Worldwide progress on the development on safe, effective vaccines against the coronavirus has been impressively fast. Three vaccines have been approved in the EU, the United Kingdom or the United States (Table 1). Given the lack of effective medication against the virus, the vaccine provides by far the most promising means of getting the pandemic under control.
Approval, however, is only one of many steps that need to be taken. For the vaccines to have the desired public health effects, they need to be procured, distributed to national governments, and administered to the population.
The European Union has been slower than the US or the UK to approve vaccines. This is, however, only a matter of days so far, and is thus a secondary concern. The far greater concern is that to date, the EU appears to have neglected to procure adequate supplies of the only vaccines that are likely to obtain near-term approval. This could lead to delays of many months, causing needless human suffering and economic loss.
In the long run, supply of multiple vaccines should be more than adequate. In the near to medium term (say, six to nine months), however, there could be serious shortages. How serious is the near-term problem in the EU and what can be done about it?
EU-wide procurement of coronavirus vaccines
Table 1 shows the vaccines approved to date in the United States, the United Kingdom and the European Union, together with the quantity already procured by the EU (US approval for another vaccine, developed by Johnson & Johnson could be applied for in February). The US approvals are provisional Emergency Use Authorizations (EUAs), as are the UK approvals, while the EU approval is a full authorisation. Table 1 also shows the number of doses ordered by the EU and additional doses provisionally ordered by the EU. For each of these three vaccines, two doses per person are required.
A major success of the European institutions to date has been to enable joint EU-level procurement. The use of the Emergency Support Instrument (ESI) helps to ensure a fair distribution across the EU, avoids the risk of individual member states bidding the price up by competing with one another, and also reduces the risk of larger EU countries hindering smaller countries in accessing vaccines. The ESI has €2.7 billion in funding from the EU budget – far too little in our judgment, particularly when one considers that the funding has to cover multiple programmes, not just the acquisition of vaccines.
According to the European Commission, in addition to the orders for approved/approval pending vaccines shown in Table 1, supplies of other vaccines have been secured under the ESI:
- Sanofi-GSK: 300 million doses
- Johnson & Johnson: 200 million doses
- CureVac: 225 million doses
Together, the EU orders of current and future vaccines would appear to be more than enough in the long run. However, in the shorter term, only 380 million doses are available of the two vaccines that are approved or, in the case of Moderna, that can be expected to be approved shortly (with an option for 80 million more doses from Moderna, thus a total of 460 million potential doses). Even for these, delivery schedules are uncertain, but are thought to be in the first half of the year. In the near term, there appears to be a serious shortfall.
Vaccinations to date
Vaccinations in most of Europe are only just beginning. What can we learn from experience elsewhere?
The United States, despite a generally ineffective national response to the pandemic, has placed great weight on a fast immunisation programme; even so, there have been delays. Instead of vaccinating 20 million Americans by the end of the year as planned, as of 31 December just under 2.6 million people had been vaccinated (less than 1% of the US population), even though 12.4 million doses had been distributed. A range of logistical and teething problems appear to have emerged, which is hardly surprising in a programme of this size, and the December holidays did not help.
A number of small countries have made much faster progress (Figure 1), including Israel and Bahrain (neither of which is much impacted by the December holidays).
Still, it should be noted that no large country has yet vaccinated substantially more than 1% of its population. Even in countries where the vaccination programme is most effective, this process will extend over many months. Too much weight should not be placed on results to date. Committed and effective delivery over many months will ultimately determine which countries and continents achieve an effective vaccination response to COVID-19.
How many doses are really required?
What the EU’s medium to long-term goal ought to be seems fairly clear: every EU person for whom a vaccine is authorised, and for whom there are no medical contra-indications, and who is willing to be vaccinated, should be vaccinated as soon as possible.
