Lianne Kraemer knows she’s in a precarious position. She has stage-4, terminal breast cancer that has spread to her brain. To keep it from penetrating further into other organs, she’s currently on chemotherapy, which weakens her body and puts her at high risk for a number of health problems—including infectious diseases like COVID-19.

Soon after public health officials identified SARS-CoV-2 virus as the cause of a previously unknown illness in the Hubei province of China, doctors realized that the novel coronavirus seemed to zero in on certain demographics: older people, as well as those with existing chronic diseases and weakened immune systems like Kraemer, 43. “From very early on, I felt vulnerable and scared,” she says. “People were saying ‘everybody don’t worry, this will only affect the sick people.’ I felt vulnerable and expendable.”

Now another specter threatens Kraemer’s already tenuous hold on health. If she develops COVID-19, and needs intensive care and help from a ventilator to breathe, she might be denied access. “With my diagnosis of metastatic terminal breast cancer with brain metastases, doctors will go, ‘Wow, she’s on her way out anyway,” she says. “The fact that I could go to the emergency room with COVID-19 and they might deem me not worthy of care is mind-alteringly terrifying.”

Anywhere from 3% to 19% of COVID-19 patients will wind up in an intensive care unit (ICU), often requiring ventilators to breathe. The number of COVID-19 cases in the U.S. is, as of writing, nearing 1 million, which means roughly 150,000 additional patients for the country’s already-stressed ICU system. If the flow of patients continues to outpace the supply of ventilators, doctors will be making heart-wrenching decisions about how to distribute those precious few devices.

This is the harsh truth that administrators at Henry Ford Health Systems in Michigan tried to address in a controversial letter that leaked on social media on March 27. The letter was intended for patients (and their families), explaining the facility’s approach to rationing ICU services: “Some patients will be extremely sick and very unlikely to survive their illness even with critical treatment. Treating these patients would take away resources for patients who might survive.”

The letter, which was part of the hospital’s preparations for emergency actions if supplies to treat patients dwindled, went on to provide a bulleted list of the types of patients who may be “not eligible for ICU or ventilator care,” including those with severe heart, lung, kidney or liver failure; terminal cancers; and severe trauma or burns. “We crafted our policy to provide critical guidance to healthcare workers for making difficult patient care decisions during an unprecedented emergency,” Dr. Adnan Munkarah, executive vice president and chief clinical officer of the Henry Ford Health System wrote on Twitter in response to questions about the letter. “It is our hope we never have to apply them, and we will always do everything we can to care for our patients, utilizing every resource we have to make that happen.”

“Being in the midst of a public health emergency triggers a change in the ethical principles that we use to make care decisions,” says Dr. Douglas White, director of the program of ethics and decision making in critical illness at the University of Pittsburgh. “A public health emergency shifts focus from individual patients and trying to maximize their well-being, to looking at outcomes for an entire population.”

Which means that people with a cancer prognosis like Kraemer might be considered ineligible to receive a bed in an ICU because their chances of survival are low. To Kraemer, this feels like a primitive approach. “I am more than just words on a paper. I am more than a diagnosis of stage-4 cancer,” she says. “There has to be a more sophisticated way than categorically excluding entire populations.”

Doctors’ Coronavirus Dilemma

In 2015, New York Governor Andrew Cuomo asked the state’s Department of Health to develop an algorithm to anticipate the need for ventilators in the event of a flu pandemic like the H1N1 outbreak of 2009. The commission responded with a 272-page report prudently calling for buying up ventilators sufficient for a flu outbreak.

Them, COVID-19 made a mockery of all of that planning.

With its dense population and popularity as an international travel destination, New York has become ground zero for the U.S. wing of the current pandemic and has completely annihilated whatever formulas the state had in place to allocate resources for prevention and containment of such an outbreak.

