 — up 19 percentage points.

• Ordering accounts: 88 in 2026 versus 76 in 2025 — up 16%.

• Revenue per ordering account: $40,000 in 2026 versus $24,000 in 2025 — up 68%.

Look at where the growth came from. Revenue nearly doubled while the account base grew 16%. The quarter was not driven by new hospitals signing on — it was driven by existing accounts going deeper, from $24,000 to $40,000 each. That is exactly what a threshold model predicts: the hospitals already above the line expand, the hospitals below it stay quiet, and the count of participating institutions barely moves until new coverage pushes more of them across.

Which is what makes the two payers the whole ballgame. Each one doesn't add revenue in proportion to its lives. It flips some number of hospitals from below 70% to above it, and those hospitals don't add incrementally — they activate. Carrico's image was a bucket with holes: every policy fills holes, and the bucket fills faster as more get filled. Convexity, not linearity.

The prior authorization rate is the metric we will track hardest from here. One in three submissions now clears, versus one in eight a year ago — and the two-thirds still denied are not absent demand. Each denial is a patient whose physician wanted to treat. That is demand sitting in a queue, waiting for the same coverage the two payers would bring.

The Unit Economics

A thoughtful critique of NeurAxis has circulated recently, and it deserves engagement because it is the best-constructed bear case anyone has made: CPT 64567 carries 1.50 work RVUs, and at an assumed 45 minutes per placement, that yields roughly 2.0 wRVUs per physician hour — against roughly 10.0 for a diagnostic colonoscopy. On that math, no rational specialist gives the procedure schedule time, coverage is a smokescreen, and the real constraint is physician incentives. Show me the incentives, show me the outcome.

We put the argument to Carrico directly. His response:

“IB-Stim placements are taking 20 minutes across the board according to consistent market feedback. Additionally, the physicians are often billing a level 2, 3, or 4 with the placement depending on education. We do not get involved in these conversations because we are only involved in the 64567 CPT code.”

Two corrections to the bear math, and they compound.

First, the denominator. The critique's calculation is a fraction with procedure time on the bottom. At 20 minutes rather than 45, 2.0 wRVUs per hour becomes roughly 4.5 before anything else is counted. The headline number isn't slightly off; it's off by better than half. And we have our own primary data point here — a family member of this publication completed a third placement this month. These are quick procedures. They are not 45-minute events.

Second, the model. The critique treats IB-Stim as a standalone transaction competing for a slot on its own merits. It isn't delivered that way. The placement happens inside a clinical encounter the patient was having anyway — evaluation, education, care planning within a pediatric GI program. That is why physicians legitimately bill an office visit alongside it: the evaluation genuinely occurred. The right question is not “does this beat other procedures per hour” but “does adding this to an existing encounter generate incremental revenue at near-zero incremental cost.” It obviously does. A 1.50 wRVU procedure at roughly $1,000 to $1,200 non-facility reimbursement, taking 20 minutes, billed four times per treatment course, embedded in encounters that carry their own evaluation value — that is not a procedure physicians tolerate. It is one hospitals should be building programs around.

And the field evidence says they are. The single most telling line of the call:

“The patients in some of our best accounts are waiting for multiple months for care due to capacity issues, which should not be the case.”

Waitlists, not empty slots. If the economics were unattractive, the symptom would be unused clinic time and programs that never launch. The actual constraint in the best accounts is that demand exceeds allocated capacity. That is the observable signature of a procedure clinicians want to do more of, gated by scheduling infrastructure — not by willingness.

NeurAxis's hiring plan confirms the diagnosis. A VP of Healthcare Economics and Policy for upstream payer work. The existing VP of Market Access shifting downstream to program and procedure economics. A Director or VP of Provider Economics whose explicit mandate is presenting the economic case to hospital administrators “to gain exponentially more IB-Stim clinic time.” Three roles aimed at one problem — and the problem is not persuading doctors the procedure is worthwhile. It is persuading administrators to give it more room.

Two honest caveats belong in the file. Same-day E/M billing alongside a procedure requires a separately identifiable evaluation and is among the more audited patterns in medicine; the first visit of a course is clearly defensible, the fourth is a harder case, and the durability of that revenue layer is worth monitoring. And if real-world placement time runs well under the surveyed time behind the current RVU valuation, the code is generous relative to work — which is a gift today and a revaluation risk in some future review cycle. That is a 2028-or-later concern. It is not a 2026 one.

The VA: The Channel With No Payer Problem

Everything above depends on commercial coverage. The VA does not — the Federal Supply Schedule is the payer — and this quarter the plan got concrete enough to model.

