It was the middle of the night in the intensive care unit (ICU) of Island Hospital in Anacortes, Wash., and a patient was hallucinating and in withdrawal from alcohol. The Ativan he’d been given wasn’t enough to lower his heart rate, and Joyce Selph, a longtime nurse, was called over to assist.
The protocol, she knew, was to call the attending physician and figure out next steps, but there was a problem. The hospital used a teleICU model, in which doctors staffed the ICU during the day, but only remote doctors were available overnight, via computer screen. Selph and the other nurses had tried to reach the teleICU doctor, but no one answered the calls.
There was no one else to ask, and the patient was rapidly deteriorating. So Selph, who had been a nurse for more than 30 years, did what she assumed a doctor would do; she started the patient on Precedex, a medication that can help sedate a patient.
Giving medication without a doctor’s orders risked her license, Selph knew, but not doing anything risked the patient’s life. The patient survived. Selph was so disturbed by the incident that she complained to her nursing supervisors and to the hospital’s administration that having a doctor on screen rather than in person was bad policy, but they decided to keep it for financial reasons, she says.
Soon after, Selph quit her job at Island Hospital, she says, because she felt that the teleICU was endangering patients’ safety and making her practice outside of her scope.
“They did it as a cost-saving measure because they were having trouble getting hospitalists to want to be there,” Selph says. “But I felt very strongly that it was unsafe.” (Island Health said, in a statement to TIME, that it launched a teleICU program in 2021 to strengthen the level of critical care available in its rural community. It has used a vendor to run the program, the spokeswoman said, and “any technical challenges that have arisen have been addressed directly with [the vendor] to ensure continued safe, effective patient care.”)
More hospitals are implementing teleICUs, especially overnight, as they struggle to recruit enough providers while keeping costs low. Having a remote physician who can staff multiple hospital ICUs at once can keep ICUs open and revenue coming in, even in the midst of an ongoing shortage of doctors. Though the model was initially pitched as a way to help rural systems deploy experts at all hours, it is frequently used in more urban and suburban areas as well. One 2025 paper surveyed 574 adult ICUs across the country and found that about one third implemented some form of telemedicine.
The teleICU model has been gaining attention recently after Conor Hylton, a 26-year-old dental student, died in 2024 after being transferred to an teleICU at the Milford Campus of Bridgeport Hospital in Connecticut, part of Yale New Haven Health. In March 2026, his family filed a lawsuit over his treatment. The complaint included an inspection report from the Centers from Medicare and Medicaid Services (CMS) that found, among other failures, that no physician had overseen Hylton’s care after he was transferred to the ICU, and that the ER physician who had been called up to intubate him did not know the way to the ICU and got lost, delaying Hylton’s treatment by crucial minutes.
William Hylton, Conor’s father, told TIME about the experience of watching his son die as a doctor on a big screen tried to give nurses instructions to adjust his ventilator. He’d brought his son to the emergency room with stomach pain the night before and was told he’d be hydrated inpatient and released shortly after. Conor was transferred to the ICU after William had left for the evening, because his pancreas was inflamed and his heart was racing. The hospital violated its own policy by not informing Conor’s parents that he had been transferred, the CMS report found. His parents got a call at 4:30 a.m. alerting them that their son was near death and in the eICU; they raced to the hospital where they watched their son die, 11 hours after he had arrived at the hospital. “I started thinking about the guy in the screen, and thinking, ‘What the hell was that?’” Hylton says. “Why wasn’t there someone at his bedside?”
The hospital in Connecticut isn’t the only teleICU where patient care appears to have suffered because of the remote setup. TIME reviewed hospital inspection reports from the last six years by the Centers for Medicare and Medicaid Services and found three other instances in which patients died or received substandard care after interacting with a teleICU doctor. In one 2024 instance in Los Banos, Calif., a teleICU physician ordered a medication for a patient whose heart rate was rapidly dropping, “but the record did not indicate it was administered,” the inspection says, and the patient died. In 2021 at a medical center in Reading, Penn., a teleICU physician ordered that a patient with COVID be intubated, but no in-person doctor capable of inserting a chest tube was available for two hours; the patient coded while the chest tube was finally being inserted. And at a hospital in Louisiana, a critically ill patient could not be transferred to another facility for care because hospital staff “failed to follow the appropriate process for after-hours when there was no in-house physician on call.” They did not know they were supposed to call the teleICU doctor, and the patient died while waiting to be transferred.
