Tragic Triage, Thinking About – and Planning for the Unthinkable.

What if…

What if we don’t manage to “flatten the curve?”  What if the health care system, your hospital is swamped with very sick patients needing access to ICU beds and ventilators?  What if the usual systems of management, sharing resources, transferring patients out, are also overburdened?  What if there are way more patients who need access to, e.g. ICU beds than there are beds available – what will/should we do?

Current practice

The system is currently designed to handle (more or less) the peaks expected as a result of, for instance usual demand plus seasonal influenza.  A patient will be admitted to an ICU if their clinical condition warrants it – if they will benefit from the admission.  The key point is that this is a clinical decision – to be made by intensivists, the clinical experts, according to their established standards of clinical practice.  The expectation is that if I need the care, I will get it, and non-clinical factors are irrelevant.

There is some fuzziness at the edges of this clinical practice.  If the objective of ICU care is to save my life does it matter for how long or of what quality? For instance, would it be a “successful” ICU intervention if I recovered to be discharged from the ICU, only to die a week, or a month later?  Would it be a successful ICU intervention if I survived but never regained consciousness?  Should ICU interventions be used to “prolong the dying process?” For the most part, these are considered “clinical” questions left to the clinical experts – and there is room for some variation in practice, especially in the face of sometimes highly vocal demands from families and loved ones.  But it is easy to see that these “clinical” judgments involve a value component, particularly in a publicly funded system, where any decision to offer one type of care necessarily entails some corresponding decision not to do something else.   (At the system level resources put into, for instance, ICU beds, are not available to be used to provide mental health services or…)

So, back to our what if scenario, imagine we have exhausted the usual clinical criteria and we still have way more patients who need ICU beds than we have beds available.  What then?

The first step seems to be to expand the fuzziness of the “clinical” criteria.  In a situation of crisis, it is a lot clearer that it would be inappropriate to offer ICU care to someone with other underlying conditions that mean they are unlikely to live long anyway.  If, even in the best case, the patient’s “prognosis” is poor, the tragically scarce resource (the ICU bed) is better used by someone else.  This is a judgment about expected time-to-live, it is a clinical judgment(estimate/guess) but that part is clinical, that it matters, or that it is relevant however, is a matter of value, more life is better than less.  And even if we all agree that this is a value we share what should the cut off be? A month, six months a year…And of course, it is worth making obvious point that the longer time period we allow here the more we are biasing the decision-making against the relatively old.

The second value judgment that lies behind a clinical assessment of “poor prognosis” is the idea of quality-of-life.  What if I survive but never regain consciousness?  What if I was not conscious in the first place? What if I am currently cognitively impaired (for instance I have dementia) such that I do not know myself as myself or recognise those who love me as loved ones?   What if I am tortured in my dementia? (Or what if I am not – what if I am the one who is “cheerfully confused?”)  Again, there is room for variance in current clinical practice and patient or sometimes family choice.  The Supreme Court of Canada ruled that life-sustaining ventilator support could not be removed from an incapable patient without the consent of the substitute decision-maker.  (Cuthbertson v. Rasouli 2013 https://scc-csc.lexum.com/scc-csc/scc-csc/en/item/13290/index.do) This seems to embody the idea that the quality of a person’s life is not relevant to clinical decision-making, but its duration is.  But the judgment speaks to withdrawing life sustaining treatment not initiating or offering it in the first place.

Patient choice (to refuse a treatment)

In both of the above examples, however, a patient’s (or substitute decision-maker’s) refusal of an offer of treatment usually means that the treatment will not be performed.  That is, the consent to treatment doctrine means that if a treatment (admission to an ICU bed for instance) is offered, then it is open to the patient, if capable, or the substitute decision-maker if not, to accept or decline that offer.  It is open to my substitute decision-maker to say “Doctor, if you think the best he can hope for is a couple more months in the hospital, then no – he would not want that.”  Or, “He told us that he would not wish to continue to live if he could not recognise us and enjoy the things he loved.”

