Mediation Analyses for the Effect of Antibodies in Vaccination

We partition the total ratio effect (one minus the vaccine effect) from a vaccine trial into indirect (effects through antibodies) and direct effects (other effects). Identifying $p$, the proportion of the total effect due to the indirect effect, depends on a cross-world quantity, the potential outcome among vaccinated individuals with antibody levels as if given placebo, or vice versa. We review assumptions for identifying $p$, showing that, unless the effect of adding antibodies to the placebo arm is equal in magnitude to that of subtracting antibodies from the vaccine arm, there are two versions of $p$. We focus on the case when the placebo is unlikely to induce needed antibodies, and in that case if a standard assumption (given confounders, potential mediators and potential outcomes are independent) is true, only one version of $p$ is identifiable, and if not neither is identifiable. We propose alternatives for identifying and estimating the other version of $p$, without making the standard independence assumption and instead experimentally modeling to identify the formerly cross-world quantity. First, a three arm trial with the extra arm being passive immunization (administering monoclonal antibodies), and using a model of antibody level amongst vaccinees. Second, combining information from a placebo-controlled vaccine trial with a placebo-controlled passive immunization trial.