Without analytics, it's impossible to achieve population health goals. Future problems within a health system can't be addressed without looking at data in the present and creating care models based around real-time information.

That's according to Betsy McVay, executive director of strategic analytics at UnityPoint Health, one of the most integrated health systems in the U.S. The organization connects more than 3,500 providers and 33 hospitals in nine regions throughout Iowa, Western Illinois and Southern Wisconsin.

UnityPoint uses Explorys for its predictive analytics technology, which draws upon data from more than 300 hospitals within the network. IBM acquired Explorys in 2015.

>> SPECIAL REPORT CASE STUDY:  Children's Hospital of Pittsburgh develops customized prediction indicator for children with PeriHealth

"Predictive analytics sounds pretty great, but we've tried to look at it as one of the tools in our toolbox," McVay said.

Its biggest results from predictive analytics have come from its nationally-recognized patient blood management tool, or PBM, which was implemented three years ago, she added: This tool helps us better implement evidence-based medicine practices.

PBM was one of UnityPoint's four critical initiatives to improve patient care. Executives sought to reduce known and unknown risks with blood transfusions, and to reduce costs. The tool tracks key data across multiple locations to meet these goals.

"Descriptive analytics compiled all of this data to bring this change together so we could provide better care for our patients," McVay said.

At UnityPoint analytics projects are mostly related to population health management and care management.

Learn more at the Healthcare IT News Big Data and Healthcare Analytics Forum in San Francisco, June 14-15, 2016. Register here.

According to McVay, predictive analytics is crucial to making sure the right patients are targeted with the right data. UnityPoint has improved clinical effectiveness, care quality, patient experience by looking at risk analytics and how to become fully utilized as an organization, she said.

In the beginning, analytics was introduced as a way to better understand our population and how to manage care, she said. We wanted to be able to intervene early and "certainly mature our responsibilities into more descriptive and predictive actions.

"We had to ask, 'What can we change today, so we can improve that future outcome?'" McVay said.

Early on, "many departments were using analytics, but nothing was localized," she said. We recruited key members from technicians to scientists to create the ideal team for predictive modeling. We wanted to create awareness throughout the organization to answer the tough questions.

"It's sort of like evangelizing," she said.

In fact, that was one of UnityPoint's biggest challenges: uniting staff across the care continuum to understand the importance of predictive analytics and how it's approached.

"We needed to connect the business side to the more technical component," McVay said. "We know that sometimes analytics won't provide the answer, but there are discussions that happen to address these needs."

Another challenge was to keep the scope narrow enough, so as to not "boil the ocean and try to be everything to everyone," she said. Goals should be realistic while ensuring adoption is adequately supported at local sites.

"From a day-to-day perspective, in terms of large amounts of data, we look at the strategic uses at the clinical level and ensure we provide value," said McVay. "In this way we're not just moving data around, but rather focusing it where maybe just a report could be needed to drive that change."

The Healthcare IT News Special Report: Bridging the big data-analytics gap
Data-Analytics Gap: How real-time strategies can improve care quality and efficiency
Case Study: St. Joseph Healthcare sees dramatic improvement serving high-risk population with HealthInfoNet
Case Study: Children's Hospital of Pittsburgh develops customized prediction indicator for children with PeriHealth
Next Step: Quality and depth important to develop predictive analytics solutions