Stephen Wall, an emergency room doctor in Manhattan, could see the surge on the horizon. In calmer times, he might need one ventilator per shift—for a patient suffering from a trauma, usually, like a fall. But as the Covid-19 patients arrived last week, the intubations were happening every two hours, he says, “like clockwork.” He had watched the situation unfold elsewhere in New York, like at Elmhurst Hospital, in Queens, where patients reportedly died waiting for staff and beds. At the current rate, Wall knows his own ER is headed in the direction of a similar shortage. “We’re going to have to ration,” he says. “We’re just a few weeks behind.”
But what would rationing a ventilator actually mean? To Wall, who also researches clinical bioethics at New York University, the answer was opaque. Would he be asked to prioritize by age, turning away the elderly, as some hospitals in Italy did? What if the choice was between an old person and a younger one who was far less healthy and more likely to die, even with a ventilator’s aid? What if that patient was a health care worker who might, in the weeks ahead, get better and save more people? The guidance was unclear.
“This is being discussed all over America right now,” says Nancy Berlinger, a researcher at the Hastings Center, a nonprofit bioethics think tank. The situation is for now most pressing in New York, with its dramatic surge in cases. But for medical ethicists elsewhere, the scenes there are a warning to prepare.
“[Deborah] Birx said it’s not time to be talking about rationing, that people shouldn’t worry there is no bed for them. I completely disagree with that,” says Arthur Caplan, head of the medical ethics division at New York University Medical School, referring to the response coordinator for the White House Coronavirus Task Force. “It’s time to get ready for surges. Some little guys like Elmhurst Hospital are already overwhelmed.” Caplan is leading efforts at NYU to develop a rationing plan that will help guide doctors like Wall. He expects to release it within days.
Hospitals are expected to prepare for this kind of thing. They have ethical guidelines drafted; they conduct simulations of mass casualty events like a mass shooting or a bombing. But there’s a big difference between these scenarios: In a bombing or shooting, a wave of critical cases crashes down upon a few hospitals. The damage is swift and cataclysmic, forcing difficult decisions. Then, whether through mutual aid agreements or at the direction of government disaster planners, other hospitals step in to help. Pandemics are more like a tsunami. They put the entire health system underwater, and it stays that way for a long time. “You’re used to thinking about one patient at a time,” Berlinger says. “All of a sudden you need to think about the whole community, including people who may get sick in a week or two weeks from now.”
And in the US, pandemic-specific planning is a patchwork. While some states and hospital systems got more serious about writing guidelines in the past two decades, after outbreaks of SARS and Ebola, those plans are rarely circulated widely, and they often lack details for practical implementation. For most places, the most relevant touchstone is the Spanish flu, from over a century ago. “It’s all just literature,” Caplan says. “The preparation is poor.”
While hospitals frequently make rationing decisions—who gets a critical organ donation or an experimental drug; who gets seen on the basis of insurance—the pandemic could expose a far broader part of the population to those realities, Caplan says. Health care workers who never would have expected to make rationing choices will be forced to take part.
The goal for most hospitals is to maximize the number of lives saved, and do so as fairly as possible. But it is far from obvious how to do that. A review put together by Thomas Cunningham, director of bioethics at Kaiser Permanente West Los Angeles, of dozens of hospital and state policies shows stark differences in how doctors are told to behave in a crisis—who would be excluded if ventilators or beds or staff runs short, and what criteria they use to decide who would benefit most, or if benefit should be figured in at all. Some places use scoring systems, but the algorithms they use to determine outcomes differ based on the evidence available when they were designed. Some exclude people based on a patient’s age or condition. In Alabama, state guidelines issued in 2010 tell physicians to deny ventilators to those with AIDS or certain mental disabilities. Others had no particular plans to ration resources at all.
Cunningham downloaded reams of research and existing hospital and state policies with the goal of developing a more actionable plan for his hospital. Eventually, he uploaded it all to a Google drive and shared it with his colleagues around the country. Now, in the midst of a pandemic, clinical ethicists are trying to iron out those differences as best as the evidence allows. “What do people agree on? What rules are the total outliers?” Berlinger says. “If we had national guidance, it would be simpler.”
Much of that debate has been taking place on a listserv of the American Society for Bioethics and Humanities. The discussions kicked off in early March, when Janet Malek, the director of medical bioethics at Houston Methodist Hospital, sent a message asking for her colleagues to share their guidance. She had gone to her superiors asking for pandemic plans, but it turned out the hospital only had mass casualty guidance, which she says was far from adequate for the present emergency.
Since then, the listserv has been like “a 24/7 virtual philosophy seminar,” says Cunningham—a place where clinical ethicists drafting guidelines debate the latest evidence. At Kaiser, he at first had trouble finding the system’s disaster guidance. Beyond that, “it was a bit thin,” he said, speaking over the phone last week through a mask—he was awaiting test results. (His test later came back negative.) On the one hand, the existing guidelines lacked much practical guidance that doctors could easily follow, but they also referred to New York State’s ventilator allocation guidelines, issued in 2015, a 270-page tome that details who gets saved first in crisis settings. “Can you imagine walking down to the ICU right now and saying, ‘Hey, read this book?’” he asks.
