Hospitals and health systems nationwide are struggling these days to create a healthcare system of improved care quality and patient safety, efficiency, cost-effectiveness, accountability, and transparency. Yet a small number of pioneering organizations are already charting a bold, clear course—among them, North Memorial HealthCare, a two-community-hospital health system based in Robbinsdale, a suburb of Minneapolis-St. Paul, which encompasses the 300-bed North Memorial Medical center in Robbinsdale and the 100-bed Maple Grove Hospital.
Leaders there have launched themselves on a path that will position their organization beautifully to adopt the principles of the new healthcare, and it is one that is already reaping rewards. The core of what the leaders at North Memorial are doing is to leverage healthcare IT to facilitate enterprise-wide continuous quality improvement, not only for its own sake, but to achieve clinical transformation across diverse patient care service lines in the organization.
As a result, the leaders at North Memorial HealthCare were named the co-second-place winning team in the Healthcare Informatics Innovator Awards program. They will be honored at the Healthcare Informatics Innovator Awards reception in Orlando in February, during the annual HIMSS Conference.
All the members of the leadership team provided HCI with extensive interviews. Below are excerpts from HCI Editor-in-Chief Mark Hagland’s interview with Jeffrey Vespa, M.D., medical director for clinical quality for the health system. Dr. Vespa shared his perspectives with Hagland on the work that he and his colleagues are doing at North Memorial. Previously, an interview with J. Kevin Croston, M.D., the organization’s chief medical officer, was published; additional interviews with other leaders at North Memorial will be published forthwith. Below are excerpts from the interview with Dr. Vespa.
In my interview with him, Dr. Croston offered his perspective on the enterprise-wide quality initiative at North Memorial, from his CMO perspective. As medical director for clinical quality for the system, what do you see as the factors driving you and your colleagues forward in this broad initiative?
A lot of it is centered around the value proposition that’s out there for healthcare. Considering that quality-cost equation, we’ve elected to pursue the quality part of it, assuming that we will gain the cost benefit along the way. And historically, when it comes to quality improvement work, we’ve employed a traditional model where you build a team and goals, and try to implement a project, and that of course, is not new. The opportunity we saw was that, without good, real-time data, it was hard to show improvement and then provide the feedback to individuals and teams about their performance. People may try to impalement change, but then the question is, did you really do it?
Jeffrey Vespa. M.D.
And we’ve found that, after an initial year where you have a focus on a process, and you do well, you would sort of lose your gains over time. And the flame would sort of dwindle over time on quality processes. But the data lives forever. And so we can put that data back into leadership’s hands, to feed that back to their frontline staffs, even after a team may not be assembled anymore, you can sustain change. Another factor involved in traditional quality work has to do with turnover of staff. You might have a great physician leader; for example, we had a great physician leader helping to guide a program around readmissions work, and he started some really great processes, but then he left the organization, and the gains evaporated. So by building these data systems, even if someone were to leave, we knew we could sustain work that we’ve done.
What timeframe has been involved in this broad initiative?
We really started the new, conceptual approach in 2011. And that was in a partnership with Health Catalyst, an electronic data warehouse company. They’re a vendor. And that was a very important piece of all of this, building that electronic data warehouse.
This was your first enterprise data warehouse?
And when did that go live?
In 2011, under the management of Greg Kern, the manager, measurement and reporting, for the health system. And, at the same time, Actually, we also formed COLT, the Clinical Operations Leadership Team. And the intention there was to ensure that our leadership was aligned between and among physicians, nursing, finance, informatics, measurement and reporting, and quality. All those leaders are involved in COLT. Currently, there are probably about 25.
What was it that you all decided?
The big goal was to reduce variation in care; that was the overarching goal. So, much like the Intermountain type organization of activities. And to do that, we needed the EDW to build the measurement and reporting systems and facilitate the work and the feedback. So one of the first things we did was that we built a Key Process Analysis Tool. What that did was that that took information from our finance database, married it with our clinical databases, and from that, we could find where the largest volume of services was, and the variation contained within that. And we don’t have exact internal cost accounting; we unfortunately don’t know the exact cost of everything that we do, so we used charges as a surrogate. So if we had big-ticket services that involved a lot of variation in charges, that was a key indicator for us.
So what areas did you isolate as the biggest areas to look at?
The first good opportunity on high volume and variation was women and newborns, and specifically, the pre-39-week elective induction in women.
Can you share metrics on the before versus the after?
