Don’t let the ethics of despair infect the intensive care unit

By David Shaw, Dan Harvey and Dale Gardiner.

Coronavirus is a killer, and most countries have implemented measures to reduce this mortality. On the one hand, public health measures aim to limit the spread of the disease, and hence limit the number of people requiring hospitalisation; on the other, healthcare professionals working in intensive care in areas where infection rates are high must make extremely difficult decisions about resource allocation. This is because the supply of intensive care beds and mechanical ventilators is limited, and yet the most seriously ill patients infected with the virus will not survive without an intensive care bed. The coronavirus crisis thus intensifies the difficult ethical issues around resource allocation and patient care in the ICU.

Some countries have developed specific guidelines to deal with this issue. The official guidance of the Italian College of Anesthesia, Analgesia, Resuscitation and Intensive Care (SIAARTI) states that “An age limit for the admission to the ICU may ultimately need to be set. The underlying principle would be to save limited resources which may become extremely scarce for those who have a much greater probability of survival and life expectancy, in order to maximize the benefits for the largest number of people.” Recent NICE guidance suggests that that intensive care is of uncertain benefit to patients with a Clinical Frailty Score > 5, perhaps with an implicit assumption that local decisions might be taken to deny such patients admission. The civil protection department in the Piedmont region went further, developing a protocol which states as a rule that “The criteria for access to intensive therapy in cases of emergency must include age of less than 80 or a score on the Charlson comorbidity Index of less than five.”

This means that those over 80 years of age with comorbidities will be denied an ICU bed, and thus almost certainly die. Those under 80 could have several serious comorbidities but still be given a bed, even if their prognosis is worse than someone over 80 with a score less than five. This is clearly ethically problematic: age can act as a rough surrogate of clinical outcomes, but should not be applied in such a crude way that potentially discriminates against patients who are more likely to recover.

Both Italian guidelines represent a utilitarian approach; they aim to maximise the benefit from each intensive care bed. Utilitarianism is often seen as the go-to theory in high-pressure triage of limited resources, and philosophers have not been slow to offer arguments in its favour; for example, Savulescu and Wilkinson argue that “If one person needs a ventilator for four weeks, and four others need it for one week, your choice is to save one or four. So, doctors should take duration of use into account.” This is more ethical than the Piedmont protocol, but still too crude and unrealistic. Doctors will not know at the point where one person needs a ventilator that four patients will turn up who could use it instead. And even if they did have this perfect future knowledge, it is an oversimplification of clinical and ethical realities to simply claim that it boils down to “save one or four”.

Simplistic broad “cut offs” for critical care admission represent a poor understanding of pathophysiology and prognosis and neglect the patient-centred and individualised nature of clinical decision-making. This means that decisions based on such cut-offs risk appear arbitrary and unjust to the wider public, because they often are unfair, like the Piedmont protocol: families will thus respond with anger, frustration, concern, complaint and ultimately attempts to undermine the “rules” by manipulation. For example, families might try to subvert the protocol by ignoring the advice of the community team and present to hospital early in the pandemic or when less severely ill; lie to the doctors about age and co-morbidity, or refuse to trust the team when the situation subsequently becomes hopeless due to an assumption they are just trying to “free the bed”.

A more nuanced approach is offered by the “MORAL balance” process designed by two of the authors (DH,DG), intensive care doctors working in the UK. This framework sets out several steps that healthcare professionals should go through in making decisions in intensive care:

Make sure of the facts; list Outcomes of Relevance to the Agents involved; Level out the arguments in a Balancing box; and Document the Decision. It builds on the well-established, ethical principles of Beauchamp and Childress, which are already known to doctors. This process has been recommended by the Faculty of Intensive Care Medicine as a part of their guidance in managing end of life and critical care, and can be applied to difficult decisions both at a patient and an organisational level during the COVID-19 pandemic.

One example provided on the MORAL Balance website explicitly addresses the case of Mrs Taylor, who is 83 and presents with possible COVID-19 infection, severe respiratory impairment, and significant co-morbidities. The decision that has to be made using the process is whether to offer treatment or palliation, and the balancing of the different factors can be seen on page 3 here. Ultimately, the decision is documented as follows: “I have had discussions with Mrs Taylor’s family, and have explained that even with infinite resource the benefits vs burdens of escalation are finely balanced, but in the current situation it cannot be justified. We do not think she will survive and we will adopt a comfort approach. With family agreement I have completed a DNACPR.” This is a much more realistic approach to the simple example given by Savulescu and Wilkinson; future patients are factored into the balancing, but the focus is on the immediate patient and her family.

While the ultimate decisions made using the MORAL Balance or other ethical frameworks may tend to coincide with those made using utilitarian policies, the process of decision-making could not be more different. Using such frameworks respects the autonomy of patients, the relationships of the family, and the moral resilience of healthcare professionals. Despite the dangers of the novel coronavirus, decision-making on the ICU must not be infected by one-size-fits-all utilitarian policies that neglect the personal and individualised nature of clinical ethics.


Where utilitarianism could be helpful is in helping families to understand the nature of the difficult decisions facing ICU staff. For example, hypothetical cases could be given as illustrations. Take Agnes and Dorothy. Agnes (age not given as not directly relevant) has a 10% chance of surviving the virus if she gets an intensive care bed; if she does survive, she will live for perhaps another 10 years. However, she will probably need 1 month on the ICU. 0.1*10=1; 1/1 = a “probability-adjusted benefit per month” (PABM) of 1. In contrast, Dorothy has a 50% chance of surviving, and a life expectancy of 50 years if she does; this will require perhaps two weeks on the ICU. 0.5*50=25; 25/0.5 = a PABM of 50.  This is a crude example, but a simple calculation is all that is required to give families an idea of the factors that clinicians have to consider.


Authors: David Shaw, Dan Harvey and Dale Gardiner


DS: Assistant Professor, Care and Public Health Research Institute, Maastricht University and Senior Research Fellow, Institute for Biomedical Ethics, University of Basel

DH: Intensive Care Consultant, Nottingham University Hospitals NHS Trust and Honorary Associate Professor at the University of Nottingham

DG: Intensive Care Consultant, Nottingham University Hospitals NHS Trust, Nottingham and Chair of the Ethics of Clinical Practice Committee, Nottingham University Hospitals NHS Trust

Competing interests: None.

(Visited 483 times, 30 visits today)