During a pandemic, it’s not only a matter of who gets a ventilator; it’s also a matter of how long they get to use it. In a truly desperate situation, when is the right time to reevaluate if a ventilator is truly doing a patient any good or should be taken away to help others? Some guidelines say 48 to 72 hours. But early data suggests that for Covid-19 patients, that might be too soon to tell whether there’s been meaningful improvement. Doctors might be taking ventilators away from patients who would soon improve. “I don’t think there is a general agreement on those timelines,” says Malek. She says Houston Methodist, which recently finalized its pandemic guidance, gives doctors flexibility to amend their plans as more evidence arises.
And while ventilators and ICU beds have gotten the most attention, they are just one piece of solving a complex ethical puzzle that unfolds over time. For example, what if there are not enough respiratory specialists to run the machines, or ICU nurses to attend to beds? A report released last week from the Society of Critical Care Medicine warns that there are not enough—and that’s before one considers staff falling ill themselves. Hospitals preparing for surges are already doing plenty of rationing, from Covid-19 tests and protective equipment to determining which surgeries are most necessary for limited staff to perform.
Then there are the specifics of this particular virus. The data gleaned from places like China and Italy—and it is sparse—suggests some traditional triage methods might do harm in the pursuit of fairness. Much of the existing guidance relies in some form on a sequential organ-failure assessment score, or SOFA, a common metric that tries to predict patient outcomes. If a person suffers from poor liver or heart function, for example, they might be less likely to come off a ventilator alive. It’s one stab at fairness—a way to decide who is most likely to benefit.
But earlier this month, a group of emergency preparedness specialists wrote a paper for the National Academy of Medicine, warning against relying upon SOFA scores too much for rationing decisions. They pointed to research conducted during the H1N1 outbreak that suggested the scores did little to predict outcomes for lung damage due to pneumonia. (SOFA scores were originally developed for sepsis, a different condition.)
Another hotly debated issue on the listserv: whether health care workers would be first in line for treatment. Much of the disaster guidance developed for other kinds of emergencies holds that they shouldn’t be, for practical reasons: The disaster would be over before they were able to get better and save more people. In a pandemic, however, they might have enough time to recover. “But it’s also a matter of justice,” Malek says. “That we treat everybody the same and we don’t prioritize or show favoritism to our own people.”
And finally, disability rights advocates have pushed for a first-come, first-served model. Ari Ne’eman, a health researcher at Harvard, argued recently in The New York Times that rationing plans would discriminate against the disabled and force some people—especially those for whom ventilators are an essential part of daily life—to avoid seeking treatment for risk of losing their life-saving equipment. “Equity would have been sacrificed in the name of efficiency,” he wrote.
Ethicists drawing up rationing plans say they must balance disability rights with a call to save as many lives as possible. Cunningham would like to see health systems move away from so-called “exclusion criteria,” found in many rationing plans—rules that, as a last resort, would result in automatic denial of life-saving support on the basis of particular characteristics such as age or disability.
Cunningham’s model guidance would use a composite score developed by Doug White, a critical care researcher at University of Pittsburgh. It includes SOFA scores, but only as a contributing factor. It also takes into account other data such as a surviving patient’s anticipated life span once they’re outside the hospital. While conditions that affect those scores may correlate strongly with age, Cunningham says, there would be no particular cutoff or disqualifying medical condition. Ideally, the decisions would be made by a diverse triage team that has no relationship with the patient, and would contain an appeals process and opportunities to reevaluate individual cases.
Cunningham’s research is contributing to draft guidelines for Kaiser hospitals, but those plans are still under review. The Kaiser system’s guidance would be deeply influential, because it sprawls across eight states and Washington, DC.
At the very least, Cunningham says, some kind of regional cooperation among hospital systems is “paramount,” so that patients don’t simply move from place to place seeking care. “If we had dramatically different guidelines, you can see how people would start hospital shopping,” agrees Malek. She says Houston Methodist is in discussions with other area hospitals to make sure standards are at least complementary.
But it remains to be seen how that will work in practice. For Caplan, whose hospitals in New York are on the precipice of having to ration equipment, implementation of any new set of guidelines is at the top of his mind. “The hardest thing is not to write a policy,” Caplan says. “It’s emotional support and psychological support for the people who are going to do it and for the families of people who won’t get on the lifeboat.”
The situation at any hospital that runs short on resources will be chaotic, Caplan says. Despite everyone’s best intentions to follow specific guidance, there may not, for example, be time to calculate a SOFA score. Some triage decisions will be made by people at the top of the funnel: by the paramedic who must decide if someone should be carried to the emergency room, or the ER staff faced with an onslaught of patients and forced to make snap decisions. But at the very least, he hopes to provide them a rationale for the hard choices they are making—including people like Wall, the New York emergency room doctor.
“It’ll still be psychologically damaging, and we’ll all have PTSD,” Wall says. “But at least I can lean on that when I talk to my residents, who are going to say, ‘I killed someone because I couldn’t save them.’”
- It's time to do the things you keep putting off. Here's how
- What isolation could do to your mind (and body)
- Bored? Check out our video guide to extreme indoor activities
- Blood from Covid-19 survivors may point the way to a cure
- How is the virus spread? (And other Covid-19 FAQs, answered)
- Read all of our coronavirus coverage here
