EpiDefend is an algorithm recently designed by the Australian Department of Science and Technology (DST). Its initial purpose was to scan disease outbreaks and decide whether they were caused naturally or were a product of ‘bio-terrorism’. But EpiDefend is now being used to directly model and monitor natural influenza outbreaks in the state of Victoria in Australia.
The technology was developed by Tony Lau, a research scientist at the DST. It uses patient records and the results of diagnostic test to model the spread of a disease. Lau claims that his algorithm incorporates environmental factors, like moisture, into the calculations and even uses particle filtering to make more accurate and helpful models. Lau remarked, “Our team’s goal is dual-purpose, we want to fulfil our defense charter, protecting our forces against intentionally released biological agents, but disease forecasting will also support the national security and public health areas.”
The team at DST collaborated with Melbourne School of Population and Global Health to get their hands on digital data from healthcare sector, and helped to design and hone the algorithm.
Particle filtering is a technique which helps us close in on the degree of uncertainty by the help of information gathered from particular situation. In other words, it helps the algorithm churn out more precise readings. EpiDefend can predict an influenza epidemic two months before it spreads, thus giving the health department of Victoria plenty of time to prepare for upcoming sweep of disease.
The Use Of Algorithm As A Diseases Outbreak Scanner To US Global Bio-Surveillance System
Fortunately for the people of Victoria, the algorithm is planned to be in place ahead of next year’s flu outbreak. Influenza vaccination programs are severely delayed by the fact that each year a new vaccine has to be produced because different strains of the virus tend to show up each year. How much EpiDefend will actually help Victoria’s heath department is yet to be seen, but if the virus epidemic patterns are accurately projected then distribution of vaccines can be prioritized to the predicted hotspots of the virus.
In a rather recent development, the US Department of Defense has granted one million Australian dollars to Tony Lau’s team. The funding is a part of Coalition Warfare Program set up with the aim of making EpiDefend a part of their global bio-surveillance system. The funding makes it appear as if EpiDefend might eventually be used on a global scale to forecast influenza epidemics. In the lab, this funding is going to be used to employ two post-docs who will work to develop ways in which healthcare professionals can make the most use out of the predictions/models made by EpiDefend and take appropriate steps.
If Tony Lau’s claims about the technology are correct, EpiDefend is going to be a revolutionary technology in a way that could possible change how we deal with epidemics. A greater awareness about future epidemics will without a doubt make our interventions more effective and efficient. As more and more electronic health records (EHR) become available, better algorithm and programs are anticipated to show up that can better simulate disease outbreaks.
However, it should be noted that all efforts to make use of computational power, in making mathematical models out of EHR, depend awfully on the reliability of the EHR itself. Even if all the fresh medical data somehow gets streamlined into these algorithms, the fact that a huge number of people don’t visit a doctor, let alone get their records uploaded, is going to be a hurdle for the success of any such algorithm trying to tackle outbreaks on a global level.