To understand the effects of vaccines, it is critical to distinguish between three things: their ability to block infection, to prevent clinical symptoms (which may require hospitalisation) and to protect against life-threatening illness.

Different immune responses, such as antibodies and T-cells, work in different ways and result in different types of immunity. Focussing on a single immune response can distort our understanding of the epidemic process, especially as we are constrained by what we can easily measure. For example, the antibody studies conducted in the UK last year are likely to have significantly underestimated the degree of exposure to Covid-19 prior to lockdown, partly because antibodies wane rapidly and also because many people fended off infection without developing antibodies, as shown by recent studies on T-cells.

While it is not that useful to speculate on whether one vaccine is better than another, we should recognise that infection-blocking immunity is temporary, while immunity against severe disease and death is likely durable. This is in line with our understanding of the transmission dynamics of other coronaviruses with which we exist without complaint because they do not cause a significant burden of serious disease.

It is now widely agreed that Covid-19 is close to becoming endemic, in a way we enjoy with other seasonal coronaviruses, where the reduction in risk to vulnerable people is primarily achieved through the maintenance of high levels of infection-blocking immunity in the population (otherwise known as herd immunity) through regular infection. The vaccines that we currently employ appear to be highly effective in preventing life-threatening illness but do not meaningfully contribute to the maintenance of herd immunity.

Now that we have protected the majority of the population – including those who did not need it – against severe disease and death from Covid-19, we should focus our attention on those who are vulnerable and yet remain unprotected. This applies to the global population of vulnerable people, especially those in poorer nations, rather than just your sceptical neighbour.

It is also important to recognise that our long history of previous exposure to seasonal coronaviruses may well have protected many of us from severe Covid-19. It is my opinion that past exposure to related influenza strains helped to prevent Covid-19 reaching the scale of the influenza pandemic of 1918. Prior to 1918, influenza most likely died out completely within inter-pandemic periods. Since the previous influenza pandemic had occurred in 1890, most individuals under the age of 30 had no experience of it and they are the ones who died.

Based on these principles, a shift in focus towards vaccines which prevent death but not necessarily infection is long overdue. Before vaccines, the only viable solution was to protect the vulnerable population through state-sanctioned shielding. The vaccines should have changed that, but instead we find ourselves trapped by the superannuated conviction that vaccines must block infection as well as disease.

Sunetra Gupta is professor of theoretical epidemiology at the Department of Zoology, University of Oxford