Mass General’s secret to population health: workflow, workflow, workflow

IT-optimized changes created big improvements in chronic care management, hospital executive says.
By Mike Miliard
10:51 AM
Mass General population health

(L-R) Gregory Weidner, MD; Shafiq Rab, MD; Adrian Zai, MD; and Thomas Lundquist, MD, at the Pop Health Forum 2016.

A better understanding of how information technology can optimize workflow results in improved use of data analytics and decision support, and leads to better patient outcomes, according to Adrian Zai, MD, clinical director of population informatics at Massachusetts General Hospital.

Speaking at the HIMSS Pop Health Forum this week, Zai said technology is tightly wound with actual workflow, and that close attention should be paid to how it actually interfaces with clinical processes. But it's important to ensure IT works in service of good workflow, not vice-versa.

"You need to develop new protocols and workflows before you talk to vendors," said Zai. "Technology should be there to wrap around workflow you came up with."

Zai presented a series of case studies from Mass General that showed how rethinking workflows and IT capabilities led to improvements in chronic care management and population health.

One focused on two different strategies for managing chronic disease: with and without central coordination.

Perhaps unsurprisingly, the program that depended heavily on IT to provide coordination through a simple, centralized interface significantly increased quality measures for patients with diabetes, cardiovascular disease and hypertension over a six-month period. The technology enabled centralized population health coordinators to work closely with medical practices and led to improved outcome measures.

"It's important to have (targeted) decision support built into your system before you present it to your physicians," he said. "The idea here is to have physicians practicing at the top of their licenses, so try to take that noise away by improving CDS to reflect true standard of care" while moving as much coordination activity to the back office as possible.

"Now that we've learned these lessons, we've adopted these systems and have care coordination everywhere," said Zai. "It all makes sense."

A second Mass General initiative developed a collaborative care model to manage patients with diabetes. It put together a multidisciplinary care team comprising primary care doctors, specialists, behavioral health professionals and others.

The project depended heavily on incorporating evidence-based strategies into the care process, and made full use of practice-based teams to facilitate staff and patient engagement to improve numbers for those patients not at goal.

One benefit was a greater clout for nurses and other clinical staff. 

"We were able to engage non-MD staff, and their role has really risen," said Zai. "Nurses are now members of the care team and the physicians are completely dependent on them."

Even better, the number of patients with high A1c counts fell dramatically. But it all depended on well-considered IT strategies to link e-prescribing with patient education and lifestyle change training, as well as improved communication and documentation across the multidisciplinary care team.

A third project at Mass General involve the use of patient navigators to enable comprehensive cancer screening. Zai pointed out that such assistance can improve rates of cancer screening in vulnerable populations, but most cancer navigator programs are located in community health centers and focus on screening for a single cancer. Many patients may be overdue for more than one screening test, he said.

To build a better patient navigation program that can help patients obtain comprehensive cancer screenings within a large academic primary care network, the initiative leveraged population health IT to identify high-risk patients and improve patient-centered navigation.

By ensuring the technology enabled the hoped-for workflow, Mass General was indeed able to increase screening rates in that patient cohort, while better empowering navigators to decrease disparities in care and improve outcomes.

Zai said that tools that can help address equity of care should be built into population health management systems, such as algorithms to better identify vulnerable patients and improve referral and tracking of those at highest risk.

For those health systems looking to explore similar population health projects, he said that close attention to detail was essential.

"It's really important to understand that when you drive innovation, you can do it in an kind of ad hoc way without scientific rigor to do it really quickly," said Zai. "But there are potholes along the way: If you take that route, you may be measuring differences that aren't there. It's really important to get it right – to get your study, your controls, all of that correct.”

There's also nothing wrong with doing quick assay cycles, he said, "but pick your measures carefully. Having a control is always a good idea. And having many little pilot initiatives that start with small populations, just to test the concept of what you want to do before you steer the ship in that direction, is really important. That will save a lot of time and money."

Twitter: @MikeMiliardHITN