archive api

The 2,158-Mile Day

Medicaid's non-emergency medical transportation benefit is small, invisible, and structurally prone to fraud. In 2025 and 2026, state attorneys general finally started enforcing like they meant it.

K3 Analytics · April 21, 2026

Medicaid's non-emergency medical transportation benefit — "NEMT" in the acronym-choked world of health policy — is the thing that gets low-income patients to dialysis, to methadone, to oncology, to the pediatrician. Federal law requires every state to arrange it. Roughly 3 to 4 million beneficiaries use it each year, which is about four or five percent of the Medicaid population. Total national spending on the benefit lands somewhere in the low single-digit billions of dollars — a rounding error in a $957 billion program.

The benefit is also, by any reasonable measure, one of the most reliably fraud-stained line items in American health care. The Government Accountability Office counted roughly 200 convictions, civil settlements, and judgments against NEMT providers across 25 states in just the five years from FY2015 to FY2020. Every audit the HHS Office of Inspector General has ever published on state NEMT programs — Michigan, Texas, Oklahoma, Indiana, Massachusetts, most recently a Texas subcontractor called SafeRide Inc. — has found improper payments ranging from hundreds of thousands of dollars to tens of millions. The first comprehensive OIG study of NEMT fraud, published in 2006, catalogued the same fraud typologies — ghost rides, mileage padding, upcoding, kickbacks to beneficiaries — that state AGs are still prosecuting in 2025.

The fraud is not a mystery. It is a feature of the structure.

How the money moves

Most states contract NEMT out to a private "broker" under a capitated per-member-per-month payment. The broker arranges rides through a network of small subcontractors — sedans, wheelchair vans, stretcher vans — most of which are tiny operations with five to twenty vehicles and thin margins. The broker keeps the spread between the capitated rate and what it pays subcontractors. The subcontractor keeps what is left after paying the driver.

Four companies now dominate the broker layer: ModivCare (formerly LogistiCare), MTM Inc., Verida (formerly Southeastrans), and — until very recently — Access2Care, which MTM finalized its acquisition of in 2024. On April 1, 2026, Georgia handed its entire statewide NEMT benefit — serving more than two million beneficiaries — to a single broker, Verida. Three companies now effectively run Medicaid transportation in most of the country.

The structure sounds unobjectionable on paper. In practice, GAO's 2021 report quoted state officials conceding that capitation "shifts much of the risk of improper NEMT payments from the State to the broker." Which is a tidy way of saying the broker now owns the integrity problem. The broker then pushes subcontractor rates down to protect its spread. Subcontractors with single-digit-percent margins respond by billing for trips that did not happen, or that happened but not the way the claim says they did, or that were performed by drivers who were not supposed to be behind the wheel in the first place. This is not a mystery. Every OIG NEMT audit for the last two decades has described it.

The price of NEMT varies wildly across states — by a factor of more than twenty on a per-member-per-month basis. Milliman's 2025 analysis of seven Medicaid clients found a range from $0.72 PMPM in states where NEMT is carved into MCO capitation to $16.57 PMPM in states that run a dedicated broker contract. The spread reflects utilization, network density, urban-rural mix, and the degree to which MCOs are willing to spend on a benefit that, in many cases, they do not control. It also reflects how much oversight each state actually exercises.

The fraud, concrete

On June 30, 2025, New York Attorney General Letitia James announced settlements with 16 transportation companies and lawsuits against seven more, resolving more than $13 million in alleged Medicaid fraud. The operative facts in her press release are the kind of thing that would embarrass a television writer as too heavy-handed:

• One driver at American Base No. 1 was alleged to have logged 96 trips and 2,158 miles in a single day. The settlement was $4.78 million.
• Dutchess Black Car Service was sued for, among other things, billing 16,203 identical $50 tolls on routes that did not pass through toll roads.
• Angel Medical Transportation in Schenectady ($1.1M), Buffalo Taxi Services ($1.69M), and others settled allegations of billing trips that never occurred.
• SMI Transportation of Buffalo was sued for using a driver who had previously been excluded from the Medicaid program after a prior fraud conviction.

