A few years ago, Massachusetts General was starting down the path of implementing a new revenue cycle when they hit a bit of a roadblock. The parent organization, Partners HealthCare, had decided to go “soup to nuts” with Epic, meaning that CIO Keith Jennings would have to construct a detour. But when your “world changes,” that’s exactly what a leader needs to do, and so he created in a new roadmap in collaboration with other Partners hospitals. Now, Mass General is in “a much better place,” according to Jennings. In this interview, he discusses the balancing act of keeping the goals of the hospital and Physicians Organization aligned, why his team is planning a staggered go-live, and the work they’re doing with population health. He also talks about how Mass General is working to make innovation and process improvement part of the culture, and what it was like stepping into Jim Noga’s shoes.
Chapter 2
- Role of case managers in pop health
- ACO pioneer — “We built a lot of history and protocol and function in this area.”
- 90/10 rule
- Creating a “hub” for best practices
- Life as a “recovering programmer”
- Fostering innovation — “That’s my dessert.”
LISTEN NOW USING THE PLAYER BELOW OR CLICK HERE TO SUBSCRIBE TO OUR iTUNES PODCAST FEED
Podcast: Play in new window | Download (Duration: 17:56 — 16.4MB)
Subscribe: Apple Podcasts | Spotify | Android | Pandora | iHeartRadio | Podchaser | Podcast Index | Email | TuneIn | RSS
Bold Statements
What’s key is the resources and protocols that you’re going to use with your target population, and then having the data 1) back up what you’ve done, and 2) eventually allow you to improve and create new protocols.
The truth is, it’s not one-size-fits-all. The things we can do and the services we can offer here at an academic medical center with a large on-campus physicians organization are slightly different than a community hospital or a primary care or specialty provider. But they can interrelate in certain fashions — some of it is reusable, and some of it needs to be specific for the location and the type of disease.
I’m that a recovering programmer, and so I have a viewpoint of, ‘any problem we can probably build something to fix.’ I really like working here at an academic medical center where there is research and innovation.
The transformation lab has got great support across the institution, and for my folks, sometimes it’s a lot of work, but it’s dessert. They get to play a little bit with something different in support of these folks, and they get to rub shoulders with world-class clinicians trying to solve the next generation set of problems, so it’s exciting and really enjoyable.
We are sprinkling trained process improvement folks throughout the institution and let them identify problems, and then they have the tools to create and monitor, and, if necessary, share the improvements that they’ve proposed.
Gamble: I want to talk about population health. I know that your organization has done some work, but then I also wanted to look at where you’re going in the future. So I guess the best place to start is to talk about where things are now with your population health strategy.
Jennings: Sure. So there may be some clinicians who have a slightly different take; I have a very IT-centric view here, but I think one of the important things to note is that we here at Mass General, and the rest of Partners as well, were getting to the population health business relatively early. We were one of the early participants. And again, I’m sure other things have some input into the formation of the ACO strategy, but there was a multi-year Medicare high-cost beneficiary program that ran for six or seven years prior to the creation of the ACO legislation. We were early adopters or participants in that program and we learned an awful lot as part of that program. I think we proved some of the models that the government is trying to enforce or offer in the industry as part of that, but we also learned a lot about how to do population health management regardless of government regulation or intervention. We’ve been at this at a large scale for a long time, and so this is second-nature or regular business to us at this point. There’s a continuum of things we have to learn, but we started early and kept going.
So as I mentioned, we were in that Medicare high-cost beneficiary program, and we were very successful. We showed that by spending a little more money upfront, there were benefits. And I don’t want to minimize anybody’s effort, but we really discovered that having case managers upfront working with your at-risk patients really seemed to be the secret sauce, along with all kinds of protocols and other activities. But it was upfront case management watching over your at-risk population really seemed to be the thing that bent the cost curve. Now, you have to pay a little more upfront for those kind of activities because you have a case manager who might actually be working on a patient who hasn’t come in for a visit, but from my non-clinician viewpoint, that seems to be part of the component. Then you have to back up everything else, those protocols and things that they do with information systems and other protocols that assist them in either preventing, minimizing or improving the care that happens after that.
