‘Fair innings’ and ‘life projects’ principles

The ‘fair innings’ and ‘life projects’ principle were strongly rejected from the outset and throughout the deliberative process. Participants found the ‘fair innings’ principle arbitrary and unnuanced, as well as unfair. They felt that age alone does not provide sufficient information about someone’s medical condition and that the lives of older people are important too: ‘We should get all equal treatment, young or old, we’re all the same’. Some participants also mentioned the contributions of the elderly to society, stating that ‘older people have just as much to give to society as younger people do’. The ‘life projects’ principle was equally firmly rejected, on the basis that it was normalising, favouring existing societal norms that not everyone meets: ‘It’s very discriminatory and not right. There are late developers. There are people who bloom later or earlier in life’. It was also emphasised that retirement was a time in which, after a life of work, people are finally free to start and pursue their life projects: ‘When you get older, that’s when you want to start projects. […] There are a lot of people almost having second lives doing all the things they couldn’t do previously’. Dismissing this period, therefore, seemed counterintuitive.

Egalitarian principle

The egalitarian principle was accepted, though a number of concerns about it were raised throughout the study. Initially, this principle was received as the most straightforward and fairest principle, but as discussion progressed, worries emerged about its practical application. First of all, participants rejected a randomised ‘lottery’ approach, preferring a ‘first come, first served’ version of this principle: ‘lottery doesn’t feel like a good system when it’s people lives. It’s inappropriate’. But even the latter approach raised concerns. Participants were mostly worried about hidden inequalities, stating this approach would not redress, and even risk reinforcing, existing inequalities (eg, people with better access to the hospital may get there sooner). One participant said that ‘first come, first served isn’t egalitarian and you have the socio-economic challenges because, if you are in a particular class, you’re in a better position to be able to take care of yourself and get to the doctors first’. There were further concerns that a ‘first come, first served’ approach would waste valuable resources, when patients with a worse prognosis happen to arrive earlier. Finally, some participants felt uneasy that, on this approach, resources would not necessarily go to those who need them most: ‘On the face of it, it looks good, but I think means that those that come in later who are in greater need haven’t got access’. A few participants remained in favour of an egalitarian approach, though all accepted that, if a patient’s prognosis is extremely poor, they should not be escalated for treatment: ‘if you were following the egalitarian principle but you have someone in front of you who the evidence would suggest is highly unlikely to survive treatment and you’ve got someone who is highly likely to survive, as unfair as it may seem, it feels like it would be an important consideration […] I’m only thinking about extreme cases where you’ve got someone who is extremely frail and therefore extremely unlikely to survive’.

‘Maximise life years’ principle

When the ‘maximise life years’ principle was introduced, immediate concerns were raised about the accuracy of medical judgments about life expectancy: ‘Nobody knows how long anybody is going to live for. There are some assumptions, even if you’ve got two people in front of you, one who is 40 and one who is 60’. Furthermore, in discussing this principle, participants spontaneously distinguished survival chance from life expectancy in the deliberations and strongly favoured the former. They supported maximising the number of lives saved, rather than the amount of life years saved: ‘There’s a logic in maximum number of lives you save irrespective of the number of life years they have’. The underlying reasoning seemed to be that every life is of equal value: a majority of participants agreed that ‘a life is a life’.

It was thus widely felt that a patient’s immediate medical condition was a very important factor in triage, insofar as this informed their chances of survival. In this context, participants recognised frailty as a key factor: though it was not initially understood as a medical term, it was eventually accepted as a relevant prognostic variable for predicting survival chances.

Some participants questioned the survival chance-based approach, though. For example, a small number of participants expressed concern about the disproportionate effects it could have on groups that may be more vulnerable to COVID-19: ‘By virtue of prioritising survival of the fittest, it will discriminate and people are uncomfortable with this because it means older people will be less likely to be escalated, people in wheelchairs, people in BAME communities’. Another more widespread worry was that this approach failed to allocate resources in accordance with need. These concerns led some participants to formulate a new, vulnerability-based allocation principle, which is discussed further below.

Quality of life

The notion of quality of life was initially treated with suspicion, seen as inviting unconscious bias and too subjective: ‘I don’t know if professionals can really confirm how somebody’s well-being is’. Throughout the study, it was increasingly accepted, though mostly as a secondary factor when patients’ medical conditions are highly similar, in which case those with a higher quality of life would be prioritised. Caveats were that it should only be applied in extreme cases and that quality of life assessments should, where possible, involve ‘input of the person, their family, carers and that kind of stuff’ to avoid biased assessments.

However, one participant said those with a lower quality of life should be prioritised, so that their quality of life may be improved. Some also noted that quality of life may be strongly influenced by socioeconomic factors, indicating a danger of exacerbating existing inequalities: ‘I do worry with quality of life, the more money you have, the better quality of life you tend to have […] your health is defined by your class and how much money you have’.

Vulnerability

Throughout the study, concerns were expressed about vulnerability, especially in reaction to the utilitarian approach. In these discussions, participants struggled to formulate an additional allocation principle. This had two aspects, though these were not always clearly differentiated. One aspect concerned vulnerable groups (eg, age, disability or ethnic groups) who may be disproportionately affected by the virus itself or the social response to it (eg, unconscious bias). One participant said: ‘we know it affects the elderly at higher rates than the youth. […] It makes the most sense to prioritise the elderly over the young, just on the basis of the percentages of old people vs young people dying. Young people are more likely to survive’. There was, however, some disagreement over whether positive action for these groups should indeed be taken to mitigate the vulnerability or whether this was itself a form of discrimination.

The other aspect concerned individuals in need (eg, those presenting to hospital as sicker) and whether a humane principle was to prioritise those in greatest medical need: ‘The more help somebody needs, the more they should get’. Some suggested to prioritise those least likely to survive: ‘I think the most vulnerable should be prioritised. […] If you think you can save them, then prioritise them’. Reasons given for such an approach were that ‘the true measure of any society is how it treats its most vulnerable members’. But, again, it was accepted that if treatment was unlikely to succeed, patients should not be escalated: ‘you give the resources to the people that most need it, in my opinion, up until the point where the giving of resources is next to useless, where it’s ascertained that they will die anyway’.

Other participants rejected this need-based approach altogether, out of a concern for efficiency: ‘Does that mean, if those people are most likely to die, you’re directing your resources at people who are weaker? So resources could be going to a group who stand the least chance of surviving? That doesn’t feel like a great use of resources’.

Implementation

During the final workshop, participants were asked how their recommendations should be implemented. We found strong support for discretion (applying recommendations as guidance rather than a mandatory policy), and participants felt groups of doctors, not individuals, should make decisions as this could reduce burden and bias. Thus, guidelines should not be binding but instead guide expert deliberation, and this deliberation is ideally executed by teams rather than individuals, so that different perspectives can be considered.

Strengths

We achieved a purposeful sample, there was a high level of participant engagement, participants showed they could think through complex ethical topics, a triad consensus emerged from a very diverse South-East London group, indicating a degree of robustness and there was the ecological validity of doing this study in the social window in between two COVID-19 waves.

Limitations

The South-East London sample may not generalise to other areas, findings may not generalise to other triage contexts (eg, pandemics effecting children) and some elements, for example, vulnerability, remained underexplored, indicating a need for further research.