Planning Ahead for Pandemic

​THE 2009 H1N1 swine flu pandemic provided yet another reminder that even in this age of vaccines and emergency planning, fast-spreading diseases can still wreak havoc across the globe. Experts say the odds are high that another pandemic will occur in the not-so-distant future. With that in mind, a team of researchers has been working to develop software modeling tools that could help state and federal authorities predict, prepare for, and respond to such an event.

That work has recently yielded results in the form of the Texas Pandemic Flu Toolkit. The effort is being led by Lauren Ancel Meyers, who is director of the Division of Statistics and Scientific Computation and professor at the University of Texas-Austin (UT-Austin). A team from the Texas Advanced Computing Center also worked with UT-Austin on the pandemic toolkit project.

The toolkit includes three different tools for decision-making. The first is a simulator that can map the spread of a pandemic based on where it starts and how strong it is, taking into account different demographics, transportation patterns, and the use of vaccines. Second is a tool that provides strategies for state stockpiling of ventilators based on expected demands.

David Morton, of UT-Austin’s Cockrell School of Engineering, worked on the ventilator tool. Morton says the researchers took the information available from the 2009 pandemic and scaled it to different types of pandemics.

The tool allows planners to figure out not only whether they will be short on ventilators and supplies but also how they will dole out resources. The tool can help pinpoint where the ventilators need to be stationed for maximum effectiveness to avoid the most patient deaths.

The simulation and ventilator tools can help officials plan for the next pandemic by modeling what could happen and projecting “how many resources they should have stockpiled…and how to distribute things like antivirals and vaccines,” Meyers says.

The third component of the kit is a forecasting tool. This feature can be used once there’s an event to predict the anticipated level of hospitalization based on a number of factors.

Morton hopes that in the future the toolkit will help the Centers for Disease Control and Prevention (CDC) and others charged with pandemic planning and response make more nuanced decisions with regard to vaccine distribution and related issues.

For example, he points out that Texas is such a large state that it takes more time for supplies to get from one part of the state to another. Morton raises the question of whether that should be taken into consideration when vaccines are distributed to states, rather than simply apportioning them on a pro-rata basis according to the population.

The Texas Department of State Health Services is already using the toolkit for planning and training purposes. Looking ahead, Meyers says researchers will continue to try to improve the modeling to make it more accurate.

Another project Meyers has worked on is designed to find ways to better use the CDC’s influenza-like illness network (ILInet). For ILInet, health providers are tasked with reporting flu-like symptoms in patients.

Meyers’ team used computational modeling to assess whether the information from ILInet adequately predicted the number of people who ended up with the flu virus and the number of people who were hospitalized because of the flu in past events. The researchers then developed guidelines on how responsible parties could better monitor for flu-like symptoms in the future and detect emerging new strains as early as possible.

Meyers says there are several other projects in the works that would expand the functionality of the toolkit. One of these projects would help authorities charged with emergency response to prioritize the distribution of antivirals, much as the ventilator stockpiling tool helps with distribution of that equipment. Additionally, she would like to develop a tool that could actively assist in public health exercises.