Written by Karl Beidatsch Tuesday, 26 May 2009 00:00
The effectiveness of these plans relies on the data that they are based on being accurate, otherwise there is no way of knowing whether or not the action plan will even work.
In a step towards providing this surety, a study funded by the WHO is seeing UWA’s Professor George Milne, Doctor Joel Kelso and Doctor Simon Huband create a finely-detailed virtual replica of the city of Albany to run pandemic simulations.
Prof Milne and partners used a topographical map of Albany as their starting point. Combining this with information from the WA Government, they are marking the location of homes, schools and businesses, and allocating the accurate number of virtual citizens to the range of daily activities.
“We’ve broken every bit of the spatial Albany into census collection districts, each of about 200 people,” says Prof Milne.
“We model each individual and attach them to a household, and a school or workplace. People who aren’t attached to a school or workplace bubble out into the community at random.
“We believe that our Albany model is the most detailed simulator there is.”
Each day in the simulation is broken up into two cycles; in the morning, the people go to work or school or just out into the community, and in the evening they return to their homes.
In each cycle, the people move around, and each time they come into contact with an infected individual the probability of transmission is calculated.
The probability that the disease will be passed on depends on a number of variables, including the age of the people involved.
Every time the simulation is set up, it is run forty times and the results used to find the probable outcomes. One infected individual is randomly assigned within the community each day to feed the epidemic.
The initial details of the simulation, including the transmissibility of the virus, the government responses to it and the level of immunization, can all be altered to model real life conditions as closely as possible.
The team is now working on how to model the effect of antiviral drugs, like Tamiflu and Relenza, on the virus in their model.
