This is one in a series of reports from hospitals responding to the Covid-19 pandemic.
The patient was around 80, an immigrant from the Dominican Republic, his medical record filled with the accumulations and erosions of old age. His blood pressure was high, his spinal cord narrowing to a pinch. He had a history of alcohol addiction, but was no longer drinking. A prostate cancer had been burned out with radiation, and he had yearly appointments to make sure it wasn’t coming back. He had type 2 diabetes.
Now, all of that was a backdrop for something more pressing. “Malaise, diarrhea, accompanied by daughter whose co-worker tested positive … cough, chest pain,” said Mary Pennington, a nursing director, reading the notes a colleague of hers had taken at the bedside. “Right now, he’s presenting with a fever of 102.7, heart rate 110, pressure 149 over 68.”
When you pressed a stethoscope to the left side of his chest, you could hear a crackle, as if the deep windy sound of his breath were coming in on a call that kept breaking up. On the right, there was just less to hear altogether. Both could be signs of fluid in the lungs, of a coronavirus infection getting bad.
He needed a ventilator, to push in air that his lungs weren’t strong enough to pull in on their own. But he wasn’t the only one, and the hospital had just one unused machine, posing a question that Pennington and her colleagues wished they didn’t have to answer. Who should get their last vent?
To figure that out, they had to translate the complex story of this man’s symptoms and underlying ailments into a number — something straightforward enough to allow for a head-to-head comparison between one patient and another. His diabetes, though, was complicating their task. His “priority score” might fluctuate a little depending on whether his diabetes was a major underlying issue with a substantial impact on his life expectancy. The man’s high glucose had damaged the blood vessels in his eye. Was that enough to count as major? Some huddled around the computer said yes, emphatically; others weren’t so sure.
A high voice broke up the argument from the far end of the room: “You’ve got to bring it back to the triage officer, and say, ‘Here’s my discrepancy.’” It was Sunil Eappen, the chief medical officer at Brigham and Women’s Hospital in Boston, interrupting the simulation the way a coach might stop a drill to give notes.
A few minutes earlier, his colleague, Eric Goralnick, the hospital’s medical director for emergency preparedness, had set up the exercise with a chilling sort of simplicity. “Team, you’ve got two patients, and you’ve got one ventilator, and you’ve got to make a decision.”
Italian hospitals had faced that same scenario, but for real, and with an even more crushing amount of need. Before there was time for any official guidance, the coronavirus had put clinicians in the sickening position of deciding which patients to treat and which to let die. Now, public health departments across the U.S. were racing to finalize policies about how hospitals should make such traumatic choices, if catastrophic shortages were to arise. Some states, such as Alabama and Tennessee, had published guidelines that simply excluded those with certain illnesses or disabilities from the calculus entirely. There were public outcries. Documents disappeared from websites, replaced with promises that they’d be revised.
“It really matters that states lead on these issues; there’s no other body with comparable legitimacy,” explained Douglas White, the University of Pittsburgh’s chair for ethics in critical care medicine, whose framework for these kinds of policies has been widely used. “It’s also incumbent on states not to F it up.”
Massachusetts wouldn’t be automatically excluding anyone, hospital leaders knew that much. They still weren’t sure whether frontline health care workers would get any sort of advantage in the triage method, but by April 6, with no official guidelines published by the state, they’d gotten enough of the gist to start training decision-makers in earnest, so they’d be ready should they run out of critical equipment or medication.
Boston wasn’t there yet. The Brigham had started the week with 53 intensive care beds open, 179 vents available, and about 12 days’ worth of their scarcest opioid, with a much bigger cache on offer from contract pharmacies. Managers had converted cardiovascular beds into Covid-19 areas and were eyeing operating rooms as overflow ICUs. Researchers were racing to make automatic vents out of the bulbs that emergency responders usually hand-squeeze to push air into failing lungs.
Still, the hospital knew a surge of coronavirus cases was coming. Putting principles on paper was hard enough; enacting them, with the deadline of patients deteriorating, was unthinkable. That’s why the triage team was in a conference room on a Tuesday morning, discussing this one man’s diabetes, trying to figure out the priority score for real patients from the previous weeks, who’d been de-identified so they could serve as examples. It was a rehearsal, for a performance they hoped never to give, with the highest of stakes.
