Talk of a Covid-19 vaccine by spring of 2021 is just too soon

Trials involving hundreds of thousands of people are needed to produce a truly safe vaccine

Everybody is tired of Covid-19, and wants to know when normal life can resume. We are told this will depend on the availability of a vaccine. There have been promises from companies, universities, and governments that this will be by the end of 2020. But how sure are we that we will ever have a vaccine, and if so when will our population be protected from Covid-19?

The good news is that the World Health Organisation estimates there are 165 vaccine programmes under way. This is unprecedented, as is the diversity of approaches to vaccine development. Shots on goal matter. Indeed, it is likely we will need more than one effective vaccine if we are to reach the numbers globally that are necessary for effective immunity.

The most advanced vaccine programmes have completed the first two stages of clinical development. They have shown in small numbers of volunteers – first in tens, then in hundreds – that vaccines can evoke an antibody response that neutralises the virus. They have collected some information on immediate side effects – dose-related flu-like symptoms that usually last for hours at most.

These studies also provide some information on dose selection, and whether a second booster shot is necessary. In parallel, studies in model systems – mice and non-human primates infected with the virus – have provided crucial information. They can show that the antibodies evoked by the vaccine shut down both the virus and the clinical consequences of infection.


However, animal models are not humans, and studies large enough to test whether any of these vaccines are protective against Covid-19 are only just under way. How big do these studies need to be? That depends on many issues, but certainly tens and possibly hundreds of thousands of people need to be included in each of these trials.

These trials are randomised and double blind. The odds are even that you will get the vaccine or a dummy, and neither you nor the investigator knows which you have received. This is crucial to a reliable interpretation of outcome. Much sloppy science has been performed in the rush to get answers during the pandemic.

Phase three

Several factors will influence how informative the outcome of these phase three trials is likely to be.

First, they must be performed in a susceptible population: if there are no cases in the group getting the dummy vaccine (the control group), it is impossible to determine whether the vaccine has conferred any benefit. The more enriched the control group is with cases, the easier it will be to see a benefit from the vaccine in a smaller sample size. The results also need to generalise, hence their performance in regions and countries with high rates of disease.

Depending on the shifts in the virus, vaccination provides variable effectiveness each season. Sometimes it is very protective

So, what is the benefit? Here we need to see that serious illness from the virus is markedly reduced in the vaccinated group compared to the controls. Thus the trial must be large enough to examine the effectiveness of the vaccines in subgroups, such as by age or ethnicity.

Second, these trials need to be large enough to provide reassurance about long-term safety. Obviously the larger the trial that fails to detect a unwanted side-effect the more reassuring is the information. Established vaccines are remarkably safe, and claims that link them to syndromes like autism are demonstrably bogus. However, disturbing numbers of people believe in vaccine hoaxes, and the trials need to deliver answers on safety convincing to the population at large.

Third, the reaction to the outcome of these trials will need a clear description of the benefits and risks they infer. Let us take the flu vaccine as an example. Depending on the shifts in the virus, vaccination provides variable effectiveness each season. Sometimes it is very protective, others it just dampens the symptoms of flu.

Masks, distancing, and hand-washing will be the rule for the foreseeable future

It is certainly possible that a vaccine for Covid-19 will be something similar; it may be variably effective in different people. So far we have not seen mutations in the virus that clearly relate to the severity of the illness but that may well come. Then we will need to tweak the vaccine over time, just like we do for flu.

On the adverse event side, the trials (or more likely an overview analysis of multiple trials) need to provide assurance on the likelihood of rare but serious risks. For example, Guillain Barre Syndrome (GBS) is a neurological disorder that occurred in one in 100,000 patients who received a swine flu vaccine. A complicating issue is that GBS can also be caused by viral infections and indeed cases have complicated Covid-19.

To exclude the likelihood of a side effect of such rarity attributable to a vaccine one would need a trial of hundreds of thousands of patients. The Centres for Disease Control monitor GBS after flu vaccination in the US. It fluctuates, but sometimes there are increased rates, but only 1-2 per million vaccinated. They estimate that you are more likely to get GBS from the flu itself than from the flu vaccine.

Sufficient doses

Finally, if we do establish that benefit from a vaccine clearly outweighs risk, we must have access to sufficient doses, convince (or require) people to take it and distribute and administer it to millions.

It is likely that vaccination rates will differ markedly between countries and amongst segments of the population within each country.

Given the political, cultural, and logistical challenges, attaining the goal of population immunity of 60-80 per cent, necessary in our connected world, is likely to take some time, maybe years.

Population protection in advanced societies with an assured supply of a vaccine by late summer 2021 is a reasonable bet. For example, Moderna, maker of an RNA vaccine, estimates that it will have 500,000 doses available at the conclusion of its 30,000-patient trial.

Despite politically and commercially-driven promises that we will be protected from Covid-19 by vaccines early next year, this is extremely unlikely. Mass vaccination based on trials of a few thousand patients – as suggested in Russia – would be reckless. Declaring victory based on such evidence would be similarly foolhardy but perhaps predictable in the US before November 3rd.

Masks, distancing, and hand-washing will be the rule for the foreseeable future. In the meantime there is encouraging progress in the development of drugs to curb the severity of Covid-19 and this, as with AIDS, may lead us to a semblance of normality even sooner than an effective vaccine.