I had the privilege of moderating a panel at ATA Nexus this week with leaders who are actually doing this work, not pitching it, not theorizing about it, building multi-site virtual care programs inside real health systems with real constraints.
A few things stuck with me.
Someone in the audience asked what the biggest barrier to scaling virtual care still is. I expected technology to top the list, or integration challenges, or the still-evolving reimbursement landscape.
The answer, across the panel, was people… their questions, their processes, their comfort with change. Ultimately, my big takeaway was this: The case for virtual care is made. The operating manual is still being written.
That is where we are in 2026. ATA confirmed it. The conversation has moved past the premise, and the field is now doing the harder work of figuring out how to build programs that actually hold.
What Changed, and What That Proves
In 2008, when Eagle entered the telemedicine space, the objections were predictable: you cannot evaluate patients without a physical exam, the elderly will not accept it, malpractice will not cover it, Medicare will not pay. Most were wrong, and the ones that were right eventually gave way as the clinical model proved itself in practice.
COVID did not validate the clinical logic of virtual care. That was already established. What the pandemic did was compress the adoption curve by roughly fifteen years. Before 2020, fewer than one percent of patients had experienced a telemedicine visit. When offices closed, care moved to wherever patients were, and the adoption that advocates had spent a decade trying to engineer happened in a matter of weeks. By the time normal operations resumed, the institutional memory of “we don’t do it this way” had been significantly disrupted.
That shift is proving durable in meaningful places. Behavioral health largely never returned to in-person delivery. Payment policy has evolved, with continued uncertainty but real structural progress. The legitimacy question has been settled. No one at ATA 2026 was debating whether virtual care works. The industry has moved on from that.
What Hasn’t Changed
The pandemic acceleration did not resolve the structural failures it exposed.
Rural hospitals are still operating without specialists. Physicians are still burning out and leaving medicine earlier than they should. Patients in rural communities are still making hundred-mile drives for care that technology could bring to them. Tight operating margins still make every new expenditure a difficult conversation, even when the value case is clear. These were the problems telemedicine was designed to address in 2008. They remain unresolved in 2026.
The conference surfaced specific gaps worth naming. Rural health leaders acknowledged they are leaving federal funding on the table: reimbursement mechanisms, grant programs, and support structures that could make virtual care more financially viable than their current models reflect. CPT codes for asynchronous e-visits and physician-to-physician e-consults are another underutilized area. The reimbursement pathways exist. Adoption has not caught up. That gap between what the system makes possible and what health systems are actually capturing is where a meaningful portion of access improvement is still sitting unused.
The case for virtual care may be made. The infrastructure to fully use it is not.
What I Heard from the Stage and the Floor
The panel’s clearest consensus: programs that scale successfully establish governance first. Not governance as process documentation or compliance overhead. Governance as operational infrastructure. A framework with enough structure to hold across sites, and enough flexibility for a rural critical access hospital and a 400-bed tertiary center to operationalize the same program differently. That balance is harder than it sounds. Get it wrong in either direction and you either have chaos at the edges, or you have built something that only works for your flagship site and cannot replicate.
In other words, governance is not a bureaucratic exercise. It is the work.
Change management is where programs actually live or die. The leaders on that stage were direct about this: resistance is predictable, and the teams that succeed identify it early, name it honestly, and address it head-on. The failure mode is not the technology. It is the human infrastructure that was not built before go-live.
The cliche that keeps being true: people before process and technology. Go slow to go fast. Gather your stakeholders early. Do the alignment work before you build the workflow. It sounds obvious until you are six months in and trying to retrofit buy-in into a program that already went live without it.
Champions and early adopters are not peripheral. They are the mechanism. They carry a program across the gap between “we went live” and “this is how we work now.” Losing a champion mid-rollout is a different category of risk than a technology failure. It deserves the same level of contingency planning.
Hospital leaders want to own care delivery. That instinct is right, and it is also incomplete. One of the most honest conversations I heard at ATA was about the pull toward in-house solutions. It makes sense: hospitals want control, accountability, and continuity. But managing virtual care across specialties and facilities at scale is a different operational problem than most health systems have solved before. The complexity compounds quickly.
The path forward is not in-house versus partner. It is hybrid, combining what health systems do exceptionally well with what a partner with breadth, depth, and years of subspecialty experience brings to the table. That combination, built on a real governance foundation, is what actually moves the needle on access and outcomes. Scale alone is not a differentiator anymore. You cannot bolt on growth and call it a virtual care program.
Where This Is Going
Our Chief Physician Executive, Dr. Mac McCormick, framed something on stage that I keep returning to during his session Telemedicine Then, Now, and Next: Lessons from the Frontlines. For most of medicine’s history, care required a doctor and a patient to be in the same place at the same time. Telemedicine broke the spatial constraint. What is emerging now is a further disruption of the time constraint: asynchronous e-consults, physician-to-physician consultations that do not require real-time interaction, and care workflows that flex across time zones and shift structures in ways synchronous telemedicine cannot fully address. That evolution in how care is delivered is already underway. The CPT structure is catching up to it.
Then there is AI.
The AI conversation at ATA was fixated and unresolved, which is about where you would expect it to be given where the industry actually is. We are getting calls from hospitals asking us to ensure our physicians are not using AI in clinical decision-making: they have not established a governance framework for it, they cannot yet account for how it performs across their patient populations, and they are not prepared to own the accountability it introduces. We are also getting calls from hospitals that have adopted specific tools and want us to incorporate them into our clinical workflows. Both responses are reasonable given the uncertainty, and both reflect the same underlying gap: the governance infrastructure for AI does not yet exist at most organizations.
Virtual care at scale is not a technology problem with some change management added on top. It is an organizational design problem that technology enables, when everything else is in order.
That framing applies directly to AI. AI is not a technology challenge with governance bolted on afterward. It is an organizational design question that technology can serve, when the governance infrastructure is already in place. The organizations building that infrastructure now are not doing administrative work. They are building the capability that determines who will be able to deploy AI responsibly across a distributed physician network, and who will not.
Governance is the competitive advantage. It was true before AI entered the conversation. It is more true now.
ATA 2026 confirmed what the pandemic set in motion: the industry has stopped debating the premise and started doing the harder work.
What is still being written is the operating manual. The governance structures, the change management frameworks, the hybrid architectures that allow health systems to own their care delivery while accessing the subspecialty depth and operational infrastructure they cannot build alone at the speed the access problem demands.
For the leaders doing that work inside real systems with real constraints, ATA offered something useful: you are not alone in it, and the people building the programs that actually hold are asking the same questions you are.