Zyrixe HealthCoverage & Gap IntelligenceReclaim 10%–30% in lost research revenue from your CTMS.
In 5 business days. Free. We cite NCD 310.1 and the exact CTA clause on every line — so your team can actually bill without FCA exposure.
In January 2024, Moffitt paid $19.56M to settle a False Claims Act case. They'd billed federal programs for items that should have gone to the trial sponsors. The OIG cited NCD 310.1 by name.
Most sites read that and pull in the opposite direction — they quietly under-bill the sponsor to stay safe. So they trade a $19M exposure for a 10–30% annual revenue leak. Different problem, same root cause: nobody can see the whole picture clearly enough to act.
Sponsors pay what they're invoiced.
Most of the leakage at sites isn't sponsors paying slow. It's research-only procedures that got done, got checked off in the CTMS, and then never made it onto an invoice — because nobody could tell if Medicare was already being billed for them.
Under Medicare NCD 310.1, routine costs in a Qualifying Clinical Trial go to the payer with a Q1 modifier, ICD-10 Z00.6, and the 8-digit NCT number per MCPM Ch. 32 §69. Research-only costs go to the sponsor per the CTA budget with a Q0 modifier. Billing the same procedure to both is a False Claims Act event — treble damages, per-claim penalties, and personal liability for the person who certifies the claim.
- 01The CA gets written once, at protocol activation — then it drifts. Amendments come in. Lab orders change. Nobody re-runs the analysis.
- 02When a procedure looks ambiguous — SoC or research-only? — your staff can't quickly check whether the practice RCM side already billed it. So it gets dropped from the sponsor invoice. Duplicate-claim risk beats revenue every time.
- 03Visit completion lives in your CTMS. Sponsor invoicing lives in a different system. Epic/Beacon lives in a third. Nothing stitches the three together at the procedure level.
- 04Your Director of Clinical Research already knows the gap is there. What they don't have is a paper trail clean enough to bill against without inviting the FCA back into the conversation on every line.
Most community oncology sites we've looked at are leaking 10–30% of their sponsor-billable revenue per active trial. Procedures that could've been invoiced inside the CTA's 30-day window quietly expire after quarter close. One oncology infusion can be $5K–$15K.
Sponsor-payment platforms tell you what the sponsor paid you.
We tell you what they should have paid you — and whether the legal basis to bill is clean enough to act on.
Unified CTA-to-pay platform. Just launched AI invoice-to-budget matching (April 2026). Built for the sponsor or CRO — not for you. Tells them if the payment they sent matches the CTA.
Launched Sept 2025, GA this quarter. Reads CTAs, pulls out budget terms, flags anomalies on the invoices the sponsor generates. A sponsor tool — not yours.
~24 people, Lake Worth FL, 11 years old. Proves the demand — sites pay humans to do this today. Constrained by headcount: slow, expensive at scale, and they can't surface amendment drift without re-scanning every protocol by hand.
Built for you — the independent community oncology site. Our Coverage Analysis cites NCD 310.1 and the specific CTA clause on every line. Our Gap Detection reconciles your CTMS activity against the invoices you sent out — not the money coming back in.
Two capabilities, one workflow.
Coverage analysis that survives amendments. Reconciliation that ties CTMS activity to sponsor invoices at the procedure level.
Coverage Analysis, line-by-line
We look at every CPT/HCPCS on your schedule-of-events and sort it into Q1 (routine → payer, with Q1 + Z00.6 + NCT) or Q0 (investigational → sponsor, per CTA). Every line ships with:
- ▸NCD 310.1 basis — QCT criteria + routine-cost rule mapping
- ▸CTA clause citation — the exact budget/obligation paragraph
- ▸Protocol section anchor — SoE row + amendment version
- ▸Claim-modifier suggestion — no more spreadsheet wars with your RCM team over what goes to Medicare vs. the sponsor (Q0 / Q1 + ICD-10 Z00.6).
- ▸Calibrated confidence score + reason code — so your reviewer knows which rows need a human and which don't (no probabilistic theatre).
- ▸Amendment drift, solved — protocol amendment comes in Friday? The CA re-runs itself and you see what changed in one screen.
| Visit | Procedure (CPT) | Classification | Modifier | NCD / CTA | Conf. |
|---|---|---|---|---|---|
| C1D1 | CT chest w/ contrast 71260 | Routine (SoC) | Q1 + Z00.6 | NCD: QCT met · CTA §4.2(a) | 0.94 |
| C1D1 | PK blood draw 36415 ×6 | Research-only | Q0 · bill sponsor | NCD: non-routine · CTA §4.3(c) | 0.99 |
| C2D1 | ECG 12-lead 93000 | Research-only | Q0 · bill sponsor | Protocol-mandated freq. · CTA §4.3(f) | 0.88 |
| C3D1 | Tumor biopsy 38505 | Research-only | Q0 · bill sponsor | Exploratory · CTA §4.3(h) | 0.96 |
This is an example. Your actual CA runs against your protocol and CTA, and gets reviewed by a CCRP- or ACRP-CP–credentialed biller before anyone acts on it.
