How population health gaps can be filled using analytics

Making headway on population health initiatives is an integral part of care delivery in the switch to value-based care models, but it requires real-time, actionable data. The lack of it can cause frustration when it comes to gauging the effectiveness of various pop health efforts.

For example, data that was available to the University of Texas Medical Branch, a safety-net organization with 125,000 Medicaid patients, was often typically delayed by at least six weeks, which naturally slowed the improvement process. The data also didn't convey the scope of potential opportunities, which hindered chances for improvement on Delivery System Reform Incentive Payment (DSRIP), a value-based care reimbursement model for Medicaid and low-income uninsured patients.

But with the help of data analytics, UTMB Health was able to focus on improvement efforts for these populations and boost reimbursement based on DSRIP performance.

Craig Kovacevich, associate vice president of community and population health at UTMB at Galveston, said in a HIMSS20 digital presentation that data has allowed the organization to look at patient health in a more holistic manner, and spurred it to do things it hadn't done previously, such as pre-visit planning and following up with certain patients for specific, targeted conditions. This ability to clearly see its patient population allowed UTMB to drill down and focus on specific population health initiatives.

Analytics on its own, however, wouldn't have done the trick; the key was pairing analytics with dynamic reporting, which allows people on the care teams to study the data and determine how they could do a better job at treating their patients.

"There's a lot of data," said Kovacevich. "We can all get our hands on reports and statistics, but having data is not good on its own. You have to take that data and understand what it means and how to use it."

Trend modeling was also very useful in UTMB's case, since it serves both rural and urban areas. With a dynamic patient mix, patient X may be in UTMB's patient panel, but receives a certain care service somewhere outside of the network. Rarely would that patient have the wherewithal to notify care teams that their chart should be updated. UTMB needed to have a way to pull that data in to create a complete picture of the patient.

"At-risk contracting and pay-for-performance arrangements are providing enhanced financial incentives for delivering preventative services, and tracking patients across the care continuum," said Kovacevich. That's where data becomes key for managing patients, especially in light of UTBM's participation in Texas' DSRIP, which adds more than 32 measures in five areas of care, worth more than $57 million in incentive dollars per year.

"We are paid for performance through metrics, and that is where the data becomes very important," he said.

UTMB's DSRIP measure portfolio includes adult diabetes, adult cancer screening, adult heart disease, adult prevention and pediatric primary care. The organization honed in on those measures and determined who should review the metrics, what data is needed and how the data could be relayed to the right people.

To facilitate this, the organization established UTMB Health Population Metrics, a proprietary Discover application that takes scores of data and translates it into something that's measurable in real time. It generates on-demand reporting that allows staff to dive down into the metrics they're trying to meet. Staff can view the performance of a specific clinic in UTMB's system, see how a specific provider is doing and track how patient X is faring health-wise over a period of time.

Through the application, staff can select a specific measure, apply filters and export lists of potential actions they can take.

"Dynamic reporting is giving us data on demand," said Kovacevich. "That has always been there, but it wasn't there in a way that we could best use it to navigate the health of patients. Now we have that data at hand. We can get proactive and reactive with it, planning for long-term and short-term solutions to a topic like diabetes care. We can drill down to patients' long- and short-term goals, and communicate that to the care team to deliver the best care."

Wellness teams now perform pre-visit planning, so if patient X comes in for one thing but he or she also needs attention in other areas, the care team can address those other issues in the same visit, augmenting the impact each visit can have. The team puts their notes into the electronic health record – Epic, in this case – to facilitate a conversation between Patient X and the nurse. Enhanced data is included in the chart before the patient even walks into the office.

In addition to enabling a boost in DSRIP-related reimbursement, UTMB Health also earned $2.1 million in pay-for-performance dollars, achieved after the analytics application was implemented, and demonstrated improvement on 23 of 32 performance measures – nearly 72%.

In 2018, the organization realized 85.4% of its performance dollars. In 2019, that number grew to 91.7%.

Kovacevich said his key recommendations to organizations looking to implement a similar approach include knowing who the stakeholders are, knowing the unique analytics reporting needs that come into play, measure often, and implement real-time analytics for value-based care.

"The work we're doing could be replicated in any setting," he said. "It's something that will grow and continue to be part of the culture at our organization for the foreseeable future."
 

Twitter: @JELagasse

Email the writer: jeff.lagasse@himssmedia.com