TAP THE POWER & POTENTIAL OF BIG DATA
Joseph Colorafi, MD, is a board-certified pulmonologist. Early in his career, he often worked with sepsis patients, seeking to provide the care that would help them return to health. In his current role as vice president and chief medical information officer at Dignity Health, however, Colorafi doesn’t provide much direct patient care. Curiously, though, he now feels as if he is having a greater impact on patients than ever before.
“In the past, I would see the sepsis patients after they had cascaded downstream and the disease had progressed some. Here I am years later and I am having a greater impact by helping to prevent sepsis from developing in the first place,” Colarafi said.
How so? Colorafi is leading efforts to use big data and analytics to first identify patients who show a high risk for developing sepsis and then get them the treatment and services required to stop the infection in its tracks. More specifically, the San Francisco-based health system is using Hadoop — an opensource software framework that stores and runs applications on large clusters of data — to support a bio surveillance program that focuses on sepsis prevention. With this system in place, the health system is identifying sepsis as soon as possible, making it possible to have a positive impact on survival rates, infection and intensive care unit length of stay. Indeed, when clinicians are alerted to a suspected infection, they are able to take key actions outlined in a “sepsis bundle” within three hours, greatly improving care. In fact, for every 10% increase in compliance with the sepsis bundle, mortality is reduced by 3%.
In addition to relying on data analytics to improve sepsis care, Dignity is leveraging data analytics to prevent harmful events such as over sedation with opioids and to reduce readmissions.
Jumping on the bandwagon
Dignity is just one of many healthcare organizations seeking to move the clinical care needle forward through the expanded use of data and analytics. Centura Health, for example, is leveraging data to support more effective employee wellness initiatives.
“We are using claims data and consumer data to predict and model how we can encourage employees to get needed preventative screenings – mammograms, cervical cancer screenings and colorectal screenings. Basically, we are looking to do what the airlines and other retail organizations do and meet consumers where they are at. Our hope is to drill down and get more personal by using data and analytics to offer services that are relevant to individual employees,” said Angie Villamaria, director of associate wellness.
For example, the Colorado-based health system is analyzing claims, biometric screening and health assessment data to identify employees who have a high body mass index and who would be receptive to participating in a free weight loss coaching program. Centura’s population data is combined with consumer data that includes information on 270 million Americans to create consumer insights and build predictive models to stratify populations based on individuals’ needs, receptivity and likelihood of participating in a targeted program.
“In healthcare, we have traditionally managed people when they are sick. But it is important to shift thinking to the consumer to buy into their health and take some ownership. By providing services that are appealing to them we are looking to bend the curve and improve health outcomes,” Villamaria said.
While many healthcare organizations are leveraging data analytics to improve compliance with current best practices, some professionals are looking to discover more effective treatments. For example, representatives from a variety of healthcare organizations recently came together at a conference convened by the Susan G. Komen breast cancer organization to explore breast cancer research and improve treatment through the use of big data.
“We are working for multidisciplinary collaborations that can help us apply big data to improve breast cancer treatment and outcomes,” George Sledge, MD, Komen chief scientific advisor and a professor of medicine
“Our hope is to drill down and get more personal by using data and analytics to offer services that are relevant to individual employees.”
at Stanford University said in a press release. “We plan to identify opportunities to work together on the very promising, but complicated, use of big data.”
In fact, conference leaders were looking to leverage big data applications in an effort to improve the focus and, hopefully, speed of research, making it possible to more quickly develop treatments that will improve how patients feel, function and survive.
Investing in solutions
With all of this interest in big data and analytics, organizations are increasingly gravitating toward more sophisticated technologies, said David Devine, a managing director with Huron, Chicago. “Big data and analytics typically fall into two categories – present the current data on hand and to predict events. The majority of the technology that exists now is focused on amassing and aggregating data and presenting it in dashboards to show how it measures up against pre-mapped out performance indicators. But we are starting to see the move toward using the data to predict what will happen. So, we are working with more organizations to put the technology in place that will help them predict what will happen with patients,” Devine said.
Indeed, many organizations are looking to turn their electronic health records systems from “systems of record” into “systems of action,” Devine said.
“We are starting to see the move toward using the data to predict what will happen.”