It has been seen that in the US, seasonal influenza can cause thousands of deaths as well as hundreds of thousands of hospitalizations. If forecasting is done, then it can able to improve the prevention, planning as well as it can care to reduce the human toll of severe seasonal as well as pandemic influenza.
As per the researchers, they have now developed a computer model which will help in forecasting about the new upcoming flu season which will be like the results that are often said to be not accurate at all. The main challenge about this is by choosing the right kinds of data to feed the models.
As per professor Lauren Ancel Meyers as well as the postdoctoral researcher Zeynep Ertem, he has developed a method for evaluating the hundreds of data sets which will be found with the most predictive and how to combine them with it to get the most accurate forecasts. When it comes to the mathematical parlance, it is known as an optimization problem.
It has been seen that about more than 600 flu related data sets have been evaluated and they have found that this best prediction will be coming from electronic health records which can get collected by athenahealth. It is the company that provides cloud-based services for the health care providers. These type of data can be collected across the US as it included the information which is how many patients have received the flu vaccinations as well as some positive flu results as and flu related prescriptions. It will combine with athenahealth’s data with traditional surveillance data that was collected by the Centers for Disease Control and Prevention (CDC). These predictions are said to be about 15 percent more accurate with the combined data sets than the CDC data that are used usually in this.
The study of the athenahealth’s data that was provided to the University of Texas at Austin shows that it will be worthy to cross some of the hurdles as the data can be powerful. Meyers said it in a statement.
The results of this research study were got published in the journal PLOS Computational Biology. This method can be able to apply to any type of geographical region along with infectious diseases like mosquito transmitted dengue or chikungunya, as said by Ertem.