Five weeks earlier, in the Eastern District of New York, Adnan Arshad pleaded guilty to leading an approximately $20 million Medicaid ambulette fraud operated through five co-owned Long Island entities — 668 MTK Taxi, All-Star Taxi, Apollo Transportation, Sunrise Taxi, and Transportation Solution NY Corp d/b/a A1 Transport. According to the EDNY indictment, the scheme paid cash kickbacks to methadone-clinic patients, routed them to distant addiction-treatment centers in Manhattan and the Bronx to inflate mileage, billed for trips that never occurred, and used drivers who were not authorized. Roughly $19 million in luxury vehicles and real estate were subject to forfeiture.

In Toledo, Abdul Haji Faqi, owner of Blue Line Express Taxi, was convicted by a federal jury in October 2022 of more than $800,000 in upcoding — billing wheelchair-van rates for patients who were ambulatory, over a seven-year period. In Florida, Attorney General James Uthmeier's office announced arrests in May 2025 in a $345,000 case in which GPS data placed the operator at home during more than 8,000 purportedly billed trips. In Western Pennsylvania, Rex Barr pleaded guilty to a $9 million Medicaid transportation scheme.

The typology library is remarkably stable across every case. It is the same library OIG described twenty years ago:

The broker layer is not immune

Industry narrative tends to locate NEMT fraud entirely at the subcontractor layer: a few bad apples in a big orchard. That narrative is harder to sustain after 2023, when Modivcare Solutions LLC — the former LogistiCare, the largest NEMT broker in the United States — paid $3.75 million to the federal government to resolve False Claims Act allegations brought by three whistleblowers from a downstream Ohio subcontractor called Mobile Care. The qui tam complaint alleged that Modivcare submitted false Medicare and Medicaid claims by failing to ensure the NEMT services it arranged in Ohio between December 2009 and March 2023 were medically necessary. Modivcare settled the allegations without admitting liability. Mobile Care itself settled separately for $475,000, also without admission of liability.

It remains the only public federal broker-level NEMT fraud settlement of that size. That is either a sign the broker layer is cleaner than the subcontractor layer — or that the broker layer is harder to investigate because it sits on the capitated side of the contract and its fraud exposure is more often a matter of inadequate oversight than direct false claims. The HHS OIG's newly opened work plan item, "Using Targeted Reviews to Reduce Fraud, Waste, and Abuse in Medicaid Nonemergency Medical Transportation" (project OEI-02-25-00360, announced October 15, 2025), is the federal government's bet that provider-level data mining will reveal what state-level audits have not.

The biggest takedowns

Below are the largest publicly announced NEMT enforcement actions from 2022 through early 2026. Every entity listed has a public enforcement action — an indictment, guilty plea, conviction, or settlement — and is referenced only in that context.

The state audit archive reads like a geological record

The HHS-OIG has been publishing state-level NEMT audits for more than a decade. The findings are remarkably similar across states and across time:

Every audit has identified roughly the same set of control failures: inadequate driver screening, missing documentation on medical necessity, claims for trips to non-covered services, inadequate oversight of subcontractors. Every audit has resulted in CMS recommendations, state corrective action plans, and modest refunds to the federal government. None has resulted in a structural rewrite of how the benefit is administered.

What is new about 2025–2026

Three things are different this cycle.

First, state attorneys general are running the enforcement. New York's Letitia James, Florida's James Uthmeier, and Pennsylvania's AG office have each produced multi-million-dollar NEMT cases in the last 18 months. Their Medicaid Fraud Control Units are using GPS data, odometer audits, and cross-reference of driver logs against LEIE exclusions to produce the kind of concrete evidence — the 2,158-mile day, the 16,203 impossible tolls, the 8,000 trips billed while the operator was sitting at home — that federal prosecutors historically struggled to secure. In 2024–2025, state-AG NEMT recoveries appear to have materially exceeded federal recoveries.

