Texas runs one of the most disciplined Medicaid programs in the country. The state's overall payment error rate (1.31% against a 5.09% national average) reflects years of investment in program integrity infrastructure. The Texas Health and Human Services Commission (HHSC) has the oversight tools, the enforcement posture, and now a direct Governor's Office mandate to use them.
Which makes what is happening in non-emergency medical transportation (NEMT) particularly conspicuous.
The Texas Governor's Office directed HHSC to conduct a targeted policy review of non-emergency medical transportation, confirmed publicly at the April 8, 2026 Senate HHS Committee hearing. The review is on track for completion by June 2026. As of June 29, 2026, no public report or findings have been released. When those findings land, encounter data quality is likely to be among the first areas scrutinized.
For health plans that currently rely on a broker to manage their NEMT benefit, the question worth asking now is a simple one: how confident are you in the encounter data your broker is submitting on your behalf?
Texas Medicaid is almost entirely managed care. More than 93% of Medicaid beneficiaries are enrolled in managed care plans, with NEMT carved directly into managed care organization (MCO) contracts. That structure places the encounter data submission obligation squarely on the MCO, not the broker.
But in most Texas health plans, the broker is the one generating the NEMT encounter data. The MCO pays the broker, the broker manages the trips, and the broker produces the encounter records that the MCO submits to HHSC. The accountability flows in one direction; the data flows in the other. When HHSC has questions about encounter data quality, they come to the MCO. The MCO has to go back to the broker. The broker has to reconstruct records from a system the MCO cannot see.
That is a structural accountability gap, and it exists in virtually every broker-managed NEMT program in Texas today.
The reason most NEMT programs struggle with encounter data is not a lack of effort. It is how the data is assembled in the first place.
In legacy NEMT operations, encounter records are reconstructed after the trip. A member calls or submits a trip request through the broker's benefit management system. That trip record lives in the broker's platform. The transportation provider accepts and executes the trip in their own dispatch system. GPS data, if captured at all, lives in yet another system. When it comes time to generate an encounter file for HHSC, the broker is stitching together data from multiple sources that were never designed to connect.
Every gap in that reconstruction is a point where data degrades. A pickup timestamp that does not match the GPS record. A mileage figure entered manually that differs from what the dispatch system recorded. A prior authorization number not captured at scheduling and added retroactively. A service code misaligned with what was authorized.
None of these discrepancies require bad intent. They are structural outputs of a workflow where benefit management, trip execution, and billing are disconnected systems. But from HHSC's perspective, and from a federal audit perspective, a discrepancy is a discrepancy regardless of how it got there.
Two developments are converging that make encounter data quality a more urgent issue for Texas MCOs in 2026 than it has ever been.
The HHSC policy review of NEMT will use encounter data as a primary analytical source. The review will examine whether services were authorized, delivered, and billed in accordance with program requirements. Plans whose encounter records contain the kinds of discrepancies described above, including mismatched timestamps, manual billing overrides, and missing prior authorization documentation, will have exposure that is visible in the findings. Plans whose encounter data accurately reflects verified trip activity will not.
A recent Texas Officer of Inspector General (OIG) audit found that one of the state’s NEMT brokers failed on five specific documentation and billing standards, generating more than $515,000 in improper payments from a single health plan in a single year. The failure categories, missing prior authorizations, inaccurate billing, unauthorized cost overrides, and failure to resolve member complaints, trace directly to the data disconnects this blog describes. That audit covered one program, one plan, and one fiscal year. The question every Texas MCO should be asking is what a similar review of their program would find.
Under 42 CFR 438.242, MCO contracts must require that each managed care organization maintain a system capable of collecting and submitting complete, accurate enrollee encounter data. Under 42 CFR 438.818, states must validate encounter data for accuracy and completeness before submitting to the Centers for Medicare & Medicaid Services (CMS). If a state cannot bring its submissions into compliance after CMS notification, CMS can defer or disallow federal financial participation on MCO contracts. States in turn have authority to withhold a percentage of capitation payments from MCOs that fail to submit timely, accurate encounter data. The accountability chain runs directly from the broker's data quality to the MCO's contract and financial standing with the state.
For NEMT programs that rely on manual data assembly, that means every submission cycle is an opportunity for documentation gaps to surface. And when the HHSC policy review is examining exactly this area, the timing could not be more consequential.
Meeting HHSC's encounter data standards is not a reporting challenge. It is an infrastructure challenge. Every data point in the encounter record has to be captured and validated at the moment it occurs inside a closed-loop system, not assembled from disconnected sources after the fact.
GPS data is where that standard becomes most concrete:
GPS, captured directly from the dispatch system in real time. In a disconnected model, GPS data is extracted from a dispatch system, transferred to a broker's billing platform, and manually mapped to a trip record. At each transfer point, data can be altered, delayed, or lost. Defensible GPS data requires a connected system where status updates flow automatically from the provider's dispatch system into the trip record at the moment they occur, without manual intervention. When that connection exists, the GPS record in the encounter file is the same record the dispatch system generated. There is no gap, no manual transfer, and no opportunity for the data to be altered between execution and submission.
The same principle applies across every data point in the trip lifecycle:
Prior authorization, verified before dispatch. The authorization number is in the record before the trip happens, not added retroactively when the encounter file is built.
Provider credentials, validated at assignment. Confirmed when the trip is assigned, not looked up during billing or an audit.
Pickup and drop-off, confirmed at the point of service. Captured as the trip occurs, not reconstructed from a separate log at month-end.
Billing, drawn from the connected trip record. Generated from verified trip data, not reconciled against separate systems after the fact.
When the infrastructure connects benefit management, dispatch, and billing in real time, the encounter record is not something built at the end of the month. It is a natural output of running each trip correctly. This is what closed-loop infrastructure delivers, and it is not achievable in a model where those systems operate in isolation.
There is a version of this problem that health plan leaders sometimes assume is the broker's to solve. The broker submits the encounters. The broker manages the providers. If the data is wrong, the broker fixes it.
That assumption does not reflect how HHSC or CMS assign accountability.
The MCO holds the Medicaid contract.
The MCO submits the encounter data to HHSC.
The MCO receives the corrective action plan when encounter acceptance rates fall below threshold.
The MCO is exposed to capitation withholds when encounter data is untimely or inaccurate.
The broker is a subcontractor. The accountability stays with the plan. That gap, between who manages the data and who is responsible for it, is exactly what the Governor's directed policy review will expose for plans whose NEMT encounter data does not hold up to scrutiny.
The policy review findings will land this summer. The plans that will be positioned well are the ones that can trace every NEMT encounter record back to a verified, source-data trip: GPS-confirmed in real time, prior authorization documented before dispatch, provider credentials validated at assignment, billing drawn from the connected dispatch record rather than assembled manually after the fact.
If you are not certain your NEMT program can produce that documentation today, now is the time to close that gap.