Experts say that while teleICUs can be useful in providing care, many hospitals don’t use them effectively and instead use them primarily to cut costs. “It can definitely be misused if the goal is to make the spreadsheet look good and staff as many beds as possible with one doctor,” says Dr. Eric Scholten, who works as an teleICU doctor in a system in the American West where he says the system functions well. (His hospital did not give him permission to speak on its behalf.)
Selph says that nurses are often the ones who have to take on the duties of examining a patient and assessing them in person because a teleICU doctor can’t. That seems wrong to her. “These patients are being charged for care, thinking they are getting a physician monitoring their case,” she says. “But in reality they weren’t even assessing the patients themselves; it was the nurse assessing and giving information.”
It’s an open secret among many doctors and nurses that some hospitals with teleICUs provide poor quality care. One doctor whose practice contracts with a teleICU in North Carolina and provides services there—who wished to remain anonymous—told TIME that none of the doctors in his practice send their families to the hospital where the teleICU is used because they all think it’s sub-standard care. “The running joke is, ‘Don’t send me to that hospital if I’m sick,’” he says.
The rise of the teleICU
TeleICUs first became popular in the mid-2000s after research suggested that patients in ICUs would benefit from having what’s called an “intensivist,” an expert in treating ICU patients, available 24-7, says Deena Costa, a professor at the Yale School of Nursing who has studied teleICUs. At the time, many hospitals didn’t have intensivists available overnight.
A company called Visicu, founded in 1998 by two intensivists at Johns Hopkins Hospital, became the leading company to provide this teleICU technology, advertising that it could cut costs by 25% while making an intensivist available around the clock and reducing the number of ICU deaths: a win-win-win.
Costa, who was an ICU nurse before going into academia, says she remembers the buzz around the Visicu system at the time, but that nurses she interviewed had mixed feelings about its effectiveness. Some told her they felt micromanaged and surveilled. Since then, some studies have suggested that the round-the-clock presence of an intensivist isn’t as essential as researchers once thought because it doesn’t have an impact on mortality, she says.
But many hospitals had already embraced the teleICU system, and for some, there was no going back. By 2016, Philips, which had acquired Visicu in 2008, was touting the financial benefits of teleICUs, citing a study that found that they increased case volume by 44% and helped save hospitals tens of millions of dollars by reducing the need for doctors on the premises.
ICUs on their own don’t generally bring in a lot of revenue for hospitals, says Dr. Nandita Nadig, an associate professor at Northwestern University’s Feinberg School of Medicine and a critical care doctor. But some hospitals that have other specialties that may bring in revenue, like a stroke center or a surgeon who performs bypass surgery, are required to have ICUs if they want to have those specialties, she says.
By 2018, about 18% of total ICU beds in the U.S. used telemedicine critical care, up from less than 5% in 2010, one study found.
Then came the pandemic. As ICUs became filled with critically ill people and hospitals faced a shortage of critical care doctors, more facilities adopted the technology.
By October 2020, Philips said it was seeing “increased interest in telehealth solutions like Tele-ICU.”
The benefits of teleICUs
For hospitals, the big benefit of a teleICU is that it can provide the perspective of a critical care doctor without having to pay for one to be at your facility at all hours. This is especially beneficial for rural hospitals that can’t afford to staff a critical care specialist around the clock, or for a larger health system where one virtual specialist can check in on patients at multiple hospitals.
“The reality is that not every place is desirable to raise a family or build a life, so you do need some of this backstop—I do believe there’s a place for it,” says Dr. Allison Harriott, a critical care and emergency medicine physician who has researched teleICUs. “But can it fully supplant what I do at the bedside? I don’t think so.”
Having a remote intensivist available to lend expertise can be useful for patients in rural areas, says Nadig. Intensivists may see a red flag that a patient should be put on a heart-lung machine, for instance, or that they need a lung transplant. They can also help patients who are in small, local hospitals stay in their communities, she says, rather than be transferred hundreds of miles to a facility with an intensivist.
The U.S. has a shortage of intensivists, Nadig says, and as patients get older and their care becomes more complex, teleICU doctors can be better than not having any intensivist available.
One 2025 study funded by Philips found that from 2012 to 2020, hospitals with telecritical care capabilities had lower patient death rates in the ICU and shorter ICU stays. Philips customers use the technology “to extend access to intensivist expertise, support bedside clinicians, monitor patient populations and help identify patients who may need earlier intervention,” says Martijn Hartjes, an executive at Philips who leads hospital patient monitoring.
But both Hartjes and Nadig agree: a teleICU doctor should not replace bedside physicians. “The value of a teleICU is really how well it’s applied,” says Nadig. Hartjes echoed this sentiment. “The Philips eICU program,” he says, “is intended to supplement, not replace, bedside care.”