The EU population as of 1 January was about 448 million, so one might expect that 2 x 448 = 896 million doses will be required to vaccinate all European with the current vaccines, each of which requires administering of two doses. But do we really need nearly a billion doses? In terms of immediate need, it turns out that this simple analysis is a bit too simple for at least three reasons.
First, the vaccines have not yet been tested on the young. The Phase 3 trial of the BioNTech Pfizer vaccine did not include subjects younger than 16. The Moderna Phase 3 trial did not include subjects younger than 18 (though a younger age group study has started). The Oxford AstraZeneca Phase 3 trial did not include subjects younger than 18 or older than 55. It is unlikely that any Western country will be vaccinating individuals younger than 16 in the near term. Thus, for now, no vaccine is needed for roughly one sixth of the EU population.
Second, a significant fraction of Europeans are expected to refuse vaccination, and there is evidence that resistance to vaccination is growing. The fraction of adult respondents who said that they would take a vaccine for COVID-19 dropped from 69% to 67% in Germany between August and October 2020, from 67% to 65% in Italy, from 72% to 64% in Spain, and from 59% to 54% in France. It is unfortunate that public acceptance of vaccination is not higher, but the practical reality is that vaccine doses for these individuals will not be needed until and unless there is a substantial change in sentiment.
These two factors suggest that about 500 million doses are needed to vaccinate all EU adults who wish to be vaccinated in the near term, assuming no wastage and that the number of adults willing to be vaccinated remains stable.
This number is a lower bound, however. If the vaccination programme produces good results with few side effects, and as the public witnesses the roll-out, public support could grow. Older individuals, or those at risk because of pre-existing conditions or frequent contact with possibly infected individuals, are likely to place a greater priority on being vaccinated (see for example here on individual risk perception during the pandemic). There is thus a strong argument for procuring an additional buffer of doses, either as firm orders or else on a contingent or optional basis.
Third, some phasing over time is going to be necessary in any case. If all 500 million doses were available today, it would still take many months to administer them because of logistics and staffing limitations. The crucial question is not, then, whether sufficient vaccine is available today, but rather whether sufficient vaccine can be made available as quickly as it is feasible to administer it?
In the context of global demand for supplies of the few vaccines that are approved and effective, it is far from certain the EU can secure additional supply quickly enough. The EU’s collective purchasing power is very substantial, but the bloc must compete with countries that are also willing to make substantial investments.
As previously noted, the EU has secured supplies of 380 million doses of the BioNTech Pfizer and Moderna vaccines, or 460 million if the optional 80 million Moderna doses are included, compared to an eventual need of at least 500 million doses.
The overall delivery schedule has not been published, but Germany has been forthcoming with some key details. To put these numbers in perspective, and following the logic we have sketched out, it would take about 90 million doses to vaccinate all adult Germans who are willing to be vaccinated. German Health Minister Jens Spahn has said that in the near term there would probably only be 400,000 vaccine doses for Germany, with another 11 to 13 million to follow by March. The German government will vaccinate those aged over 80 first, along with high-risk healthcare workers, followed in a second group by those over 70 and various high-risk individuals. There are 7.7 million Germans between 70 and 80 years old, and 5.4 million Germans older than 80 years, for a total of 13.1 million. A large fraction of these can be expected to be willing to be vaccinated. Thus, the 11 to 13 million available doses of BioNTech Pfizer and Moderna vaccine are unlikely to be sufficient to vaccinate both the elderly and the other high-risk groups currently targeted, including care-givers and those with other health conditions. Delay can be expected – delay that could have been avoided with better planning.
Press estimates suggest that supplies of the vaccines available to Germany for the first half of the year should be some 45 million BioNTech Pfizer doses and an additional 15 million Moderna doses, for a total of about 60 million. This represents only about two thirds of the 90 million doses that we estimate will be needed to vaccinate all adult Germans who wish to be vaccinated, which once again represents the minimum that should be sought.
