Penny Wheeler, MD is president and CEO of Allina Health of Minneapolis, MN.
You’re a brand new health system CEO, unusual in that you’re female, you’re a practicing physician, and you’re not afraid to get your hands dirty digging into IT and data issues. What are the challenges Allina is facing and how does IT fit into those?
The challenges aren’t atypical to other healthcare organizations. Our biggest challenge is that what we want to accomplish for our mission isn’t necessarily the way that our incentives are lined up right now. Our financial success doesn’t always equal the success for our patients.
For example, right now we’re having, unfortunately for the community, a bad flu season. A lot of people are getting sick. That’s meaning good things for us financially. That’s not the way it should be. We want to help drive an incentive system that rewards us for good health for the community. That’s our biggest challenge.
The challenge to get to that is the second biggest challenge … to transform into a whole different business model. What we have to do to reduce costs and have the right information to do that.
The third thing would be to have the infrastructure set up to be able to do that successfully, which is why this information and turning data and information and consolidating that and organizing that in a way that actually can move towards better better outcomes is so important.
Regarding the announcement that just came out (Allina Health and Health Catalyst Sign $100 Million Agreement Creating Model for System-wide Outcomes Improvement), most people would see Health Catalyst as an IT tools and services vendor. What gave you the confidence to let them get so involved in quality projects going beyond the technology?
We have a history with Health Catalyst. We were their first customer out of the chute when they were a two-man shop, then known as Health Quality Catalyst. At that point, I was leading the quality agenda for the organization. I saw that we had insufficient information to know what outcomes we were getting at what price for the patients we served.
We had to integrate dozens of databases to be able to show what that looked like, to show our outcomes. The electronic medical record was not enough. It was one source input, but it wasn’t nearly enough for us to do what we needed to do to get all the — you can call it big data, you can call it whatever you want — integrated in a way that it allowed us to focus our resources correctly for patients.
We engaged the two founders of the Health Quality Catalyst, now Health Catalyst, Steve Barlow and Tom Burton. We were their first customer because we realized we wanted a way to set up a data in an integrated way akin to what Intermountain Health had done, and they came from Intermountain Health.
There’s the question of whether organizations don’t have enough data or whether they don’t have the willingness to act on the data that may already have available to them. How do you quantify an organization as to whether they’re ready to be data driven?
I’d say it’s neither of those two things. We have tons of data, but our previous data was just dumped, like a dumping ground, into what was a data warehouse. There was no way to get it out in any meaningful ways nor have it be actionable on the back side.
I would say that it’s actually the integration of the data in a way that’s usable and meaningful for the people who know the work the best. That is the biggest challenge for folks. Everybody and their mother might say they have a data warehouse and they do. We had one before, too. It was a big dumping ground and we couldn’t get anything meaningful out of it.
Now we have a data warehouse that actually integrates over 27 different databases and shows us an outcome of what our variation is in the way that we’re caring for people. What outcomes at what cost. It puts an overlay on it that gives us a dashboard so that people can make some use out of the numbers.
For example, through all these databases, we were able to develop using hundreds of factors a predictive model that told us who was at greatest risk for readmission. Now we have a predictive model that predicts with about 80 percent effectiveness who’s likely to be at greatest risk. They’re flagged right when they’re in the hospital. We know we can put care management resources right to those people because they’re flagged as being high risk for readmission.
It’s those kinds of things that we can do. Our caregivers can use that information directly.
Other than readmissions, what impactful results have you seen from the use of data analytics?
A lot of its success has been cultural. We’ve been able to engage physicians much better in better care improvement activities because they know that we have measures of performance that are meaningful and accurate. A huge part of this has been cultural and engagement of physicians and caregivers because they’re the ones who really can improve care process, which is where most of the wasted cost is now is … on care processes. A lot of it’s been on the engagement because they know it.
I’ll give you an example that shows you about an initiative and shows you about that engagement. We had an early initiative about reducing, like many organizations have, early inductions of labor. We had integrated all the data that shows us where they were happening, by what physicians, at what gestation, for what reasons. We could get it in moments where other health care organizations were taking months and months to get that information. We could get it in moments.
Then I was at a conference table where somebody said to another doctor, "Well, that can’t be right. That can’t be what my induction of labor is." The other person looked at them and says, "Nope, I’ve looked at the data. I trust the data and it is right." We were able to, in that case, reduce our elective inductions of labors from 14 percent — which it should be next to zero — to about 1 percent. We have maintained that for the last two and a half years. That results in about 250 fewer women having Cesarean sections per year and many fewer babies being in the intensive care unit.