We ran a big program here at Mass General as part of that high-cost beneficiary project. We succeeded after three years, rolled it out to Brigham and North Shore Medical Center, another community-based but very large facility, and then Newton Wellesley, a smaller community hospital, and into our primary care network, which used to be called PCHE but is now PCPO, where they are doing similar things. We knew it might work in a central core with concentrated resources, but we wanted to see if we could do these things at scale in less central locations, and it turned out we can.
So we built a lot of history and protocol and function in this area. We became a pioneer — I guess we still are a pioneer ACO — once those programs came on board. When you look at our structure organizationally, we have many of those programs that started as part of our high-cost beneficiary work, but we have established, at the enterprise or health system level, a departmental program for population health management that is now driving enterprise-wide programs, policies, data warehousing, those kind of activities, to provide analytical capability in support of programs at our major facilities here at Mass General, Brigham, our community hospitals and out in the physician practice network.
Gamble: There’s a lot of talk right now, and rightfully so, about having the enterprise data warehouse and all of that in place, but it seems like it’s also really important too to kind of go through what you guys did to be able to have that knowledge as far as training and staffing and the other aspects of population health and what actually gets results.
Jennings: I really think that’s right, Gamble. You probably need both. Data gives you certain amount of things and I don’t want to dismiss that — I think the data warehouse in that sense really provides you with additional value, but you really need that core set of services, whatever they’re going to be. Depending on the type of care provider you are, those may be different things, whether you’re a tertiary care facility or a primary care network, you might approach population health in somewhat different fashion. But I think what’s key is the resources and protocols that you’re going to use with your target population, and then having the data 1) back up what you’ve done, and 2) eventually allow you to improve and create new protocols.
Back in the day when I was doing some research and trying to figure out how we might take our early steps, I read an article by someone who said, ‘if you ask any of your clinicians, you can identify your sickest patients. You don’t need a big system; you can probably use a spreadsheet to track a handful of them, and you can start a dent in the curve just by addressing those guys or those patients first.’ And I actually believe that’s true. The 80/20 or 90/10 rule really exists in this space. There’s a handful of patients that you can go after and provide a set of services around and start bending the curve right away and get some successes, and then in parallel or even in serial, you can build some more of that structure and data-heavy services later.
Gamble: With having a department dedicated to population health, is that something where you hope to really keep that training going and applying all the lessons learned to keep learning more about how to approach the problem?
Jennings: Yes, absolutely. I don’t want to speak for them, but I believe that’s really their core value — they’re providing data and then they’re also providing a recipe book. But the truth is, it’s not one-size-fits-all. The things we can do and the services we can offer here at an academic medical center with a large on-campus physicians organization are slightly different than a community hospital or a primary care or specialty provider. But they can interrelate in certain fashions — some of it is reusable, and some of it needs to be specific for the location and the type of disease. Whether you’re really looking at higher-risk older patients or younger patients with cancer, you need a set of protocols both for the disease data or patient types that you’re working with, as well as for the physical location and the density of either patients or providers. But that’s the reason we have this hub that can take best practice and distribute it across the organization. One thing will work in one place and you got to try something else in another, but I think that’s the whole purpose behind the program.
Gamble: It’d be interesting to see in the next few years how that turns out.
Jennings: We’re succeeding. There was an article in the Globe — we are saving money and providing better healthcare.
Gamble: Right. So I feel like it’s safe to say there’s definitely at a culture of innovation at Mass General and at Partners. Some of the things that I’d read about that really strike me are things like the Mobile Health Hack, I think that that’s really interesting and maybe something that doesn’t require a huge effort but could get results. Is that just kind of an example of trying to think outside the box a little bit?
Jennings: Sure. In truth, one of the things that I really like about working here at Mass General, a large academic medical center, is I’m that a recovering programmer, and so I have a viewpoint of, ‘any problem we can probably build something to fix.’ I really like working here at an academic medical center where there is research and innovation.
Clearly, we do medical and basic science. We have a very large research enterprise that looks at basic science, and then we have clinicians developing new protocols and new drugs and artificial hips, all kinds of things. It’s a very exciting environment that fosters this kind of activity. We also have folks who research and explore and invent in the IT space in healthcare, and a lot of that — not all of it — but a lot of that happens actually in our clinical domains and not necessarily in corporate IS. At the minimal, we’re exposed to them. We often help and support them in those activities, and in truth, in corporate IS, we do a reasonable amount of that as well. For me, that’s my dessert. I spend a fair amount of time dealing with complaints about the email system being slow, but then I also get to work a fair amount of time on next-generation and cutting-edge programs, processes, and protocols, and my staff does as well. I’m actually kind of excited, someone in my corporate IS department will be an author on two peer-reviewed research studies. I think those are fantastic opportunities for folks in an IT department.
