In May 2026, Congressman Dunn of Florida issued a press release about a new Enhance CLIA bill (here), and there was a policy article from the McDonald Hopkins law firm, Campbell et al. (here). Around June 11, a full version of the 45-page legislation has been posted. Find it here:

https://www.congress.gov/119/bills/hr8890/BILLS-119hr8890ih.pdf

Track updates to HR 8890 here.

Download as a 9-page white paper here.

###

SUMMARY (CLAUDE OPUS 4.8)

H.R. 8890, the Enhancing CLIA Act of 2026 (Rep. Neal Dunn, M.D.), codifies that laboratory developed tests are regulated by CMS under CLIA, not by FDA — statutorily carving LDTs out of the device definition and cementing the ACLA/AMP v. FDA vacatur. It creates a self-determined analytical-and-clinical-validity standard, a voluntary third-party "supplemental affirmation" (with FDA, MolDX, and New York as recognized pathways), a public LDT database, centralized error reporting, and for-cause CMS enforcement. It expands scope to digital laboratory data, aligns Medicare coverage and companion-diagnostic pathways, and modernizes CLIA specialties. Most provisions phase in two years after enactment.

###

Contents

The jurisdictional core: CLIA in, FD&C Act out
The validity standard: real, but self-determined and evidence-flexible
Voluntary third-party affirmation — and the special status of FDA, MolDX, and New York
Transparency: the centralized public database
Safety signal: centralized error reporting
Enforcement teeth: the for-cause validity review
Digital laboratory data: the frontier definition
Medicare coverage and companion diagnostics
CLIA modernization (Section 4)
Transition and the device-to-LDT conversion glidepath

Sidebar 1 — Red flags the lab industry should be watching Sidebar 2 — Advocates, opponents, prospects, and the (negligible) score

###

The Enhancing CLIA Act of 2026 (H.R. 8890): What the Full Bill Text Actually Does

An analysis for CLIA and laboratory-industry professionals, based on the 45-page introduced text (Dunn, FL-2; introduced May 19, 2026; referred to Energy & Commerce and Ways & Means).

Now that the legislative language is public, the bill turns out to be considerably more architecturally complete than the five-bullet press release suggested. The earlier Dunn release and the McDonald Hopkins alert captured the headline ideas — CMS not FDA, a public database, voluntary third-party review, error reporting — but the statutory text fills in the load-bearing details that determine whether this is a genuine framework or a slogan. Below is a plain-English walk through everything the bill changes, followed by two sidebars: one on the red flags the industry should be watching, and one on the politics, the coalition map, and the (probably trivial) budget score.

The bill operates on three statutes at once: it rewrites the core of CLIA (PHS Act §353 / 42 U.S.C. 263a), surgically amends the Food, Drug & Cosmetic Act to pull LDTs out of FDA's reach, and touches the Social Security Act to align Medicare coverage plumbing. That three-statute structure is the tell that the drafters were trying to settle the jurisdictional question for good, not just gesture at it.

1. The jurisdictional core: CLIA in, FD&C Act out

The first operative move strikes and replaces CLIA §353(a) with a flat declaration of scope: laboratory operations are to be regulated by the Secretary under CLIA and not under the Food, Drug, and Cosmetic Act. This is the legislative codification of what the Eastern District of Texas handed the industry in ACLA v. FDA and AMP v. FDA (the March 2025 vacatur), and what FDA conceded when it rescinded the 2024 LDT Final Rule in September 2025. The court said FDA lacked the statutory authority; the bill removes any doubt by amending the device definition itself.

The decisive companion edit is in the FD&C Act: §201(h)(1) is amended so that the term "device" does not include laboratory developed tests as defined here. Without that carve-out, the CLIA-side declaration would be a turf claim; with it, FDA's hook is statutorily severed. Two further FD&C edits (the adulteration provision at §501(f)(1)(C) and the misbranding provision at §502(o)) add carve-outs so that a device lawfully distributed for use with an LDT isn't dragged into adulteration/misbranding liability simply because it's deployed inside a lab's home-brew test. And the transition section orders FDA to revise the IVD definition at 21 C.F.R. 809.3 within 30 days to say the same thing in regulation: LDTs are out, and a protocol becomes an IVD only when it's commercially distributed to labs outside common ownership (or distributed by a non-laboratory) — with the useful clarification that merely publishing or patenting a protocol is not "commercial distribution."

