Population Health IT Gaining Momentum

The shifting landscape of healthcare reimbursement means providers must prepare for a time when they are rewarded for meeting quality objectives for their entire patient population—not just those in a clinical setting.

Payers, accountable care organizations (ACOs), integrated delivery systems, government agencies, IT companies and consumer groups all are putting the pressure on for better tools to manage population health.

Population health technologies are still in their infancy, but the industry always follows the payments, says Grace Terrell, MD, CEO and president of Cornerstone Health Care, a North Carolina-based multispecialty group that became a Medicare Shared Savings Program (MSSP) in 2012. As more providers share risk with payers, these tools will grow in sophistication and use.

A Multi-pronged Approach

As a renowned research center, Johns Hopkins University launched its Center for Population Health IT (CPHIT, pronounced “see-fit”) to meet the demand for technologies that support value-based payment models. The center focuses on the application of EHRs, mobile health and health IT tools targeted at communities and populations.

“Population health is now in vogue,” says Jonathan Weiner, DrPH, director of CPHIT, “but most [providers] have not done population health. They have focused on their patients but not their denominator.

“We’re fairly unique as we view our target as an entire community, including people without access to healthcare and technology,” he adds.

Housed in the Johns Hopkins Bloomberg School of Public Health, CPHIT brings together expertise from the school’s engineering, medicine, nursing, applied physics laboratory units and its health system. Through interdisciplinary research and industry collaboration, CPHIT hopes to build viable population health IT for the greater good. “The nexus of population health, healthcare delivery and health IT really depends on successes at CPHIT and organizations like it.”

Cutting-Edge Research

Working with four large integrated delivery systems, all of which have had EHR systems for more than a decade, CPHIT is developing real-time integrated decision support using predictive modeling analytics. The project’s goal is advancing the state-of-the-art of EHR-based, advanced predictive modeling tools for high-risk case detection and management for populations actively engaged in outpatient care, as well as populations with select chronic conditions.

The initiative expands upon the Johns Hopkins ACG [Adjusted Clinical Group] System, which has been performing risk measurement and case-mix categorization for more than 25 years, and measures accuracy and fairness in evaluating provider performance, identifying patients at high risk, forecasting healthcare utilization and setting equitable payment rates. 

The new project, known as the “e-ACG” project, is a collaboration between Johns Hopkins School of Medicine and the Department of Computer Science. It incorporates EHR elements such as vital signs, lab values, cardiovascular data, clinical notes and patient reports such as health risk assessment and functional status surveys. 

“We are trying to blend clinical data with the claims-based predictive modeling environment. It’s an exciting project in the next couple of years,” Weiner says.

In other research, CPHIT is partnering with CRISP (Chesapeake Regional Information System for our Patients), Maryland’s state health information exchange (HIE), to develop real-time predictive modeling to identify patients at high risk of readmission. The HIE returns the organization’s data, allowing them to develop warnings about who is most likely to be readmitted.

The center also is pursuing computer science methodologies to advance the application of non-structured data and big data to population health interventions. More than half of the information within an EHR is unstructured, Weiner says, so better methods for capturing and sorting data will make it more actionable.

Johns Hopkins engineers, who primarily work on Department of Defense technologies, are transferring their knowledge of natural language processing for use in the school’s health system. The engineers are mining clinical notes to help the obstetrics department, health plans and public health agencies do a better job of identifying high-risk mothers early in their pregnancy. “It’s classic population health, sending out nurses to moms who are at higher risk than others,” Weiner says.

ACO Forerunners

While CPHIT seeks to advance population health IT on a broad scale, Cornerstone Health Care and Baylor Quality Health Alliance, a Texas-based clinically integrated ACO of the Baylor Health Care System, are leading the pack in harnessing population health IT for their own patient populations.

Cornerstone, encompassing more than 380 physicians at 90 sites, shifted to accountable care with the belief that is where the market is headed. “We saw profound changes in the market. We went ahead and aggressively moved to value and population health instead of playing the volume game, knowing it was better for us and better for our patients,” says Terrell.