Their task was totally new. Normally, these practitioners zeroed in on the patient in front of them, using whatever machines, drugs, and maneuvers were necessary — American academic medicine in a nutshell. Now their focus was on the common good, to save as many lives and years on earth as possible, even at the expense of the few.
To ask the bedside caregiver to suddenly shift philosophies, balancing what each person needed against the gaping need of everyone else, didn’t seem right to ethicists: For a physician, that would mean determining both that someone needs a ventilator to keep breathing, and also that they wouldn’t get one — a horrifying case of cognitive dissonance.
Instead, those questions would be put to a tribunal of sorts: A trio of initial scorers — one nurse, one physician, and one administrator — overseen by a doctor-turned-triage-officer. There would be someone filling each of these four roles at the hospital, 24/7, but for now, they were in one room, physically or virtually, to practice. There was an algorithm, but it was in their hands.
For each patient, there were two initial scores to double check. The first was a careful mix of numbers revealing how sick you are, predicting your chances of survival — the amount of cellular garbage that builds up if your kidneys aren’t filtering at their best; the level of clot-forming platelets in your blood; an assessment of eye tracking, verbal response, and movement used to detect awareness.
Pennington, who was playing the triage nurse, leaned forward in her chair to calculate these and a few other measures. She could pull them from the electronic medical record, but they needed verification, one by one, before being used to make life or death decisions. The chart would use the worst platelet count seen over the last day or so. Did she want to use that, someone asked, or the number the patient had now?
“I think it’s the one that’s the most recent,” Pennington said, her masked face close to the computer, scanning its array of numbers.
“Team, you’ve got two patients, and you’ve got one ventilator, and you’ve got to make a decision.”
Eric Goralnick, medical director for emergency preparedness, Brigham and Women's Hospital
The other score came with other challenges. This one was a number representing the seriousness of your health conditions overall and how many more years they’d give you — and that’s where the question of the man’s diabetes came in. Was the injury in his retina bad enough for it to be labeled “end organ” involvement, meaning damage in one of the most important of tissues?
“That comorbidity could tip you in any direction,” Goralnick explained, from where he hovered, on the edge of the group.
Even the smallest of medical definitions, easy to take for granted — just another entry in a professional’s inner dictionary — now carried an enormous moral weight. You could hear a hint of anxiety in Pennington’s voice.
“I’m not feeling great about retinopathy being end organ,” she said, shifting in her seat. To her, “end organ” sounded like something serious and life-sustaining, like the kidneys or the heart.
“My rationale would be that that implies advanced and longstanding diabetes,” James Rathmell, head of the anesthesiology department, replied.
“A sign of damage in other organs,” someone added.
There were murmurs of agreement, and the man’s diabetes officially became a major health issue.
This man’s place in the queue for the one remaining ventilator depended on the combination of those two scores: how sick he was right at that moment, predicting short-term survival, plus his underlying health landscape and what it said about how much longer he had to live in general. All of that, crunched into a number between one and eight.
If two patients’ were tied, the triage team would resolve the impasse by prioritizing youth, and by giving attributing extra life-years to women who were at least 24 weeks pregnant. There would be continuous reassessment as people’s conditions worsened or improved. As with a court, determinations could be appealed.
Everyone there knew that it was an attempt to make as fair as possible a situation that would, if it arose, inherently end up unfair. As White, at the University of Pittsburgh, put it, “The prioritization of patients in the face of not enough ventilators — there’s no perfect answer. There are only bad options, and the goal is to choose the least bad option.”
In Massachusetts, teams at the Brigham and other hospitals were learning this sort of structure, built on the solid foundation of numbers, to exclude no one from the possibility of care, no matter their race of ethnicity or ability to pay — an attempt at making a moral choice empirical. But devising a way of distributing care during a catastrophe couldn’t erase the long history of pre-pandemic inequalities. If being poor or of color meant you were less likely to have access to health care, then chances were that your chronic illnesses would not be kept in control. This man’s diabetic eye damage was a case in point: Some researchers have reported that the condition is nearly twice as prevalent among Hispanic patients than it is among those who are non-Hispanic white.
