HIMSS15: Data-informed post-acute care choices save Advocate Health Care $200M

The conventional approach to achieve the Triple Aim is to identify and allocate resources to patients who are most at-risk for readmissions, complications or adverse events--that small group of patients who account for a disproportionate share of resources and cost.

But what if the conventional wisdom is wrong?

Execs from Advocate Health Care outlined a different approach to data analytics and risk stratification in a session at the annual meeting of the Health Information Management and Systems Society (HIMSS) in Chicago this week.

Advocate, based in Chicago, has 13 hospitals and two medical groups and, as the largest Accountable Care Organization in the U.S., prides itself on clinical integration.

One secret to its success: Data-enabled resource allocation and using data to risk stratify the entire population. That's a departure from the typical focus on using data to identify the patients who are most expensive to treat.

"It's a good place to start, but it's not where you want to go," said Rishi Sikka, M.D., senior VP of clinical transformation at Advocate Health Care. "We talk too much about cost--we need to turn the conversation back to care."

A focus on finance lumps patients into big groups that aren't really all alike, he said. Some patients have modifiable risk factors, while others don't have those options.

Meanwhile, of those patients who are in the high-cost bucket one year, a significant number would be in the lower-cost group the following year regardless of interventions. "They naturally regress to the mean," he explained. "When you just focus on the tail, you'll never move the mean or the average."

Advocate has implemented a new method of risk stratification--what Sikka called a "targeted big data approach." They look at many data points from sources including medical history, claims, demographics and patients' own descriptions of their health.

Tina Esposito, vice president of Advocate's Center for Health Information Services, offered two case studies to illustrate the approach.

The first focused on acute transitions of care and post-acute care infrastructure resources.

Early on, Advocate used data to identify those patients at risk for readmissions. But clinicians suggested they instead focus on post-discharge care and use data to identify the most appropriate location for patients transitioning out of acute care.

For example, the data showed that the risk of medical instability at long-term facilities was high and that Advocate was over-using them. Meanwhile, home care and home health services had lower risks and Advocate was under-using that resource.

By making data-informed shifts in post-acute care, Advocate saved an estimated $200 million in cost of care.

The biggest opportunity was in skilled nursing care, the data showed. Advocate owns two skilled nursing facilities (SNF) and partnered with about 36 others to maximize that opportunity.

"We wanted to see ... if that program had a difference in cost and length of stay," Esposito said. Sure enough, on average, patients saw about a $2,000 decrease in SNF costs and a five day decrease in length-of-stay.

The second case study focused on human capital and outpatient care management. Allocating care managers to the "top of the pyramid is over-simplifying," Esposito said. Instead, the organization targeted interventions to patients who were "most impactable" based on the data.

"I like to think that we're going to start a trend here," Sikka said. "The big idea to take away … is impactability … It doesn't matter who's high-risk. It's who's high-risk that you can do something about."

The organization focuses on clinician-identified preventable events most appropriate for care management, events where intervention can reduce hospital encounters within a 120 day time period and impactable in a measurable way with defined outcomes.

All those parameters must be there for the biggest impact on care, Esposito said. 

Additionally, she said, an effective outpatient care management program is short term, focused on potentially preventable events and evidence-based. It should measure not only outcomes but also process and the impact on patients.

"Our thinking is evolving," Sikka said. "Care management is the backbone to what you do in population health [but] those patients need something different than care management."

To learn more:
- here's the session handout

Related Articles: 
Blue Shield-Advocate ACO lowers cost trends
Hospitals must consolidate, merge to manage population health
Chicago healthcare systems earn nearly as much from investments as operations

Read more on