A two-year telehealth runway is long enough to build on and too short to bet on

The week

Medicare’s telehealth flexibilities, the temporary rules that since the pandemic had let patients see their doctors by video or phone from home, were about to run out again in the week ending January 30, 2026. For patients and clinicians, it had a familiar, uneasy feel.

They had already lapsed once, during the fall 2025 government shutdown, and been restored only through January 30, 2026. At the end of that week they expired again. So did the Acute Hospital Care at Home waiver, the permission that lets hospitals treat certain acutely ill patients in their own homes, with daily clinician visits and remote monitoring, and be paid as if those patients were in a hospital bed. The lapse coincided with a brief partial government shutdown that began on January 31.

Under Medicare’s permanent rules, most telehealth is tightly restricted, often to patients in rural areas who travel to a clinic to connect with a distant doctor. The pandemic flexibilities removed those limits. They let patients be seen from home, allowed audio-only visits by phone for people without video, and widened the list of practitioners who could bill for telehealth. Each time they lapse, clinics must decide whether to cancel virtual appointments, convert them to in-person visits, or keep going and risk not being paid.

This time, the uncertainty lasted only days. On February 3, the President signed the Consolidated Appropriations Act, 2026, the government funding law. It extended the telehealth flexibilities, including coverage for patients at home, audio-only visits and a broader set of practitioners, through December 31, 2027.

It went further for hospital at home. The waiver was extended through September 30, 2030, and the law funded a study comparing that model with traditional inpatient care. It also allowed cardiopulmonary rehabilitation, the supervised exercise and education programs for heart and lung patients, to be delivered at home through 2027. And it directed CMS, the federal agency that runs Medicare, to create billing codes or modifiers, the labels attached to claims, for services delivered through contracted third-party telehealth platforms.

After years of extensions measured in weeks or months, health systems suddenly had dates far enough away to plan around.

Serial short extensions cost health systems in scheduling, billing and patient trust. Every few months, staff had to plan for a cliff, warn patients, prepare to reschedule visits and then, often at the last minute, stand down. The February law changed that planning problem. For the first time since the public health emergency ended, Medicare telehealth had a horizon measured in years rather than weeks, and hospital at home had one of nearly five years.

That is enough time to justify real investment, though not enough to assume permanence. A health system leadership team will be tempted to treat the law as either a green light (telehealth is here to stay, so scale it) or another reprieve (nothing has really changed, so keep everything provisional), and both readings are wrong.

The useful response is to sort virtual services by the legal footing they stand on, because each footing supports a different kind of commitment. There are three: services that rest on permanent Medicare rules, services that depend on the end-2027 extension, and hospital at home on its 2030 horizon. First, what these services actually are.

Concrete floors and scaffolding

Telehealth is care delivered at a distance: a video visit with a physician, a phone check-in with a nurse, a therapy session over a secure app. Technically it is simple. What makes it complicated is payment. Medicare pays for a telehealth visit only if the law says that type of visit, by that type of practitioner, to a patient in that location, is covered.

Some telehealth services stand on concrete, written permanently into law; others stand on scaffolding, temporary flexibilities with an end date. Building a permanent structure on scaffolding is a risk, however sturdy the scaffolding looks today.

Audio-only visits are conducted by phone without video. They matter for older patients, people with poor internet access and people uncomfortable with video. Remote patient monitoring uses connected devices, such as blood pressure cuffs, scales or glucose meters, that send readings to a care team, and it is billed separately from telehealth visits.

Hospital at home is the most ambitious model. A patient who would otherwise be admitted, say with pneumonia or heart failure, is treated at home instead. Nurses and physicians visit in person and virtually, monitoring devices send vital signs to a central command center around the clock, and the hospital arranges meals, medications, imaging and transport.

Finally, third-party telehealth platforms are outside companies that provide clinicians, software or both under contract to a health system. The new requirement for billing codes or modifiers means Medicare will be able to see which claims came through these arrangements.

How a telehealth rule change travels through a health system

From the outside, a telehealth extension looks like a single decision in Washington. Inside a health system it touches dozens of people and systems, which is why short extensions have been so costly.

Start with scheduling. When a patient books a video visit, the scheduling system has to know whether that kind of visit will be paid on the date it happens. If a flexibility is due to expire, schedulers must either stop booking virtual visits past the date or book them with a warning attached. Many health systems spent the past two years doing this by hand, with spreadsheets and email reminders, because their systems were not built to change the rules by date.

Then comes the visit itself. The clinician must document where the patient was, whether video or audio was used, and why. Those details are not paperwork for its own sake. They determine how the claim is coded and whether it is paid.

Next is billing. A claim for a telehealth visit carries specific codes describing the service, the location and the technology used. If the rules changed between the visit and the claim, billing staff must decide whether to submit, hold or write off the claim. Holding claims delays cash. Submitting claims that turn out to be ineligible creates repayment risk later.

