Limits...
Comparative analysis of the effectiveness of three immunization strategies in controlling disease outbreaks in realistic social networks.

Xu Z, Zu Z, Zheng T, Zhang W, Xu Q, Liu J - PLoS ONE (2014)

Bottom Line: Present stochastic strategies are mainly evaluated based on classical network models, such as scale-free networks and small-world networks, and thus are insufficient.The results show all the strategies have decreased the coverage of the epidemics compared to baseline scenario (no control measures).These results could have important significance for epidemic control research and practice.

View Article: PubMed Central - PubMed

Affiliation: Center for Biosecurity Strategy Management, Beijing Institute of Biotechnology, Beijing, P. R. China.

ABSTRACT
The high incidence of emerging infectious diseases has highlighted the importance of effective immunization strategies, especially the stochastic algorithms based on local available network information. Present stochastic strategies are mainly evaluated based on classical network models, such as scale-free networks and small-world networks, and thus are insufficient. Three frequently referred stochastic immunization strategies-acquaintance immunization, community-bridge immunization, and ring vaccination-were analyzed in this work. The optimal immunization ratios for acquaintance immunization and community-bridge immunization strategies were investigated, and the effectiveness of these three strategies in controlling the spreading of epidemics were analyzed based on realistic social contact networks. The results show all the strategies have decreased the coverage of the epidemics compared to baseline scenario (no control measures). However the effectiveness of acquaintance immunization and community-bridge immunization are very limited, with acquaintance immunization slightly outperforming community-bridge immunization. Ring vaccination significantly outperforms acquaintance immunization and community-bridge immunization, and the sensitivity analysis shows it could be applied to controlling the epidemics with a wide infectivity spectrum. The effectiveness of several classical stochastic immunization strategies was evaluated based on realistic contact networks for the first time in this study. These results could have important significance for epidemic control research and practice.

Show MeSH

Related in: MedlinePlus

The topological characteristics of the social contact network.(a) log-log plot of the degree distribution. (b) log-log plot of the vertex strength distribution. (c)  distribution. (d) heat map of .
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4008523&req=5

pone-0095911-g002: The topological characteristics of the social contact network.(a) log-log plot of the degree distribution. (b) log-log plot of the vertex strength distribution. (c) distribution. (d) heat map of .

Mentions: AI applies to the scenario where large heterogeneity in contact structure is observed. The analysis of Portland contact network shows both the degree distribution and the vertex strength[24] distribution are highly heterogeneous (see Figure 2), so AI could be used to control the epidemics in this network. AI strategy is defined as follows: first, selecting n random nodes from the network; second, for each chosen node, immunizing a random neighbor. A node with k connections will be targeted as the immunization node with the probability [15], where is the average degree. Therefore, for any susceptible individual, it will enter into the controlled set and proceed to class V with the probability q. For each infectious individual, it will enter into the controlled set and proceed to class D with the probability (second-order detection). Notice that the probability of direct detection is r, so class I individual will proceed to class D with the total probability . For individuals in other states (V, R and D), no influences were demonstrated on the spreading of the epidemics, so no processing was made.


Comparative analysis of the effectiveness of three immunization strategies in controlling disease outbreaks in realistic social networks.

Xu Z, Zu Z, Zheng T, Zhang W, Xu Q, Liu J - PLoS ONE (2014)

The topological characteristics of the social contact network.(a) log-log plot of the degree distribution. (b) log-log plot of the vertex strength distribution. (c)  distribution. (d) heat map of .
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4008523&req=5

pone-0095911-g002: The topological characteristics of the social contact network.(a) log-log plot of the degree distribution. (b) log-log plot of the vertex strength distribution. (c) distribution. (d) heat map of .
Mentions: AI applies to the scenario where large heterogeneity in contact structure is observed. The analysis of Portland contact network shows both the degree distribution and the vertex strength[24] distribution are highly heterogeneous (see Figure 2), so AI could be used to control the epidemics in this network. AI strategy is defined as follows: first, selecting n random nodes from the network; second, for each chosen node, immunizing a random neighbor. A node with k connections will be targeted as the immunization node with the probability [15], where is the average degree. Therefore, for any susceptible individual, it will enter into the controlled set and proceed to class V with the probability q. For each infectious individual, it will enter into the controlled set and proceed to class D with the probability (second-order detection). Notice that the probability of direct detection is r, so class I individual will proceed to class D with the total probability . For individuals in other states (V, R and D), no influences were demonstrated on the spreading of the epidemics, so no processing was made.

Bottom Line: Present stochastic strategies are mainly evaluated based on classical network models, such as scale-free networks and small-world networks, and thus are insufficient.The results show all the strategies have decreased the coverage of the epidemics compared to baseline scenario (no control measures).These results could have important significance for epidemic control research and practice.

View Article: PubMed Central - PubMed

Affiliation: Center for Biosecurity Strategy Management, Beijing Institute of Biotechnology, Beijing, P. R. China.

ABSTRACT
The high incidence of emerging infectious diseases has highlighted the importance of effective immunization strategies, especially the stochastic algorithms based on local available network information. Present stochastic strategies are mainly evaluated based on classical network models, such as scale-free networks and small-world networks, and thus are insufficient. Three frequently referred stochastic immunization strategies-acquaintance immunization, community-bridge immunization, and ring vaccination-were analyzed in this work. The optimal immunization ratios for acquaintance immunization and community-bridge immunization strategies were investigated, and the effectiveness of these three strategies in controlling the spreading of epidemics were analyzed based on realistic social contact networks. The results show all the strategies have decreased the coverage of the epidemics compared to baseline scenario (no control measures). However the effectiveness of acquaintance immunization and community-bridge immunization are very limited, with acquaintance immunization slightly outperforming community-bridge immunization. Ring vaccination significantly outperforms acquaintance immunization and community-bridge immunization, and the sensitivity analysis shows it could be applied to controlling the epidemics with a wide infectivity spectrum. The effectiveness of several classical stochastic immunization strategies was evaluated based on realistic contact networks for the first time in this study. These results could have important significance for epidemic control research and practice.

Show MeSH
Related in: MedlinePlus