The doctor one screen away
In 2022 one emergency department in thirteen lacked an attending physician around the clock. Every rural one should have an emergency physician on video at all hours, paid for by Medicare rather than grants.
Listen to this special topic
read by an AI voiceDeaths after stroke ran higher at the seven community emergency departments that could reach a neurologist only by telephone, if at all, in a study of 52 departments published in July. Risk-adjusted death rates a month after stroke were nearly two percentage points higher than where a neurologist came in person. By video, the gap could not be told from chance. The study compared hospitals, not patients, and cannot show cause. It asks the right question of rural emergency care. Can a physician see the patient?
The three-bed emergency room in Ekalaka, Montana, two hours' drive from the nearest hospital with more advanced services, has not had a physician in at least 30 years, a physician assistant there, Carla Dowdy, told KFF Health News last year. When a patient arrived in June 2025 needing more than the room could have given even with a doctor, Ms. Dowdy called for a plane and called out drugs and doses while a paramedic and nurses gave them. She can phone a physician, a doctor across the state reviews every chart, and the emergency room was working on virtual reality glasses that would let remote physicians see what its staff see. It is not enough. Every rural emergency department should have an emergency physician on video at every hour, paid for by Medicare as standing care rather than by grants that run out.
Ekalaka is an extreme case of a common gap. In a national survey, one emergency department in thirteen reported no attending physician on duty around the clock in 2022, and in the Dakotas more than half did. Nine in ten of those were in critical access hospitals, which federal rules let rely on an on-call clinician, not necessarily a physician, who can arrive within half an hour, or an hour in frontier country. Rural emergency hospitals, which no longer admit patients for acute care, need someone competent in emergency care on site at every hour, the same kind of on-call clinician as a critical access hospital, and a physician reachable by radio, telephone or electronic link.
The evidence is better on speed and judgment. At 14 rural Midwestern hospitals, patients for whom the remote service was called met a clinician six minutes sooner than matched patients, mostly because the remote clinician came first, though they stayed longer in the department. In a randomized trial in 15 rural and community emergency departments in northern California, video rather than telephone consultation about acutely ill children meant fewer transfers. The sharpest result came from 29 rural Australian departments without emergency specialists. A suspected heart attack set off an automatic alert to a referral hospital's coronary care unit, which read the tracing and blood test remotely and phoned the local doctor. Usual care missed about one in three of the heart attacks that need an artery opened at once. The alert system missed none.
On whether patients live longer, evidence is thin. The Australian trial was too small to settle it. In 23 hospitals of an upper Midwestern network that already offered a board-certified emergency physician by video around the clock, patients with sepsis seen through it did no better on days out of hospital or on deaths than matched patients. In Medicare claims for sepsis, patients at hospitals with the service had higher total costs of care and similar outcomes. For patients treated on site only by nonphysician clinicians, a subgroup named in advance, the 23-hospital study suggested fewer deaths with video. It is a reason to measure, not yet a finding.
Software is sold to fill the same gap. Since 2018 the Food and Drug Administration has allowed an app from Viz.AI that reads a brain scan for a blocked artery and texts a specialist. A randomized trial of the company's software at four Houston stroke centers, whose staff the company trained, found it sped the start of clot removal by about 11 minutes, but excluded transfers from other hospitals, the patients a small emergency room sends. The Australian system that worked was an alert that summoned a person. An alert is only as good as whoever answers it, and ten departments with no physician on duty could not reach any physician by voice at every hour.
A payer could read the sepsis studies and refuse. Why pay every month for a service that raised costs and showed no fewer deaths? Fair, as far as it goes. But sepsis is one condition, and the two trials, of video consultation for sick children and of remote review of heart tracings by telephone, found fewer transfers and fewer missed diagnoses. Neither tested round-the-clock video access to an emergency physician. Standby is the condition for treatment, not a treatment to be proved like a drug, and Medicare already pays for it. A critical access hospital can include the cost of a clinician on call who can arrive within 30 or 60 minutes, and a rural emergency hospital gets about $295,000 a month this year for its operation and services, telehealth among them. Neither requires an emergency physician.
So pay for the standby itself. Smaller rural emergency departments are, researchers wrote in 2021, "the most likely to benefit from telemedicine, but the least likely to have it." Departments with it mostly carry the subscription as overhead. Grants fill the gap. In May Kansas awarded $5 million of federal money to a hospital alliance that, with partners, says it will expand emergency and other telehealth to more member hospitals. The federal money behind it is available through fiscal 2030. Medicare should make round-the-clock video access to a board-certified emergency physician a condition for critical access and rural emergency hospitals, with Congress where the law requires, and pay for it every month at every rural hospital. The networks should report how often they are called, how fast a physician appears and what becomes of the patients.