Finally there is the patient. Every lapse means calls to explain that a visit must move, be converted to in-person or be rescheduled. For a patient with limited mobility or a long drive, that can mean the visit simply does not happen.

Take a community clinic that runs about a fifth of its follow-up visits by video or phone. In the week before a lapse, its front desk spends hours calling patients, its billing team freezes claims, and its physicians are unsure whether to keep their virtual sessions. When Congress restores the flexibility days later, the clinic must unwind every one of those steps. Multiply that by several lapses and the hidden cost of uncertainty adds up. A horizon of nearly two years removes most of that churn, if the health system uses the time to automate the date logic rather than return to spreadsheets.

Services on permanent footing

Some Medicare telehealth is not temporary. Congress made home-based telehealth for behavioral and mental health permanent in 2020 legislation, and telehealth for substance use disorder treatment from the home was made permanent earlier, in the 2018 SUPPORT Act. The main open condition is the in-person visit requirement for telemental health, a rule that patients receiving mental health care by telehealth must also be seen in person periodically. Congress has repeatedly delayed it, most recently to the start of 2028.

These services can carry long-term commitments. A health system can build dedicated virtual behavioral health teams, sign multi-year platform agreements and recruit clinicians into remote roles with reasonable confidence. The operational work is to prepare for the in-person requirement rather than hope it disappears. That means tracking when each telemental health patient was last seen in person, building scheduling rules that trigger an in-person or exception review before the date, and documenting the clinical reasons when an exception applies. Systems that build that tracking now avoid a scramble in late 2027.

Workforce planning follows the same logic. Behavioral health clinicians are among the hardest roles to recruit, and many have chosen remote or hybrid work. Offering them stable virtual positions is a recruitment advantage the permanent rules support. The constraint is licensure and supervision, not Medicare payment, so the planning effort belongs with credentialing and clinical leadership as much as with finance.

Services on the 2027 date

Most other Medicare telehealth depends on the extension. That includes general medical visits to patients at home, audio-only care outside behavioral health, and services by physical and occupational therapists and other added practitioner types. These services have twenty-three months of footing from the February signing. That supports investment that pays back within the window or survives a lapse.

Three kinds of investment pass that test.

• Infrastructure that serves any channel. Integrated video inside the electronic health record, remote patient monitoring that is separately reimbursable, and scheduling and billing logic driven by date-effective rules, meaning systems that know which rules apply on which date and switch automatically.
• Clinical models that would still work in a reduced form. Virtual follow-up visits, for example, that could convert to clinic visits without redesign.
• Staffing that stays flexible. Clinicians whose schedules mix virtual and in-person sessions, rather than fully remote roles that depend on home-based Medicare visits.

Commitments that fail the test include long platform contracts sized to today’s Medicare telehealth volume, and service lines whose margin depends on audio-only visits for non-behavioral care. Those are bets on another extension.

The new requirement for codes identifying third-party platform services also matters here. Health systems that contract with outside telehealth companies should expect those encounters to become more visible in claims data. They should check that their agreements cover documentation, quality reporting and audit support, because a claim submitted under the health system’s name is the health system’s responsibility.

The window also needs a decision date. A health system that waits until December 2027 to learn whether Congress acts will repeat the scramble of the past two years. A better practice is to set an internal review in mid-2027, with a defined contingency for each service on this footing: continue, convert to in-person, or move to a permanent pathway such as remote monitoring. That review should include Medicare Advantage and commercial contracts, which often follow their own telehealth terms and may soften the effect of a Medicare lapse.

Hospital at home on a 2030 horizon

The nearly five-year hospital-at-home extension is the biggest change in planning terms. A program that serves acutely ill patients at home needs capital and contracts: monitoring equipment, logistics partners for meals, supplies and imaging, paramedic or nursing services, pharmacy delivery and round-the-clock command center staffing. Earlier short extensions made those commitments hard to justify, because the waiver could expire before a program matured.

A horizon to September 2030 changes the payback calculation. It is long enough to recruit and train a dedicated team, negotiate multi-year logistics agreements, and build patient selection protocols that improve with experience. The congressionally funded study comparing quality and cost with inpatient care also raises the stakes. Programs that collect clean outcome, readmission and cost data will shape what Congress does next. Programs that cannot show their results will be poorly placed in that debate.

Hospital at home also has workforce consequences beyond the program. It draws on the same nurses, pharmacists and transport capacity as inpatient units. Planning should treat it as part of total bed capacity, with explicit decisions about which patients and service lines it serves, rather than as a pilot running alongside the hospital.

One health system, three footings

Suppose a mid-sized health system with three hospitals and a large outpatient network. After the February law, its leadership team lists every virtual service and sorts it.

Its virtual psychiatry and addiction medicine programs sit on permanent footing. The system decides to grow them, offers its therapists permanent remote roles and signs a multi-year agreement for its behavioral health platform. It also asks its informatics team to add a field to every telemental health patient’s record showing the date of their last in-person visit, with an alert months before the 2028 requirement takes effect.

