It takes a lot more than analytics to make population health work

UMass Memorial exec says getting providers on board is a good first step, but workflow adjustments, a data-driven culture and coordination with community resources will deliver results.
By Mike Miliard
02:40 PM

BOSTON – How can analytics help improve population health? Good data and smart use of information technology are important, said Tom Scornavacca, senior medical director, population health at UMass Memorial Health Care. But enthusiastic buy-in from care providers is essential.

"No matter how good the data is, if you have providers who are tuned out disinterested or misaligned, it's hard to get the Titanic to turn," said Scornavacca, speaking Monday at the HIMSS Big Data and Healthcare Analytics Forum.

UMass Memorial has notched some important wins in recent years, leveraging analytics tools to identify and intervene with the patients who need it most. From diabetes management to immunizations and beyond, it's been able to spot gaps and in care, find innovative ways to address them and, crucially, engage clinicians and care teams to intervene with positive steps for better health.

[Also: Change management is IT's greatest challenge in delivering needed innovation]

For instance, as part of the American Medical Group Associations' Analytics for Improvement collaborative, UMass was able to increase vaccination rates across all age groups over the course of a year. It required an all-hands-on-deck approach that entailed a lot more than a snazzy data dashboard.

The health system incorporated immunizations interventions into its existing pop health management and quality improvement efforts, using care registries to identify evidence-based gaps for primary care panel; employed outreach coordinators to help schedule patients for care when it was required, and used performance improvement facilitators to work alongside physicians and practice staff on workflow redesign.

On the technology side, UMass made use of clinical decision support at the point of care and deployed performance reporting tools and customized analytics that integrated clinical and claims data.

"Providers want actionable data," said Scornavacca. "The more predigested you can get the data, the better."

It's important to show, not just tell. As part of its larger population health management efforts, UMass creates action plans for discrete periods of time and then uses data to show its providers how certain quality measures looked before and after that window.

It also developed a patient risk matrix, a quadrant that groups populations according to the urgency required for interventions, from healthy (a focus on wellness and prevention) to rising risk (may need more complex risk management) to chronic disease to urgent illness.

But effective and lasting population health and quality improvement efforts need much more than a series of charts and graphs.

Scornavacca said 33 percent of patients in the "rising risk" corner eventually end up in the "urgent" top right, requiring complex and resource-intensive intervention.

"You have to figure out who will bubble up to the top and that's where we devote our time and energy," he said.

Pop health is hard work: "You can't just rip the Band-Aid off and say, 'Take care of readmissions,’" said Scornavacca. It requires intensive interventions from care teams, "windshield time" on the road to visit the patients most in need, alignment with an array community-based resources, close collaboration in the hospital, smart use of sophisticated analytics tools and, often, disruptive changes to workflow and well-established routines.

Buy-in from all stakeholders comes with huge rewards, of course. But "if you don't have provider engagement, this becomes a very difficult task," he said.

Twitter: @MikeMiliardHITN
Email the writer:

 Read our coverage of HIMSS Big Data & Healthcare Analytics Forum in Boston.
⇒ Change management is IT's greatest challenge in delivering needed innovation
⇒ Healthcare is swimming in data, but what to do with it?
⇒ Go figure: The key to big data is actually small data

Want to get more stories like this one? Get daily news updates from Healthcare IT News.
Your subscription has been saved.
Something went wrong. Please try again.