“It will de facto preference some people over others. If you’ve been able to get good health care all your life, you’re not going to have a lot of premorbid problems,” said Martha Jurchak, executive director of the hospital’s ethics service, in an interview a few days before the exercise. She knew that putting this plan into practice would be traumatic for everyone involved. But in its most basic expression, the idea wasn’t new; rather, it was a staple of American health care. “We can talk about this in terms of rationing, although the word is anathema in U.S. culture,” she said. “We have been doing rationing, even though we don’t call it that. You know, who’s got health care, and who doesn’t — that’s a form of rationing.”
She worries about how longstanding ethical failures of American medicine will creep into such a highly charged moment, but also knows that the country hasn’t been able to fix them before the pandemic. “That’s not going to happen before the Covid-19 wave crashes over us,” she said. “The best we can do is work with the system we’ve got.”
Nobody had consulted the phlebotomist. People seldom did. She wished they would, whether or not there was a pandemic going on. Still, she knew her work was important. “A lot of the time, the blood has the answers,” she said, with a hint of a Caribbean accent.
She was strolling along the sidewalk, on a break between blood draws, enjoying the sun near the conference room where the triage team’s training took place. Though she didn’t want her name used, she did want to be heard. She needed to be here at the hospital, not only because her work was considered essential, but also to feed her kids. She worried she would get sick herself, and bring the virus home to them. She knew the hospital hadn’t reached a point where there weren’t enough ventilators, but she worried about how those decisions would be made if things got that bad.
“If a health care worker falls sick caring for a sick patient, absolutely they should be prioritized for things like a ventilator,” she said. “What’s going to happen if everyone falls sick? Who’s going to take care of who?”
The question had come up again and again over that weekend, as the Brigham had begun training clinicians on the proposed rationing system. Health care workers around the country were already scared about their risk of being infected with the coronavirus, especially given shortages of protective gear. They wanted to know how they would rank if they fell ill.
But hospital leaders couldn’t give a definitive answer, not until the state adopted and released its guidelines for allocating scarce medical resources, which had been under discussion for some 14 years.
“If a health care worker falls sick caring for a sick patient, absolutely they should be prioritized for things like a ventilator.”
A Brigham and Women's phlebotomist
In 2006, the Massachusetts Department of Public Health had mustered up a pack of experts in medicine, law, and ethics to talk over the most moral way to guide the commonwealth through a disaster. What to do if there was a shortage of antiviral medications? How to prioritize who got critical care? Should the state seize private assets? By the end of 2009 — the year of the H1N1 “swine flu” pandemic — the group had pinpointed some goals and principles to help steer how scarce medical supplies should be allocated. A draft of the initial guidance emerged, and then appears to have sat around for years.
Work resumed, Goralnick recalled, around 2016 or 2017. There were more consultations, with bioethicists and nurses, doctors’ associations, and hospital councils. There were more recommendations woven into a new iteration.
“We came up with a pretty robust draft,” said Jurchak, who took part in one of the focus groups about these guidelines. “But despite urging, they were never enacted as public health policy in Massachusetts.” (Goralnick, who sits on the state committee developing guidelines for the Covid-19 pandemic, directed questions about its work to the state health department, which did not respond to repeated requests for comment.)
You could imagine why officials might be reticent. It wasn’t so different from the middle-aged mulling over end-of-life wishes, somehow never getting around to signing the forms. For state agencies, there was added complication of potential blowback: If government agencies were planning for potential shortages, why not just forestall the shortages?
“It’s emotionally very difficult for everyone to move this forward,” said Eappen, the chief medical officer. “You almost don’t even want to think about it, that we’re not going to have enough resources. ‘Why would you even want to plan for something like that where you don’t want to ever get there?’”
Then, Covid-19 hit. What had seemed unimaginable to some was, all of a sudden, frighteningly real. Committees again coalesced. They debated recommendations anew, specifically for the current crisis. One of the thorniest questions that the Brigham and other hospitals talked about was whether health care workers should be bumped up in the scoring system, to give them a higher likelihood of receiving treatment for an illness they may have contracted on the job. On the one hand, saving their lives could allow them to eventually save more lives in turn, if they themselves survived. On the other, that meant prioritizing some who were privileged enough to be employed, who had training enough to work in a hospital.