That last point matters for academic and translational labs: presenting or patenting a method does not, by itself, convert your LDT into a regulated kit.

2. The validity standard: real, but self-determined and evidence-flexible

The bill creates an actual substantive bar — this is not pure deregulation. Beginning two years after enactment, no lab may perform an LDT unless it meets the "applicable standard." For clinical-use tests, that standard is a reasonable assurance of both analytical and clinical validity; for investigational-use tests, analytical validity alone.

Crucially, the standard is self-established by the developing laboratory, not gated by premarket review:

Analytical validity is met when the developing lab sets performance specifications supporting the test's ability to identify, measure, or detect its targets — and, if a different lab performs the test, that lab verifies the specs before use.
Clinical validity is met when the lab holds documented evidence supporting that the test reliably achieves its stated purpose. The bill enumerates a deliberately broad menu: peer-reviewed literature, clinical guidelines, bench studies, case studies/histories, consensus standards, reference standards, data registries, postmarket data, real-world data, one or more clinical validation studies — or "other evidence deemed appropriate by the Secretary."

The design philosophy is unmistakable: a flexible, file-based evidentiary standard that a competent high-complexity lab can satisfy from the literature and its own validation work, rather than a 510(k)/PMA-style dossier and queue. The trade-off — uneven rigor and enforcement ambiguity — is discussed in Sidebar 1.

3. Voluntary third-party affirmation — and the special status of FDA, MolDX, and New York

Layered on top of the self-determined standard is an optional "supplemental affirmation" that a CMS-approved third party agrees the test meets the standard. A lab may, but is not required to, obtain one. Once granted, it does not expire (unless the third party later loses its approval, with a 90-day wind-down and a 60-day window to migrate to another reviewer).

Approved third parties must apply standards CMS judges "sufficiently risk-based, rigorous, and not overly burdensome," and must agree to notify CMS of adverse findings (false/deceptive claims, or a probable risk of serious adverse health consequences) and of changes to their methods. Review runs on a 60-day clock, with structured "meets / doesn't-meet-but-resolvable / lacks-credible-evidence" outcome tracks, teleconference rights, and cure periods.

Three features will draw immediate attention from this readership:

Deemed affirmations. A test is deemed to carry a supplemental affirmation if it is (i) approved by the New York State Department of Health (CLEP), (ii) found reasonable and necessary under Medicare via a favorable MolDX technical assessment, or (iii) within a category CMS designates by regulation. Folding the MolDX TA and NY CLEP into the framework as deemed pathways is the most consequential operational choice in the bill for the molecular and genomic community — it means the review apparatus labs already navigate can satisfy the new standard without a parallel submission.
FDA as one approved third party among many. FDA is deemed an approved third party, and a lab may ask FDA for a supplemental affirmation — but FDA must review only against the CLIA standard and the third-party procedure, may not apply device standards or any other FD&C requirement, and may charge no more than the 510(k) premarket-notification user fee. The delegation provision (§353(o)) is amended to bar the Secretary from delegating the for-cause validity determination to FDA at all. Net effect: FDA is boxed into an optional, standard-constrained reviewer role and explicitly kept out of the enforcement seat.
No accreditor back-door. The certification/accreditation standards at §353(e) are amended so they may not require an LDT to carry a third-party affirmation. This locks the voluntariness in against accreditation bodies that might otherwise make it a de facto condition.

4. Transparency: the centralized public database

Beginning two years after enactment, every lab running clinical-use LDTs must register each test in a CMS centralized database, with a public-facing interface on the CMS website and a secure submission portal. Required fields include the lab and (if different) developing lab and certificate numbers; test name; purpose (analytes, target conditions, and intended use — screening/diagnosis/prognosis/etc.); specimen types (which may include digital laboratory data); examination type (molecular, genomic, flow, pathological, etc.); a summary of performance specifications; whether the test modifies an FDA-marketed device; and whether it is deemed or third-party affirmed. For new, non-affirmed tests, the lab must also post a brief summary of the clinical-validity evidence.

A corporate parent may file on behalf of commonly-owned labs. The portal must protect trade secrets, confidential commercial-financial information, national-security-sensitive material, and identifiable patient data. Submission timing: new tests within 30 days of first clinical use (or 60 days after the database launches); legacy tests within three years of enactment (or 60 days after launch). Updates/corrections are due within 30 days.

