The Hybrid Network Era

Why the future of healthcare mobility is about orchestration, not just vehicles

This is Part 2 of a three-part series on the evolution of healthcare mobility. [← Part 1: The Rideshare Revolution Comes to Healthcare] [Part 3: When the Driver's Seat Is Empty →]

Picture a typical morning at a PACE center in Southern California. Mrs. Alvarez, who uses a wheelchair and has early-stage dementia, is picked up at 7:45 by a dedicated fleet driver she's known for two years — someone trained to wait patiently at her door and gently remind her where they're going. James, a 28-year-old in an I/DD day program, is picked up at 8:10 by a contracted NEMT provider in a vehicle tagged to his specific Regional Center billing code. And Mr. Tran, ambulatory and cognitively sharp, gets a Lyft to his 9:00 cardiology appointment across town.

Three passengers. Three different provider types. Three different sets of capabilities, credentialing requirements, and dispatch systems. One program responsible for making sure every ride is safe, appropriate, and on time.

This is what a hybrid network looks like in practice. And it's now the operational reality for virtually any program serving older adults or people with disabilities at scale.

What "hybrid" actually means

As we discussed in [Part 1 of this series], the combination of TNC integration and pandemic-era disruption pushed healthcare transportation programs toward diversified supply. But "diversified" understates what's actually happening. Today's healthcare mobility programs draw from a remarkably varied pool of transportation resources.

There are in-house and dedicated fleet vehicles — wheelchair-accessible vans, stretcher vans, sedans — staffed by drivers who are trained in the specific needs of the populations they serve. There are contracted NEMT providers, some dispatched through transportation brokers, others engaged directly. There are TNC partners: Lyft now operates NEMT in 27 states and the District of Columbia, covering 70% of the Medicaid population and works with all 10 of the largest U.S. health systems and all 10 of the largest NEMT brokers. There are volunteer driver programs, especially important in rural areas where professional transportation supply is thin. There's public transit — both fixed-route and paratransit — for passengers who can use it. And there are specialized providers for bariatric transport, behavioral health, or other high-acuity needs.

On the brokerage side, consolidation has accelerated. MTM acquired Access2Care in October 2024, expanding to 8.6 million annual rides across all U.S. states. Modivcare serves 30+ states. These large brokers are themselves managing multi-provider networks, subcontracting to dozens or hundreds of local operators within each state.

The result is a transportation ecosystem of considerable complexity. And the programs that are thriving in this environment are the ones that have stopped thinking about individual provider contracts and started thinking about network orchestration.

Why orchestration is so hard

Adding a TNC to your provider network is relatively straightforward. Managing a hybrid network so that every passenger consistently gets the right ride — that's where things get difficult.

The first challenge is technological fragmentation. Different providers use different dispatch platforms, different data formats, different communication channels. Despite years of progress, interoperability and real-time data exchange remain core gaps across broker-administered networks. Many operations still rely on spreadsheets, manual status updates, and disconnected driver files.

The second challenge is quality consistency. A TNC driver and a dedicated fleet driver may both be picking up passengers within the same program, but they bring vastly different levels of training, accountability, and situational awareness. The TNC driver doesn't know that the passenger has a seizure disorder. The dedicated driver does — because that information lives in a system the TNC never touches.

The third challenge is visibility. When a trip is handed to a third-party provider, the referring program often loses real-time insight into what's happening. Is the vehicle on time? Did the passenger board safely? Was there an incident? The answer might live in the provider's system, but it doesn't flow back automatically.

And the fourth challenge is cost efficiency under growing complexity. The numbers paint a sobering picture: across 16 key metropolitan areas, paratransit productivity has decreased 13% since 2015 while cost-per-trip has risen 37.5%. Programs that eliminated shared rides during COVID saw operating miles increase 50 to 70 percent. More providers in the network doesn't automatically mean better outcomes — not without the operational infrastructure to manage them as a single system.

What skilled orchestration looks like

The organizations getting this right share several common capabilities, even if they implement them differently. These aren't product features — they're operational principles that define what it means to run a modern healthcare mobility program.

