A False Claims Act case built on 50,000 claims does not require the government to prove each claim false one at a time. In United States ex rel. Martin v. Life Care Centers of America (E.D. Tenn. 2014), the court let the government use a 400-claim sample drawn from more than 154,000 submitted claims to establish liability across the full set, not merely to calculate damages once liability was already proven. Once a court accepts sampling as proof of liability, the fight moves to whether the sample was built correctly, and that now turns on Daubert-style methodology challenges as much as billing facts.
Sampling as Proof of Liability, Not Only a Damages Shortcut
Courts have long accepted extrapolation to calculate damages once falsity is established claim by claim. The harder question is whether a sample can prove falsity and liability in the first place, standing in for individualized review of every claim in a lookback period. Martin answered yes for a nursing home operator accused of inflated therapy billing across 82 facilities nationwide. Extrapolation now regularly enters a matter at the liability stage, not only the damages stage, and a compliance officer's first exposure is often a government witness's report, not a courtroom ruling. Our companion piece, Statistical Sampling in FCA Cases: Where Courts Draw Limits, covers how far courts have let sampling reach into individualized clinical judgment; this piece focuses on the methodology challenge itself.
The Daubert Fight Over Sample Size and Qualification
A sample offered to prove liability has to survive scrutiny under Daubert v. Merrell Dow Pharmaceuticals and Federal Rule of Evidence 702 before a judge weighs the underlying claims. In United States ex rel. Grant v. Zorn (8th Cir. 2024), the defendant challenged a certified public accountant's extrapolation from a 31-file sample, arguing the sample size was never proven statistically valid and the witness lacked a trained statistician's credentials. The Eighth Circuit upheld the extrapolation anyway, holding that a bench trial allows a more flexible evidentiary standard, and it expressly left open whether the same sample would survive a jury. The same court separately vacated the damages award on Eighth Amendment grounds, a reminder that surviving the sampling challenge does not end the exposure fight. Inside an FCA Investigation: The Defense Timeline walks through the government's steps that lead up to this point.
Where the Circuits Have Not Agreed
No circuit court has adopted a uniform rule for when extrapolation can prove liability. In United States ex rel. Michaels v. Agape Senior Community (4th Cir. 2017), the court took an appeal specifically to decide whether sampling could establish liability across more than 63,000 hospice and long-term care invoices, then dismissed it as improvidently granted. The Fifth Circuit reached the opposite practical outcome in United States v. Hodge (5th Cir. 2019), affirming a verdict of nearly $300 million built on a sample of loan files, but only because the defendant had agreed to the sampling protocol during discovery and could not challenge its reliability on appeal. The split is really a timing problem: objections raised early carry weight, and the same objections raised after a verdict rarely do.
Building the Record Before the Sample Becomes the Case
A defense to an extrapolated demand starts with the sampling plan, not the trial. Counsel and an independent statistician should test whether the sample was drawn randomly from a comparable population, whether the sample size supports the claimed margin of error, and whether the government's witness holds the statistical qualifications the methodology requires, not just familiarity with the billing codes. A matter resolved on an extrapolated demand can carry a negotiated corporate integrity agreement or trigger an OIG exclusion for individuals tied to the conduct, so the sampling fight sets the exposure ceiling before settlement talks start. An extrapolated finding can also start the clock on repayment duties, covered in The 60-Day Overpayment Rule: Identification, Quantification, and Repayment.
A sample a court accepts to calculate damages once liability is proven is not automatically the same sample a court will accept to prove liability in the first place.
Why Early Legal Counsel Is Critical
It is critical that healthcare providers promptly retain experienced healthcare defense counsel upon receiving a subpoena, audit notice, civil investigative demand, or other government inquiry referencing a claim sample or an extrapolated demand. Early legal intervention can protect the provider's rights, ensure appropriate responses to government requests, avoid inadvertent admissions that could validate a flawed sample, and preserve methodology objections before they are waived. Delaying legal representation can significantly affect the outcome of a matter and expose the provider to unnecessary risk.
How Health Law Alliance Can Help
Health Law Alliance defends healthcare providers against False Claims Act allegations built on statistical sampling, from the first civil investigative demand through negotiated resolution or trial. Our bench includes a former federal prosecutor who has challenged government sampling methodology in qui tam and government-initiated matters alike. If your practice faces an extrapolated demand, contact us today for a free consultation.
