Homophily impacts the success of vaccine roll-outs

Abstract

Physical contacts do not occur randomly, rather, individuals with similar socio-demographic and behavioral characteristics are more likely to interact among them, a phenomenon known as homophily. Concurrently, the same characteristics correlate with the adoption of prophylactic tools. As a result, the latter do not unfold homogeneously in a population, affecting their ability to control the spread of infectious diseases. Focusing on the case of vaccines, we reveal that, provided an imperfect vaccine efficacy, three different dynamical regimes exist as a function of the mixing rate between vaccinated and not vaccinated individuals. Specifically, depending on the epidemic pressure, vaccine coverage and efficacy, we find the final attack rate to decrease, increase or vary non monotonously with respect to the mixing rate. We corroborate the phenomenology through Monte Carlo simulations on a temporal real-world contact network. Besides vaccines, our findings hold for any prophylactic tool that reduces but not suppress the probability of transmission, indicating a universal mechanism in spreading dynamics.

Herd immunity and epidemic size in networks with vaccination homophily

“…Recently, we became aware of two other research works [31,32] that report results in line with what we have described here. They found qualitatively similar effects of homophily on epidemic size for scale-free networks [31] and empirical contact networks [32].…”

“…Recently, we became aware of two other research works [31,32] that report results in line with what we have described here. They found qualitatively similar effects of homophily on epidemic size for scale-free networks [31] and empirical contact networks [32]. This further corroborates the generalizability of our theoretical findings to networks with heterogeneous degree distributions [33].…”

Ethnic homophily affects vaccine prioritization strategies

“…Homophily describes the tendency of people from a particular demographic group to interact more frequently with people from the same group, and homophily among ethnic groups as well as age groups is well documented (McPherson et al, 2001; Currarini et al, 2009; Mollica et al, 2003). Homophily is known to affect disease dynamics (Kadelka and McCombs, 2021; Burgio et al, 2022; Hiraoka et al, 2022; Salathé and Bonhoeffer, 2008), and while the 4-phase COVID-19 vaccine allocation strategy implemented by the CDC in 2020 accounted for some differences in occupational hazards (healthcare workers first, then frontline essential workers, etc. ), it did not account for ethnic homophily (Dooling et al, 2020).…”

Spreading dynamics in networks under context-dependent behavior

“…To this subclass belong, among others, epidemic models where the individuals can adopt or not permanent prophylactic tools (e.g., vaccines, digital proximity tracing systems) to avoid transmission [24,25], and information spreading models where agents stably divide into active sources/spreaders and passive consumers.…”

“…The local (spreading) dynamics -and the global one emerging from it-is thus determined by both, the behavior of the types and the way they mix. For instance, it has been recently shown that an assortative (or homophilic) mixing with respect to prophylactic adoption can be either beneficial or detrimental to epidemic control [24,25,[29][30][31].…”