“COVID-19 is not like the flu at all,” says Dr. Chen Fu, clinical assistant professor of medicine at New York University Langone Medical Center, who has spent the pandemic working in the hospital’s critical-care ward. “We’re finding that these patients who are intubated with COVID-19 lung disease are extremely hard to take off the ventilator. It’s almost like a black hole.”

It’s largely this fact that is contributing to the nationwide shortage of ventilators, forcing doctors to make the kinds of existential choices—picking who gets to live and who must die—that both they and their patients would have been spared in the past. “In normal times we practice a sort of hopeful, benevolence-based medicine,” Fu says. “Whereas now we have to practice a sort of holistic justice-based medicine. I have to think about how to help society as a whole.”

Spotlight Story

U.S. Health Care Workers Are Dangerously Ill-Equipped to Fight COVID-19

“We are out of everything.”

That involves not just denying care to patients like Kraemer, but sometimes withdrawing it after it’s been provided. Fu and other members of the hospital staff recently faced one such mortal choice when they were treating a terminally ill patient on a high-flow ventilator—a step down from full intubation, and which is also in short supply as a result of the pandemic. The ventilator offered the dying patient not just comfort, but also a certain measure of end-of-life dignity. On another floor, however, a COVID-19 patient who had a greater chance of survival was panting for air and in desperate need of a high-flow ventilator. One machine, two patients, and Fu’s group had to make the call.

The algorithm—which was also a part of the 2015 New York State report—made it for them: “We took that machine away from the patient who was dying,” he says.

The patient did die, and the experience—for the doctors, the patient and the family—was harrowing. “Whenever you apply statistics to an individual it’s always very messy,” Fu says, “because individuals have their own desires, they have their own history.”

The People Behind the Machines

Ventilators are currently scarce, but when industry steps up—as the Ford Motor Company and General Motors recently did, committing to build tens of thousands of ventilators each—the gap can be filled relatively fast. Health care workers can’t be ordered up so quickly.

“A critical care physician takes a total of 10 years to train,” says Fu, “four years of medical school, three years of residency and three more years of critical care fellowship.”

At Boston University’s Boston Medical Center, this has led to on-the-fly staffing decisions. “In the internal medicine units where I work,” says Dr. James Hudspeth, medical director of the COVID-19 response team at Boston Medical, “usually we’re staffing with internists [but] at this point we are pulling in a subset from other specialties to help us out.” The hospital’s nursing team is being similarly stretched, Hudspeth says, with floor nurses, who are not specialists in critical care, being pressed into service alongside ICU nurses.

Supplies of medications being called up to alleviate some of the most severe symptoms of COVID-19 are also dwindling. For example, there are developing shortages of interleukin-6 (IL-6) inhibitor (which can reduce inflammation and help relieve infection), a fact that’s forcing other tough decisions.

Dr. Bushra Mina, chief of pulmonary medicine at Lenox Hill Northwell Hospital in New York begins the day with 10 to 20 vials of IL-6 inhibitor. Every time he thinks about using one of these vials, he has to judge whether it might be wiser to save it for a later, more desperate patient. “It’s like military triage,” he says.

Hudspeth worries that if an existing drug does prove effective against SARS-CoV-2, there will be a stampede for it—especially if it’s rare or expensive. “It’s instantaneously on the market and then at that point we may also be forced to do more rationing,” he says.

Then too there is the COVID-19-related rationing that has nothing to do with COVID-19, at least not directly. Doctors pressed into emergency medicine may no longer be available to their regular patients for appointments to check up on chronic conditions like diabetes and heart disease. That, in turn, increases the risk of an at-home emergency. During the 2014 Ebola outbreak in West Africa, epidemiologists recorded a spike in deaths from preventable causes, as the epidemic swallowed the health care system whole, leaving the medically vulnerable uncared-for.

“I’m confident we’re going to see that play out again [with COVID-19] globally and potentially within the United States,” says Hudspeth, noting that the number of heart attack and stroke patients coming into Boston Medical Center in the last few weeks has been lower than they would normally expect. “They’re probably suffering those at home.”

