The Zika virus recently claimed its first victim in the Continental US, taking the life of an as-yet-unidentified pensioner in Salt Lake County, Utah. Although Zika has been around since the 1940s, it is only during the last few years that it has really exploded, and its spread across Americas has been a tremendous cause for concern, particularly with the Rio Olympics coming up.
As with all contagions, one of the most pressing challenges for its containment is understanding where it will spread. Obviously, it is not enough simply to deal with a disease once it has infected an area. Infectious disease physician at Toronto-based St. Michael’s Hospital, Kamran Khan notes that one thing is true of the spread of infectious diseases: ‘If you start to analyze the situation when an outbreak occurs, you’re already too late.’
This is particularly true of Zika, as there is still so little known about the disease. The disease is often symptomless, with just 1 in 4 of those with the disease developing them. The most worrying aspect of the virus is the birth defects it causes, such as abnormally small heads and brain damage. From what we know about the disease so far, it is transmitted by the Aedes aegypti and Aedes albopictus mosquito, neither of which are found in Utah. The majority of cases in America have been travel related, which means finding a pattern to its spread is exceptionally difficult. The only treatment available at the moment is also ‘mosquito management’ - an indiscriminate, costly, and wasteful program of insecticide spraying in areas with a large population of the mosquitos in question, the environmental impact of which is hard to ascertain.
Data science is now one of the primary tools health experts have at their disposal for attempting to control an outbreak. Big data has a checkered history when it comes to spotting disease trends. Google’s Flu tracker, for example, was a spectacular failure that is often held up as a warning of the hubris of data practitioners. On the other hand, an algorithm from Nashville-based health analytics firm WPC Healthcare was able to predict the spread of another virus spread via mosquitos, the West Nile virus, with 85% accuracy.
The data necessary to track a virus like Zika comes from a variety of sources, including clinical trials and flight patterns. University of Miami researchers, for example, have previously studied risk maps to combat mosquito-borne illnesses like malaria and dengue fever. They want to apply the same tools to Zika, and researchers from the university are using data from several local databases to develop maps that county officials could use to forecast where the disease may strike in future. They have already managed to garner a number of insights that could prove useful, including that affluent neighborhoods are more likely to have Zika mosquitoes, though they have yet to find out why.
Obviously, when you are dealing with a disease like Zika in which it appears that travel is playing a central role in its spread, one of the first things that requires analysis is data streams like flight itineraries. By blending this information with clinical information, you can better understand the points of origin. Organizations are learning a lot from their recent experience dealing with the ebola virus, which utilized big data in a number of ways to quell its spread. The Centre for Disease Control (CDC), for example, applied real-time mapping software and telecommunications masts to track the disease across Western Africa, sending resources to anywhere that it could see the threat of infection raised as quickly as possible.
It is through this sharing of data that organizations and governments will be able to best understand the spread of diseases, and organizations need to work together to ensure this happens. Big data is not a cure in and of itself, but if it allows companies to respond in a timely fashion, containment is a far easier proposition.