URL
https://doi.org/10.1101/2021.02.26.21252553
Type d’article
Preprint
Thème
Immunité Stratégies de vaccination
Que retenir de cet article, en 1-2 phrases ?
Taking into account population heterogeneities may lead to a lower vaccine-induced herd immunity threshold (HIT) than the one obtained from a corresponding model with a homogeneous population.
Objectifs de l’étude / Questions abordées
The goal of this paper is to investigate the effect of heterogeneity on the values of the HIT by considering a multi-group vaccination model.
Méthode
The authors consider a standard susceptible-vaccinated-exposed-infected-recovered (SVEIR) Kermack-McKendrick multigroup model for the transmission dynamics of the disease. The total population is divided into m groups with similar characteristics, such as age, contact patterns, infectious period, or social, cultural, demographic, or geographic factors. It is assumed that the vaccine is not perfect and the protection wanes over time. It is also assumed that recovery induces permanent immunity and there is no disease-induced mortality. The analysis of the dynamics is done only for the case of two groups (m=2). The basic (vaccination) reproduction number Rv is computed. This number depends on the proportions (v1 and v2) of vaccinated individuals in each group, on the local basic numbers without vaccination and on the interactions between the two groups. By considering different types of mixing (rates of exchanges from a group to another) and two different scenarios (vaccinating only one group or vaccinating the two groups), the authors investigate the existence of values for v1 and v2 such that the proportion of vaccinated individuals in the total population is minimal and such that the basic vaccination reproduction number Rv is less than one which ensures the extinction of the epidemic.
Résultats principaux
Under biased random mixing (transmission occurs more because of mixing between groups than because of mixing within groups), to achieve herd immunity, it is optimal to exclusively optimize vaccination coverage among the group that results in relatively more prevention of per-capita disease transmission. Under biased assortative mixing assumption (transmission occurs more due to mixing within groups rather than due to mixing between groups), it is optimal to vaccinate more than one group. The authors show that population heterogeneities tend to result in lower HIT values, compared with the homogeneous population case. This is true under both proportionate and other types of mixing among heterogeneous populations. Using realistic numerical examples and parametrization (e.g. assuming biased assortative mixing with vaccine efficacy of 95% and basic reproduction number of 2.5) they illustrate this finding, where the HIT value considering heterogeneity is shown to be significantly lower (40%) compared with a HIT value assuming a homogeneous population (63%).
Commentaire / brève évaluation, limites, ouvertures possibles
For many diseases (SARS-CoV-2, for instance), the disease-induced mortality cannot be neglected and recovery only induces a temporary immunity. In this situation making the basic (vaccination) reproduction number Rv less than one is no more sufficient to eradicate the disease because one may have an endemic equilibrium even if Rv is less than one.