This is the provision most likely to reshape competitive dynamics — see Sidebar 1.

5. Safety signal: centralized error reporting

The bill builds a lightweight, MDR-flavored adverse-event system. Beginning two years after enactment, a lab must file an individual report within 5 calendar days of becoming aware of an undetected inaccurate result for a clinical-use LDT that reasonably suggests the test caused serious harm resulting in death, or an imminent threat to public health. Lesser serious-harm events (short of death) are captured in quarterly reports. "Serious harm" and "serious injury" are defined in device-recall-adjacent terms. A reassuring clause specifies that filing a report — or CMS releasing it — is not an admission that the test caused harm.

6. Enforcement teeth: the for-cause validity review

This is where the "voluntary" framework acquires a backstop. Beginning two years after enactment, the Secretary may issue a written request for information to a developing lab when credible and verifiable information indicates either (a) the test lacks the applicable standard and has no affirmation, or (b) it is affirmed but is being represented for a purpose the affirmation doesn't support. The process is procedurally generous — teleconferences, 45-day windows, cure periods — but its endpoint is sharp: if CMS concludes the criteria apply, the lab must cease performing the test (and instruct all commonly-owned labs to cease) until CMS confirms the concern is resolved.

The catch — and it's the most important structural feature in the bill — is how a lab gets the green light to resume: CMS confirms only when the lab shows it has obtained a supplemental affirmation from an approved third party and that the criteria no longer apply. In other words, the "voluntary" affirmation becomes effectively mandatory the moment CMS raises a credible concern. The safe harbor is conditional, not optional, once you're in the crosshairs.

7. Digital laboratory data: the frontier definition

One of the most forward-leaning pieces — and the one the press release flagged as "innovative analyses of digital laboratory information" — is a broad new statutory category of digital laboratory data. It expressly includes whole-slide images derived from glass slides, flow cytometry plots, cytogenetic karyograms, chromatographic/mass-spec/clinical-chemistry/immunological/hematological/microbiological data, electropherograms, gel images, expression/array/sequencing data — and, critically, subsequent analyses of such data.

Three downstream consequences:

The definition of "laboratory" itself is expanded to include the analysis of patient-specific digital laboratory data. Re-interrogating data — e.g., re-analyzing an already-sequenced genome for a new indication — is squarely "laboratory testing."
The LDT definition explicitly survives software being executed outside the laboratory (clause iv). Cloud compute and remote bioinformatics pipelines do not break LDT status.
Together, these codify that digital pathology and the re-analysis of existing molecular data are LDTs performed under CLIA — clarity the digital-path and bioinformatics field has lacked. The double edge (the FDA SaMD/AI boundary, and whether commercial software vendors are swept in) is in Sidebar 1.

8. Medicare coverage and companion diagnostics

Two quieter provisions align the payment world with the new framework:

NCD equivalence (SSA §1862(l)(1)). For all existing and future national coverage determinations for a clinical lab test, a third-party supplemental affirmation is to be treated as equivalent to an FDA approval or clearance. Many coverage constructs reference FDA status; this keeps affirmed LDTs from being stranded by coverage language that assumes an FDA pathway.
Companion diagnostics. Where a diagnostic result is required for a drug approval (§505) or biologic licensure (§351), that result may be furnished by an affirmed LDT, not only by an FDA-approved companion device. This cracks open a space historically reserved for FDA-cleared CDx.

9. CLIA modernization (Section 4)

Beyond LDTs, the bill finally addresses the long-stale machinery of CLIA itself:

New specialties. Within 180 days, CMS must propose regulations establishing new examination types reflecting current science — explicitly molecular diagnostics, digital pathology, and next-generation sequencing — and then evaluate whether new proficiency testing programs are needed. CLIA's specialty list has never had a molecular or genomics category; this is a genuinely overdue fix.
Sub-regulatory transparency. CMS must publish a report and take public comment at least 90 days before issuing new or revised sub-regulatory guidance — including State Operations Manual changes — a direct response to years of industry complaints about policy-by-memo.
Annual open-door forums with labs, and a five-year regulatory review cycle with a Federal Register RFI and a 180-day public docket.

10. Transition and the device-to-LDT conversion glidepath

Jurisdiction flips immediately on enactment, but the operative obligations phase in. CMS must propose regulations within 180 days, stand up the database within two years, finalize conflict-of-interest rules within two years (more on that in Sidebar 1), and revise 21 C.F.R. 809.3 within 30 days. No regulations take effect until the two-year mark.

