ICU triage: How many lives or whose lives?

By Angela Ballantyne

Bioethicists around the world have been asked to advise on the goals and methods of triage protocols. Estimates suggest 5% of COVID19 cases will require ICU care. The key ethical tension is between utility and equity. There are other relevant principles of fair allocation such as reciprocity for frontline workers who have taken extraordinary risks; and a “fair innings” to prioritize those who are younger on the grounds that they have lived a shorter proportion of their natural life span. I leave these aside for now, because the major tension is between utility and equity.

The World Health Organization defines health equity as preventing remediable differences among groups of people, whether those groups are defined socially, economically, demographically. We cannot simultaneously prioritize utility (saving the most lives) and equity (avoiding unjust discrimination).

These values are direct trade-offs.  We can save more lives or we save a more diverse group of lives. There is no perfect algorithm but we need to identify an acceptable balance.

Several influential bioethicists have argued that is it simply obvious that the “many” matters more than the “who”:

According to these positions, the value of saving lives is a substantially more weighty ethical consideration than the value of equal outcomes. It is unsurprising that guidelines based on utility have already elicited legal challenge on the grounds of disability discrimination. Implicit cognitive biases will likely lead clinical decision-makers to unconsciously and unfairly discriminate against older persons or persons with disability. Some of the proposed triage tools add explicit barriers. Jackie Leach Scully argues that current triage guidelines are an “…indicator of our ongoing inability as a society to consider people with disabilities as equal members of the community, with equal human and civil rights, equal claims to citizenship, and equal moral agency.”

In this blog I consider the impact of a utility-focused triage approach on social and ethnic health disparities and question why equity has been so quickly de-prioritized.  ICU triage tools cannot be expected to fix the problems of longstanding health disparities. But that doesn’t mean we should uncritically accept the mantra of utility.

There will be a point at which the prospect of a patient’s survival in ICU will be so low that admittance would amount to a failure to responsibly steward public resources. But there is significant ground between maximizing ICU survival and senselessly throwing away resources. This middle ground should include at least some space for deprioritising utility in favor of giving a wider range of people a shot at ICU. Contrary to the dominant position described above, I argue that equity and justice are critical public interests that must be protected.

Of course those justifying prioritization of utility don’t endorse discrimination. But in reality, an ethical approach aimed at maximizing lives saved results in prioritizing certain social groups. The easy lives to save will be those of people who already enjoy social privilege. As a population, younger, white, wealthy people will be more likely to derive benefit from the ICU resources and survive because they enjoy, on average, higher baseline health status.

Comorbidities are not evenly distributed amongst the population. Underlying health status correlates with privilege based ethnicity and wealth. Current evidence suggests that people with hypertension, cardiovascular disease, chronic obstructive pulmonary disease (COPD), and cancer are more likely to die from severe COVID19.  In New Zealand for example prevalence of multimorbidity (two or more conditions) is higher for Māori (13.4%) and Pacific ethnic groups (13.8%) than for NZ Europeans (7.6% prevalence). Multimorbidity is also more common for people in areas of higher socioeconomic deprivation. Māori aged 45 and over have a COPD hospitalisation rate over 3.5 times that of non-Māori.  When you plug the clinical status of individuals into an ICU triage tool aimed at maximising population survival rates you will get results that reflect these social and ethnic divisions.

Exposure to coronavirus may also vary by socio-economic status. The ability to socially isolate is a luxury. According to US cell phone data, high income demographic groups have reduced their geographic scope of movement more significantly than lower income groups.

Prior epidemics in New Zealand have had disproportionate impact on Māori and Pacific peoples. This wasn’t as a result of differential access to ICU, but probably does reflect vulnerabilities relating to underlying health status and access to medical care. The first ICUs were established in the late 1950s so access to ICU was not a relevant factor in the 1918 pandemic; and ICU access was not triaged during the 2009 pandemic in New Zealand.  During the 1918 influenza pandemic, Māori in New Zealand died at eight times the rate of Europeans. During the 2009 H1N1 pandemic the mortality rate was 2.6 times higher for Māori and 5.8 times higher for Pacific peoples than European New Zealanders.

Maximising the number of lives saved is a function of (1) prognosis (chance of survival) and (2) length of time to benefit (how long each patient is likely to need ICU care).  A key problem for utilitarianism has always been the accuracy of predictions. Standard scoring systems for ICU mortality are designed to predict capacity to benefit from ICU care. These include APACHE-II, sequential organ failure assessment (SOFA), and Simplified Acute Physiology Score. SOFA for example assesses the severity of the patient’s clinical state based on function/dysfunction of six major organ systems. But how reliable are these scoring systems in predicting survival of severe COVID19 patients in ICU?

In China, the overall case-fatality rate (CFR) was 2.3%; but the CFR was higher for patients with pre-existing comorbid conditions—10.5% for cardiovascular disease, 7.3% for diabetes, 6.3% for chronic respiratory disease, 6.0% for hypertension, and 5.6% for cancer. The CFR was 49% among critical cases, but detailed information about critically ill COVID19 patients is scarce.

Analysis of emerging UK data published a week ago found that 7 COVID19 patients with severe comorbidities survived ICU (41.2%) whereas 10 died (58.8%). Is this data sufficient, either in quality or differential outcome, to suggest that those with even severe comorbidities should be denied access to ICU? If we are going to deny specific patient’s access to ICU on the grounds of population utility we need a high degree of confidence in the clinical evidence.

Given this uncertainty, what can we do? I am not a health provider and not in a position to interpret the emerging clinical evidence or speak to the reliability of different ICU scoring systems. But as bioethicists are invited to consult on these triage tools we need to be asking hard questions about the reliability of the screening measures used to exclude certain patients.

Where there is uncertainty, we should err on the side of broader rather than stricter clinical criteria; meaning a wider range of patients get a shot at ICU, even where this is likely to be less efficient overall. Screening measures that include Quality Adjusted Life Years (QALYs) must be avoided as these are inherently biased against persons with disabilities. When resources are limited we need to allow time constrained ICU trials, and generate protocols for withdrawing ICU care. These are important mechanisms to give more people access to ICU. The following approach goes some way towards limiting the impact of comorbidities on access criteria:

“To address concerns about compounding injustices associated with systematic disadvantage and the arbitrariness of comorbidities, and unlike other approaches that have been proposed for ICU triage in a disaster, the scoring system described here affects patients whose comorbidities are so serious that they are expected to live no more than 12 months even with successful ICU treatment.”

Regardless of the specific number of patients who survive severe COVID19, we will need to tell a story about the values that guided us as a community in a time of crisis. The narratives that emerge during this pandemic will have a lasting legacy in bioethics. Public interest requires the responsible use of resources. But this does not exclusively entail maximizing utility.  There is public interest in re-affirming human rights, the value of diversity, promoting fairness, and resisting entrenched privilege.


Author: Angela Ballantyne


Visiting Senior Research Fellow, Centre for Biomedical Ethics, National University of Singapore

Associate Professor of Bioethics, Department Primary Health Care and General Practice, University of Otago (currently on leave)

Competing interests: I have no competing personal or financial interests with regards to the content of this post.

Acknowledgments: I would like to thank the following people for valuable discussions that informed this post and for constructive comments on earlier versions: Wendy Rogers, Owen Schaefer, Jackie Leach Scully, Karen Bartholomew, Cindy Towns, Ruth Cunningham, Donna Cormack, Jason Gurney, James Stanley and Nic Aagaard.


(Visited 9,791 times, 1 visits today)