Authority and shared decision-making

The logic of consent in health care decision-making is usually as follows:

1. The patient presents with the problem – disease, illness..
2. The physician (health care provider) diagnoses
3. The physician recommends or offers a treatment
4. The patient or substitute decision-maker accepts or declines the treatment

We have been discussing the clinical decisions made by physicians at step 3 and acknowledging the way in which a values element can creep into those clinical decisions.  We have also identified the patient’s right to refuse a treatment.

What if we have done all this and we still have far more people who need access to ICU beds than we have beds available?

Two fundamentally different approaches

There a re two fundamentally different approaches to this problem.  One lies in randomization – first come first served, or a lottery of some sort, the other looks to non-clinical criteria.

Randomization

This approach is conceptually simple, but rests on a significant ethical premiss.  The foundational idea is that the state – or the community should treat each life equally.  We each count for one – and only one, we each have an equal claim to health care based on our need.  Everything else is irrelevant, so when you are faced with two or more people with an equal claim to the scarce resource you randomise – first come first served or draw lots – we preserve fairness through chance.

It could be argued that this approach is enshrined in the Canada Health Act value of “Universality.”

10 In order to satisfy the criterion respecting universality, the health care insurance plan of a province must entitle one hundred per cent of the insured persons of the province to the insured health services provided for by the plan on uniform terms and conditions.

https://laws-lois.justice.gc.ca/eng/acts/c-6/page-2.html#docCont

In addition, Ontario’s Ethical framework for dealing with drug shortages explicitly states:

Stage 3. Apply Secondary Allocation Principles to Ensure Fair Access to Needed Care

When decisions must be made between patients within a level of priority as described in Stage 2, prioritize patients using a fair and unbiased procedure that does not discriminate between patients based on factors not relevant to their clinical situation (e.g., race, social value, sex, age) such as:

• First come, first served (where queuing is consistent with regular clinical practice), or
• Other procedure that is developed and sanctioned by affected stakeholders (e.g., dividing dose among more than one patient, random selection)

www.health.gov.on.ca/en/pro/programs/drugs/supply/docs/ethical_framework.pdf

Supplementary non-clinical criteria

An alternative is to use supplementary non-clinical This approach has intuitive appeal.  Surely it is better to use criteria that maximise the benefit that comes from utilizing a scare resource.  We must be able to do better for the community than mere chance.  To this end in 2009 a group from Hamilton published a paper called: Development of a Critical Care Triage Protocol for Pandemic Influenza: Integrating Ethics Evidence and Effectiveness.  (Frolic et al.) https://pdfs.semanticscholar.org/1a53/3c9d049358dd5a9ad4778f66eb6edcd537fb.pdf

They proposed using the following supplementary criteria in the event of a catastrophic pandemic:

• Those whose skills and knowledge can be used to save others (multiplier effect)
• Those who contracted the disease as a result of their occupation
• Caregivers
• The relatively young (fair innings/life-cycle principle)
• Those with a better prognosis

The idea is that each of these criteria can be scored easily and the end result can be used to supplement initial triage decisions.  There is much to say about this paper and the general approach it embodies.