Ten 1099 territory representatives, each covering one to three VA hospitals. Offers went out the last week of July; the team is expected trained and selling by September 15, led by a VP of Market Development who spent his career at Zimmer. Notably, from just three small territories worked part-time by reps simultaneously covering children's hospitals, multiple VA facilities are already ordering and reordering.

Recall the framework from our May analysis: roughly 7 million annual VA patients, a 3% functional dyspepsia rate, giving an addressable population of about 210,000 veterans at roughly $4,000 per four-device course. Typical VA launches penetrate 0.1% to 0.5% of addressable population in year one and 1% to 5% in years two and three. The ten-rep structure lets us translate those bands into workload:

• 0.25% penetration: 525 annual courses; $2.1M annual VA revenue; 53 courses and $210,000 revenue per rep.

• 0.50% penetration: 1,050 annual courses; $4.2M annual VA revenue; 105 courses and $420,000 revenue per rep.

• 1.00% penetration: 2,100 annual courses; $8.4M annual VA revenue; 210 courses and $840,000 revenue per rep.

• 2.00% penetration: 4,200 annual courses; $16.8M annual VA revenue; 420 courses and $1.68M revenue per rep.

At roughly 20 covered facilities, 0.5% penetration requires about one patient course per facility per week — a single half-day clinic slot. Two percent requires four per week. The year-two band that typical VA launches occupy is reachable with clinic commitments measured in sessions, not departments.

The cost structure makes this close to free optionality. The reps are commission-only: zero fixed cost if they sell nothing, and at typical device commission rates against an 85%+ gross margin product, roughly 55 to 65 cents of every VA revenue dollar falls to operating income with no committed expense. For a company deliberately increasing burn elsewhere, a variable-cost second growth engine is exactly the right structure.

Timing expectations: reps selling from September 15 contribute two weeks to Q3. Q4 is the first full period and will be a ramp quarter — anything above $300,000 validates the trajectory against the electroCore Quell comp, which did $114,000 in its launch quarter. The number that matters is the Q4 exit rate, because it sets the 2027 base. Carrico said internal per-territory expectations exist but declined to disclose them, committing to break out VA revenue beginning next quarter or with the Q4 report. Federal channel revenue is structurally lumpy — budget cycles, contract administration, facility-level purchasing — so a soft quarter inside a good trend should not be read as a broken thesis. Our table is our framework, not company guidance.

The Spending Decision

CFO Tim Henrichs said plainly that burn rises in the second half. After two years of unusual capital discipline, that is a deliberate reversal — made because two quarters under the CPT code now show exactly where spending converts.

We pressed on operating leverage. The year-to-date answer is genuine: revenue up 98% against much slower opex growth. But Q2 opex grew 53% against 116% revenue growth, and both executives confirmed spend accelerates from here. Translation: leverage is real but deferred, and cash flow breakeven is a 2027 conversation contingent on the payer wins. The going concern qualification stays until breakeven; Henrichs was straightforward about that, and equally straightforward that it plays no role in payer decisions.

Carrico's answer on profitability was the most revealing moment of the call:

“Do I want to be profitable? Of course, yesterday. But I don't want to be profitable more than I want to drive revenue.”

In a threshold-function market where being first into a newly unlocked hospital compounds, that is the correct priority — provided the payers land.

The Final Word

Here is the quarter in one paragraph. Two of the largest remaining payers are in direct, dated discussions, and each one that signs flips a cohort of hospitals across the 70% line where programs activate rather than accrete. The hospitals already above that line grew revenue per account 68% this year, which is what activation looks like. The procedure driving it pays physicians well — 20 minutes, program-embedded, billed four times per course — well enough that the binding constraint in the best accounts is a months-long waitlist. And a second channel with no commercial payer dependency turns on September 15 with zero fixed cost attached.

Carrico said on the call, of a record quarter, that they are basically treating no one. He meant it as a statement of how much is left. Eighty-eight ordering accounts against every children's hospital in America. Two-thirds of prior authorizations still denied. A hundred million covered lives producing a fraction of what two hundred million unlocks.

We are up over 100% from our $3.18 entry, and the market's reaction to this print — indifference — is the same reaction it had at every prior stage of a thesis that has since doubled. The gap between what this quarter documented and what the price reflects is as wide as it has been since we entered.

Holding.

---

Disclosure: Microcap Opportunities holds a position in NRXS. This is not investment advice. Positions may change without notice. Do your own work.