Sponsor-Invoice Gap Detection
We match every completed visit in your CTMS against the invoices your team actually sent the sponsor. Any research-only (Q0) line that got done but never got billed shows up as a gap — priced at the CTA rate, with a recovery path and the full evidence chain attached.
- ▸Visit-level match: CTMS completion ↔ sponsor-invoice line ↔ CTA budget rate
- ▸Windowed flags — every gap you can still recover before the CTA's 30-day invoicing window closes, ranked by dollar value.
- ▸Invoiceables vs. pass-through vs. fixed-fee per-visit awareness
- ▸Sponsor-portal exports (Medidata Rave Site Payments, IQVIA CTFS, Greenphire eClinicalGPS)
- ▸Aging view — unbilled research AR by trial, sponsor, CRO, holdback status
- ▸Approval queue — Director of Research signs off, nothing auto-sends
See exactly what you'll get back.
Every free review comes back as a written gap report. Below is a redacted sample from a mock 3-trial review.
| Trial | Visit Type | CTA Clause | Gap Flag | Est. Value |
|---|---|---|---|---|
| ██████-301 | C1D1 · PK Draw | §4.3(c) | Not invoiced | $██,███ |
| ██████-5 | Screening · ECG | §4.3(f) | Not invoiced | $█,███ |
| ██████-2 | C2D15 · Biopsy | §4.3(h) | Window closing | $█,███ |
Human-in-the-loop by design.
Every gap gets eyes from a CCRP- or ACRP-CP–certified research biller before it turns into an invoice recommendation. The model never submits anything. Your team always approves.
- 01
Send us what you've got
Your protocol, CTA, existing CA, schedule-of-events, and CTMS/ERP exports. No clinical-system integration needed for the review — CSVs and PDFs are fine. De-identified data only; no PHI required.
- 02
We classify
We run every procedure against NCD 310.1 and the CTA budget, and suggest the Q0/Q1 modifier. Anything the model isn't sure about goes to a human reviewer with the ambiguity spelled out ("frequency exceeds SoC," "CTA silent; protocol-mandated," and so on).
- 03
We reconcile
We match your completed visits against the sponsor invoices you've already sent. Gaps get priced at the CTA rate and ranked by recoverable dollars and days-to-deadline.
- 04
A human signs off
A CCRP- or ACRP-CP–credentialed biller on our side reviews every flagged gap. Your Director of Research approves. Every action carries a full audit trail — reviewer name, reason code, timestamp — ready to drop straight into the sponsor's portal.
Under the hood.
Expanded detail for finance and compliance readers. Skip unless you're evaluating for adoption.
How Coverage Analysis is generated
Step 1: Protocol + CTA + schedule-of-events ingested and normalized. Each procedure-visit pair becomes a row.
Step 2: Each row is classified under the three mandatory QCT criteria (benefit-category, therapeutic-intent, eligible-population) with the "deemed" pathway checked (federally funded, IND, IND-exempt per 21 CFR 312.2(b)(1)).
Step 3: Rule-based and LLM-based reasoners independently produce a classification; disagreement is escalated. Each output carries a calibrated confidence score and a reason code keyed to the specific CTA clause and NCD section.
Step 4: A CCRP / ACRP-CP–credentialed human reviews low-confidence rows (configurable threshold, default 0.85) before any downstream action is taken.
How Gap Detection reconciles CTMS to sponsor invoices
Source set: CTMS visit-completion log, CTA budget (with amendments), sponsor invoice history, payment remittance file, holdback ledger.
Match logic: each completed visit is resolved to a CTA line (invoiceable, pass-through, or fixed-fee per-visit). Each CTA line is then matched to an outbound invoice; unmatched research-only lines become candidate gaps.
Scoring: gaps are priced at the CTA rate, aged against the CTA invoicing window, and ranked by recoverable value × time-to-deadline. Evidence chain (CTMS record + CTA clause + invoice absence) is attached to every gap before it reaches a reviewer.
How we handle edge cases (the things that cause the leakage)
Frequency exceeds SoC. Routine-type procedure ordered at a higher cadence than clinical practice would justify — the excess is research-only (Q0).
