First, we try to prove the pattern innocent.

Stage 01 · Detect

Working from claims data

A claim records who billed, for whom, under what code, on what date, and for how much. One record rarely shows whether the work happened.

Reviewed together, claims describe behavior, and behavior can be compared: to a provider's own history, to a peer group, to the size of the population being served, to what is physically possible in a working day.

We work from public and licensed sources: program claims and payment data, provider enrollment and identifier records, exclusion lists, ownership and corporate registries, court dockets and regulatory filings, and public reporting on payments between manufacturers and clinicians.

What we look for

Outliers

A provider, supplier, or plan sitting outside every reasonable peer group on a measure that should not vary much.

Implausible volume

Billing that cannot be reconciled with hours in a day, staff on site, or patients in the area.

Migration

Billing that leaves one program after enforcement and reappears in another. One documented example is set out below.

Velocity

The bust-out pattern: open fast, bill heavily, close before the billing is reviewed. HHS OIG has described suppliers shifting into exactly this shape, sometimes by buying an enrolled supplier and redirecting its payments.4 Source 4, see sources page

Quiet zones

Expected activity that is absent. A denial rate that should exist and does not. A category of service that stops being billed the month a rule changes.

Structure

Shared addresses, shared owners, transferred billing numbers, and referral relationships that do not match the geography of the patients.

Missing identifiers

Gaps in the data itself. HHS OIG has found that identifiers for ordering providers are largely absent from managed care data, which limits the government's ability to detect patterns of inappropriate billing.4 Source 4, see sources page

See the sources

How the government uses models

CMS operates a Fraud Prevention System that uses hundreds of models to monitor provider and supplier outlier billing.6 Source 6, see sources page HHS OIG data scientists monitor claims data to flag outliers and anomalies, and agents then follow up in the field.4 Source 4, see sources page The published record of that work, and what it says about the gap it is working against, is set out in the analysis.

Stage 03 · Litigate

Turning a pattern into a complaint

A False Claims Act complaint has to plead fraud with particularity, which means identifying specific conduct.

The third stage converts the pattern into the elements the statute requires: a false claim or statement, knowledge as the statute defines it, materiality to the government's payment decision, and federal money paid out.§ Source, statute, see sources page It then has to withstand a motion to dismiss, a motion for summary judgment, and, if the case goes that far, a jury.

What that involves

The complaint

Drafting the complaint and the written disclosure of material evidence that is served on the United States.§ Source, statute, see sources page

The seal period

Working with the government while the case is under seal, which usually means answering questions, producing documents, and making the relator available.

The intervention decision

Preparing for the government's election, and for the case continuing if the government declines.§ Source, statute, see sources page

Discovery

Custodians, systems, claims extracts, policy documents, and the people who wrote them.

Expert work

Statistical, clinical, coding, and damages.

Damages and penalties

The statute trebles damages and carries a civil penalty for each false claim.2 Source 2, see sources page,§ Source, statute, see sources page

The relator's share

Determined by the court within the range the statute sets.§ Source, statute, see sources page