Gamble: Definitely.
Jennings: To give you a couple of examples here at Mass General, we actually have a group called the Laboratory of Computer Science. We know it internally as LCS and in fact, if you’re old enough, people may know the language called MUMPS that is now InterSystems Caché. The first ‘M’ in mumps is for Massachusetts General (in Massachusetts General Hospital Utility Multi-Programming System). It was invented here by the Lab at Computer Science 20 or 30 years ago.
In fact, somewhere in my messy office I have some of the original specification documents that have been passed down by CIOs for when Octo Barnett and others created that language, so we have a rich history of those kind of things here in the labs. Computer science is still here doing amazing things in research, both down to the usability level, but also new techniques for presenting data and analyzing data, and they occasionally spin out startups that go off in other places.
Another exciting program or department here is our heart center has created a transformation lab, which has a reasonable amount of funding both from internal and external supporters. They are looking at ways to use information to create an incubator to enable clinicians who have good ideas about tools that we could offer to patients or tools we can use to provide better care, to provide an incubator space for them. This is one of those cases where what I’ve done is I’ve ensured that they are in full contact with my corporate resources, both developers and data analysts, and others who can support them as they go after these endeavors. It’s turned out to be a fantastic combination of exposing clinical folks to IT folks who can help them turn their thoughts or dreams into practical applications.
And it’s really exciting. The transformation lab has got great support across the institution and for my folks, sometimes it’s a lot of work, but it’s dessert. They get to play a little bit with something different in support of these folks and they get to rub shoulders with world-class clinicians trying to solve the next generation set of problems, so it’s exciting and really enjoyable for myself and the staff.
Gamble: Right. It seems like there needs to be this understanding that anybody can present an idea, because you really just never know where the next great idea is going to come from, even something simple that turns out to be really beneficial.
Jennings: Yes, absolutely. I’d like to mention two other things — again, these are ones that in IT we tend to support as opposed to lead, but we have 39 clinical services here at the hospital. Patient care services, in conjunction with their physician colleagues and others, have identified 12 of those floors or services as innovation units where for a variety of reasons, either because of their patient flow or because of just the temperament and the way the flow works, that are willing to embrace and can tolerate the change and the risks that happen when you try and innovate. They run this fantastic program where they have units that are very accepting of trying new things, and once we give it a try and it turns out it’s a winner, we quickly then can roll it out to the other innovation units, and then once it’s battle-tested, roll it out across the hospital.
It’s just a fantastic, fantastic program, and to your question about does anyone have a good idea, it’s this type of program that allows anyone to say, ‘we can move patients faster through our burn unit,’ or ‘here’s a way that we can turn beds over quicker and get the next patient out of the ED faster,’ and they roll these through the innovation units. And then it’s very inclusive as far as who gets to participate.
One other program we have, I forget who we copied it from, but there’s a gentleman out at Intermountain Health who used to run a performance improvement seminar. We at Partners, not just Mass General but Partners, sent a couple of people out there, and then we liked it so much we sent 10 people the next year and 50 one year. Eventually the gentleman out there said, ‘you know, you guys could do this back home. You send enough people — why don’t you stand up you own program?’ And so we set up our own, we call it CPIP (clinical process improvement program). It’s a formal program where the clinicians — doctors, PA, mid-levels, etc. — can apply with a project in mind. We set up a fairly intensive, two or three-month program with meetings two days every other week or so to teach clinicians who are interested in doing small or large scale (but mostly small scale) process improvement projects, methods for doing that—fishbone studies, how do you solicit feedback, what kind of statistical process modeling do you want to use so you can evaluate whether your intervention has changed anything. It’s a very rigorous model and process. And we keep running two classes a year of clinicians through this, and so we are sprinkling trained process improvement folks throughout the institution and let them identify problems, and then they have the tools to create and monitor, and, if necessary, share the improvements that they’ve proposed.
Share Your Thoughts
You must be logged in to post a comment.