Second, the broker layer has consolidated to the point where a single vendor failure is a statewide failure. Georgia's handover of its entire NEMT program to Verida on April 1, 2026 concentrates the benefit for more than two million beneficiaries under one private contractor. Mississippi made a similar swap to ModivCare in 2024. MTM's absorption of Access2Care means three companies now deliver the benefit in most of the country. This is good for standardization and bad for resilience — and OIG has not yet published a dedicated audit of the broker layer.

Third, CMS is finally enforcing provider-exclusion screening at the NEMT level. In July 2021, CMS issued guidance implementing §1902(kk) of the Social Security Act, requiring every NEMT provider and every driver to be clean on the HHS-OIG LEIE. This was written specifically because the industry had been ignoring it — the SMI Transportation case above involved exactly this pattern. Enforcement of the rule depends on cross-walking broker networks against the LEIE, something that requires either the broker or the state to actually do the lookup. HHS-OIG's new targeted-review project is expected to test this.

What is worth watching

Four things, between now and late 2027.

One: the HHS-OIG targeted-review project (OEI-02-25-00360). If the project produces provider-level referrals in the way the urinary-catheter fraud pipeline did in 2023, it will generate a new wave of federal cases against subcontractors that state AGs have not yet reached.

Two: the broker-contract rebids queued up in Arizona, Connecticut, Maine, Nevada, and Tennessee. Rate precedents set in these contracts will shape subcontractor margins — and therefore fraud incentives — for the next five-year cycle.

Three: whether ModivCare's 2024 liquidity squeeze and associated securities-litigation pressure produces a second broker-level FCA action. The Modivcare Solutions LLC settlement resolved allegations for 2009–2023 conduct in Ohio; it is not structurally implausible that similar allegations exist in other states, and the qui tam whistleblower incentive is strong.

Four: the Minnesota model. Minnesota has taken an approach distinct from the AG-driven playbook — its Department of Human Services Inspector General, James Clark, has imposed a moratorium on new Twin Cities NEMT enrollments and conducted hundreds of unannounced provider site visits. If the Minnesota approach materially reduces fraud rates relative to comparable states, it will become the template other state Medicaid agencies adopt. If it does not, the federal OIG's provider-level data-mining bet will be the only remaining play.

The industry footprint vs. the exclusion footprint

A FraudGraph cross-reference of the NPPES provider registry against the HHS-OIG List of Excluded Individuals/Entities (LEIE) puts the enforcement record in perspective against the industry base. 35,828 providers hold an NEMT taxonomy code in NPPES across the United States. 47 have a direct NPI match on the LEIE. Ohio — the state from which the Modivcare/Mobile Care FCA whistleblower case originated — accounts for 15 of those exclusions, more than any other top-10 market despite being only the third-largest by provider count. New York, site of Attorney General James's June 2025 sweep, accounts for another 8. California and Texas, the two largest NEMT markets with roughly 4,300 and 3,400 registered providers respectively, have one to two LEIE-excluded providers each. Virginia, North Carolina, and Georgia — all top-10 markets — have zero direct NPI matches on the LEIE. The picture is consistent with the article's central claim: the recent enforcement wave is state-AG-driven and geographically uneven.

How we did it

This report cross-references public HHS-OIG audits and work-plan items, federal and state DOJ press releases, FCA qui tam settlement announcements, state attorney general press releases, CMS reports to Congress, GAO report GAO-22-105447, MACPAC's 2021 mandated NEMT report, and SEC filings of publicly traded brokers. The provider-level cross-reference in the final chart joins NPPES (HHS-NPI registry) to HHS-OIG's LEIE via the K3 Analytics FraudGraph warehouse. All named individuals and entities appear in connection with publicly filed indictments, guilty pleas, convictions, or civil settlements — each linked to the originating press release or audit report. Where civil settlements resolved allegations without admission of liability, that fact is stated. Source: FraudGraph cross-reference of DOJ, HHS-OIG, state AG, and GAO public records, April 2026.