Its primary care video visits, physical therapy telehealth and audio-only chronic disease check-ins depend on the 2027 date. It keeps investing in the video tools built into its electronic health record, because those serve every channel. It declines a vendor’s offer of a five-year contract priced on current Medicare telehealth volume, choosing a shorter term instead. It schedules a formal review for mid-2027.

Its small hospital-at-home pilot now has a horizon to 2030. The system decides to scale it, signs multi-year agreements with a home logistics partner and a mobile imaging provider, and builds a reporting dashboard for outcomes, readmissions and cost from the first patient onward.

Not one of these decisions requires predicting what Congress will do. Each matches the length of the commitment to the length of the legal footing.

Who loses when a flexibility lapses

The stakes go beyond the virtual care team. For finance, telehealth revenue that disappears overnight leaves fixed costs, such as platforms and staff, uncovered. For risk and compliance, claims submitted after a flexibility lapses can create repayment exposure. For the workforce, clinicians recruited into remote roles on the promise of stability will leave if the rug is pulled. For patients, especially older adults and people in behavioral health treatment, an abrupt switch back to in-person care can mean missed appointments and interrupted treatment. And for reputation, a health system that cancels hundreds of visits on short notice is remembered for it.

There is also a strategic dimension. Virtual care changes where a health system competes. A patient who can see a specialist by video is no longer limited to the hospital down the road, and national telehealth companies compete for the same patients and the same clinicians. A health system that pulls back every time a deadline looms cedes ground to competitors who planned around the dates. One that matches its commitments to the footing can invest with confidence where the law is stable and stay nimble where it is not.

Finally, regulators and auditors will look harder at telehealth claims as the volume grows and as the new third-party codes make those claims easier to identify. A health system that can show clean documentation, clear location and technology coding, and a record of how it handled each lapse is in a far stronger position than one that improvised.

Owners, dates and a one-page board view

The target is simple to describe. Every virtual service is tagged with its legal footing and end date. Contract terms and staffing commitments match those dates. Systems apply date-effective rules automatically. A mid-2027 review is already on the calendar. Hospital-at-home programs collect clean data from day one.

It also needs ownership. Each footing needs a named executive who is accountable for watching its date, maintaining the contingency plan and reporting to leadership as the deadline nears. Without an owner, dates drift until they become emergencies. Communication matters too: patients and clinicians who are told plainly which services are stable and which depend on a future decision are far less likely to be blindsided.

The board has a role as well. It does not need to understand billing codes, but it should see, at least once a year, a one-page view of virtual care revenue and staffing by legal footing, with the relevant end dates. That turns a regulatory detail into a strategic exposure the board can actually oversee.

The usual errors are treating telehealth as one program with one plan, signing long contracts sized to temporary volume, forgetting the in-person requirement for telemental health until it arrives, and running hospital at home as a side project that competes with inpatient units for the same nurses without anyone deciding how to share them.

Which commitments outrun their dates

• Which of our contracts or staffing commitments run past the date their underlying flexibility expires?
• How will we know, for each telemental health patient, when they were last seen in person?
• What is our plan for each telehealth service if the 2027 date passes without another extension, and when will we decide?
• Is our hospital-at-home program collecting the outcome, readmission and cost data that will matter when Congress revisits it?
• Who decides how nurses and pharmacists are shared between hospital at home and our inpatient units?

With hindsight

CMS reportedly treated the brief January lapse as if it had not happened for claims purposes, which limited the damage of that week. Visits delivered during the gap were not left stranded, and the scramble many health systems had prepared for did not materialize.

The end-2027 date remains the next Medicare cliff for most telehealth. A nearer clock also deserves attention. The DEA and HHS extended telemedicine flexibilities for prescribing controlled substances only through December 31, 2026. Controlled substances are medicines with potential for misuse, such as many pain, anxiety and ADHD medications and some addiction treatments, and prescribing them remotely runs on a separate set of rules from Medicare payment. That question returns more than a year before Medicare’s, and it touches some of the same behavioral health programs that otherwise sit on permanent footing.

Sorting by legal footing still looks more useful than a single telehealth plan. Behavioral health programs could commit for the long term, provided they tracked the in-person requirement. General medical telehealth stayed worth investing in where the infrastructure would serve any channel. Hospital at home gained a horizon long enough to build real programs, and the funded study gives well-run programs a chance to shape what comes after 2030.

Health systems that treated the extension as permanent, or as another short reprieve, misjudged one of the three. Regulatory dates are planning inputs, not background noise. Each virtual service should carry its own date, owner and fallback, and the calendar should hold not one cliff but several: the end of 2026 for remote prescribing of controlled substances, the end of 2027 for most Medicare telehealth, the start of 2028 for the telemental health in-person requirement, and September 2030 for hospital at home. A leadership team that can name all four, and the services that depend on each, is ready for whatever Congress decides.