Triage – this is the sorting of patients in groups to be treated preferentially or subordinately when the mass of patients vastly exceeds the available resources. It has always been a delicate and stressful procedure. The Italian Society for Anesthesia, Analgesia, Resuscitation and Intensive Care Medicine (SIAARTI) has recently given recommendations to intensive care physicians who are currently no longer able to provide lung ventilators to all Covid-19 patients in need. The aim of those recommendations was to relieve the practitioners of the burden of taking personal responsibility for the selection decisions and to make the criteria explicit and communicable. The criteria, or so the authors propose, should also be made accessible to those affected and to their families to maintain trust in the public health service.
These concerns are understandable. If I were a patient or family member in Italy currently, I would not find it appropriate to involve the exhausted medical staff in discussions about their allocation decisions. If the criteria – more precisely: recommendations on criteria – are now made public, one must however be allowed to comment on them. This is what I intend to do here. My trust in the (Italian) public health system has not been raised by this publication. Rather, it suffered. According to Italian media reports, there are physicians on the ground who have publicly refused to follow the recommendations. I was relieved to read this. I wonder, though, how many of them did.
The central passage of the paper reads as follows: Resources that could get very scarce should first be reserved  for those who are more likely to survive, and secondly  for those who can gain more years of saved life („più anni di vita salvata“), with a view to  maximizing the benefit („dei benefici“) for the greatest number of people. In the third part of this formulation (numbered by me) one can easily detect the maxim of utilitarianism. The second part specifies the relevant utility values as years of saved life, i.e. the years of life which a patient is expected to live if she receives treatment. Finally, the first part of the formulation presumably does not refer to the absolute probability of survival, but to the increase in probability associated with the treatment. Otherwise you would have to favor mildly sick patients whose survival would be certain with lung ventilation, but who already have a good prognosis without.
One should be appreciative of uncertainties and disagreements within the medical profession regarding questions of distributive justice. The topic, as I said, is delicate and there is a lot of controversy in the specialist literature. In addition to medical experts, professions that publish on the topic include the normative disciplines, i.e. ethics and law, but also health economists if they choose to give normative advice, which they often do. All of this is not easy to keep track of. However, the medical group responsible for the SIAARTI recommendations has not expressed any uncertainty. It referred to „disaster medicine“ for which „ethical reflection“ had developed specific instructions.
The cited formulation can, however, not be seen as following an established consensus on dealing with existential scarcities. This holds true not only for the debate on the allocation of life-saving resources in general which also includes the problem of allocating donor organs. It also holds true for the interdisciplinary literature on triage in disaster medicine, which differs in certain respects from the more general debate. Not every scarcity, even if it is existential, is classified as a disaster. Only in the event of a disaster – a suddenly emerging situation in which masses of people in need exceed by far the amount of resources regularly kept in place – the procedure known as „triage“ has established itself. This also includes a mass influx of wounded persons in wartime. The traditional rule here is to use resources in such a way that as many people as possible survive. This is not the decisive criterion when it comes e.g. to organ allocation. You can see that from the fact that people with double transplant needs are also treated when it is their turn according to the usual rules.
Why did the Italian medical professionals replace the criterion of maximizing the number of survivors with the criterion of maximizing the number of years of life saved? I don’t know. Perhaps they chose the formulation by themselves, perhaps the passage is backed by recommendations of an Italian or foreign medical association that I am not familiar with. Or perhaps the authors considered some casually perceived contributions from the non-medical professions as authoritative which are in fact highly controversial. In any case, the change of the criterion is disquieting. Not only might a triage-untypical distribution result if the new rule were implemented consistently (2). Above all, the change indicates that the complex reasoning behind triage has been misunderstood (3).
If the rule is that scarce resources are to be used in such a way that as many affected people as possible survive the disaster, the traditional triage procedure follows. It provides for a division into four groups: priority is given to severely affected patients who will certainly or very likely die if they don’t get the treatment but who have a good prognosis if they do. In the second row, patients are treated whose chances of surviving the disease are significant even without treatment but would increase significantly under treatment. Not treated are lightly affected patients who have a good prognosis even without treatment. Likewise not (or only palliatively) treated until the situation relaxes are those seriously affected patients who have a poor prognosis even under treatment.
If the rule is to maximize saved years of life, the picture changes in parts. One would have to differentiate within the groups mentioned according to age. That is not part of the traditional procedure. And in the case of large age differences, older people in the first group would have to be put aside in favor of younger people in the second group. Specifically, and with a completely fictitious exactness of the probability data: In the case of a 60-year-old with a statistical remaining life expectancy of 20 years who would without treatment definitely die (0% probability of survival) and with a 70% probability survive if treated, the use of resources would arithmetically yield 14 years (70% of 20). This patient would have to give way to a 20-year-old with a remaining statistical life expectancy of 60 years who would already likely (70%) survive without but certainly (100%) with treatment. For this patient, the use of resources arithmetically yields 18 years (60 minus 42, i.e. 70% of 60).
I hope the example is clear enough. It shows that the criterion of maximizing the years of life saved might require physicians to radically break with all they have been trained in terms of medical need. The deferred patients and their relatives would be expected to give up their own significant chances of survival in favor of augmenting the chances of people who already have substantial chances without this act of solidarity (or whatever one should call this). Why should they do that? And how is it that something like this can be presented as a result of „ethical reflection“?
This brings us to the matter of reasoning. As to the traditional procedure, it should, first, not be overlooked that patients for whom the treatment in question is indicated and to whom it offers an (albeit small) chance of survival (group 4) are expected to give up their chance in favor of other patients as well. However, in contrast to the 20-year-old in our example, the prioritized patients in group 1 are not significantly less in need. The traditional procedure however also includes the deferral of group 4 patients compared to group 2 patients. This imposition cannot be justified by the idea of solidarity with significantly more seriously affected people. It can (and should, if one wants to stick to the triage practice) be justified with the thought that as long as a catastrophe has not yet occurred and no one knows which group he ends up in, we can all supposed to be interested in maximizing the number of survivors. We are then held to this presumed consent to the rule even when the disaster puts us in group 4.
This is not a utilitarian justification. A utilitarian justification of the same rule would argue that a saved human life is „valuable“ and that therefore two (other) saved human lives are even more valuable. From this sort of reasoning, and only from this one, one can easily move on to the modification of the maximand which the SIAARTI recommendations undertake: If (and as long as) life is valuable, a longer life is more valuable than a shorter one. Are not the resources used in the most value-productive way („most efficiently“) if you maximize the number of years of life instead of the number of survivors? This way of speaking is suitable for commercial companies that have owners who own the produced values. It is not suitable for healthcare. Humans have no owners. For the public sector, two human lives are not „more valuable“ than a single other one, and of course twenty-year-olds aren’t more valuable than sixty-year-olds.
These maxims should and can be maintained in the Corona crisis. This presupposes, however, that in the case of a triage becoming necessary one publicly adheres to the non-utilitarian logic of justification. The best way to do this is to describe distributional issues in health policy, as legal scholarship has always done, not in terms of utility or values but in terms of rights. As every lawyer knows, the rights of individuals do not automatically give way just because they are pitted against the rights of several others. Rights work „non-aggregatively“. In times of shortage, they do not have to be maximized, but rather specified in a just manner. This, also, is burdensome. But one can do it without getting involved in reflections on how much residual value a human life might have.