Modeling contact networks and infection transmission in geographic and social space using GERMS.

BACKGROUND

Stochastic models of discrete individuals and deterministic models of continuous populations may give different answers to questions about infectious diseases.

GOAL

Discrete individual model formulations are sought that extend deterministic models of infection transmission systems so that both model forms contribute cooperatively to model-based decision making.

STUDY DESIGN

GERMS models are defined as stochastic processes in continuous time with parameters analogous to those in deterministic models. A GERMS model simulator was developed that insured that the rate of events depended only on the current state of model.

RESULTS

The confidence intervals of long-term averages of infection level in simulated GERMS models were shown to contain the deterministic model means.

CONCLUSION

GERMS models provide a convenient framework for testing the sensitivity of model-based decisions to a variety of unrealistic assumptions that are characteristic of differential equation models. GERMS especially facilitates making more realistic assumptions about contact patterns in geographic and social space.