By using computer simulations and modeling, an international group of researchers including scientists from the Virginia Bioinformatics Institute at Virginia Tech’s Network Dynamics and Simulation Science Laboratory have determined how a pandemic influenza outbreak might travel through a city similar in size to Chicago, Ill.

This information helped them to determine the preferred intervention strategy to contain a potential flu pandemic, including what people should do to decrease the likelihood of disease transmission.

The new results, based on three different computer simulation models, are described in a paper published in the Proceedings of the National Academy of Sciences by scientists involved in the Models of Infectious Disease Agent Study.* It is a collaboration of research and informatics groups supported by the National Institutes of Health to develop computational models to examine interactions between infectious agents and their hosts, disease spread, prediction systems, and response strategies.

The global epidemic of avian influenza in bird populations, as well as the risk of a virulent form of the bird flu virus being transferred to humans, has made influenza pandemic preparedness a top public health priority in the United States, Europe, and other countries. The great influenza pandemic of 1918 resulted in 40 to 50 million deaths worldwide. If a pandemic were to occur today, it could cause widespread social and economic disruptions.

In the paper, “Modeling Targeted Layered Containment of an Influenza Pandemic in the USA,” members of the Models of Infectious Disease Agent Study Working Group on Modeling Pandemic Influenza concluded that a timely implementation of targeted household antiviral prevention measures and a reduction in contact between individuals could substantially lower the spread of the disease until a vaccine was available.

The groups coordinated efforts to each create individual-based, computer simulation models to examine the impact of the same set of intervention strategies used during a pandemic outbreak in a population similar in size to Chicago, which has about 8.6 million residents.

Intervention methods used were antiviral treatment and household isolation of identified cases, disease prevention strategies and quarantine of household contacts, school closings, and reducing workplace and community contacts. Although using the same population, each model had its own representation of the combinations of intervention. All of the simulations suggest that the combination of providing preemptive household antiviral treatments and minimizing contact could play a major role in reducing the spread of illness, with timely initiation and school closure serving as important factors.

“[The Virginia Bioinformatics Institute’s] computer simulation models are built on our detailed estimates for social contacts in an urban environment,” said Professor and Network Dynamics and Simulation Science Laboratory Deputy Director Stephen Eubank, who leads the institute’s team in the working group. “They provide a realistic picture of how social mixing patterns change under non-pharmaceutical interventions such as closing schools or workplaces. For example, when schools close young students require a caregiver’s attention. That can disrupt social mixing patterns at work if a working parent stays home or make closing schools pointless if the children are placed in large day-care settings. We can use our model to suggest the best mix of intervention strategies in a variety of scenarios, taking factors like these into account.”

Bruno Sobral, executive and scientific director of the institute, remarked: “Transdisciplinary science, which is the foundation of the way we do research. [It] requires a special type of collaborative framework at the very outset of a project. The highly detailed social-network models that underpin this international research project arise from transdisciplinary science that removes disciplinary boundaries and promotes innovation. The impact of this approach to science is highlighted by the success of this research undertaking, which benefits from a very clear interface between diverse experts in high-performance computing, disease modeling, and public health practice.”

While the three different models used in the study show that timely intervention significantly impedes the spread of influenza through a population, the authors caution against over-interpretation of the modeling results. The researchers emphasize that the models are tools that provide guidance rather than being fully predictive. In the case of a future outbreak of pandemic influenza, capabilities such as real-time surveillance and other analyses will hopefully be available for the public health community and policy makers.

“These models, which are built from the best available data and with the best tools, contribute greatly to our understanding of how a pandemic could spread and what measures might protect the public’s health,” said Jeremy M. Berg, director of the National Institute of Health’s National Institute of General Medical Sciences, which supports the program. “But they are not our only resource — field work and experimental studies remain critical and will enhance the quality and reliability of these and other models.”

Along with Eubank, Professor and laboratory Director Chris Barrett, Professor and laboratory Deputy Director Madhav Marathe, Simulation Science Statistician Richard Beckman, graduate student Bryan Lewis of Lexington, Ky., Assistant Professor and Senior Research Associate Anil Vullikanti and other members of the Network Dynamics and Simulation Science Laboratory group contributed to the study. The teams contributing to the working group include researchers from the Virginia Bioinformatics Institute, the University of Washington, Fred Hutchinson Cancer Research Center in Seattle, Los Alamos National Laboratories, Imperial College London, and the University of Pittsburgh. The paper’s lead author, M. Elizabeth Halloran, is affiliated with the University of Washington in Seattle, and the Fred Hutchinson Cancer Research Center also in Seattle.

Models of Infectious Disease Agent Study is a collaboration of research and informatics groups established to develop computational models of the interactions between infectious agents and their hosts, disease spread, prediction systems, and response strategies. The models will be useful to policymakers, public health workers, and other researchers who want to better understand and respond to emerging infectious diseases. If a disease outbreak occurs, the MIDAS network may be called upon to develop specific models to aid public officials in their decision-making processes. Find more information about the collaboration.

The Virginia Bioinformatics Institute at Virginia Tech has a research platform centered on understanding the "disease triangle" of host-pathogen-environment interactions in plants, humans and other animals. By successfully channeling innovation into transdisciplinary approaches that combine information technology and biology, researchers at VBI are addressing some of today's key challenges in the biomedical, environmental and plant sciences.

*This work was partially supported by the National Institute of General Medical Sciences Models of Infectious Disease Agent Study network grants U01-GM070749, U01-GM070694, U01-GM070698, and U01-GM070708.

  • Modeling targeted layered containment of an influenza pandemic in the United States. Proceedings of the National Academy of Sciences (2008). In press. The paper is available on-line.

Share this story