Problems with implementation
The original purpose of teleICUs seems to have been lost in recent years, Nadig says. A 2021 study found that teleICUs were actually less predominant in rural hospitals than in urban and suburban hospitals.
Recent estimates suggest that about one fifth of hospitals in the country use teleICUs. Now, as more hospitals struggle financially in response to funding cuts in legislation like H.R. 1—the One Big Beautiful Bill Act—and impending cuts to Medicaid, some are considering switching to teleICUs. In 2025, the Milwaukee Journal Sentinel reported that Ascension, a nonprofit chain of Wisconsin hospitals, was replacing some of its critical care doctors with physicians from a national staffing agency, including telehealth doctors. (Ascension said in a statement that it had incorporated telehealth capabilities “as a supplement to our providers and care teams, a model used broadly around the world.”) A Montana-based medical center told NBC in June 2026 that it planned to transition its pediatric intensive care services to a 24/7 telehealth-supported model, despite pushback from physicians and nurses.
TeleICUs operate very differently in the U.S. than they do in other countries, says Dr. Claudia Spies, medical director of Charite Universitatsmedizin Berlin, a teaching hospital that has a teleICU. Spies, who has researched teleICUs, says that in Germany and other high-income countries, the teleICU exists to give patients access to more specialists who can weigh in on their case. But this complements in-person care; hospitals with teleICUs do not typically cut staff, she says. Instead, there is still a doctor physically present in the teleICU, who consults with the other doctors over the screen to decide how best to treat the patient.
“Here, it was never really used to save money on staff,” Spies says. “In my opinion, that’s not the way an ICU should be established.” Public health insurance in Germany won’t reimburse treatment in the ICU if there is not a doctor physically present, she says.
In Brazil, teleICUs expanded during the COVID pandemic, but were primarily used to train physicians and nurses in responding to critically ill patients, rather than replacing in-person health care. The program stayed in place after the pandemic in many hospitals that did not have intensivists on site.
Scholten, the doctor who works at an teleICU in the U.S., says that the system can work well—even without a doctor physically present in the ICU—if the virtual doctor is familiar with the hospital system and the doctors who work there. The doctor needs to intimately understand the processes of how the hospital works, like who to contact to get a patient intubated or discharged. His hospitals have flow sheets that tell nurses what to do if someone is deteriorating rapidly and there’s no cardiac surgeon nearby, for instance. Outsourcing the teleICU to third-party contractors is generally not a good model, he says. “It takes a lot of resources to make it work well,” he says, “as a service to keep people in their local community.”
In his experience, teleICUs have been helpful at getting patients off of ventilators earlier than they would have been able to if they’d had to wait for a doctor to round, he says, which cuts down on infection risk. For routine pneumonia or routine heart failure, teleICUs can help local doctors more effectively manage patients, he says.
But teleICUs shouldn’t be thought of as a way to save money, he says—because oftentimes, they don’t. “I would argue that it’s cost-neutral,” he says. “It provides high-quality care to rural sites and allows you to efficiently care for patients.”
“It can go awry when they just say, ‘We’re getting rid of bedside coverage, and you just call this person who’s a long way away and doesn’t know the culture,” Scholten says.
Unfamiliar territory
The problem of a teleICU doctor being unfamiliar with the inner workings of a hospital came up repeatedly in the public records searched by TIME. Many times, hospitals with problematic teleICUs did not have a clear process for communication between virtual doctors and inpatient doctors. In other instances, patients and their families were not told why the patient was being seen by a teleICU doctor rather than a bedside doctor.
After Conor Hylton died, CMS investigated Bridgeport Hospital’s Milford Campus and found that “the hospital failed to ensure quality medical care was provided to a patient who presented to the Emergency Department and was transferred to the intensive care unit.” Among other things, the investigation found that the hospital did not integrate the teleICU well into the rest of the system: there was no process for the in-person emergency room doctor to sign out to the teleICU doctor, for instance. And when Hylton went into cardiac arrest at 4:33 a.m., the only doctor who could intubate him—the emergency room doctor—did not arrive until 4:45 a.m. because the doctor did not know how to get to the ICU: proof, his family’s lawyer argues, of a system that did not plan adequately for emergencies. The nurses, teleICU doctor, and hospitalist were all unsure who was supposed to lead the code in the ICU when Hylton went into distress, according to the report. Yale New Haven Health said in a statement to TIME that “our priority has always been, and remains, providing safe, high-quality care to our patients. We remain committed to continuous review and improvement of patient safety practices. Because this is active litigation, we cannot comment on specific allegations and will address the claims in the appropriate legal forum.”
