Going beyond traditional data sources, Steward looks at psychosocial factors as to why patients may or may not be doing the right thing to manage their health.
While it became public last week that Massachusetts's Steward Health Care was officially dropping out of the Centers for Medicare and Medicaid Services' Pioneer accountable care organization (ACO) program, Steward's former chief operating officer (COO) gave some insight into the reason behind it at the HIMSS Connected Health Conference.
On Nov. 10, Dominique Morgan-Solomon opened her presentation at the Gaylord National Resort in National Harbor, Md. (the HIMSS Connected Conference combines the mHealth Summit from previous years with the new Cybersecurity and Population Health Summits, all under the HIMSS umbrella) by telling attendees that as of very recently she is no longer COO at Steward since she is planning to dedicate more time to her family. She then highlighted the early successes that the ACO had in the Pioneer program, which began in 2012.
Specifically, Morgan-Solomon pointed to Steward being one of of 13 of 32 Pioneer ACOs that generated enough savings to earn a gross surplus in Year 1 of the program. In the second year, she added, Steward finished as the No. 2 Pioneer ACO in the country and first in the state—an impressive feat given the highly competitive ACO market in Boston. Morgan-Solomon also confirmed, as reported last week by the Boston Globe, that the Steward ACO would be looking to join a similar federal program, CMS’s Next Generation ACO Model, set to begin on January 1, 2016.
ACOs in the Next Generation ACO Model will take on greater financial risk than those in current Medicare ACO initiatives, while also potentially sharing in a greater portion of savings. "We are actively looking into the Next Generation model," Morgan-Solomon said during her presentation. "It's not the concept of value-based payment that [made us leave] the Pioneer program. About 60 percent of our business is risk-based. But there are nuances in the Pioneer model that don't exist in the Next Generation model when it comes to downside risk," she said.
Morgan-Solomon's presentation at the HIMSS Connected Health Conferenced went extended far past Steward's ACO participation, however. She explained how Steward is experimenting with patient engagement tools that go beyond just analyzing claims and clinical data. For effective population health management, Morgan-Solomon said it's about putting data and analytics to work to change patient behavior. "Data is directional and informative, but it doesn't get into the core about what population health is about," she said. "In this population health world, data can point out how previous behaviors have not been what you wanted them to be, but the data alone won't change behavior. Data does not equal patient engagement. You have to take it a step further to know how and why to engage with the patients. You can make assumptions based on data that says you can enroll a person in a certain program, but unless you know more about that patient, you won't change the behavior and get the intended outcome," Morgan-Solomon said.
Looking back at her days working at Kaiser Permanente, Morgan-Solomon recalled analyzing the three million adult members in Kaiser's patient population over a three-year span. While it's become well-known that 5 percent of the U.S. population accounts for half of total healthcare costs, Morgan-Solomon said this is a statistic that is overplayed in healthcare circles. "We found out that by the time you identified these 5 percent of patients, they have already spent the money. And by the time you have engaged them, it's too late," she said. "Instead, the population that you need to focus on is the patients in the next percentile (6-20 percent) of healthcare spending. This is the rising population that will move into the next tier if you don't do anything. This is where you want to expend the energy," Morgan-Solomon said. "These are people with chronic conditions and need their behavior changed today so you can help them and save money for tomorrow," she said.
How can behavior be changed? Steward is applying the Patient Activation Measures (PAM) model to its care management process, making PAM part of its discovery with patients enrolled in care management. The PAM model, from Portland-based Insignia Health and developed nearby at the University of Oregon, has three main components that patients are assessed on—knowledge, skill, and confidence, Morgan-Solomon explained. There are also four levels of patient activation in this model: Level 1 is disengaged and overwhelmed; Level 2 is becoming aware but still struggling; Level 3 is taking action; and Level 4 is maintaining behaviors and pushing further. PAM activation levels are mapped to hundreds of consumer health characteristics—motivators, attitudes, behaviors and outcomes.
Morgan-Solomon then gave two case study examples of how the PAM approach worked with patients at Steward. The first case was of a 44-year-old diabetic patient who was a former smoker, asthmatic, and had three different inhalers. He refused to participate in Steward's care management program twice in one year—and according to Morgan-Solomon, Steward's health coaches "will go to the end of the earth to get you involved." The clinical makeup of this patient looked like someone in his 60s, despite him being just 44. When asked about his goals, the patient said he wanted to run a half marathon, but due to his condition, felt discouraged, leading to weight gain and a sharp decline in activity levels.