by Angela Guess
Bernard Marr recently wrote in Forbes, “The World Health Organization has declared the Zika Virus a public health emergency that could affect four million people in the next year as it spreads across the Americas. Big data and analytics have played a role in containing previous viral outbreaks such as Ebola, Dengue fever and seasonal flu, and lessons learned there are undoubtedly being put to use in the fight against Zika. However while statistical modeling of vast, real time datasets is becoming ingrained across healthcare and emergency response, the support infrastructure needed to put these initiatives to work at ground level is lagging behind, experts believe. Big data research has dramatically sped up the development of new flu vaccines. By analyzing the results of thousands of tests at institutions around the world, compounds can be developed to target the specific proteins that are found to enable the virus to grow. Big data is also used by epidemiologists to track the spread of outbreaks.”
Marr goes on, “The Google Flu Trends project, launched in 2008, was one of the first to show how search engine data could be used to predict flu trends. While it often achieved high levels of accuracy (correlating with US CDC reporting of flu outbreaks at a rate of 97%), its tendency for occasionally making wildly inaccurate predictions is often attributed to its reliance on one primary (and volatile) source of data. Search engine queries, it has been established, are often influenced by other factors such as media reporting or public concern over other illness not related to flu. Jamie Powers, consultant to the healthcare industry with SAS, says that ‘This highlights why multiple, disparate data sources are needed for this type of predictive analysis. We need to connect vaccine-makers, the CDC and other national and state public health agencies, even health providers, before an outbreak is identified’.”
photo credit: Flickr/ coniferconifer