Our baseline was 1.2 percent: for a woman who was earlier than 39 weeks in gestation, we were seeing 1.2 percent elective inductions, and drove that down to 0.3 percent, over about a year. So we organized a leadership team led by one of our obstetricians, Jon Nielsen, M.D., the medical director of our Women’s and Children’s Service, and Linda Engdahl, R.N., M.S.N., director of Women’s and Children’s Services at Maple Grove Hospital, and they helped to lead an initiative to drastically reduce pre-39-week elective inductions. And the challenge here was that this was multiple, private, independent groups of OB physicians. This was not employed physicians. This was truly integrating independent groups and their work; so standardizing workflow was the first step. And that also helped us to improve our standardization around data collection.
As Dr. Nielsen and Ms. Engdahl explained to me, there wasn’t even a standardized method of determining gestational age, so that had to be put in place through consensus-based work first, and then the work on analyzing variation could take place, and the subsequent work to reduce variation. So was the key to turning this around?
It was generating the data, measuring deviations, and feeding the data back to the physicians. And part of the build, obviously, is determining some of those nuances. No one is trying to do bad medicine. A lot of the challenge comes out of having to work to standardize workflow and other processes, so that if physicians do deviate from [consensus-driven norms], they understand that they’re being measured.
In other words, this is about sustainable change created through data-driven analysis of ingrained patterns of practice?
Yes. And imagine if you have, say, 40 OB doctors—the pace at which you can build a pathway by consensus is very slow. One of the ways that we tackle that is that we do have small workgroups going through and working on definitions. They spend a lot of time codifying the clinical information, they meet weekly. In this case, it was led by our measurement and reporting data architects. The data architects or analysts are codifying the clinical work.
And that happened with the early induction initiative?
Yes. And they would take that work and bring it once a month to a large group of OB and nursing, to vet that with them, to look it over and make changes. So there was a fairly rapid pace of work. But you accomplish that with a small group, but with a larger group’s vetting it. There are five people in the small workgroup, and the larger group was about 25 people.
How many other projects like the early induction project have you initiated so far?
Within Women and Newborns, we’ve also launched a respiratory distress initiative—reducing chronic lung disease following respiratory distress syndrome in newborns.
More broadly, how many clinical projects have been data-facilitated?
We’ve done a variety of both clinical and non-clinical projects, among them, an OR workflow and efficiency project. We’ve also done a project around utilization of oral antibiotics compared to IV antibiotics, that was a potential cost-reduction project. And we’ve used this same data methodology to help us in surveying our core measures. And we’ve also used it to help us provide surveillance on catheter-associated urinary tract infections. And the other big one: we’ve also in the last couple of months launched our Community Medicine Guidance Team. And through the EDW, they’ve built a population health management module around preventive care around diabetes, vascular care, immunizations. We’re really excited about that one.
You’re obviously letting a thousand flowers bloom here?
Yes. Well, when we first started introducing the idea, many people thought that this was just more of the same traditional approach to quality improvement. But once they’ve experienced it, they’ve wanted more of it. And that’s a challenge for us, because the resources are limited.
What are the biggest lessons that you’ve learned in this so far?
Lesson number one, I think, has been making sure that you have a very clear aim. It’s pretty easy to have sort of global goals. But when you’re talking a project itself, we’ve had more success when we’ve picked a true process that has a start and stop. We’re dabbling in heart failure, currently, and how to provide a seamless care for heart failure across the continuum of care, from the clinic to the hospital and back. And that’s easiest if you can identify the elements along the way.
Is one critical success factor identifying the discrete, manageable elements, that one can work on?
Yes, and that’s part of the initial work, when you do your value stream mapping. Often, it’s so hard for physicians and nurses to identify the key processes. But that’s so important, because when you can identify the key processes, you’ll have a discrete, measurable intervention. And that will more likely generate the overall outcome you’re trying to achieve. Because you can write a big, long guideline that has a million steps to it, but you can’t measure them all. So in my mind, there are parts that are guidelines, and part that are protocols, and the protocol parts are most interesting to me: where should we never deviate?
Are physicians beginning to realize that this is where things are going, towards more accountable and transparent care delivery?
I think so. I spend half of my time as an ED doctor and half of my time in quality improvement. I think the leaders understand this; but that has to trickle down to the frontline physicians. It’s improving. I just gave a talk this morning about value-based care. And it’s still amazing to me that when we talk about the shift from discrete processes to looking at the outcomes, the light bulb starts to go off in people’s heads, and they say, I need to work with you on this!
Subsequent interviews with Jon Nielsen, M.D., and Linda Engdahl, R.N., M.S.N. will delve more deeply into the early induction initiative at North Memorial HealthCare.