The first is a unified passenger record that travels with the person across every provider and every trip. This means more than a name and an address. It means mobility level, equipment requirements, behavioral considerations, communication preferences, medical alerts, caregiver contacts, and the kind of experiential detail that matters enormously in practice — that this passenger needs two extra minutes to board, that one should never be paired with loud co-riders, that another has a toileting need that imposes a hard time-on-vehicle constraint. When any provider in the network picks up that passenger, they should have access to what they need to serve them well.

The second is intelligent trip matching that goes beyond availability and cost. A system that assigns the cheapest available vehicle to a trip isn't optimizing — it's gambling. Real optimization means enforcing constraints: this passenger requires a WAV with rear securement. This one needs same-driver consistency for behavioral stability. This route can't exceed the state-specific time-on-vehicle maximum (which varies by state and sometimes by client). That pair of passengers can't ride together. Every one of these constraints is a patient safety decision, and the system has to honor all of them simultaneously.

The third is real-time, multi-provider visibility — a single operational view where dispatchers can see every vehicle, every trip status, and every exception, regardless of which provider is fulfilling the ride. When a TNC driver cancels, the dispatcher needs to know immediately and reassign from another part of the network. When a dedicated fleet vehicle is running behind, downstream pickups need to adjust. This kind of cross-network awareness is what turns a collection of providers into an actual system.

And the fourth is a feedback loop that drives continuous improvement. Trip data — on-time rates, passenger experience, incident reports, no-show patterns — should flow back from every provider type into a common data layer. That data becomes the basis for provider performance management, for identifying passengers whose needs have changed, and for optimizing network composition over time.

The technology to support these capabilities is maturing. AI-powered demand forecasting now achieves 95%+ accuracy in predicting transportation needs. AI-driven route optimization is reducing empty miles by up to 40% and fuel costs by 15 to 30 percent. Providers adopting these approaches report 20-30% operational cost reductions and driver utilization rates above 85%. The tools exist. The question is whether programs are building the operational layer to use them.

Expanding TNC use cases — responsibly

With better orchestration infrastructure in place, programs can begin asking a bolder question: which trips that currently require traditional NEMT could be safely and appropriately served by TNCs?

The policy environment is creating room for this. Transportation supplemental benefits are now offered by 24-30% of individual Medicare Advantage plans, and that number jumps to 67-85% for Special Needs Plans. Under the Value-Based Insurance Design model, 100% of VBID plans are now required to offer supplemental benefits in at least two of three social-needs categories — food, transportation, and housing. The demand signal is clear: payers want more transportation options, and they want them to reach more beneficiaries.

But responsible expansion means having clear guardrails. It means clinical criteria for which trips can safely use TNC fulfillment versus requiring higher-acuity transport. It means establishing driver training requirements that go beyond the TNC baseline for healthcare rides, even if they're lighter than full NEMT credentialing. It means feedback loops where passenger experience data from TNC trips flows back into the system of record. And it means escalation protocols for when a TNC trip reveals something new — when the driver reports that the passenger seemed confused, or needed more assistance than expected, or couldn't manage the vehicle steps.

Programs that do this well don't think of TNCs as a separate channel. They think of them as one supply source within a unified network, governed by the same passenger-need logic that governs everything else. The passenger's profile determines the ride — not the other way around.

The system of record underneath it all

There's a concept quietly taking shape across the healthcare mobility industry, even if it doesn't yet have a universally agreed-upon name. Call it the transportation system of record: a comprehensive data layer that captures who every passenger is, what every provider can do, and what constraints govern every trip — and uses that information to make the right match, every time, across the full network.

This is the infrastructure that makes hybrid networks work. Without it, adding more provider types creates more risk, not less — more chances for a mismatch between a passenger's needs and the vehicle or driver that shows up. With it, every new supply source genuinely expands access.

And this infrastructure is about to matter even more. Because the next wave of transportation supply — the one already arriving in pilot programs across the country — doesn't come with a driver at all.

Next in this series: [Part 3: When the Driver's Seat Is Empty →] — What autonomous vehicles mean for people who need more than just a ride.