A New Framework for Health Care Rationing

Currently, the U.S. has no national criteria for allocating ICU beds and ventilators during an emergency like a pandemic. So, doctors and hospital officials have largely been relying on the guidelines of professional specialty groups, including those published in 2008 by the American College of Chest Physicians (ACCP), for making decisions about how to allocate scarce resources to treat respiratory disorders.

The ACCP guidelines hinge on the so-called “exclusion criteria,” which advise not considering ICU care for people with moderate-to-severe heart or lung disease, those with terminal cancer — like Kraemer— the elderly, and people with severe cognitive impairment. Such policies, says White, are discriminatory from both an ethical as well as legal perspective. The Department of Health and Human Services Office for Civil Rights in April conducted a review of Alabama’s 2010 guidelines for rationing ventilators based on exclusion criteria, for allegedly discriminating based on age and disability. The state has since revised its guidelines to be compliant with federal civil rights laws.

The ACCP did not clarify the reasoning behind its guidelines in its official publications, but White says it’s likely that the medical association relied on standard metrics for evaluating patients’ health and, given their medical histories, considered most of them as less likely to survive whatever brought them into the ICU than those without those underlying conditions.

After the ACCP put out its guidelines, White and his colleagues at the University of Pittsburgh felt “that was a very big overreach in terms of abandoning traditional medical ethics in favor of a public health ethic,” he says. “When the lever gets pulled and a public health emergency is declared, that indicates the need to do things differently than we normally do, and to allocate resources differently. But exclusion criteria are the wrong tool for the job. They send the wrong message that people’s lives are being judged not worth saving. The ethics of medicine are diametrically against that kind of thinking.”

So White and his colleagues proposed a different framework, which they published in 2009. Their approach assigns a “priority score” to every patient, based on a constellation of factors including existing health conditions, as well as something called “life years gained,” or the likelihood of living beyond the current crisis. Doctors score a patient’s chances of surviving and of living more years once they are discharged, each on a scale of one to four, with one representing the highest likelihood. These two numbers are combined, and the final score determines a patient’s priority for ICU care.

The system provides a “brutal reckoning” for how bad the supply shortages are, says White. “If we just exclude people, then those people disappear from the calculus.” But by giving everyone a priority score, this framework accounts for those patients who, in normal circumstances would be treated, but in times of scarce resources, are denied care. “It doesn’t allow us to brush under the rug how bad things are.”

For the last decade or so, the framework has been more of a theoretical exercise than a necessity, but that changed when patients began flooding hospitals with COVID-19. “When things began going sideways in China, and again more recently with what happened in Italy, a number of institutions [started] reaching out to me asking about our framework,” says White. “They wanted to know if we were using it [at the University of Pittsburgh] and if it had been incorporated into our hospital policy,” he says.

So, he condensed the principles from the original proposal into a “very operational, practical step-by-step recipe hospitals could cut and paste into whatever format they need,” he says, which the University of Pittsburgh published on its website on March 22.

The system is inspired by the one used in the U.S. to allocate organs for transplant, which relies on the “lifecycle principle,” taking into account where a person is in their overall life cycle, and how much more life they have left to experience. The lifecycle principle holds that if all other things are equal as far as patients’ priority score, and there is only one ventilator available, doctors could prioritize younger people, who have had less chance to live their lives, over older people who have likely enjoyed more life experiences. It also accounts for “instrumental value,” which suggests prioritizing those who may, if they survive, further the goals of improving public health—health care workers, for example, who could then go on to save more lives.

There is clearly interest in an alternative to the current standard of care. In the first three days after they were posted, the new guidelines were downloaded more than 2,000 times; Pennsylvania health authorities have already asked all hospitals in the state to utilize these guidelines, and Massachusetts, New Jersey, Colorado and Oklahoma have followed suit.