CTA silent, protocol-mandated. Procedure required by the protocol but not line-itemized in the CTA budget. Flagged for CTA amendment rather than billed to either payer.
Amendment drift. Procedure added via protocol amendment after CA finalization. Coverage re-run on amendment import; delta shown in the reviewer queue.
Possible double-bill. Procedure with a potential trace in practice RCM (Epic/Beacon, etc.). Flagged hold — nothing leaves the queue until the operator confirms the prior claim status.
Reads your stack. Doesn't fight it.
CTMS, eReg, sponsor portals, CTA. Flat-file ingest for the review; API connectors on the contract.
Works with the CTMS you already have.
We sit underneath your CTMS billing module and tell it what it missed. No data migration. No workflow change for your coordinators. No new login for your Director of Research. Flat-file ingest for the pilot, API connectors on the contract.
Estimate your unbilled research AR.
Four inputs, a rough range out. The free review swaps these guesses for actual line-level numbers from your site.
The industry leakage range is 10–50%. We use a conservative 15–30% here. Recovery assumes about half the identified gaps are still actionable inside their CTA invoicing windows. Your free review replaces all of this with real numbers from your CTMS and CTA.
Built for a regulated buyer.
We're pre-revenue, not pre-posture. This is the compliance floor your research finance lead will check for on the first call — we set it up before we needed to.
Every flagged gap passes through a CCRP (SOCRA) or ACRP-CP credentialed research biller before it turns into an invoice recommendation. Reviewer name and timestamp get written into the audit trail on every action — not buried, not optional.
Stress-tested with the people who live inside the problem.
We're pre-revenue and we say so. What we're not doing is building this in a vacuum. In the last 60 days we've spent hours on calls with cancer-center administrators, CTMS operators, and oncology data leaders who've watched research revenue slip through exactly the gaps this product closes. Every feature in the sections above has a name attached to it — someone told us it mattered before we built it.
We're not naming them publicly until they're comfortable being named publicly. If you want references, ask on the review call and we'll make introductions.
Built by engineers who lived inside this problem.
Zyrixe Health is built by engineers who came out of Triomics, the clinical trial AI company. We spent years building AI infrastructure for oncology trial operations — close enough to the site-finance workflow to see exactly where revenue slips through, far enough from the CTMS vendors to say so out loud.
Sites already know the gap exists. What they haven't had is a tool that can cite NCD 310.1 and the exact CTA clause precisely enough to be defensible, with a credentialed human in the approval path before anything gets invoiced. That's what we built.
The questions we keep getting.
Short answers here. We'll go deeper on the review call.
Does the system auto-bill or auto-invoice anything?
No. Nothing gets invoiced without a human hitting approve. A CCRP- or ACRP-CP–credentialed biller on our side reviews every flagged gap before it even reaches your Director of Research. Automating the invoice itself isn't defensible at this stage of the category — the FCA accuracy bar is too high to cut that corner.
How is this different from Ledger Run, IQVIA CTFS, or our CTMS's billing module?
Ledger Run and IQVIA CTFS are sponsor/CRO payment platforms — they automate invoices that already exist. CTMS billing modules assume the CA is already right and just invoice against it. We do the two things none of them touch: re-examine your CA line-by-line against NCD 310.1 with Q0/Q1 classification, and surface the research-only procedures that got done but never made it onto a sponsor invoice.
How does the free 3-trial review work?
Send us protocol + CTA + visit log + CA for up to three active oncology trials. We send back a written gap report: the dollar estimate on unbilled research AR, the line items, the Q0/Q1 mapping, and how to actually recover it. Usually about 5 business days. No contract, no install, no obligation. If the number is small, we'll tell you — and leave you alone.
PHI, BAAs, security?
Coverage Analysis only needs the protocol and CTA text — no PHI at all. Gap Detection runs on CTMS activity logs; for the free review we'd ask you to de-identify the export. A BAA is available for the pilot if de-identification gets in your way. SOC 2 Type II is in progress. We're US-hosted. Everything's encrypted at rest and in transit, and we only keep data as long as we actually need it.
Commercial model?
Post-pilot: a SaaS base sized to number of active trials, plus a modest flat fee per sponsor invoice processed through the platform. Predictable line item, no revenue-share, no attribution fights. Exact numbers discussed after the free review establishes the gap size at your site.
Find your unbilled research AR before your next quarter close.
Send us protocol + CTA + visit log for up to three active oncology trials. We'll send back a written gap report: dollar estimate, line items, Q0/Q1 mapping, and how to recover it. No contract, no install, no obligation. If the number is small, we'll tell you — and leave you alone.