For the end of 2021, the situation in Germany might look better if all goes as planned. The German Ministry of Health claims that a total of 136.3 million BioNTech/Pfizer and Moderna doses have been secured, almost all of which could be delivered in 2021. That should be sufficient to vaccinate 68 million out of 83 million Germans. If willingness to be vaccinated is high, a third approved vaccine might be needed to vaccinate the rest.
If it can be approved quickly, the Oxford-AstraZeneca vaccine could potentially fill this gap in the near term. There are, however, multiple concerns. The first is that the application for EU authorisation appears, at time of writing, either to have not been made, or else not made to the satisfaction of the European Medicines Agency, the agency that must assess the authorisation application. The date for EU authorisation consequently cannot be predicted with confidence, but press reports suggest that January is not possible. The second has to do with errors in the Phase 3 test, and what they imply. Results have been reported for the Phase 3 tests that were conducted in the UK and Brazil, but, according to The Lancet, “a subset (LD/SD cohort) in one of the UK trials inadvertently received a half-dose of the vaccine (low dose) as the first dose before a change in dosage quantification methodology”. Paradoxically, the cohort that received the incorrect dosage showed effectiveness of 90%, where those who received the planned dosage showed effectiveness of only 62%. It is unclear (and also somewhat inconsistent with other results from the Phase 3 trial) why the lower initial dose should lead to greater effectiveness. The combined results imply an effectiveness of 70%, which is fairly good compared to many conventional vaccines but rather poor compared the 95% effectiveness of the BioNTech and Moderna vaccines. How should policymakers decide who gets the less-effective vaccine?
In a concerning report, Der Spiegel said that the Commission concluded contracts with Sanofi, Johnson & Johnson and AstraZeneca between August and July 2020, even though it was already fairly clear since July that the BioNTech and Moderna vaccines were the frontrunners. In addition, according to Der Spiegel, the EU did not place firm orders with BioNTech and Moderna until mid-November (after both firms claimed 95% effectiveness), and even then failed to order as much as it could have. The Moderna CEO said the company could have delivered up to 300 million doses but the Commission wanted only 80 million. The magazine also alleged it was politically impossible for the Commission to order more doses from a German firm (BioNTech) than from the French firm Sanofi, though the Commission has disputed this claim (Sanofi has subsequently reported that its vaccine will not be ready for authorisation before Q4 2021).
Whether closing the gap will actually confer herd immunity remains uncertain. Estimates to date have typically assumed that 50% to 70% of the population needs to be vaccinated (or a combination of vaccinated and infected) for herd immunity to be effective, but this estimate is heavily dependent on R0, the rate of spread. If for instance the new COVID-19 variant reported in the UK has a higher rate of spread, then the percentage of vaccinations needed to achieve herd immunity will obviously be higher. Nevertheless, if a large fraction of the population can be vaccinated, and an especially large fraction of those at risk, many lives can be saved and the risk of overloaded healthcare systems will be reduced.
The populations at greatest risk are those between 70 and 80 years of age, and especially those over 80 years of age. These cohorts represent 8.6% and 5.8% of the total EU population, for a total of 14.5%, or 64.6 million individuals. Extrapolating from Minister Spahn’s statement (the allocation of vaccines to EU countries is based on population; thus Germany obtains 18.6% of available supplies), the EU expects to have between 59 and 70 million doses by the end of March, suggesting that as of March there will not be enough doses to cover both the elderly and the other high-risk groups, including care-givers. The situation will differ slightly in different countries as a function of their demographic breakdowns and the willingness to be vaccinated, but all countries can be expected to have a short-term shortfall.
What can the EU or its members do today?
A key element of the solution to the near-term shortfall would be for the EU to revisit its arrangements with Moderna. It has been reliably reported that the price of the Moderna vaccine to the EU is $18 per dose, which is the highest price of any of the vaccine prices that the EU has negotiated. The cost of purchasing the full 220 million doses of vaccine that the EU declined to purchase in November would have been about €3.3 billion at current exchange rates. When one considers that the EU is looking at a pandemic relief package of €750 billion, it becomes obvious that paying up to €3.3 billion to accelerate a return to something approaching normality by even a few months is an exceedingly good investment.