Ekalaka may never have a doctor in the building. It can have one who sees the patient at every hour, not only one who hears about them by telephone. Whether that saves lives in rooms like Ms. Dowdy's is unproven. Paying for it properly is how anyone will find out.
- The emergency room in EkalakaKFF Health News, "Experts Say Rural Emergency Rooms Are Increasingly Run Without Doctors," by Arielle Zionts, Aug. 12, 2025, datelined Ekalaka, Montana. It describes Dahl Memorial Hospital's three-bed emergency department, in a town of 400 residents, as a two-hour drive from the closest hospital with more advanced services, staffed by physician assistants and nurse practitioners. Physician assistant Carla Dowdy said the hospital has not had a physician in at least 30 years; staff can call physicians with questions, a doctor on the other side of Montana reviews all patient treatment notes, and the emergency room is working on virtual reality glasses for remote physicians. Patient numbers average one or two a day. The article reports, from the college's president, that the American College of Emergency Physicians is asking the federal government to require critical access and rural emergency hospitals to have physicians on-site or on call day and night.Source: KFF Health News
- The national survey of physician coverageCamargo, Boggs, Sullivan, Espinola, Swanton and Fletcher, "Lack of 24/7 Attending Physician Coverage in US Emergency Departments, 2022," JACEP Open 2025;6(2):100050, published Feb. 5, 2025. The 2022 National ED Inventory-USA survey, administered in 2023, asked every nonfederal ED director whether at least one attending physician was on duty 24 hours a day. Of 5,622 EDs, 4,621 (82%) answered; 344 (7.4%) said no. The highest shares were North Dakota (58%), South Dakota (56%) and Montana (46%). Of the 344, 318 (92%) had fewer than 10,000 visits a year, 307 (89%) were in critical access hospitals and 248 (72%) were rural; 266 (77%) received telehealth and 78 (23%) did not. Half lacked 24/7 two-way voice contact with any physician inside the hospital, 19% with any physician outside it, and 10 (3%) with any physician at all. Data are self-reported. The authors report no conflicts.Source: JACEP Open (PubMed Central)
- What the federal staffing rules requireUnder 42 CFR 485.618(d), a critical access hospital must have a physician, physician assistant, nurse practitioner or clinical nurse specialist with emergency training or experience on call, immediately available by telephone or radio contact, and on site within 30 minutes, or 60 in qualifying frontier or remote areas. Under 485.618(d)(3), a registered nurse may temporarily meet the requirement in a frontier or remote hospital of no more than 10 beds, on the governor's request; under 485.618(f), the hospital must, in coordination with the area's emergency response systems, arrange for a physician to be immediately available by telephone or radio contact around the clock to receive emergency calls. Under 42 CFR 485.528, a rural emergency hospital's emergency department must be staffed at all hours by someone competent in the skills needed to address emergency medical care, with a doctor available by radio, telephone or electronic communication; CMS guidance adds that it must have a physician, physician assistant, nurse practitioner or clinical nurse specialist with training or experience in emergency care on call at all times, immediately available by phone or radio and on site within 30 or 60 minutes. CMS says these hospitals were critical access hospitals or rural hospitals of no more than 50 beds on Dec. 27, 2020, provide no acute inpatient care other than post-hospital extended care, and receive a CY 2026 monthly payment of $295,051.54 after sequestration, rising to $304,492.86 in CY 2027; CMS lists telehealth among the services the payment may support.
