Rationing occurs in the United States. Anyone on a waiting list for a kidney or liver transplant knows this. But in just two weeks, the need to select which severely ill coronavirus patients will receive treatment — and which will go without — will be more acute than anything Americans have ever experienced or probably imagined. Even in the most favorable scenario, 39,727 intensive-care-unit beds will be needed on April 15, the projected peak date for the covid-19 pandemic in this country. Only 19,864 such beds will be available. There won’t be enough ventilators, either. These projections, from the Institute for Health Metrics and Evaluation at the University of Washington, optimistically assume that social distancing measures are maintained for weeks. The actual shortfalls are likely to be worse.

How should these choices be made? There is no authoritative source of ethical wisdom. But the issues have been examined intensively for two decades, initially as a response to the prospect of an avian flu pandemic. That scenario was frightening enough to prompt authorities to convene task forces at the global, national and state levels to survey available resources, identify shortfalls and set priorities. I participated in deliberations at the World Health Organization and in Massachusetts, both of which gave explicit attention not only to equipment stockpiles and treatment protocols but also to identifying and trying to resolve conflicts among values. Clinicians, epidemiologists, patient advocates and ethicists had seats at the table. We’re not bound by any of the recommendations issued by the many task forces, but we’re not starting from scratch.

Bodies like these perform valuable functions, even though they do not always reach similar conclusions. They offer a chance to think critically about difficult choices before crises occur. If adopted, their recommendations and guidelines relieve caregivers of some of the burden of making the determinations themselves, allowing them to focus on how best to implement them, as their training has prepared them to do. And though none of these bodies can claim to represent the publics they serve — they’re not elected, after all — they make clear the importance of devoting time and expertise to studying and debating these questions in depth, well ahead of the need.

This potentially legitimizing function is important because what caregivers must do in a pandemic crisis requires a fundamental shift in moral perspective: They go from dedicating themselves to the good of the individual patient to striving for the best outcomes among many patients. No clinician finds this reorientation welcome or intuitive. The Hippocratic traditions that inform a clinician’s conscience have to be suspended; if they are not, the results for many who need care could be disastrously worse than they need to be. To avoid this, we all should understand why a less individually focused perspective is required — why it is imperative.

Most advisory reports from the various task forces that assembled in anticipation of pandemics specify that the overall goal is to save the most lives possible. Their authors generally hope that if they beat the drum loudly enough, the appropriate authorities will stockpile necessities and drill their staffs so that fewer tragic choices will have to be made. But if resources turn out to be inadequate when a pandemic does occur, the save-the-most-lives goal then becomes a principle of selection among those needing care. This is not the only choice we might make, but it has great merit. What we most fear in a pandemic is death, and survival is what we hope for. It is a good that can be wished for every person without any need to decide that some lives are more worthy of being saved than others. And by choosing this as our common goal, many other decisions fall into place.

For example, the save-the-most-lives strategy explains and justifies a priority on protecting and treating essential caregivers, a shortage of whom may rival the dearth of ventilators as a cause of avoidable deaths. They must be protected, both to induce all to pitch in and to keep them in the action; and they must be treated when ill, both to restore them to the front lines and as a further incentive to keep going. Some priority can be extended for similar reasons to certain other classes of essential workers: police officers and others needed to maintain civil order; firefighters; those essential to the provision of food, safe water and public utilities, especially people who can be restored to work.

What about the rest of us? We must reject what often seems optimal in ordinary times, such as first-come, first-served, or even a lottery. These choices risk filling up ICU beds with patients unlikely to emerge alive, at the cost of the deaths of multitudes of patients who are likely to survive if given temporary care. These latter must come first, even if it means holding open an available bed on pain of death to a lower-priority patient, and being prepared to withdraw the use of a respirator from a patient unlikely to survive.

The save-the-most-lives strategy faces two plausible objections. The first questions how much survival counts, and whether a patient projected to survive for a shorter period should have the same priority as one likely to live much longer. The second asks whether, on balance, younger patients should be favored over older ones.

Neither question has an easy answer. But for the purposes of forging a rough consensus, precision may not be necessary. Few will want to flip a coin to choose between a patient unlikely to survive more than a month or two and one who might enjoy many years. Yet equality and solidarity might be best achieved by declining to make further distinctions between patients on these grounds.

Should priority be given to the young? They will generally be less likely to need intensive care, since covid-19 puts the old at greatest risk. But those young patients who do need an ICU bed generally will be favored by a save-the-most-lives strategy if they are most likely to regain health. As a group, the young are more resilient. In specific cases, however, a pure save-the-most-lives rule could favor a 65-year-old over a more severely afflicted college-age patient. Many of us would root for the younger person, both because she might live longer if she pulls through and because the older patient has already enjoyed the decades of life that may be denied to the other. This modification of the save-the-most-lives strategy might be achieved by accepting a “bias” in favor of the young, and not merely as tiebreakers. Ideally, this might be carried out without imposing formal age limits.

Choosing a save-the-most-lives strategy means we will not take into account a host of other considerations, including some that would matter in ordinary clinical decision-making. No preference will be given to those most likely not merely to survive but also to enjoy the best health, let alone the highest quality of life. Nor does the strategy excuse abandonment of the disabled. Rather, it directs resources toward providing the greatest assurance possible that each of us will achieve what we most value: our own survival and that of our loved ones and fellow citizens.

No discussion of the ethics of rationing during a pandemic should be limited to how to choose among patients who present themselves for treatment. Pandemics and other health emergencies tend to favor the well-to-do, and this will be true this time around, too. But we can and must make every effort to identify and care for those utterly unable to protect themselves from infection or to make effective demands on the system, and providers must be assured that fulfillment of this duty will not put them at risk of bankruptcy. Saving the most lives means caring about every life, even when we must choose whom to save.