Public Health and Big Data

Oct 29, 2012, 3:37 PM

"Information is the oil of the 21st century, and analytics is the combustion engine.” - Peter Sondergaard

That was a quote from panel at APHA 2012 on how new data mapping tools can help support health departments and vulnerable population. Taking a look at everything from local health department budget cuts and their effect on vulnerable populations to how “big data” has implications for public health trends, the panel made an effective argument for the fact that mining data sets has increasingly important ramifications for the public health field and practitioners.

“We’re gonna get technical,” warned Matthew Dollacker of CSC, during his presentation. “But we can come away with a real appreciation for a fundamental shift that is taking place through what big data technology can do.”

Dollacker pointed out that today, we’re in a new era of data. Users posts about 2,200 tweets per second to Twitter, and conduct 1.6 billion searches today. Originally, decoding the human genomte took 10 years to process – today, it can be achieved in one week.

“Data production will be 44 times greater in 2020 than it was in 2009,” said Dollacker. “And those who are positioned to take advantage of this data explosion are those that are aware of it, can access it, and can handle it.”

Dollacker noted that plenty of other fields have long taken advantage of massive sets of data, applying algorithms to daily issues. “Why aren’t we applying the Amazon algorithms that recommend books to us to health care?” Dollacker asked.

Dollacker highlighted a couple of recent examples that show the potential big data has for public health. One instance was the Google Flu project – merely based on what users were searching on, the search engine was able to take that data and help predict flu outbreaks at a regional and city level – ahead of official sources.

“What this highlights,” said Dollacker, “is the tremendous value that is sitting out there in data right now. Who would have thought we could have address this issue by looking at nothing more than existing search terms?”

The other case was one of two MIT physicians who were convinced that there were patterns in EKG data they were not yet seeing that could predict death and future risk. Taking large sets of discarded EKG data, they were able to find three abnormalities in the data that correlated with a two to three times higher risk of dying from a second heart attack within the year.

“What can public health do with these technologies? What kind of problems can we solve?” Dollacker asked. From the looks of it, we’re only starting to find out.

This commentary originally appeared on the RWJF New Public Health blog.