If BioNTech/Pfizer has any uncommitted capacity, then buying more of their vaccine would be more cost-effective than purchasing the Moderna vaccine. BioNTech is reportedly attempting to bring more production capacity online in various ways. One way or another, the EU would be well advised to purchase enough of one of the two highly-effective vaccines quickly enough, in order to avoid significant supply shortfalls.
If neither is possible, then the Oxford-AstraZeneca vaccine represents the best currently available alternative. Steps should to be taken on an accelerated basis to clarify the very strange results of the Phase 3 tests and to determine the most appropriate dosage to apply; to grant authorisation on an expedited basis if warranted; to determine who should receive the Oxford-AstraZeneca vaccine (perhaps populations that are at lower risk), if it is truly less effective at proper dosage.
The Johnson & Johnson vaccine may represent another alternative, but it will become available a few months later if all goes well.
Meanwhile, practical experience with the BioNTech Pfizer vaccine indicates that the amount of vaccine in each vial is often enough for six doses instead of the intended five, and that this is not a question of administering too little vaccine in the first five doses. There is simply a bit extra. Health authorities on both sides of the Atlantic appear to be concluding that it is safe to take advantage of this unexpected benefit, which may ease the shortage a bit.
The exact scale of the additional doses needed requires careful planning. Every effort should be made to ensure that additional supply can be obtained if the number of individuals willing to be vaccinated increases over time, as seems likely. Arranging for contingent supply could be the answer. The Commission has put in place options for additional quantities of vaccine. It will be important to try to ensure delivery commitments for quantities that have not been firmly committed; however, as a matter of good global citizenship, the EU should not needlessly tie up supplies that it cannot use, thus raising prices and decreasing availability for other regions of the world.
No matter what approach is taken, all vaccine suppliers must be monitored to ensure that they can meet committed delivery schedules at suitable quality.
An equally important consideration for both the EU and its members is to ensure efficient and effective administration at all levels. Procurement is only one stage in a complicated logistics chain. The vaccines must be distributed to countries, further distributed to health facilities within each country, and administered to as many people as possible.
Lessons to be learned
For future pandemics – and we should not doubt that there will be future pandemics – there are lessons to be learned from the coronavirus experience.
All indications are that the Commission was lax in its planning during the past few months. Funding a variety of vaccines early in the process in order to make sure that at least one or two successful vaccines would emerge was appropriate. Once it became clear, however, that a small number of candidates were likely to prove safe and effective, and would be ready much sooner than others, the EU’s purchasing strategy should have shifted to ensure a sufficient supply of frontrunners. This was apparently not done, and still has not been done.
In November, after the two frontrunners had been shown to be safe and highly effective, a course correction was still possible but was not made. This was a mistake – the cost of securing a sufficient supply of one of the two leading vaccines was minimal compared to the cost of delaying vaccination of a significant fraction of the population. As a practical matter, this possibility could have been anticipated, and some reserve cash set aside in advance – a consideration for future pandemics.
What this demonstrates is that the ESI, the chosen instrument for funding vaccines, was substantially under-funded at €2.7 billion (not all of which was for vaccines). This reflects poor planning on the part of the Commission. The need to ‘pivot’ once the front-runner vaccines had clearly established themselves was foreseeable, but was not foreseen.
Nevertheless, the EU has made major strides in strengthening its joint response to this pandemic. But it seems clear that continued vigilance to ensure both good planning and solidarity in the EU approach across the member states will be needed in future crises, just as it has been essential in the current crisis.
Marcus, J.S. (2021) ‘Has the European Union squandered its coronavirus vaccination opportunity?’ Bruegel Blog, 6 January