One very simple way to prioritise fairness in the second phase would allocate vaccines using a lottery procedure. This would ensure equal access to all individuals not included in Phase One, without prioritising any particular group over any other. This is one important conception of what fairness might involve. Phase One of the UK’s vaccine prioritisation strategy focused on saving the most lives possible, based on data suggesting that age is the single greatest risk factor for dying following infection by COVID-19.
The first phase of the UK’s Joint Committee on Vaccination and Immunisation (JCVI)’s guidance on vaccine prioritisation outlined 9 priority groups. Together, these groups accommodated all individuals over the age of 50, frontline health and social care workers, care home residents and carers, clinically extremely vulnerable individuals, and individuals with pre-existing health conditions that put them at higher risk of disease and mortality. These individuals represent 99% of preventable mortality from COVID-19. Prioritising these groups for vaccination will mean that the distribution of vaccines in a period of scarcity will save the greatest number of lives possible.
BAME groups
Continuing To Minimise Mortality? – Demographic Risk Factors
Males
However, it is not the only one. Considerations of fairness could go beyond aiming for equal access, and instead factor in reciprocity. Many essential workers in the pandemic have continually put themselves at risk of exposure in order to perform vital societal roles. It could therefore be argued that these workers deserve priority for the vaccine, in recognition of the role they have already played in the pandemic.
This analysis is supported by evidence from a prospective cohort study using data from the UK Biobank, which analysed the relative risk (RR) of contracting severe COVID-19 in essential occupations, compared to the baseline risk for non-essential workers. The study’s analysis suggested that, adjusting for age group, sex, ethnicity, and country of birth, healthcare workers had an RR of 7.43 (that is, they were more than 7 times more likely to become infected compared to other occupations), lending support to their inclusion in Phase One of prioritisation.
Prioritisation decisions in medicine often have to forge a balance between fairness and benefit. The Phase One strategy focused on benefit: the number of lives saved. Phase Two instead could incorporate considerations of fairness to a greater extent.
Occupational Risk
Deprivation (by IMD groups)
- Working in environments with high virus exposure.
- Working in close proximity to others.
- Coming into contact with lots of different people.
· How can we compare different societal benefits?
· Equal access is not the only conception of what fairness demands.
· Does exposure translate to mortality?
On reason to prioritise certain groups is that they are central to the functioning of society. Prioritising certain essential workers for vaccination might then be justified on this basis. Indeed, a number of EU countries included professionals who maintain critical infrastructure or security in their first phase of prioritisation, in addition to the elderly and healthcare workers. China focused their vaccine prioritisation solely on the working population; the functioning of society appears to be taking precedence in this approach to allocation.
Essential workers – The greater the risk, the greater the desert.
One long term hope for national vaccination programmes is that they may engender herd immunity within the population if a sufficient proportion of the population has been vaccinated. Such herd immunity would confer protection to those who have not received (or responded to) a vaccine. Particularly whilst vaccines are scarce, the achievement of herd immunity may not be the immediate goal of prioritisation. However, as more vaccines become available, future phases of vaccine distribution should be sensitive to the strong moral reasons to achieve herd immunity as soon as possible. This consideration will support distribution strategies that will maximize the speed of vaccine deployment, as more doses become available – for instance, this may justify prioritising children if vaccination could be performed quickly within schools.
Alternatively, the second phase could aim to prioritise fairness by seeking to mitigate inequalities that have become starkly apparent in the pandemic. Data suggests that the higher mortality risk faced by some demographic groups is at least partly explained by pre-existing inequalities or unfair prior treatment. For this reason, many prioritisation recommendations (including the JCVI guidance) suggest that one of the key aims of prioritisation must be to mitigate these inequalities.
40-49 age band
Prioritising To Achieve Societal Benefits
What other aims besides maximizing the number of lives saved should be considered for Phase Two policy development? Public debate about this question has largely focused on the potential prioritisation of certain occupations for vaccination, rather than prioritisation based solely on the above demographic risk factors. Part of the reason for this focus is that there is evidence to suggest that different professions face different degrees of occupational risk in the pandemic. Those who are at more risk might deserve priority. We discuss this next.
Prioritising essential workers may serve to ensure that prioritisation achieves key social goods (such as social order, security, and justice) as well as reducing mortality. Prioritising certain occupations sooner may also make it possible to relieve certain public health restrictions. For instance, prioritising teachers for vaccination might potentially enable schools to re-open sooner; this would not only have the significant benefit of enabling children to access education facilities, it would also relieve the significant childcare burden faced by parents during lockdown. Some prioritisation strategies may thus enable greater societal freedoms, even if they are not the most effective way to minimize mortality.
As the first phase of vaccine distribution continues to proceed, a heated debate has begun about the second phase of vaccine prioritisation, particularly with respect to the question of whether certain occupations, such as teachers and police officers amongst others, should be prioritised in the second phase. Indeed, the health secretary has stated that the government will look “very carefully” at prioritising shop workers – as well as teachers and police officers – for COVID vaccines. In this article, we will discuss moral and scientific reasons for and against different prioritisation strategies.
Prioritising Fairness – Equal Access, Reciprocity, or Minimising Risk Discrepancy?
Ethnicity
Box 1 – Key Considerations For Second Phase of Vaccine Prioritisation
Age (below 40)
· Lottery basis
Since Phase One of vaccine prioritisation targeted 99% of preventable mortality, it might be argued that other moral values ought to have greater recognition in the second phase. There are other moral reasons to prioritise certain groups for vaccination besides their increased risk of death or severe disease.
