The COVID-19 vaccine: who gets it first?

Rationing and prioritising the provision of care is not a new phenomenon for the NHS – deciding what treatments to fund, for whom, and at what cost, is at the core of any publicly funded health care service. Normally these decisions are made by the National Institute for Health and Care Excellence (NICE), the independent body that approves the funding of drugs and other treatments, such as vaccines, based on cost and effectiveness (including impact on both length and quality of life). Typically, NICE recommend funding a drug if the cost per ‘Quality Adjusted Life Year’ is less than £20,000–30,000. The NHS is also required to take into account inequalities with both the Health and Social Care Act 2012 and the Equality Act 2010 placing a duty on the NHS to have ‘due regard to’ and ‘advance’ equality.

For vaccines, the Joint Committee on Vaccines and Immunisation (JCVI) – an independent body of experts – advises the government on prioritisation. In this case, JCVI has advised an approach that prioritises those in care homes first, then those aged 80 and older and front-line health and social care staff. Beyond these groups, prioritisation is a mix of age (in 5-year intervals) and younger people who are clinically vulnerable. Sex, ethnicity, and profession (apart from health and social care workers) are not taken into account.

Figure 1 details the prioritisation plan for phase 1, and the estimated number of people in each group in England (net of those included in higher priority groups).

Figure 1

Vaccines can do two things. First, they can protect vaccinated individuals by reducing their risk of catching the virus or developing severe disease thus preventing deaths. Second, they can also protect societies, including unvaccinated individuals, by reducing transmission. Deciding who to vaccinate weighs both of these factors. Put simply: do you vaccinate individuals less likely to catch COVID-19 (for example because they are shielding), but highly likely to die or suffer severe disease if they get it? Or, those who are more likely to get it and transmit it to others (for example, university students), but much less likely to die. The former prioritises mortality, the latter controlling the spread of an epidemic.

Ultimately, the JCVI resolved this question relatively easily. We do not yet have firm evidence on the ability of these vaccines to reduce the transmission of COVID-19, therefore we should prioritise reducing mortality. This has the secondary benefit of meaning that fewer people need to be vaccinated for a large impact – achieving meaningful reductions in transmission is likely to require many millions of people to be vaccinated, each high-risk individual vaccinated has a benefit in and of itself.

If we agree that reducing mortality is the most important factor in deciding who gets the vaccine first, who should be prioritised? Below we outline some of the key factors government could consider: age, gender, ethnicity, profession, prevalence in a community and having already had COVID-19. We look at two ‘outcome’ metrics: the number of deaths avoided, and the number of years of life saved, which also considers the number of years that individual would likely have lived had they not contracted COVID-19.

We also consider some broader trade-offs – should government be thinking about ensuring they reduce inequalities (or at least do not exacerbate them)? Is fairness important, and how might we measure this?

Age: the biggest factor affecting COVID-19 mortality

The chart below shows the number of vaccinations needed to prevent one death and to save 1 year of life (assuming those dying have average life expectancy), by age group. Our model uses historical data from the pandemic to date and so gives an estimate of the number of lives that would be saved if an imagined ‘third wave’ of the pandemic caused the same number and pattern of deaths as it did in the first and second waves.

This is not a perfect model. We would not necessarily expect that deaths in an imagined ‘third wave’ would be distributed in the same way as in the first wave. For example, we might expect that better infection control would lead to far fewer care home deaths. However, this provides a reasonable ballpark figure that allows us to roughly compare the impact of vaccination on different age groups.

This scenario-based approach also highlights how significant age is as a driver of mortality. You would need to vaccinate around 1,300 60–64 year olds to save one life, compared with only 40 vaccinations to save one life among those older than 90 years. This holds even if we look at years of life saved rather than just deaths avoided. For example, 10 vaccines would save one year of life among those aged 90 but around 60 vaccines would be needed to save one year of live among those aged 60–64.