For tests that currently hold — or have pending — an FDA PMA, HDE, 510(k), de novo, IDE, or biologic license, the bill provides a conversion mechanism: such a test remains a "device" until the lab notifies FDA (within 60 days) whether it intends to keep device status or convert, with various date triggers. A test with an existing FDA approval that converts is deemed to hold an FDA supplemental affirmation that it meets the standard — so the validation work already done at FDA is not wasted.

SIDEBAR 1 — Red flags the lab industry should be watching

The bill is broadly favorable to labs, which is exactly why the friction points are worth naming precisely rather than glossing.

"Voluntary" is conditional, not optional. The cleanest way to read §353(r)(4)(G) is that once CMS opens a credible-concern review and concludes the criteria apply, the only documented route back to performing the test is a third-party affirmation. The much-praised voluntariness holds right up until the moment you actually need it — which is the moment it disappears. Labs should not treat third-party affirmation as a marketing nicety; they should treat it as cheap insurance against an enforcement stop.

A self-attested clinical-validity floor invites both gaming and ambiguity. The flexible evidence menu is a feature for legitimate innovators and a loophole for marginal actors who can paper a file with "case histories" and "real-world data." "Reasonable assurance" is undefined at the level of granularity that determines enforcement outcomes. Expect the contours to be set by the first wave of for-cause reviews and any guidance, not by the statute.

The public database is a competitive-intelligence engine. Performance-spec summaries — and, for non-affirmed new tests, clinical-validity summaries — become publicly posted. The trade-secret protection exists, but the boundary between a permissible "summary" and proprietary detail is unspecified. Competitors, payers, and short-sellers will mine this. Labs will need a disciplined posture on what goes in the summary field.

The error-reporting regime is operationally and legally fraught. Defining an "undetected inaccurate result," and attributing serious harm or death to a test rather than to confounding clinical factors, is genuinely hard. The "not an admission" clause helps, but reports are discoverable and will feed malpractice and potentially qui tam theories. Labs need a defined intake, triage, and 5-day-clock process before the obligation begins.

The conflict-of-interest rulemaking is aimed at a specific incumbent model. The mandate to mitigate conflicts among organizations that bundle accreditation, proficiency testing, and supplemental affirmation (§3(b)(1)(C)) reads as a direct constraint on entities like CAP that occupy multiple oversight roles. Who can serve as an approved third party — and whether existing accreditation relationships get unsettled — is unresolved and could reshape the vendor landscape.

Third-party capacity is the silent bottleneck. The deemed pathways concentrate demand on a few overloaded chokepoints: MolDX technical assessment (already backlogged, and with its own recent governance turnover), NY CLEP (finite capacity), and an FDA reviewer role that is both slow and now standard-constrained. If adoption is heavy, the queue, not the standard, becomes the binding constraint.

Business-model constraints favor consolidation. The "performed only within the same or commonly-owned high-complexity lab" boundary, plus the rule that commercially distributing a protocol to unaffiliated labs forfeits LDT status (it becomes a kit/IVD), constrains licensing, franchising, and distributed reference-lab models. The structural pressure is toward either keeping everything in-house/under common ownership or going the FDA kit route.

The digital-data scope is broad and double-edged. Codifying that re-analysis of existing data is "laboratory testing" gives digital pathology welcome clarity — but it also raises whether standalone AI/bioinformatics software vendors are now (a) swept into the lab framework, or forced to partner with a CLIA lab, and (b) sitting in a gray zone against FDA's separate SaMD/AI device authority. The bill carves LDTs out of "device," but a software product commercially distributed to many unaffiliated labs may still be a device. That seam runs straight through the most active part of the field.

Funding is the unspoken liability. CMS must build and run a database, an affirmation-oversight program, new specialties, and PT evaluations. The bill lets the Secretary spend "such funds as necessary" but appropriates nothing. CLIA is fee-funded — so the realistic funding mechanism is higher CLIA fees on labs. A bill that costs the Treasury almost nothing may not be free to the industry it regulates.

Two interlocking years of fog. Jurisdiction flips on day one, but the standard, database, COI rules, and specialties all land on a two-year horizon through interlocking rulemakings. That's a long stretch of "the rules exist but aren't operative yet," with real implementation risk if CMS under-resources the build.