Who is most at risk of contracting COVID-19? Analysis from the ONS has identified the following occupational risk factors:
| Minimizing Mortality | Societal Benefits | Fairness |
| Justice In their initial guidance, the JCVI also suggested that a key focus for the second phase of vaccination could be on further preventing hospitalisation, and that this may require prioritising those in certain occupations. However, they also note that the occupations that should be prioritised for vaccination are considered an issue of policy, rather than an issue that the JCVI should advise on. · Equal access for all Differences in mortality risk between groups may be driven by two main factors: between-group differences in (i) the probability of becoming infected, and (ii) probability of dying from COVID-19 once infected with SARS-CoV-2. The age-based prioritisation in Phase One was based on evidence suggesting that increasing age represents the single greatest risk of mortality from COVID-19, and that this risk increases exponentially with age. · How do these different risk factors interact? Challenge: However, it is difficult to establishing and compare the precise contribution of occupational exposure and other demographic risk factors to overall mortality risk and/or risk of severe disease across different potential priority groups. This is a hugely complex task. As we noted, the JCVI has suggested that although certain occupations could be prioritised for vaccination, this is ultimately ‘a policy issue’. However, in so far as prioritisation decisions should take into account which groups are at highest risk, the input of the JCVI is absolutely crucial to making an informed and balanced policy decision on this matter. However, it is difficult to establishing and compare the precise contribution of occupational exposure and other demographic risk factors to overall mortality risk and/or risk of severe disease across different potential priority groups. This is a hugely complex task. As we noted, the JCVI has suggested that although certain occupations could be prioritised for vaccination, this is ultimately ‘a policy issue’. However, in so far as prioritisation decisions should take into account which groups are at highest risk, the input of the JCVI is absolutely crucial to making an informed and balanced policy decision on this matter. Deprived groups (by Index of Multiple Deprivation score). Deprived groups (by Index of Multiple Deprivation score). Deprived groups (by Index of Multiple Deprivation score). Deprived groups (by Index of Multiple Deprivation score). Deprived groups (by Index of Multiple Deprivation score). On this approach, the morally relevant feature for prioritisation is not which groups have the highest absolute mortality risk (as per the “saving the most lives” strategy). Instead, what matters are the relative risk discrepancies between different groups: groups who have the greatest discrepancy in such risk compared to a relevant reference group, on this view, should be prioritised. On the data available, this might involve prioritising BAME groups, those in socially deprived areas, men, and/or those in occupations at much higher risk of exposure. However, on this approach, we might be justified in giving highest priority to disproportionately affected groups that have been also been subject to significant historical disadvantage, to partially correct for prior unfair treatment. So far we have been discussing reducing mortality by targeting those at greatest risk for priority. Such a strategy aims at reducing mortality by focusing on the direct benefit of vaccination for the recipient, in affording the recipient protection against severe disease. Of course, in many other infectious diseases, vaccination can also have the indirect benefit of preventing the recipient from transmitting the virus. When this is so, prioritising those most likely to transmit the virus may serve to significantly reduce mortality, even if those individuals themselves are at low risk of developing severe COVID-19 . However, since there is limited data about the efficacy of the available vaccines in preventing transmission, such an approach cannot yet be empirically justified. e.g: Teachers Since the age threshold for inclusion in Phase One prioritisation is 50, prioritising in accordance with age in Phase Two will only continue to be the most effective way of minimizing mortality, if the age-associated mortality risk below 50 outweighs other identifiable mortality risk factors. However, the evidence suggests that there is a substantial fall in age-related mortality risk below 50. The OpenSAFELY analysis cited in the JCVI guidance suggests that those in the 40-49 age group have a fully-adjusted mortality risk hazard ratio of 0.3, in comparison to the analysis’ reference group of 50-59 year olds (who therefore have a hazard ratio of 1). So this means, that 40-49 year olds face 30% of the age related mortality risk faced by 50-59 year olds. Elsewhere, a systematic review and meta-analysis suggests that the infection fatality rate (which assesses the probability that an individual will die from COVID-19 if they become infected with SARS-CoV-2) is 0.41% in the 50-59 age group, falling to 0.12% for the 40-49 age group. So the probability of dying if you become infected with SARS-CoV-2 is also lower for 40-49 year olds. We shall suggest that the input of the JCVI is absolutely crucial to making an informed and balanced policy decision on this matter. But what policy should be pursued? Here, we outline some of the ethical considerations that bear on this policy decision. |
There is an important caveat here. The analysis of occupational risk in this study assessed relative risk of contracting severe COVID-19, rather than occupational mortality risk per se. So, it gives us evidence about which occupations are at greatest risk of becoming severely ill – but that alone may not entail that they are at greater risk of death. Although the ONS has published mortality data by occupation, this data is not adjusted for all relevant risk factors, and thus cannot be used to establish whether differences in mortality were necessarily caused by differences in occupational exposure.
Concluding Remarks
But other occupations were also identified as being at higher risk in this study. Professionals broadly classed as “social and education workers” in the study faced a relative risk of 1.84 (that is, they were almost twice as likely to be infected as non-essential workers). Within this broad group, social care workers had the highest relative risk (RR=2.46). Collectively, professionals broadly classes as “other essential workers” had a RR of 1.6; transport workers had the highest relative risk in this other broad group (RR=2.20).
Sex High risk occupations
Age is not the only relevant factor, however. Further evidence from the openSAFELY analysis suggests that a number of other demographic features beyond age may also contribute to increased mortality risk. For instance, ethnicity, sex, and social deprivation are all identified as important mortality risk factors. Indeed, analysis from the Office of National Statistics (ONS) suggests that for all ages together, the mortality risk for people of a Black ethnic background is 2.0 times greater for males and 1.4 times greater for females compared with those of White ethnic background. We will return to such between-group differences in a subsequent section discussing fairness as a policy consideration in addition to sheer mortality reduction.
