Efficient vaccination strategies for epidemic control using network information

Abstract

Background: Network-based interventions against epidemic spread are most powerful when the full network structure is known. However, in practice, resource constraints require decisions to be made based on partial network information. We investigated how the accuracy of network data available at individual and village levels affected network-based vaccination effectiveness.

Methods: We simulated a Susceptible-Infected-Recovered process on static empirical social networks from 75 rural Indian villages. First, we used regression analysis to predict the percentage of individuals ever infected (cumulative incidence) based on village-level network properties for simulated datasets from 10 representative villages. Second, we simulated vaccinating 10% of each of the 75 empirical village networks at baseline, selecting vaccinees through one of five network-based approaches: random individuals (Random); random contacts of random individuals (Nomination); random high-degree individuals (High Degree); highest degree individuals (Highest Degree); or most central individuals (Central). The first three approaches require only sample data; the latter two require full network data. We also simulated imposing a limit on how many contacts an individual can nominate (Fixed Choice Design, FCD), which reduces the data collection burden but generates only partially observed networks.

Results: In regression analysis, we found mean and standard deviation of the degree distribution to strongly predict cumulative incidence. In simulations, the Nomination method reduced cumulative incidence by one-sixth compared to Random vaccination; full network methods reduced infection by two-thirds. The High Degree approach had intermediate effectiveness. Somewhat surprisingly, FCD truncating individuals' degrees at three was as effective as using complete networks.

Conclusions: Using even partial network information to prioritize vaccines at either the village or individual level, i.e. determine the optimal order of communities or individuals within each village, substantially improved epidemic outcomes. Such approaches may be feasible and effective in outbreak settings, and full ascertainment of network structure may not be required.

Keywords: Agent-based models; Sociocentric networks; Vaccination.

Fig. 1

Flow diagram of the village-level study design.

Fig. 2

Flow diagram of the individual-level study design.

Fig. 3

Comparison of network characteristics to predict village-level cumulative incidence across different levels of network degree truncation using fixed choice design. Numbers underlying this figure are provided in Supplementary Table 4. RMSE relates to cumulative incidence measured on (0–100) scale.

Fig. 4

Estimated cumulative incidence under different approaches to vaccinating 10% of each village. The six different vaccination methods are described in Section 2.4. Solid or dashed lines and markers are point estimates; shaded areas represent 95% pointwise confidence intervals. Cumulative incidence is calculated as the mean of each of 75 villages’ mean cumulative incidence across 500 SIR runs, i.e. $C{I}_{mean}=mean\left(mean{\left(C{I}_j\right)}_i\right)$, where i indexes villages and j indexes SIR runs. The confidence intervals are computed as $C{I}_{mean}\pm \hspace{0.17em}1.96\left(SD\left(C{I}_i\right)/\sqrt{75}\right)$, where SD is standard deviation. The High Degree method uses a cutoff of K = 6, which corresponds to the median of the 75 village median degree values. Numbers underlying this figure are provided in Supplementary Table 3.