Ventilators, Guidelines, Judgment, and Trust

Abstract

Covid‐19 confronts us with tragic choices, in which every option is unacceptable. On the New York State Task Force on Life and the Law, I worked on guidelines for such situations. We did not envision the scale or character of Covid‐19. To minimize fear that the decisions made in these situations might be unfair, we all must know what guidelines or mandates inform them. Only with transparency about how decisions will be made, by whom, and according to what requirements can we have confidence that fairness prevails. We now face many related questions about process, goals, leadership, and trust. For example, how might ethical guidelines evolve as scientific understanding advances? Should guidelines vary with different venues? And if, as I have argued, judgment is necessary even with the best of guidelines, how can we prepare clinicians to make good judgments? Are there implications for better training of personnel in nonemergency times for what they might face in the worst of times?

perspective

Among the terrors caused by Covid‐19 is the fear that one's fate, like the fates of those one cares about most deeply, is not only out of one's control or even one's influence but in the hands of unknown others whose decisions may be arbitrary, uninformed, impulsive, or biased in any of a vast number of ways—personally or by the systemic context in which the decision‐makers function. In medical crises, some choices are extremely hard emotionally, even heartbreaking, but clearly necessary. Covid‐19 confronts us with a different kind of decision—the tragic choice, in which every available option is unacceptable. The classic examples of such situations involve choosing who gets a ventilator when too few are available and deciding when it might be justified to remove a patient from a ventilator needed by someone with far greater recovery prospects.

In 2007 and again in 2015, as a member of the New York State Task Force on Life and the Law, I worked on guidelines to assist in such situations. We tried to envision all the possibilities, but our foresight fell short. We knew that serious epidemics are a recurrent phenomenon, and that sometimes they evolve to pandemic scales requiring tragic decisions. But we did not envision the scale or character of Covid‐19.

For anyone in a position to make a tragic decision, there is no escape. One must act, knowing the result will be horrible whatever one does, and that therefore profound disappointment and perhaps great anger will result. We must have deep empathy for those unlucky enough to be such decision makers; they are typically in such positions only because they courageously strive to minimize harm, even as they face harm themselves.

To minimize the fear that these decisions might be made unfairly, we must know what guidelines or mandates have informed them. We would expect such guidelines or mandates to forbid considering a patient's politics in making a medical decision. We hope a patient's wealth is also irrelevant. We wonder whether a patient's age is relevant, and if so, how. We hope disability is not relevant, and we aren't sure about a patient's religious convictions. Only with transparency about whether and how such parameters matter can we have confidence that fairness pervades the process. And if we do not have that confidence now, about Covid‐19, it will be difficult, at best, to have it for future public health crises. Such confidence requires transparency about how decisions will be made, by whom, and according to what guidelines or requirements.

A further value in making guidelines transparent is that they can then be scrutinized more broadly and revised in reaction to evolving scientific evidence and deeper ethical insights. Guidelines in Massachusetts were developed, as the Boston Globe reported this April 20th, “to help shape the decisions hospitals would make if they do not have enough life‐saving equipment, such as ventilators, to serve every patient in need.” Public reaction to those guidelines prompted revisions reflecting stronger protection of the values of equity and social justice (see A. Rosen, “State Revises Guidelines for Who Gets Ventilators in Crisis, following Complaints about Equity”). Transparency thus also enables visible improvement, which itself enhances trust. As we consider the impact of the new virus on prisons, nursing homes, Native American reservations, food processing plants, homeless populations, and more, we know that guidelines and how they apply must be revised periodically. Some of our assumptions about infectious disease may also need revision. The process of assessing guidelines in light of rapidly evolving information is necessarily dynamic.

As a strong advocate of developing and disclosing clear guidelines for the use of ventilators, I am asked by local, national, and international media what standards I favor for making tragic choices. I reply that ventilators must not be thought of in isolation, but as part of an ecosystem including trained staff and necessary materials. If any of that array is missing, the ventilators are to no avail. More importantly, I emphasize that our tragic choices cannot be dispelled or resolved by any decisional algorithm.

One reporter invoked the economist's beloved metric of quality adjusted life years, seeking agreement about this hypothetical: He envisioned an octogenarian with Covid‐19 and various comorbidities on a ventilator, with uncertain outcome. An otherwise healthy forty‐year‐old needed access to a ventilator, but none was free. He thought it clear that to maximize benefit to public health, the right decision, tragic to be sure, would be to move the elderly person to the best possible palliative care and place the younger person on the ventilator. And this would be required by guidelines that make probability of recovery determinative.

The reporter was audibly startled by my response. If the elderly person is Ruth Bader Ginsberg, I replied, a credible case can be made that maximizing public health requires doing everything possible to keep her alive, even if for just a year. I did not insist on this decision— only that a credible case could be made for it. And no algorithm, guidelines, or rules could determine the merits of that case. That would require a judgment in which one's values as well as the empirical evidence would be brought to bear. This is so in every tragic choice situation; there is always an ineliminable need at some point for judgment that no algorithm can replace.

That guidelines are sometimes insufficient does not mean they are useless or even not essential. The moment when one faces a tragic choice is no time to begin thinking about how to decide. Information is evolving rapidly about the science, as we explore reinfection, immunity, and the possibility of second and perhaps subsequent waves. As Dietram Scheufele et al. note (see “How Not to Lose the COVID‐19 Communication War,” in Issues in Science and Technology), “scientific facts and uncertainties are moving targets.” That is also true of values and guidelines for tragic choices in pandemics, and of how we should manage their interplay with evolving scientific understanding.

We face many questions. If there are to be guidelines, how much diversity and how much uniformity are best? Would we want national guidelines, rather than state‐level guidelines, given regional differences and the absence of trust in national leadership? Can we envision better processes for the revision of guidelines in response to our evolving understanding of epidemiological and societal information? Ought we to consider new standards of consent or risk assessment as experimental treatments, such as high‐flow nasal cannulas, or new drugs are made available for patients in desperate situations? What are the appropriate goals and strategies for public education about medical decisions in an era of politicized communication through social media? And if, as I have argued, judgment is necessary even with the best of guidelines, how can we prepare clinicians to make good judgments? Are there implications for better training of personnel in nonemergency times for what they might face in the worst of times? Better times will come again one day. But so might the worst of times.