- Two randomized trialsMarcin and colleagues ran a cluster-randomized crossover trial from Nov. 18, 2015 to March 26, 2018 in 15 rural and community EDs in northern California, enrolling 696 acutely ill children aged 14 or younger; EDs were assigned to telemedicine or telephone consultation with pediatric critical care physicians. Of 537 children assigned to telemedicine, 251 (46.7%) received it. Risk of transfer was lower with telemedicine (intention to treat RR 0.93, 95% CI 0.88 to 0.99; per protocol RR 0.79) (JAMA Network Open, 2023). In the MORACS cluster-randomized trial (JAMA Cardiology, 2022), 29 rural Australian EDs with no emergency medicine specialists were randomized; triage of suspected acute coronary syndrome triggered an automated notification to a tertiary coronary care unit, which reviewed the ECG and troponin remotely and phoned the treating physician. Missed STEMI: 27 of 77 (35%) with usual care, 0 of 46 with MORACS. Primary reperfusion among eligible patients: 64% against 100%. Twelve-month mortality 10.3% against 6.5% (RR 0.64, 95% CI 0.18 to 2.23).Sources: PubMed, pediatric trial, PubMed, MORACS
- The Midwestern network: speed, use and outcomesA matched cohort study at 14 rural hospitals of a Midwestern telemedicine network (2,857 telemedicine encounters) found door-to-provider time 6.0 minutes shorter, the telemedicine provider first in 41.7% of encounters, and total ED stay 40.2 minutes longer (Telemedicine and e-Health, 2018). A propensity-matched cohort of 1,191 sepsis patients (August 2016 to June 2019) at 23 rural hospitals subscribing to Avera eCARE, which offers 24-hour audio and video consultation with a board-certified emergency physician, found tele-ED cases far more often transferred (88% against 8%), with no difference in 28-day hospital-free days or in-hospital mortality (aOR 0.51, 95% CI 0.16 to 1.60). In an a priori subgroup treated only by advanced practice providers, analyzed unmatched, mortality was suggested to be lower with tele-ED (aOR 0.11, 95% CI 0.02 to 0.73), which the authors call hypothesis-generating (Annals of Emergency Medicine). A Medicare claims study found 6.7% higher costs for sepsis patients at tele-ED-capable hospitals and similar outcomes (Academic Emergency Medicine, 2024).
- Telephone, video and triage softwareShi and colleagues (Journal of Stroke and Cerebrovascular Diseases, July 2026) linked 52 community EDs in the American College of Emergency Physicians Emergency Quality Network Stroke Collaborative (2022 to 2024) to CMS risk-standardized 30-day stroke mortality. Neurologists were available in person at 48.1%, by telestroke at 38.5% and by phone or not at all at 13.5% (about seven departments); mean mortality was 13.1%, 14.0% and 15.2%. Phone-only access carried an adjusted difference of +1.7% (95% CI 0.32 to 3.08) against in-person; telestroke did not differ significantly (cross-sectional). On Feb. 13, 2018 the FDA permitted marketing, by De Novo review, of the Viz.AI Contact application, which texts a neurovascular specialist about a suspected large vessel blockage on brain CT for the specialist to review. A stepped-wedge cluster-randomized trial of Viz.AI's software at 4 comprehensive stroke centers in one Houston-area health system (2021 to 2022), with vendor staff training each site, found an 11.2-minute reduction in door-to-groin time; outside transfers were excluded, and functional independence did not differ in exploratory analysis. PubMed indexes the trial as supported by the National Institutes of Health and by sources outside the U.S. government.
- How tele-emergency is paid forZachrison, Richard and Mehrotra (JAMA Health Forum, Aug. 3, 2021) write that payers such as Medicare reimburse the remote consultant, that few ED telemedicine consultations are billed and most EDs pay subscription costs as overhead, and that a 2015 study found an average annual subscription fee of $60,000 plus $3,000 to $8,000 a year for connectivity and maintenance. In a 2016 national survey, 453 of 977 rural EDs (46%) did not use telemedicine; cost was the most commonly cited reason (37%). Medicare's CY 2026 telehealth originating site facility fee is 80% of the lesser of the actual charge or $31.85. Under 42 CFR 413.70(b)(4), as a Medicare contractor's guidance also explains, a critical access hospital's costs, paid at 101 percent, may include reasonable compensation for an emergency room physician, physician assistant, nurse practitioner or clinical nurse specialist on call, not on the premises, immediately available by telephone or radio and available on site within the 485.618(d) timeframes.
- Grants, then and nowThe Health Resources and Services Administration's Federal Office of Rural Health Policy funded the Evidence-Based Tele-Emergency Network Grant Program, under which six health systems provided tele-emergency care to 65 hospitals (91% rural) in eleven states and reported uniform data from Nov. 1, 2015 to Dec. 31, 2017 (PLOS One, 2021). On May 29, 2026 the Kansas Department of Health and Environment announced $79.1 million in Rural Health Transformation Program awards to 39 organizations, including $5,000,000 to Great Plains Health Alliance. A June 4, 2026 release from Avel eCare, Great Plains Health Alliance and Pioneer Health Network announces an effort supported through the Kansas program to expand telemedicine-enabled care across rural hospitals, including emergency, behavioral health and hospitalist telehealth services for more of the two groups' member hospitals, and speaks of strengthening clinical resources across 31 critical access hospitals in Kansas. CMS says the program makes $10 billion available each year from fiscal year 2026 through fiscal year 2030.