1. It is noteworthy that health care providers feature in two of the five criteria. The process adopted in the development of this list involved consultation with “internal stakeholders.”  That is, stakeholders internal to the hospital.  It turns out that if you ask health care providers who should be given priority access to scarce ventilators and ICU beds they say: “Me.”  This does not make the conclusion wrong – but it does give one pause.  It should also prompt consideration of the consultation process in general.  It matters greatly what gets asked, who gets asked and how?
2. The “multiplier effect” is the idea that saving some with a particular skill set is beneficial to the community because they can go on to save others. This only works if the person who is saved is able to subsequently continue, or get back to serving the community.  The multiplier effect is a good utilitarian reason for giving health care providers and other front line workers priority access to vaccines – in order to keep them healthy and working but it is simply not applicable to people who are sick enough that they need ICU care or ventilator support.  They will not be recovering enough to get back to work any time soon.
3. Who is in the category “those whose knowledge and skills can be used to help others?” Health care workers presumably, but what about vaccine researchers, epidemiologists, what about people who do not provide hands on care, equipment suppliers, IT specialists, and what about decision-makers – managers, politicians.. etc.  Are health care ethicists health care workers? The categories all suffer from intractable fuzziness at the edges.
4. When you think of someone who contracted the disease as a result of their occupation what comes to mind first would do front line nurses, emergency services and so on. But there are many other occupations that bring a person into close contact with others, what about teachers, personal trainers, sex workers and on and on.  And indeed how could one tell in the midst of a pandemic what the source of an individual’s infection might be?
5. The category of caregivers is appealingly utilitarian, but again it is not clear that society would benefit most from giving priority access to caregivers as opposed to teachers or researchers, or architects or artists…
6. The idea of social reciprocity is embedded in some of the categories. This is the idea that if we demand extraordinary things from certain members of society, we ought to reward them.  This is a perfectly legitimate idea, and we do this by providing special access to health care for veterans.  But who would have the authority to make such a decision?  Certainly not an individual hospital or health care provider dispensing public goods.  This is a public policy question that needs a well-thought-through public policy response.
7. The fair innings/life cycle argument is attractive. It suggests that people ought to be given a fair chance at living a complete life.  The young deserve a chance to be middle-aged and the middle aged deserve a chance to be old, while the old have already had their share of youth and middle age and don’t need a chance to be very old as much as others need a chance at a full life.  This looks good, but who has the authority to make this decision?  It looks pretty suspicious if the authors of these criteria are not in the category – the old, that would be negatively affected.   This is reminiscent of the issue we identified in consent or refusals of treatment.  Individuals could certainly say “I have lived a full life, do not try to save me but rather save someone who could go on to a live a full life.”   Could that be done collectively?  What consultative process would we have to go through to get to the point where the relinquishment of treatment applied to all who fell in the relevant category.

These are some of the objections to the list developed by the Hamilton group.  But the most significant opposition is to the principles behind the approach rather than the specific criteria themselves.  A utilitarian approach seeks to maximise a good.  But what good is being maximised?  We start with lives saved – the standard approach in pandemic planning and preparedness.

The purpose of Canadian Influenza Preparedness is:

First, to minimize serious illness and overall deaths, and second to minimize societal disruption among Canadians as a result of an influenza pandemic.

https://www.canada.ca/en/public-health/services/flu-influenza/canadian-pandemic-influenza-preparedness-planning-guidance-health-sector/table-of-contents.html#a2f

This approach treats all lives equally.  But then we could look at other possible goods, life stages for instance in the fair innings approach or social benefit in others.  Should we do that?  And who is the we?

Conclusion

Current Canadian health care is based on the idea of equal access based in need.  That is an admirable egalitarian principle which sends a strong message that no-one, in health care terms, is more important or valuable than anyone else.  That approach may, in certain circumstances, yield results that are less “good” on some measurement scale other than fairness and equity.  But, I would argue, so be it, the value of equality of access to health care is more important to our identity than some other competing notion of what? Social efficiency?  Indeed, I would further argue that the embodiment of notions of equity in health care are themselves constitutive of, and contribute to, our collective social welfare.  So, let’s leave the clinical (perhaps tragically difficult clinical) decisions to the clinicians and identify clearly that if push really does come to shove we are content with the judgment of clinicians and, if necessary, the fairness of chance.  If that approach seems inadequate then let us embark on a genuine community debate which clearly lays out the issues and what is at stake, this choice cannot be left to small groups making decisions at the local level.

And, let each of us lead in our own lives and in our own practice.  If you would not want treatment in circumstances where your outlook is gravely uncertain or where your expectations of future life are short or of a quality you would not want, speak up, identify this to your substitute decision-maker and empower them (require them) to follow your wishes.  Further, now is indeed the time to have those conversations with those we love and care for.

Robert Butcher March 17, 2020