Big Data holds the potential to revolutionize healthcare — improving care, reducing costs, even alerting us to the threat of epidemics before they occur.
What if we could predict disease? That long sought-after goal — and the major implications it would have for the quality and cost of healthcare — was the genesis of a recent study by MIT researchers.
The analysis of electronic health records of more than 500,000 patients turned up a promising answer: there is predictive power in Big Data. For more than four in 10 patients analyzed, the data could help predict what disease was likely to occur next.
For Carlo Ratti, a co-author of the report who is director of the MIT Senseable City Laboratory, the study shows how Big Data “has the potential to revolutionize the whole sector.”
“Data is a very powerful tool to provide better care at lower cost,” says Ratti. The study is only the beginning of exploring the potential of Big Data, he adds, in the area of research as well as care. The lab, which focuses on how data is changing knowledge about cities, sees promise in examining the intersection between our genes and the environment we live in.
“Think about it as characterizing a city’s micro-biome and potentially see epidemics before they happen.” he says in the interview:
Containing healthcare costs is a long-fought battle. How will the ability to predict disease change this?
Some level of disease prediction — which is possible through the analysis of electronic health records, as we show in this paper — can have a major impact in this sense. It can help practitioners explore medical conditions that are more likely to happen, based on a patient’s history. This can lead to a potentially faster diagnosis — better care — at lower cost. Also, sharing the information about diseases that have a higher likelihood of occurrence with patients can help prevention and early diagnosis — something that again can improve care.
What needs to happen to turn patient data into actionable information that doctors can use in making a diagnosis or prescribing a treatment?
First, I think that we need more research in healthcare data — something that has the potential to revolutionize the whole sector. Second, it is very important to invest in the training of practitioners, so that they can use information in an effective way. Finally, we need to take into account privacy policies – who has access to the data and under which conditions?
Given your lab’s focus on data and cities, how do you view the interplay between our health and where we live?
It’s a really important topic. Diseases develop at the intersection between our genes and behavioral and environmental factors. That’s why monitoring cities — as well as our behavior as we live in them — holds immense promise.
I should like to mention a recent project we started at the MIT Senseable City Lab, called Underworlds. We are sampling wastewater across several cities, and analyzing DNA from viruses, bacteria and humans. We aim to extract a new world of information on human health and behavior. The main benefits lie in the real-time aspect of the technology, providing insight into the diseases circulating in a community even before people themselves are aware of it. Think about it as characterizing a city’s micro-biome and potentially “see epidemics before they happen.”
What do you see as the potential for Big Data to transform policies ranging from healthcare to urban planning and beyond?
Big Data today is getting really big. As estimated by Eric Schmidt, every two days humans create as much information as from the dawn of civilization up until 2003 — an estimate that is already a few year old.
Beyond healthcare, Big Data is impacting on many other dimensions of society. In cities — the focus of most of our work — it can help us better understand the world around ourselves and plan its transformation. Collecting data and responding accordingly has always been a crucial endeavor. Over a century ago, Élisée Reclus stated that good “surveying,” i.e. data collection, is the first fundamental step in city planning. It is not different today — if not for the fact that we know our cities much better and can plan their transformation accordingly.