Population structure in the simulation tool 4Flu

Overview

4Flu is an individual-based tool for the simulation of seasonal influenza. Up to 1,000,000 individuals are connected by a contact network which dynamically changes over age.

Demographic turnover

Demographic data for 30 countries have been implemented, using official data sources. Each dataset contains the number of individuals by year of age for the calendar years 1992 to 2060. The individuals simulated in 4Flu are aligned with these age-distributions. Individuals continuously grow older and "newborns" are continuously created during the simulation year. If the size of a cohort decreases from one year to the next, a random selection of individuals of this cohort are chosen and are removed ("death") during the simulation year. If the size of a cohort increases from one year to the next, some individuals are created in this cohort ("immigration") during the simulation year.

Risk status

As the vaccination coverage can depend on the risk status of the individuals, each individual also is given a risk status (either "low risk" or "high risk"). On the user interface, the user can set the percentages of newborns, 0-15 year olds, 16-59 year olds and 60+ year olds who belong to the "high risk" category. As individuals age, their risk status will be adjusted to keep the status population synchronized with these values at any time.

Contact network

Individuals are connected with other individuals ("contact network"). These contact networks are based on the Polymod study (Mossong et al. 2008: "Social Contacts and Mixing Patterns Relevant to the Spread of Infectious Diseases", PLoS Medicine 2008; 5: e74). The Polymod contact matrices describe the frequency of contacts among 5-year age-groups. These matrices were extended to accomodate 101 cohorts. The contact structure of the Polymod contact matrices by nature is asymmetric, yet, a symmetric contact network can be constructed from these matrices. This was obtained by assuming that every individual initiates (on average) as many contacts with other age-groups which is given by the Polymod matrices. Finally, all contacts for transmission of infection, irrespective by which individual they were initialized.
As individuals age, the age-distribution of their contacts also needs to change (according to the Polymod matrices). Thus, some of their contacts must be removed, new ones must be added. This is being done throughout the simulation to keep the contact network synchronized with the Polymod contact structure at any time.