Similarly, in 2024, CMS investigated Memorial Hospital Los Banos in California after the teleICU doctor ordered a critical medication for a patient that may not have been administered, and the patient died. The report found no record of the in-house physician being notified of the patient’s condition, and no record of any doctor—virtual or not—proclaiming the patient dead. CMS determined that the hospital and the outside teleICU company it contracted with did not appropriately coordinate care. Medical directors of the teleICU and the hospital were supposed to be meeting quarterly to improve care in the teleICU, but they did not, according to the report. Sutter Health said, in a statement, that “we consistently take opportunities to reexamine and reevaluate policies and clinical workflows within all our care settings so as to enhance our abilities to care for our patients,” but that legal requirements prevented the hospital from providing additional information.
The lack of a clear chain of command between doctors in the hospital and the one in the teleICU came up frequently in CMS investigations into incidents involving this type of care. In Glenwood Regional Medical Center in West Monroe, Louisiana, an ICU patient was awaiting transfer to another hospital, but one doctor did not answer their phone, and staff failed to notify any other medical provider—including those in the teleICU, who were on call from 7 p.m. to 7 a.m. The teleICU was called later that night when the patient was in severe distress, according to the CMS report, and soon after, the patient died. The hospital’s chief medical officer later said in interviews with CMS that staff should have called the teleICU for the transfer, but the inspection report found that staff weren’t trained on who to contact. (Healthcare Systems of America, which owns Glenwood Regional Medical Center, did not respond to a request for comment.)
The absence of in-person staff to treat patients who need immediate care also came up repeatedly in CMS investigations involving teleICUs. At Penn State Health St. Joseph Medical Center, a patient’s lung collapsed and a virtual ICU physician ordered intubation and placement of a chest tube, but “the cardiothoracic physician failed to arrive in a timely manner,” and the patient coded when the chest tube was being inserted, according to the CMS investigation. The report found that there was no chain of command for staff to escalate the process when an in-person physician does not show up on time, the report found. Penn State Health did not respond to a request for comment.
William Hylton says that patients, like his son, have the right to know whether the hospital they’re going to has an teleICU, and whether there are nearby hospitals that may have more in-person resources available to help patients in distress.
“There was no doctor in person providing critical care,” he says. “You can’t insert a breathing tube through a television.”
He hopes that Connecticut and other states will pass laws that require hospitals to divulge whether they have teleICU and offer to transfer patients to other hospitals if they can’t handle the level of care that the patient needs.
It’s not just Connecticut where family members are seeking more transparency about teleICUs. In its investigation into the incident at Memorial Hospital Los Banos, CMS interviewers found that the family members of another patient who was in the same ICU had concerns about the teleICU. That patient’s son and daughter-in-law said that they did not like that the hospital did not have doctors present in the ICU and that patients were telemonitored. They had asked for their father to be transferred to a larger hospital with more resources, according to the inspection report, but he was not transferred. He also died in the hospital.
Joyce Selph, the nurse who was uncomfortable with how telehealth was administered at her hospital in Washington, now works at a hospital with in-person staff in the ICU, but she still thinks about the suffering that her patients had to go through because of the teleICU.
“We all know as nurses that stuff happens after hours,” she says. “But if there’s a time delay in getting care, for an ICU setting, that’s deadly.”
It bothers her that she told Island Hospital repeatedly about the problems in the teleICU and that nothing changed. She was the first to leave based on the changing model of care, she says, but since then, “there has been an exodus.” (A spokesperson for Island Health said its retention is in line with the national average and that its teleICU helps give care teams support so that patients aren’t automatically transferred to Seattle when Island Health can safely care for them near their homes.)
Selph says she remembers waiting as long as 30 or 45 minutes to hear back from a teleICU doctor at her old job when she needed to know how to handle a patient. In an ICU, time matters, she says, and a wait like that can change a patient’s trajectory. Even when the doctors did answer her calls, there was not a good system to communicate what was going on to the doctors on the screen, she says. “There’s a doctor who has never seen the patient before, has never assessed them—it’s like putting a blindfold on and trying to do medicine.”
She also believes that patients should know more about what kind of care they’ll be getting in an ICU before they check into a hospital. “The community has a right to know, if you’re going to come in and half of your care is going to be covered by a doctor who pops up on screen,” she says. “I wouldn’t take anybody I cared about to a hospital that used an eICU at night.”
.jpg?branch=production)
.jpg?branch=production)