Gender

Men are more likely to die from COVID-19 than women of the same age. Based on the scenario above, saving 1 year of life among women aged 80–84 would require 20 vaccines, compared with 13 for men of the same age, and 18 for men aged 75–79. This difference in risk has not been taken into account in most governments’ vaccine strategies – possibly because of concerns about public perceptions of the fairness in prioritising one gender over the other.

Ethnicity

Black men and women are more than four times as likely to die from COVID-19 than their white counterparts, and Bangladeshi/Pakistani men and women more than three and a half times as likely to die. This is both because they are more likely to be exposed to the virus and, once infected, are more likely to die from it. Although it recognises this inequality, the JCVI is not currently explicitly targeting minority ethnic groups. Indeed, the younger age of ethnic minority populations (only 8% are aged 65 and older, compared with 25% of white people) mean they are less likely to be vaccinated in the first round of prioritisation. Minority ethnic Britons are also more likely to have underlying health conditions – so the focus on those who are clinically vulnerable should, JCVI suggest, mitigate some of the inequalities here.

Profession and exposure to COVID-19

Unlike in the UK, the US currently has draft plans to prioritise vaccinating essential workers (eg teachers, shop staff and emergency services) after care home residents and health and social care staff. These are groups who are particularly likely to be exposed to the virus, to transmit it to others, and whose absence from their jobs has a significant effect on the American economy. A moral argument could be made here: essential workers – who are often more economically disadvantaged and often from minority ethnic communities – may have no choice but to risk exposure to COVID-19. On this basis, this group could be given priority over another – such as those in their 70s – who might be more able to shield until a vaccine becomes more widely available.

Having already had COVID-19  

ONS estimates that around 6.9% of people in England have antibodies, with higher rates in London, among some professions (such as those in patient-facing roles), and in certain age groups. When the vaccine is scarce, it would be possible to prioritise those older people who do not have antibodies over those that do. However, while it is likely that individuals with antibodies are less likely to have severe disease, or transmit it to others, we are not certain. Is it acceptable to take the risk? 

Living in a high-risk area

People living in different areas face different risks of contracting COVID-19 and, once infected, of subsequently dying. For example, in the first wave of the pandemic, those in London were around three times more likely to die as those in the South West. Prioritising vaccinations in high risk areas – for example those with the highest current prevalence – may therefore save more years of life than giving it to all areas at the same time. The risk, though, is that high-risk areas change; areas of high prevalence last month (such as the North West) have declining prevalence now.

This analysis has laid out the trade-offs that government has made in deciding who gets vaccinated against COVID-19 and when. It shows that, if the focus is on maximising the years of life saved, there is a moderately clear age-based hierarchy in who should get vaccinated first. We have also highlighted some of the complexities in making this calculation – the way age interacts with an individual’s sex and ethnicity to affect risk; the broader importance of trying to reduce inequality; and questions around what is fair or just.

But it is one thing to design an optimal approach to prioritising vaccines based on saving years of life; it is quite another to implement it. Any strategy will have to consider challenges – logistical and attitudinal – to actually getting the vaccine to the people who need it most. Take ethnicity as an example – even if we did decide to prioritise, say, black and minority ethnic men in their 60s over white men of the same age, do GPs have reliable enough data to identify and contact these individuals? And would GPs in these areas have the resources to rollout the vaccine? Will these individuals want to be vaccinated first – or will they see this as riskier and prefer to wait for more data? Will we even need to make these micro-trade-offs, or will the vaccines arrive in appropriately large batches such that this is not necessary?

We also may want to consider public acceptability. While, in the UK, we have an institutional architecture that is used to making utilitarian decisions about how care is rationed, COVID-19 is an exceptional case with far more public attention and engagement. Will the public view prioritising older people over essential workers as unfair? Would we want to see some adjustment made for the differential risk between men and women?

These issues are complex. They ask a society to decide on how, when our resources are constrained, to deliver on the promise of an NHS that offers equal care to all, on the basis of need not the ability to pay. This is not an issue confined to vaccination alone. 2021 will see the NHS begin its long road to recovery – part of which will involve making decisions about who to prioritise from historically large waiting lists, or how to weigh investment in mental over physical health care. This is not a debate that is going away.