Optimal timing for social distancing during an epidemic

Oscar Patterson-Lomba

It is important to remark that these results stem from very idealized assumptions about the course of the epidemic. The inclusion of transmission or contact level heterogeneities within the population [19] (e.g., age structure) may play a significant role in the quantitative description of the results in this paper, as would the inclusion of seasonality effects. Additionally, since we are using a deterministic mean-field approach, these analyses do not account for the effects of stochastic fluctuations (which could lead to epidemic extinction if the number of infected is low enough during the social distancing phase) and as a consequence, the findings herein apply more closely to very large populations (i.e., in the thermodynamic limit) where stochastic fluctuations are less relevant. Similarly, the model assumes homogeneous mixing among individuals, thus the contact network structure of the population is not accounted for; as a result this model cannot realistically predict the epidemic-hampering effects that social distancing measures (e.g., limiting the size of crowds) have in the context of contact structured populations [20] .