KPSGILL policy proposal · model CMS policy
Laboratory quality as payment integrity
When a laboratory’s own quality controls fail, the current instrument is a false-claims settlement years later. The cheaper instrument is a coverage condition now.
The problem
Laboratory quality and laboratory payment are regulated as though they were separate subjects. Analytical validity sits with one regime; whether Medicare should pay for a test sits with another; and clinical reliability — whether the result can actually support the decision a physician makes with it — sits nowhere in particular until an enforcement action arrives.
The record now shows what that gap costs. A programme-integrity campaign reports more than $1.6 billion in potentially improper laboratory payments stopped, with 157 providers revoked ENFORCEMENT. Separately, a company settled allegations that it billed for skin-cancer testing despite known quality-control failures. The second is the more interesting record: the alleged failure was in the laboratory, and the instrument used to address it was a false-claims settlement years later.
Documentary baseline: the CMS laboratory programme-integrity announcement of 28 August 2026; the DermTech resolution through a bankruptcy claim; and the Villages Health risk-coding settlement, which shows the same pattern in a different domain. Each with its primary source in the record layer.
The recommendation
Make clinical reliability a condition of payment prospectively, rather than a theory of liability retrospectively. If a test’s performance characteristics do not support the clinical use being billed, that is a coverage question now — not a fraud question in five years.
1. Reliability as a coverage condition. Payment for a laboratory test billed for a stated clinical use is conditioned on the laboratory maintaining, and producing on request, performance characteristics for that use: analytical validity, clinical validity where the claim is diagnostic or prognostic, and the population in which those characteristics were established.
2. Quality-control attestation. A laboratory billing more than a threshold volume of a test shall attest annually that its quality-control results for that test met its own established criteria, and shall retain the underlying records for the claims-retention period.
3. Known-failure rule. A claim submitted for a result produced during a period in which the laboratory’s own quality-control criteria were not met, and were known not to be met, is not payable. This states a payment condition; it does not establish intent for any other purpose.
4. Notification. A laboratory that identifies a quality-control failure materially affecting results already reported shall notify the ordering clinicians and the programme within thirty days, identifying the affected date range.
5. Escalation ladder. The programme shall address failures under this policy in the following order: education, corrective action plan, prepayment review, payment suspension, revocation, referral. Referral is not a substitute for the earlier steps where they have not been attempted.
6. Publication. Actions taken under section 5 shall be published with their stage identified, so that a revocation is not reported as a fraud finding and an education letter is not reported as an enforcement action.
Section 5 is the part that will be resisted from both directions: it constrains the programme’s discretion to escalate, and it removes the option of treating enforcement as the primary quality instrument. Section 6 is what would make the site’s own enforcement ledger easier to keep honest.
Who bears what
Physician
A result arrives with knowable performance characteristics for the use it was ordered for. Notification under section 4 tells her which of her patients were affected, while it still matters.
Patient
A test that cannot support a clinical decision stops being billed as though it could.
Laboratory
Real prospective burden under sections 1 and 2 — and, in exchange, a defined path under section 5 instead of an unannounced arrival at revocation.
Programme
Recovers less through settlement and prevents more through coverage. That is a worse press release and a better outcome.
Competition
Sections 1 and 2 fall hardest on laboratories whose economics depend on volume rather than on validity.
Equity
Unreliable testing concentrates where oversight is thinnest; prospective conditions reach places that a fraud investigation never gets to.
The strongest arguments against
- Clinical validity is contested for many tests; conditioning payment on it lets the programme deny coverage by definitional fiat.
- Section 3 converts an ordinary quality lapse into a payment violation.
- This duplicates the existing laboratory quality regime.
- Section 5 ties the programme’s hands where fraud is obvious.
- Prospective review will slow access to new diagnostics.
Answers
- Section 1 requires the characteristics to be produced, not to meet a threshold set after the fact. The obligation is disclosure of what is known, which is the least a payer can ask.
- Section 3 is deliberately narrow: the criteria were the laboratory’s own, and it knew they were not met. It also expressly disclaims any finding of intent.
- It does not duplicate it; it connects it to payment, which is the connection currently made only by litigation.
- Section 5 governs failures under this policy. Nothing in it limits action on evidence of fraud through existing authority.
- A real cost. The alternative on the current record is paying for years and litigating afterwards, which serves nobody, including the laboratories that do this properly.
Metrics, sunset, open questions
Metrics. Share of high-volume tests with performance characteristics on file for the billed use; median days from quality-control failure to clinician notification; distribution of actions across the section 5 ladder; recoveries through settlement as a share of total programme action, which should fall if the policy works.
Sunset. Five-year review. Sections 1 and 2 sunset if the laboratory quality regime itself adopts equivalent prospective conditions — the objective is the condition, not the venue.
Open questions. What volume threshold makes section 2 proportionate for a small laboratory? Should section 4 notification run to patients directly as well as to ordering clinicians? Does the known-failure rule need a safe harbour for a laboratory that self-discloses before the programme finds it?
Related in the record: the enforcement ledger, where all three anchoring records sit with their action types and states · pre-market versus post-market surveillance · the False Claims Act in healthcare.