This morning’s New York Times offers a panel discussion on the question of who should get vaccinated against Covid first. Broadly speaking, they take a utilitarian position: it’s interesting that none disagreed with the positions taken by panelist Peter Singer, the world’s most prominent utilitarian philosopher. And I wouldn’t either, except for one thing.
The vaccines approved by the FDA, along with those approved by other countries like China and Russia, have gone through the fastest possible testing. Tens of thousands of individuals have been placed in control and treatment groups in order to determine two things: to what extent do the vaccines reduce the likelihood of getting infected (efficiency) and how common and severe are the side effects (safety)? Meeting both criteria is sufficient for approval, which is how it should be.
But there is another crucial question, to what extent do the vaccines reduce transmission of the virus to others? The answer does not affect whether these vaccines should be employed, but they do have large consequences for other policies during this phase of the pandemic, such as rules for separation and masking, restrictions on activities and events, resumption of in-person schooling, and how much should be spent on interventions like ventilation overhauls. To the extent that vaccination reduces transmission, other restrictions and investments can be modified as the vaccinated portion of the population increases. Unfortunately, our knowledge of this issue is minimal. We don’t have any published lab results at all, and we are at least months away from meaningful epidemiological data.
A rollout that prioritizes crucial learning could change this. Some substantial portion of the early vaccines could be reserved for community trials. A number of communities could be given treatments in which a designated proportion of the population is vaccinated as soon as possible; this portion could be varied (30%, 50%, 70%) so that a variety of treatments could be tested. Others matched to them by relevant demographic, economic and other variables would be controls and would not receive any vaccines during the trial period. (Note that the lack of blinding at the community level should not be a serious problem as long as unvaccinated individuals in treatment communities are given a convincing placebo.) Everyone living in these communities would be tested regularly. We could then observe differences between community infection rates corresponding to treatment and infer transmission probabilities under real world conditions. It might also be possible to learn how transmission varies across the different viral strains that have emerged. The entire operation could be accomplished within the space of a month or less.
What is disheartening is that not a single expert on the Times panel broached this possibility. They are entirely preoccupied with the health significance of vaccination at the individual level and consider communities only in social and economic terms. To the extent they consider the need for learning at all it is in the context of individual response to vaccines, such as comorbidities and interactions with other drugs people may be taking.
Where’s the public health?