Surging On in the COVID-19 Pandemic

There would, of course, be no need for these ethically wrenching choices if there were enough ICU personnel, beds, and other equipment to handle the patient load.

In a matter of weeks after the first COVID-19 patients were confirmed, the Chinese government turned arenas into field facilities and erected two entirely new hospitals in Wuhan. In New York City, Mount Sinai Hospital and the charity Samaritan’s Purse worked together to build a 68-bed tent facility in Central Park to care for some of the city’s COVID-19 overflow; the field hospital began receiving patients on March 31. Los Angeles County went New York one better when, on April 13, it opened a 266-bed surge ward within the walls of the recently shuttered St. Vincent’s Hospital.

In ordinary circumstances—even an ordinary epidemic—surge hospitals like these wouldn’t be necessary. All hospitals in California are supposed to have a 20% to 40% surge capacity within their own walls, says Dr. Christina Ghaly, director of the Los Angeles County Department of Health Services. That capacity is mostly generated by canceling elective surgeries and converting the facilities used for same-day surgery, anesthesia-recovery and even emergency medicine into ICUs.

But COVID-19 isn’t an ordinary epidemic and free-standing surge facilities are an important—if imperfect—adjunct.

In mid-March, the Los Angeles County and California State departments of health began work on a statistical model projecting one month ahead, which sought to forecast when peak demand for hospital beds, ICU beds and ventilators would hit, and what that peak would look like. “What the model showed,” Ghaly says, “is that we had sufficient bed capacity for COVID-19 patients within our existing hospital supply, but there was a shortage of ICU beds.” That added to the impetus to repurpose St. Vincent’s, which came online a month later to fill the gap.

Perhaps, however, the U.S. could develop a system that would forestall any sort of care-rationing, even in the midst of a public health disaster like COVID-19. Hudspeth suggests one solution: building a national medical corps not unlike the Federal Emergency Management Agency (FEMA), which could mobilize fast, get on-site where needed, and have personnel and equipment ready to move when a pandemic hits, much the way FEMA mobilizes for a hurricane. “Not having a system that makes it possible to move personnel and necessary equipment between different parts of states and basically across the United States as a whole is a really glaring gap that we face,” he says. “We could have critical care physicians, infection control specialists, critical care nurses who could work their way across the country as different hot spots flare up.”

That would address not only a crying medical need, but a crying ethical one, Hudspeth says, sparing frontline caregivers from having to make the kinds of calls Fu and his colleagues made when they withdrew the ventilator from the terminally ill patient and gave it instead to someone who had a better chance to survive. It would also spare patients like Kraemer, suffering from one disease, from becoming collateral casualties of another one, simply because the resources to care for them are no longer there.

Lianne Kraemer in Chicago on Sept. 25, 2019. Kraemer, who has stage-4, terminal breast cancer, could be denied intensive care and access to a ventilator if she develops COVID-19.

David Kasnic for TIME

When Kraemer learned her breast cancer had spread, her doctor couldn’t guarantee that she would live another two years. That was four years ago, thanks to experimental therapies she tried by participating in a variety of clinical trials. That’s what she hopes doctors will remember if she does need hospitalization for COVID-19.

But as it stands, Kraemer has been forced to do the unthinkable and make a case for her life should she get sick with COVID-19. In order to ensure that emergency health care workers look beyond her medical chart to see her as an individual patient, she recorded a video of herself, at her current doctor’s suggestion, to show that, despite being a terminal cancer patient, she does not currently suffer any serious health consequences—and that she deserves potentially life-saving treatment in the ICU, were she to develop COVID-19.

“It was scary, and I thought, ‘This is what it’s come to, that I have to prove that my life is worth saving,” she says. “I laid out that I expect to receive ICU placement if I needed it, and that I expect to receive a ventilator if that’s necessary, too. To have hospital administrators essentially write us an early death sentence for something completely terrible that happened to us through no fault of our own—it’s brutal.”