Population Health Leading the Way to Better Health Outcomes

Harnessing data for better health outcomes is the next big push for health IT. Population health and analytics are critical components of accountable care and healthcare reform with the potential to truly improve quality of care while helping providers fend off federal penalties for underperformance.

With chronic disease, physicians are just beginning to scratch the surface in seeing positive health outcomes through population health management. “Our healthcare system, the best in the world, has never been well-organized to manage chronic disease,” says Mark Braunstein, MD, associate director of the Health Systems Institute at the Georgia Institute of Technology in Atlanta.

Better population health management and care coordination hold the keys to preventing and managing chronic disease, Braunstein says, citing findings from the Physician Group Practice Demonstration published in September 2012 in the Journal of the American Medical Association. This study, involving 10 physician groups representing 5,000 physicians and 220,000 Medicare fee-for-service beneficiaries that received bonus payments if they met quality targets and reduced spending on care, had helped nudge CMS to move Medicare toward the ACO model.

In the study, two relatively low-spending systems, Marshfield Clinic in Wisconsin and Park Nicollet Clinic in Minnesota, achieved particularly large savings. “In essence, it says that they used health IT skillfully and engaged in population health management. They were the sites that really saved money and earned bonuses,” Braunstein says of the findings.   

Heart Disease: Management at Duke

As the Centers for Disease Control and Prevention (CDC) report that chronic disease consumes 75 percent of U.S. healthcare costs, many providers look to population health management to rein in costs and improve quality of care.

Zubin J. Eapen, MD, medical director of the Duke Heart Failure Same Day Access Clinic, is part of that movement. Eapen and his team use analytics to help manage heart failure patients from the outpatient side at his clinic in Raleigh-Durham, N.C.

“A lot of hospitals may capture, but not know how to structure, their local data to understand whether they are meeting performance measures and quality metrics,” he says. To help get a handle on these issues, Duke joined up with the American Heart Association’s Get with the Guidelines quality improvement program, which has several registries across different disease states, including coronary disease, heart disease and stroke.

In Eapen’s case, he is part of a specific initiative, called Target Heart Failure, that helps hospitals optimize care improvements to help reduce readmission rates. The registry has more than 500,000 patient records that span several years, and captures processes of care such as appropriate discharge medications, whether education is provided and follow-up care plans.

“At a local, single center level, you can benchmark yourself against best practices and other hospitals and similar institutions to see where you are in terms of your readmission rate and the quality of your care in terms of the processes of your care,” he says. “Having had analytics around readmission rates, particularly my local patient population, and whether they’ve been receiving evidence-based therapies, has been helpful.”

Consequently, the initiative helps physicians identify patients eligible for therapies and how many are receiving them, he says. Longitudinal data are helping Eapen and his team at Duke understand what the clinical encounter looks like, including visits before and after to the access clinic or the emergency department (ED).

Because Medicare has started penalizing hospitals for excessive readmission rates, that issue has become a core clinical research focus, he says.  
“It’s affecting hospitals’ bottom lines, so we need to figure out how to benchmark and use our local data to reduce readmissions or provide some other metric to the quality of care,” Eapen says.

Focus on process outcomes

As providers get on board with utilizing registries, attention needs to be paid to what goes on in the practice setting, according to Frances Sagona, the current consumer engagement program manager at the Louisiana Department of Health and Hospitals in Baton Rouge, who previously served as director of clinic operations at Baton Rouge General Medical Center.

“We did a survey using an electronic record and identified unmanaged blood pressures,” Sagona says. To dig into the problem, she looked at how and when blood pressures were captured when patients entered the ED.  

“We found out that the nurses rushed them in and took the blood pressure, and then the doctors retook the blood pressure but didn’t document it in the right area,” she says. Because they failed to capture that information, they were unable to manage the population as a whole or be considered a controlled environment. As a result of the survey, physicians were re-educated to ensure proper documentation of blood pressure, Sagona says.

Process measures are important to keep track of, and go hand-in-hand with outcome measures, Braunstein says. For example, every diabetic must receive the hemoglobin 1 blood test once or twice per year to ensure blood sugar levels are in check. These tests need to be documented and providers must proactively reach out to patients and intervene when the patient has not been keeping up with the testing regimen.

“We can no longer pay attention to them when they are in the office but ignore them when they are not,” he says. Better care coordination also prevents duplicate tests, so there are overall better results at a lower cost.

As physicians continue to fine-tune their processes, better outcomes are sure to follow. Successes so far are the tip of the iceberg with what is to come as population health analytics becomes more the norm. “That is the great national experiment happening right now,” Braunstein says.