Who Should Get the Last Ventilator?

COVID-19 patients could overwhelm many U.S. health care facilities. Doctors and hospitals need guidelines for allocating scarce resources.

A ventilator, half of it fading away
Photo illustration by Slate. Photo by Getty Images Plus.

This piece originally appeared on the website of Issues in Science and Technology.

The COVID-19 pandemic has overwhelmed health care facilities in New York City and seems likely to do so in other U.S. cities. Health care providers are having to decide who among their patients will receive limited resources, including but not limited to ventilators.

Deborah Birx, the esteemed White House coronavirus response coordinator, said this is not the time to talk about such things, at least not in public. “There is no situation in the United States right now that warrants that kind of discussion,” Birx said. “You can be thinking about it in a hospital, certainly many hospitals talk about this on a daily basis, but to say that to the American people, to make the implication that when they need a hospital bed, it’s not going to be there, or when they need that ventilator, it’s not going to be there—and we don’t have evidence of that right now.”

Few front-line hospital workers or bioethicists would agree. Across the country they plead for guidance now, before hospitals are overwhelmed. Dozens of hospitals, states, and ethicists have proposed guidelines for resource allocation, with remarkable overlap on many issues. I myself am now chairing a committee that is advising Wisconsin on the development of guidelines for allocation of ventilators during the pandemic.

Across these many efforts, the most important point of agreement is on the need for the guidelines themselves. Decisions about who lives and who dies should not be left in the hands of the attending physician. Should the directors of intensive care units get to decide whether their spouse or the homeless man lives or dies—Bill Gates or John Doe? Guidelines are necessary to assure that such life-or-death decisions are not subject to whim, unpredictable biases, and conflicts of interest.

How should such guidelines be developed? Health care workers, hospital administrators, and bioethicists share the conviction that the process should include input from the broad community of potentially affected stakeholders, particularly groups that are traditionally disadvantaged when scarce resources such as health care are distributed. This is important not only if the guidelines are going to be perceived as fair but also if they are going to be accepted, to avoid legitimate protests.

Community involvement in developing guidelines for allocation of ventilators can be achieved through hospital or regional bioethics committees with appropriately diverse membership, through online public meetings and surveys, or ideally through in-person meetings in the community—although that is obviously not feasible in the current crisis. Involving the community will not produce unanimity. There will be winners and losers. The goal is to attract and deserve support for the claim that the process by which the guidelines were developed was fair, or at least it made a good faith effort to be fair and inclusive.

In addition to obtaining community input in the development of guidelines, the community needs to be informed of the guidelines, however they are developed. They can’t be a secret, if there is to be acceptance.

The first piece of this public education is to make clear that there must be a shift from the traditional ethical commitment to the interests and preferences of each individual patient, to a commitment to the interests of the community. In times of abundance, society can afford to offer a ventilator to a dying patient with little chance of survival, but not when another patient will die who had excellent prospects for a long life.

Those involved with developing guidelines generally agree on several key issues. First, no one should be able to buy their way into an ICU bed when others have the same need and the same prospect for benefit. This would not be perceived as fair by the overwhelming majority of the American people. There have already been justified complaints about celebrities getting priority for the diagnostic test for COVID-19.

Second, whatever the rules are, they should apply to all patients, not just those with COVID-19 infections.

Third, although the issue is complex, there is also general agreement that the goal should be to save the greatest number of lives. Thus, the first criterion for being put on a ventilator should be a reasonable expectation that the patient will survive this illness. Even when resources are plentiful, it is not sensible to institute treatments that have little or no likelihood of benefiting the patient.

This principle, that scarce ventilators should not be wasted on patients who have no likelihood of benefiting because they will die soon anyway, explains one reason for opposing what might seem to be a common-sense, first-come, first-served guideline. Such a policy would result in resources being wasted on patients with no prospects of benefiting from the treatment.

A more difficult question is whether the principle of not “wasting” a ventilator should be applied to a patient who was appropriately started on such treatment, but has now deteriorated and appears unlikely to survive. A decision to withdraw life support from such a patient might be defended on principled grounds, but in practice many physicians would have difficulty telling the family of such a patient that the ventilator is needed for a new patient—or, worse, asking them for permission to take it away.

Once physicians have determined that a patient has a reasonable chance of surviving this illness, there needs to be a way of breaking a tie, when two or more patients are both likely to survive. On this point, there are different views.

Some people would argue that the goal should not be to maximize the number of lives saved, but the number of life-years saved. In this view, if a 60-year-old patient has the same chance of surviving as a 20-year-old, the younger person should get the ventilator, as this will result in more life-years.

Also in defense of this tiebreaker is the “fair innings” principle. Just as each baseball team is entitled to get to be at bat for the same number of innings as the competing team, the 20-year-old should get the same chance to live his or her life as the 60-year-old, who has already had his opportunity.

A further ethical complexity arises, however, because for a person of a given age, life expectancy is also related to social determinants of health. Patients who were born and grew up in poverty are more likely to have a life-shortening medical problem, through no fault of their own, due to lack of access to basic health care, environmental hazards, crime, hazardous work or unemployment, and so on. These disadvantaged patients are more likely to come from racial or ethnic groups that are victims of other forms of wrongful discrimination, or to have disabilities that have subjected them to discrimination in the workplace, in access to health care, and so on. The life-years approach to allocating care cannot easily accommodate such considerations, but can potentially exacerbate inequity by penalizing a patient for past disadvantages.

To avoid this problem, some people suggest avoiding the life-years approach and requiring only the likelihood of short-term survival (for, say, one year after discharge) as a criterion for receiving treatment, and breaking the tie by assigning the ventilator randomly among those who meet the one-year standard.

“Randomly” could mean first come, first served, but that too could result in an unfair advantage for the wealthy, who are more likely to find their way to the tertiary care hospital than the less well off, due to stronger connections with an influential primary care provider or other benefits of socioeconomic wherewithal. Flipping a coin would reduce this possibility, though it wouldn’t eliminate it, since the poor patient obviously can’t be treated if he or she hasn’t arrived at the hospital, for whatever reason.

Some guidelines propose giving first responders and health care workers extra points as part of a tiebreaking system. One claim here is that this will help save more lives, if the doctor or nurse can return to duty. It’s unclear how many survivors of a COVID-19 infection severe enough to require a ventilator will return to work in a timely way. There are also concerns about whether police, ambulance drivers, and others should be included in this privileged group.

A separate reason for giving priority to health care workers is the claim that they deserve this support, as a reward for courageous and often heroic service and as an incentive to encourage them to stay on the job or come out of retirement.

What should be obvious by now is that there is no right answer, no ethically optimal solution. In turn, there will never be unanimity on the details of any particular allocation system. The process of developing guidelines will thus be subject to posturing, to inflammatory language about “death panels,” or the inevitable and inappropriate comparisons to the Holocaust, and there will be the possibility of subsequent litigation.

But if the guidelines are developed in an open way, with good faith efforts to consider the perspectives of the wider community, few people will disagree that it’s better to have guidelines than to leave decisions to individual physicians. Nor is every hospital on its own in coming up with an allocation process; already there are numerous examples of basic principles and guidelines to work from, such as those recently catalogued by the Hastings Center, a nonpartisan, nonprofit bioethics group. But each institution anticipating the arrival of critically ill patients with COVID-19 has to decide what its plan will be. Having no plan will be more difficult to defend, and is more likely to be inequitable and divisive, than having an imperfect one.

Future Tense is a partnership of Slate, New America, and Arizona State University that examines emerging technologies, public policy, and society.