Integrating a population health management program using data found in their electronic health record (EHR) and certain registries, Massachusetts General Hospital (MGH) was able to focus on vulnerable populations and significantly improve care quality for patients with diabetes, cardiovascular disease and hypertension over the initial six months of the program. Using centralized population health coordinators (PHC), MGH allowed its physicians the freedom to work more closely with patients to manage chronic diseases and saved money to boot.
The project began in 2011 with a preventive cancer screening program using a health information technology (IT) tool already in place to identify patients who had not had a mammogram within two years, or who were overdue for a pap smear or a colorectal cancer screening. But that program was done independently by each practice in their network.
MGH wanted to compare “two different forms of population health management, either centralized or leaving it to the practices,” said Steven J. Atlas, MD, director of primary care research and quality improvement at MGH, during a presentation at HIMSS17 in Orlando, Florida. “We hypothesized that having a central focus within our network to do this type of work would do a better job than training practices to do it themselves.”
When initiating this population health initiative, MGH left the cancer screening program in place as a control, and then in 2014, across eight practices, brought in PHCs to manage the day-to-day processes and help physicians identify the patients at highest risk. Once those patients were identified, the primary care physicians could focus on making sure those patients received proper preventive measures to address their conditions, whether diabetes, cardiovascular disease or hypertension.
“We were creating a central population [health] program and we didn’t have enough money to roll it out to all of our practices,” Dr. Atlas said. “[Allocation] was based on practical things in our network—interest from practice leaders, baseline quality scores, practice size, the nature and location of practices. Were they community health centers or were they on-campus practices?”
Turning a study into standard of care
The goal was to distribute the available resources to many different practice types “because we wanted to see if we were giving central resources to different types of practices if it would help,” he said. Eight practices were chosen for the trial and 10 practices acted as controls without the PHCs.
In the MGH network, there were approximately 13,000 patients with diabetes, 10,000 with cardiovascular disease and 45,000 with hypertension.
“We did this using validated algorithms, not just using billing claims, but basically all the data that we had in a health care system—lab tests, problems, medications or procedures, depending on the registry,” Dr. Atlas said. “We had a control in that we were focusing this time on rolling this out for chronic disease, but we had already established procedures for doing preventive health for cancer screening, so we continued that at the same time.”
“In effect, we didn’t touch that program that was the same in all our practices,” he said. “It became a control group for us to see if that was behaving differently than what we were doing with the chronic disease population.”
“We worked with network leaders to create what was important for our clinicians,” Dr. Atlas said. “Not just measuring things, but measuring things that were clinically relevant for them.”
The PHCs were nonclinical staff integrated into the practices to huddle with doctors and identify patients at risk. “We would go through [the patients] on a regular basis to take action,” Dr. Atlas said. The PHCs took over some administrative tasks such as appointment scheduling, ordering overdue laboratory tests, chart reviews, obtaining home blood-pressure values and outside laboratory tests to free up physician time with patients.
Results show a healthier patient population
For almost all measures, there was improvement over the six month period, Dr. Atlas said. “Not just for those that had the essential coordinators, but also for those that didn’t. … The key point was that there was … greater improvement across the board for each of these different measures for practices that used the coordinators.”
Looking at hemoglobin A1c tests for patients with diabetes, there was a 6 percent improvement in patients at goal in the intervention group, he said. “We did conclude … that the population health program using the IT system significantly increased quality measures overall for these conditions.”
“The central coordinators helping these practices did a better job than those [practices] who did it themselves,” Dr. Atlas said. The results “support using essential personnel with the practice staff, and that became what is now our standard of care using this data—rolling it out to the entire network.”
The rollout to all practices in the MGH network moved ahead. And health outcomes improved, thanks to physician-led team-based care, patient education, lifestyle and behavioral changes, and continuous care team training and coordination. Improvements include:
- More patients eligible for insulin are now being treated.
- Aided practices improved their diabetes team care.
- Higher engagement of non-MD staff with diabetes care.
- More resources available to primary care practices.
- 14.8 percent increase in cancer screening among underserved patients.
- Patients with A1c greater than 9 percent saw a greater than 1 percentage-point drop in two years.
The network recorded savings attributable to the program of $1.6 million in just the first six months.
The AMA’s and the Centers for Disease Control and Prevention’s (CDC) joint effort, Prevent Diabetes STAT, reflects long-term dedication to prevent or delay the onset of type 2 diabetes through prediabetes awareness and education. The website holds vital information and tools for physicians, health professionals and patients.
Target: BP™ features numerous tools and resources to help practices and patients control hypertension, including a treatment algorithm developed by the American Heart Association, the American College of Cardiology and the CDC.
