Three experts on analytics and health IT gathered at the Becker's Hospital Review 4th Annual Health IT + Revenue Cycle Conference in Chicago Sept. 20 to discuss how hospital leaders can develop robust data programs that see a return on investment.
Here are three pieces of advice they shared for hospitals looking to build successful data analytics teams:
1. Build the program over time, says Deborah Chasco, DNP, director of nursing informatics at University Medical Center of El Paso (Texas):
"[Building a data analytics team] has been a challenge for us, but I'm very proud to say that within the four years we have had six data analytics associates working on part of that team," she said, highlighting that University Medical Center of El Paso is a 350-bed safety-net hospital. "We're now expanding into a department. I kind of hate saying 'department' because it almost centralizes the importance that data analytics has into one department, when it shouldn't be that — It should be the entire organization as a whole," she added. "But it's a good start."
2. Bring together diverse perspectives, says Jason Joseph, CIO of Spectrum Health (Grand Rapids, Mich.):
"One of the things we're heavily debating on an ongoing basis is whether data analytics is one team or multiple teams. Right now we have multiple teams, because you really can't ask a group that has its expertise in population health about market share … so we have pockets of analytical teams, but they're all coordinated through a center of excellence," Mr. Joseph said. He added the team includes employees with diverse backgrounds — some in public health, statistics, computer science, business or even clinical areas.
"The way we typically see it structured is those heavily scientific backgrounds are the core of the team, and they might do the deeper technical work," he said. "But they're going to be surrounded by people who really understand the subject matter."
3. Realize the value data scientists bring, says Christopher Caggiano, MD, clinical principal at Allscripts:
"In a mid-size hospital, you might have two or three people [on a data analytics team], and two of those are going to be the people trying to create the data warehouse and maintain the data dictionary," he said. "If you're lucky, one of those people will also be a data scientist. When you're shifting to becoming a data-driven organization, you also need someone to talk about how the behavior of people and relates to data, and … you need a specialist to set up data governance [for this purpose]. I've found the best type of person [for this role] is a behavioral data scientist."