Aaron Black calls it “hand-to-hand combat”: the time-consuming, painful battle to convert a Tower of Babel of information into life-saving guidance for doctors and researchers.
Black is chief data officer at the Inova Translational Medical Institute, a nonprofit that gathers genomic information to support precision healthcare. He and his colleagues analyze hundreds of datasets to identify correlations that can help doctors make better diagnostic and treatment decisions. But the chore of reconciling those databases—terabytes of information, often in incompatible formats—can clog the gears. Black says that data-management chores might typically saddle a bioinformatics expert with 40 hours a month of spreadsheet wrangling.
These days, however, data-visualization technology is helping Black’s team find shortcuts to the information they care about. Inova works with Tableau (data), a Seattle-based analytics software firm that specializes in converting “dirty data” into graphics and dashboards that even data science newbies can interpret more quickly.
Read more on this topic in our feature, Tech’s Next Big Wave: Big Data Meets Biology.
Data visualization as a concept, of course, is hardly new (ask anyone who used crayons to color in a pie chart in grade school). But easy-to-understand representation of data has become a more urgent need in the digital era, as the speedy processing of information has become more central to all kinds of professional decision-making. “Companies now want analytics to be used by tens of thousands of their employees, rather than just the specialists,” says Adam Selipsky, a veteran of Amazon’s cloud-services division who became Tableau’s CEO in August 2016.
Tech giants like Microsoft, Google and SAP have all expanded their data-visualization offerings in recent years; among independent players, Tableau is one of the biggest. Launched by two grad students and a professor at Stanford in 2003, the company now claims 70,000 customer accounts, including 90% of the Fortune 500 along with multiple government agencies; it had $877 million in revenue in 2017.
While they’re used in milieus ranging from Major League Baseball to aerospace manufacturing, Tableau’s visualization and pattern-recognition capabilities have had a particularly big impact in the medical and public health communities. In Cincinnati, for example, the city government has used Tableau to build dashboards that help it recognize geographical and timing patterns in heroin-overdose emergency calls—which in turn helps them deploy police and EMTs in targeted ways to enable faster responses. In southern Africa, Tableau software powers disease-surveillance systems that have helped bring about sharp declines in malaria infections.
The software has also won fans within big hospital systems, where doctors and nurses use its visualizations to spot the kinds of problems that can worsen patients’ health problems and drive up costs. Massachusetts General Hospital says it used Tableau analytics to reduce rates of hospital-acquired infections—the hospital system was able to reduce catheter-related urinary tract infections, for example, by 85%. And the prestigious Cleveland Clinic used the company’s tools to identify the patients who were most at risk of potentially avoidable emergency-room admission, a step that enabled them to reach out to those patients proactively and provide them with care less expensively and more safely.
Tableau has been widening access to its software by transitioning to a subscription model, which means a smaller up-front investment for customers. “Instead of marrying their software, they get to date it,” jokes Selipsky. And a recent upgrade to Tableau’s platform, using a technology called Hyper, has sped up by an average of 5x the time it takes for its software to ingest and analyze new data. (The advance involves a process called parallelization; the company explains it here.)
The result is that Tableau users can update their visualizations almost instantaneously to reflect new information. At Inova, that kind of speed has been enabling doctors to quickly diagnose genetic diseases in newborns, and makes it easier for pharmacists to prescribe medications that match a patient’s genetic makeup. “They light up when they see the technology,” Black says of the clinicians in his network. “You’ve changed how effective they can be.”