First, the company-wide committee at the Brigham, Massachusetts General Hospital, and other affiliated medical centers suggested giving extra weight to health care workers, Eappen said; then, after seeing the reaction from others in the hospital, they took that provision out. The final determination, though, had to come from the government.
So that was what leaders told some 2,500 clinicians and others during live webcasts throughout the first weekend of April. “We are awaiting guidance from the state,” said Nicholas Sadovnikoff, co-director of the Brigham’s surgical ICU, during one such training. “Some states have developed policies which do provide a benefit to those who are on the front lines; others do not. It’s a very complicated question. It’s not at all straightforward. But for the time being we are not designing a system which favors its designers.”
The hospital wanted to have everything ready, to be prepared for the worst. If there were shortages, there would be security guards near the patients’ rooms, in case triage news caused agitation. There would be a mandatory mental health check-in at the end of every score-calculating shift, to make sure the volunteers weren’t becoming withdrawn or angry or, in an extreme situation, suicidal.
Being picked for the triage teams was both a source of honor and trepidation. Physicians, nurses, and administrators felt flattered to be entrusted with such an important task, but horrified that things might come to that. “In this pandemic, I have not been on the frontlines,” said Christopher Fanta, 72, explaining why he volunteered. To him, it felt akin to a moment his grandchild might return to, the way other generations had asked, “Where were you during the war?” The thought that he wasn’t contributing to the anti-Covid effort was getting to him.
As a pulmonologist and critical care specialist, he’d had plenty of discussions with patients about withdrawing care, but they’d been very different from what he was contemplating now: “Those are conversations that we’ve had at the point where we’ve done everything and tried everything, and death seems inevitable and is just around the corner.”
The trainings and exercises had their fair share of workplace camaraderie — masks momentarily lowered so clinicians could take a sip of coffee, laughs when someone took a break from the videoconference and the room filled with flute-heavy hold music. But their task was a painful one to even think about, let alone rehearse.
The whole scenario, it turned out, was a trick question. The triage teams had been charged with scoring two patients — a 23-year-old with epilepsy, and this elderly man with diabetes — but it turned out that only the older man actually needed a ventilator. “You could have saved a lot of time by just intubating the 85-year-old,” Sadovnikoff said.
Yet even that wasn’t straightforward. “If we have one vent, and we anticipate we’re going to get five more admissions during the course of the day, then we don’t want to just intubate the patient if he’s the only one who needs it right now, right?” another doctor interjected, from behind the huddle at the computers.
“Well, you might want to intubate him and then see what happens, because otherwise he’s going to die,” Sadovnikoff replied.
“A bird in the hand is worth two in the bush,” said another physician. “You’re going to treat the patient that’s in front of you, you’re not going to hold resources for hypothetical —”
“That’s not how we’ve been discussing it, though,” Eappen said, cutting him off.
Rather, the triage team would look at a dashboard showing projections of how many patients would arrive that day, how many would need an ICU bed, how many a ventilator. Taken together with the equipment they had available, they would know where the cutoff was. That didn’t mean those who didn’t make the cut wouldn’t get care. They’d get everything the hospital could offer — except for whatever had become scarce and subject to this calculus. Still, it was hard to stomach. Some triage team members couldn’t help imagining themselves running the numbers on their colleagues, wondering if they’d be able to go through with it.
Later that very day, the state guidelines would come out, finally providing an answer to the question that had stirred so much angst among hospital employees. “Individuals who perform tasks that are vital to the public health response … should be given heightened priority,” the document would say. That included health workers.
But the release would prompt other questions. Some lawmakers would echo the worry expressed by the hospital’s own ethicist: That the state’s scoring algorithm would only deepen the discrimination against those who already had trouble getting health care, that it would turn their poorly treated conditions into a mark against them if they ended up needing machines or medicines that had grown scarce.
Within less than a week — on Monday, April 13 —the state committee would re-convene to reconsider these policies, just in case, even as the Brigham still had plenty of open ICU beds and vents.
That was all in the future. For now, the triage teams were huddled in a conference room, zeroed in on this one patient, with his diabetic eye damage, his high blood pressure, the labored crackling of his breaths, trying to untangle the nightmare scenario of not being able to give him the care they wished they could. Somewhere — elsewhere in the hospital, or at home, or in the grave — was the man himself, unaware that arriving at a different moment might’ve entailed such a different sort of treatment.