Population Health Management

Henry Ford Health System Builds a Population Health Management Platform

May 11, 2017 8:23 am

Streamlined reporting and standardization of clinical workflows are important components of the health system’s population health strategy.

Over the past 24 months, Detroit-based Henry Ford Health System has been actively preparing for value-based contracting, says Katherine Scher, director of clinical integration and population health management, Henry Ford Physician Network. In particular, leaders at the health system, which includes the country’s third-largest group practice, are gearing up to support specific patient populations and improve clinical and efficiency outcomes.

Henry Ford’s population health management team has established tight connections to the health system’s four hospitals, as well as its post-acute care facilities, home care services, and other assets. “What makes Henry Ford especially effective is its ability to collaborate to improve the patient experience of care,” Scher says.

The fact that the entire health system shares the same electronic health record (EHR) promotes the kind of cooperation needed for effective population health management. Specifically, it allows for more streamlined reporting and standardization of clinical workflows, says Matthew Hussmann, system director of population and practice management analytics.

Developing a Platform to Manage Populations

A strong analytics foundation drives population health management at Henry Ford Health System. Using population analytics—which includes data integration, patient stratification, analysis, and reporting—clinicians target the specific needs of various patient populations.

Until 2015, the health system’s analytics group was decentralized. “Each group that needed an analyst would hire one,” Hussmann says. But then the system decided to create a centralized group of analysts who would be “loaned out” to key areas of the health system. To support this new structure, leaders at Henry Ford created four analytics departments — population health, clinical, revenue cycle, and strategy—under one leader. Today, Hussmann and several other analysts are embedded in the population health department but report to the head of analytics.

This new structure has given the analytics team more latitude to work on major projects, such as building a data warehouse, integrating claims data, and pulling in data from state health information exchanges (HIEs). “Those are major initiatives that take many hours to accomplish and that we couldn’t do if we just had a few analysts as part of the population health team,” Hussmann says. “We need to work through our centralized group to achieve that bigger vision.”

The health system also centralized its performance improvement and project management teams. “Now, the teams have standardized tools to work with so they can make change happen at the local level,” Scher says. “Before, everyone was building clinical programs in a vacuum, and that’s not happening anymore.”

Improving collaboration has been a primary objective, and today, leaders on the clinical, analytics, and performance improvement/project management teams work together on key projects and share the same performance goals, including those that reflect clinical outcomes, not just process goals, Scher says.

Preparing Data for Population Health

For Hussmann, one of the most challenging projects was creating a patient and provider matching process. Now, the health system can seamlessly integrate patient claims data sets into its data warehouse and couple that with clinical data from the EHR—rather than simply “park” the claims data there, as many organizations do. “Being able to match those patients has set Henry Ford Health System apart,” Hussmann says. Analysts have created a “fuzzy logic” algorithm that combs through external data and matches medical record numbers for all patients so analysts can send the patient data back to physicians. “What this means is that we can put patients at the center and look at their clinical information in the EHR and also see what information we can gain from their claims data set,” he says.

With the matching process behind them, the analytics have been able to tackle other major projects, such as batch-loading data from the state HIE into their warehouse. “We are able to take any data and completely embed it with any of our data sets, so data are not just sitting in different silos,” Hussmann says.

Engaging Physicians

To empower physicians with more actionable data, leaders at Henry Ford created a physician scorecard that reflects the Triple Aim. “We tried to build the scorecard to be extremely flexible so physicians could view data on different populations and also view data appropriate for the executive, physician-in-charge, and the primary care physician,” Hussmann says.

These scorecards are examples of tools that analysts can build to help physicians integrate the work of population health into their daily routines. “Part of our job is to make sure physicians don’t see population health as extra work,” Hussmann says.

See related web extra: Henry Ford Health System Sample Physician Dashboard

Hussmann makes it a priority to ask physicians to vet scorecard metrics when analysts identify areas of opportunity. “You get a lot more buy-in when physicians have helped build it,” Hussmann says.

Lessons Learned

Leaders at Henry Ford offer the following advice to organizations that want to leverage analytics to improve population health management.

Develop an engaged workforce. “Everyone should understand the goal of population health management and what you are trying to accomplish, including improving the patient experience,” Scher says.

Facilitate collaboration. Every quarter, Henry Ford’s analytics, clinical, and performance improvement/project management teams—which comprise more than 100 people in population health—come together to work on clinical initiatives. Besides targeting observation stays in the emergency department (ED), recent interventions have centered on diabetes management, case management, and post-acute surveillance.

Pursue only a few clinical interventions at a time. Henry Ford Hospital’s project to avoid observation stays leverages navigators in the ED to help support patients at risk for observation. This might include having navigators set up urgent ambulatory appointments or arrange home health referrals. The program has been so successful— the ED intervention achieved a 7 percent reduction in observation stays—that it is now spreading to other system hospitals.

Spend time cultivating buy in. When developing the ED program, leaders in population health management discussed the observation day reduction goal with the ED chair. “Through that partnership, we were able to create the initiative. But had we done it ourselves, we probably would have come up with something that wasn’t as successful,” Scher says. “So the big takeaway for us was that we cannot do this without our physicians.”

Take the time to work with your data. EHR data tends to be “vertical,” illustrating just what happened in a particular service line. But population health is “horizontal,” following the continuum of care, Hussmann says. Getting vertical data to follow a more horizontal stream that follows patients along their care journey can be challenging because it requires integrating data from multiple sources, including various payers. But it is essential to population health management. During this process, leaders should ensure that the raw data they have is “clean,” or accurate, Hussmann says.

Use vendors but recognize their limitations. “If you are a complex organization like Henry Ford, no single analytics vendor can fully slice and dice the data to meet all of your needs,” Hussmann says.

Promoting Population Health

“For 2017, we’re focused on making sure we have the appropriate staffing, the right interventions for the staff to carry out, as well as the reports that staff need to understand how they are doing,” Scher says.

Scher and Hussmann also plan to educate physicians, the finance team, and the C-suite on what population health means to the organization. Analysts have already been exposed to this tpe of education.

“We want the analysts to understand the big picture,” Hussmann says. In particular, many analysts may find it difficult to appreciate the financial ramifications of reducing inpatient volume under value-based contracts. Some of the analysts also come from a quality reporting background, which tends to emphasize the “end” of initiatives, rather than the beginning of initiatives. For example, in their case management intervention, analysts at Henry Ford might supply case managers with a list of patients who should be targeted at the start of an intervention. This proactive application of analytics is new for some analysts. They are learning that effective population health management analytics at the front and back ends of initiatives are equally important, Hussmann says.

Despite the health system’s early progress, Scher is quick to point out that the clinical, analytics, and performance improvement/project management teams are still building their core competencies in population health. “I don’t think any organization has fully figured it out yet, but we have a focus and an engaged group of people to work on these efforts,” Scher says.

This article is based in part on a presentation at the September 2016 Accountable Care & HIT Strategies Summit in Chicago.

Interviewed for this article:

Katherine Scher, RN, CCM, is director of clinical integration and population health management at Henry Ford Physician Network, Detroit.

Matthew Hussmann is director of population and practice management analytics at Henry Ford Health System, Detroit.


googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text1' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text2' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text3' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text4' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text5' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text6' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text7' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-leaderboard' ); } );