Essentials in Population Health—a four-part webinar series for children's hospital leaders and teams—recently hosted the third installment, "Role of Data Analytics in Pediatric Population Health." Here's a glimpse into that session—and the move toward this emerging approach to health care.
It was just a small line on a graph, but it told a big story: smoking among teens in Philadelphia has dropped by more than 50 percent over the last 15 years.
Harm Scherpbier, M.D., M.S., a lecturer of Health Informatics and Population Health Analytics with the Jefferson College of Population Health and Lead Strategist of Banter Health IT, presented this graph during the CHA webinar "Role of Data and Analytics in Pediatric Population Health." The drastic decline in teen smoking in Philadelphia is in line with similar results nationwide, which is a victory from a population health standpoint.
But how do we even get this data point on teen smoking? How do we get a clear picture of the social determinants that drive population health? From big data.
Making sense of the data
The widespread adoption of electronic medical records (EMRs) has increased the availability of patient data for healthcare organizations. That data drives all aspects of the population health movement, from identifying social determinants of health to measuring healthy outcomes to the calculation of value-based payments.
But how do children's hospitals go about harnessing these vast amounts of data already available to help guide their decisions? Scherpbier suggests following the advice of Thomas Davenport, a big data expert.
A pioneer in data analytics, Davenport has identified a detailed six-step process around the use of data analytics as a basis for decision-making. But Scherpbier breaks it down even further, with a focus on three key priorities:
- The problem: First, recognize the problem you need to solve or the question you need to answer, and frame up the problem so you can identify the right path to a solution with the data you have available
- The players: Although the number crunchers will handle many of the details around gathering and analyzing the data, remain engaged with them and ask questions along the way. As the subject matter expert, your involvement is key to ensuring accurate data
- The presentation: Once you have all the pertinent data, make a clear presentation for decision makers and stakeholders with a recommendation for appropriate action and next steps
It's all in the presentation
No matter how compelling the data may be, its significance is often lost if not properly presented. Scherpbier recommends tapping into expert resources to enhance your presentations. A good starting point, he says, can be found in the books written by data visualization guru, Edward Tufte.
Finding the right solution
Not surprisingly, as the need for hospitals to invest IT resources into population health management systems has increased, so too has the selection of vendors in the marketplace. To find the right fit for your organization, it's essential to identify your requirements upfront and clearly communicate your needs to vendors, so you can select the right partner. Involve stakeholders across your organization, including physicians, in the process. The key, Scherpbier says, is to start small and build from there because change happens fast.
"Don't try to do everything. Today you cannot pick an approach that will be your end-all be-all approach," Scherpbier says. "Instead, go with one solution, and work with it for three to five years, so at least you learn how to build analytics and use them for value-based care. And as you evolve, the market will evolve, and that will lead to your next decision."
To learn more, listen to the replay of the "Role of Data Analytics in Pediatric Population Health" webinar or view the whitepaper.
These webinars are part of the Essentials in Population Health, an educational series through CHA that explores changes under health care reform and how to operationalize a population health strategy.
Send questions or comments to email@example.com