SIDEBAR 2 — Advocates, opponents, prospects, and the (negligible) score

The coalition map is essentially the VITAL-vs-VALID map, redrawn. This bill is the affirmative legislative descendant of the CLIA-oversight camp — the philosophy of Rand Paul's VITAL Act — now riding the tailwind of the ACLA/AMP court victory and FDA's 2025 rescission of the Final Rule.

Likely advocates:

ACLA is the clear champion; President Susan Van Meter issued a supportive statement on introduction, noting the bill reflects CLIA-modernization concepts ACLA has long recommended. The bill essentially codifies ACLA's litigation win.

AMP, co-plaintiff in the vacatur and a long-standing opponent of FDA LDT regulation.

The CLIA-oversight coalition — ADLM (formerly AACC), NILA, and AAB — which led or joined amicus briefs and has argued for years that oversight belongs in CLIA, not FDA.

Large reference labs (Quest, Labcorp), academic medical center and hospital labs, genomics/precision-medicine labs, and digital-pathology vendors, who gain long-sought clarity.

The sponsor profile helps: Dunn is a physician who also introduced an "Improve LCDs" bill — a recognizably lab-friendly agenda.

Likely opponents and skeptics:

The VALID Act coalition — AdvaMed/AdvaMedDx and its historical public-health allies (Pew Charitable Trusts, Friends of Cancer Research, the American Cancer Society Cancer Action Network) — who favor a single, risk-based FDA framework spanning both kits and LDTs, and who view the where-the-test-is-made distinction as the wrong organizing principle. They will frame Enhancing CLIA as cementing an unlevel playing field: IVD kits face FDA premarket review while competing LDTs self-attest.

IVD manufacturers, on the same level-playing-field logic.

Patient-safety voices and academics in the Theranos-cautionary tradition, who consider self-attestation plus voluntary affirmation too thin.

FDA institutionally, though its posture is materially weakened post-vacatur and post-rescission.

CAP sits in an awkward middle: supportive of CLIA-based oversight in principle, but directly exposed by the conflict-of-interest provision targeting bundled accreditation/PT/affirmation roles.

Prospects. Two structural headwinds: the bill carries a dual referral (Energy & Commerce and Ways & Means), which means jurisdictional negotiation and the practical need for movement in both committees; and LDT legislation has a long graveyard — VALID never reached enactment across multiple Congresses despite repeated attempts to ride must-pass FDA user-fee and year-end vehicles. As of now this is freshly introduced, with a thin cosponsor list and no Senate companion or markup. The realistic path is multi-Congress, most plausibly as a rider on a larger FDA/health vehicle rather than as a standalone enactment. The counter-tailwind is real, though: with the Final Rule vacated and rescinded, the acute crisis that the VALID camp leveraged is, for the moment, defused — which both lowers the urgency to act and clears the field of a competing must-do.

The budget score — and why "negligible" cuts both ways. Expect a minimal-to-negligible CBO score. This is a regulatory restructuring, not a spending or revenue program: no new mandatory outlays, no benefit expansion on the face of it, and a database/oversight build that is modest and, because CLIA is user-fee funded, largely recoverable through fees rather than appropriations. A scorekeeper might probe whether smoother Medicare coverage (the NCD-equivalence and CDx provisions) marginally lifts Part B test utilization, but that's a second-order effect, not a headline cost.

Politically, a near-zero score is a double-edged asset. On the upside, it removes the single biggest obstacle to attaching the bill to a vehicle — no offsets to find, no budget point of order, no PAYGO fight.

On the downside, it gives the bill no fiscal constituency: there are no scoreable savings to make it attractive as a pay-for for something else. A negligible score means low friction but also low leverage — the bill will have to move on stakeholder muscle and policy merit, not on its budget math.

Note on sources: the analysis above is drawn from the introduced text of H.R. 8890 itself. The Dunn press release and the McDonald Hopkins client alert preceded the public text and are used only as background; where the enacted-as-introduced language differs from those summaries, the bill text governs. Political and coalition context (the ACLA statement, the ACLA/AMP v. FDA vacatur, and the VALID/VITAL history) reflects public reporting and stakeholder statements current to mid-2026.

Discoveries in Health Policy

Friday, June 12, 2026

Congress Releases Full Text of 45-page CLIA Reform Bill