A gap analysis revealed inadequacies in Cornerstone’s health IT, so the organization developed a five-part strategy for population health management:

• Medical Home Transformation, by improving capabilities across care continuum (i.e., clinical pharmacy services, extended hours, medical home profession, outpatient infusion center, patient care advocates).
• Clinical Integration of primary care practices, hospitals and specialists
• Information Integration:  Integrating disparate platforms enable required reporting and data sharing to support population health management
• Organizational Realignment 
• Reimbursement Model Transformation

To achieve clinical integration, “We needed capabilities, analytics, health information exchange, registry function and the ability to really stratify patients by looking at both clinical and claims data,” she says. Cornerstone thus built tools into the EHR–some of which were vendor agnostic—that live inside the platform. This all resides in Cornerstone Health Enablement Strategic Solutions, a consulting arm of Cornerstone that advises physician groups interested in moving to a value-based model, she says.

Cornerstone also invested in a data warehouse and “best in class” tools from vendors.

Remarking on the slow uptick of available tools, she says, “Clearly, the industry is starting to understand that there needs to be a focus on population health.” However, she adds that these tools predominantly remain siloed.   

From a clinical integration standpoint, Cornerstone not only broke down health IT silos but business structures as well. Cornerstone leaders “started to think along service lines,” and implemented broader clinical integration. For example, it formed a co-management arrangement between oncology and cardiology, among many others. 

In 2012, all of its then-29 primary care practices had obtained National Committee for Quality Alliance (NCQA) Level 3 recognition under the 2008 standards. In February, the NCQA announced that 15 of its practices have been granted Patient-Centered Medical Home recognition for the 2011 criteria, which are more rigorous. Cornerstone is working on obtaining recognition for more sites during the next year.

Since its population health transformation, Cornerstone has seen the following promising results:

• Patient satisfaction scores have significantly improved.
• Heart failure patients are being admitted 69 percent less frequently, according to manually extracted data, and they are going to the emergency room less often.
• Payer data indicates that Cornerstone is meeting high quality performance targets, including hemoglobin A1C testing for patients with diabetes, cholesterol screening and avoidable emergency room visits.
• Among Medicare ACOs, Cornerstone appears to be decreasing higher cost services, including hospital stays, ED visits and ED visits that lead to hospital stays and CT scans.

“Our hospital readmissions continue to decline. Our ER visits continue to decline and our heart failure readmission rate is much lower than average,” Terrell says.

Predictive Analytics

Baylor Quality Alliance is another system focused on populations. Leaders sought to figure out ways to gather population-level data and create an environment where it’s useable, says Michael Sills, MD, CMIO and vice president of informatics technology, speaking at the Health Information and Management Systems Society’s 2014 annual conference.

The Medicare Shared Savings Program  looked at several areas to improve care: utilizing a narrow network of high performing providers, data aggregation and integration of clinical and claims data, and financial modeling.

“Our system has 700 employed doctors and 42 hospitals. It gives us the ability to approach care in an integrated way that hasn’t happened before,” Sills says.

Through analytics, the system identified the 880 chronically ill patients who compromised 60 percent of the costs. “We were really surprised by the number of people who had high HCC [hierarchical condition categories] scores,” he says. Of the 44 patients at highest risk, five hadn’t been seen in three months. “That was startling.”

Analytics also revealed that more than 20 percent of patients had high blood pressure—people not previously identified with hypertension so there is a lot of opportunity for improvement, he notes. Once identified, care coordinators reached out to the patients at highest risk and scheduled appointments with their primary care providers. The care coordinators attended these appointments and showed the algorithms to the primary care providers to better inform them of their patients’ risks.

“We need to take information into the structure that allows us to come up with information that is useable and actionable without threatening anyone. When we do that, we change outcomes and improve access by finding patients who need to be seen the most and creating an environment in which to see them,” says Sills.

The Maryland Experiment

Cornerstone and Baylor combined health IT and population health concepts within their systems’ population to improve care, but imagine doing that for all of Maryland?

Since the 1970s, Maryland has operated the only all-payer rate-setting system for hospital services in the U.S. An independent commission sets a rate structure for each hospital and all payers—public and private—are required to pay for services according to these rates dollar for dollar.

But in January, the commission announced that over the next five years, it will seek to shift hospital revenue away from fee-for-service models into population-based payment models that reward providers for improving health outcomes, enhancing quality and controlling costs.

“It’ll make things happen that couldn’t happen anywhere else,” says Weiner. CPHIT is meeting with the state health secretary to develop a statewide hospital payment network so all hospitals can work together on the whole state’s population health.

“It’s a long way off before we have the tools ready,” he says. “But Rome wasn’t built in a day, and true population health wasn’t built in a day.”