The Theory of Herd Immunity, or the Ayatollahs of Public Health

Herd immunity, what are we talking about?

Initially, this theory – much discussed in the epidemiological literature – was developed to show that it is not necessary to vaccinate 100% of a population to fully protect it from an infectious risk or even to eradicate it. Total immunity is ensured as soon as a fraction x of the population is vaccinated. This threshold x depends on two variables: the famous “reproduction number” R0, which indicates the number of people that an infected subject can contaminate on average, and the reported vaccine effectiveness, e.

For example, for a “classic” seasonal flu with an average R0 of 1.5 and a vaccine effectiveness of around 60%, we obtain an x of 56%. Thus, vaccinating 56% of the population would ensure complete and “efficient” protection of the population. According to France’s Group for Expertise and Information on Influenza (Groupe d’Expertise et d’Information sur la Grippe), the average rate was 24% in France during the 2017-2018 season! Still a lot has to be done…

Herd immunity and Covid-19, what’s the link?

Some experts, through their science, have attempted to extend the concept of herd immunity to include immunity acquired, after recovery, by individuals infected with SARS-CoV-2. Contact with the infectious agent would thus play the same role as vaccination, by stimulating the immune defences against the pathogen. However, widespread contact can be disastrous. If the R0 is estimated to be 2.5 (between 1 and 3 under current assumptions) and if it is assumed that the contact is immunizing (e=1), then the x is of the order of 60%.

If 60% of the population is contaminated, then:

1. The epidemic disappears;
2. The population is immune to a rebound epidemic, relapse or new infection by a similar pathogen.

Thus, the “public health” argument is that herd immunity provides effective, efficient and definitive immunization. Except that 60% of a population of 60 million inhabitants represents 36 million people, and even if the lethality rate of infected people is low, say in the order of 1 to 1.5%, that makes between 360,000 and 540,000 deaths! In fact, it would probably be less because these lethality rates refer to proven cases when many subjects are asymptomatic carriers. A generalized serology would be needed to determine the “true” rate. But even divided by 10, the figure of 36,000 to 54,000 potential deaths is considerable (at the time of writing, there are 860 deaths in France). What vaccine would be authorized with such a benefit/risk ratio?