That’s an example of where we’ve been able to engage the physicians. They believe the data. We’ve been able to drive towards an outcome that has meaning in terms of better health for the individuals at a lower cost.
For a health system that doesn’t have a physician CEO, what would you recommend as a structure to take that information and convince physicians to act on it?
We try not to invent what has already worked well other places. The structure we’ve used is looking at our clinical service lines across the organization for specialty care and also our primary care base. Then having content expert groups around a particular care conditions of the patients focus on what they wanted to measure and what data and information they needed to make sure that the care was the best.
For example, in the oncology clinical service line, we have a breast program committee. That breast program committee decided the 31 things that they wanted to measure around quality. That breast program committee includes doctors, but it also includes administrators and nurses and radiation folks and all kinds of multidisciplinary physicians from oncologists to surgeons to radiation oncology people and general medicine. I think focusing around a condition of the patients was important and getting the multidisciplinary team together to do that and to find what was important.
One other thing I’ll add is in the best of our groups, adding some patients to it also has helped. The patients — I’m just thinking about that breast program committee — said, "You’re measuring how long it takes me to get a diagnostic mammogram done. That doesn’t matter to me. What matters to me is when you find a problem, how long it takes me to get in for the next test, because that’s when I don’t sleep at night."
That’s where we’ve been able to connect that dot with the patients in our best forms. We don’t have that everywhere, but that’s been best. That changes the conversation and makes it assured that it’s patient focused because, as I said when you were talking to me about challenges right now, right now one of our biggest challenges in healthcare is that the waste that we’re trying to reduce in healthcare is somebody’s revenue these days. That’s where it gets very difficult.
The agreement with Health Catalyst must have been complicated to negotiate since Health Catalyst is taking financial risk along with Allina, but then Allina gets partial ownership of Health Catalyst. Is the agreement that you’ve signed going to be difficult to manage and measure?
We spent a lot of time on that part. I think it’s a great partnership because Health Catalyst has incredible tools, but most people like them just hand you the software tools and they say, "OK, now here do." Through this partnership, we can have these incredible tools, but then we have the arm that shows us how to implement those and how to engage caregivers around implementing them in the right way for the right reasons for the right patient. We think it’s a good marriage that marries the knowing and the doing in a big, important way, so that’s huge.
We spent a lot of time, to your point, about what do we each look at as measures of success. We spent just painstaking time saying, “What looks like success in this? What key process indicators do we have? What do we have to make sure is maintained in this agreement to make sure it’s successful?” I think we’ve done that enough up front so that it won’t be very difficult to administer now that we’ve defined it on the front end of these negotiations. We’ve aligned our interests in a very profound way.
There’s a lot of discussion about the need for patient empowerment, getting patients involved in their care and having some control over their care episodes. Are there any projects that you’re working on that address that?
That’s a really great question. That’s where we’ll have to evolve to. Right now we’re evolving in ways that we can have trusted outcomes information and an implementation arm in terms of management to move those outcomes in a better way.
But you’re absolutely right. The holy grail is, how can we make patients the principle agents in their own health so that we move even further upstream instead of just reacting to how we can better care for those who are ill? How can we better support them to be well? I don’t think we’re all the way there yet, but we’re talking about tools and ways in which to do that better.
How do you draw the line between the healthcare system and social services delivery, where the health system has responsibility for managing populations but can only go so far into the community? What are the challenges or opportunities?
That’s another really great insight and question. There will surely be partnerships and walls broken down in ways that we never envisioned them to be broken down. I certainly don’t think that we as a healthcare organization can do things like fixing the cracks in the sidewalk. We need partnerships with social services agencies. But a lot of what we do will become more an more analogous to social work services and partnerships with those social service organizations than we do today.
I’ll give you an example. We did actually merge with a organization called the Courage Center for people with disabilities. We had all the acute care services for those people who had had a recent stroke and needed physical rehabilitation. We had some sports and PT clinics. But we didn’t have the in-betweens of having post-acute care for those patients, transitional care. What the Courage Center brought us was vocational training, activity-based training, and sports and activity programs where they were able to see how they could drive, where they could do some of those community-based programs. That’s the full continuum.
In this case, we did it through a merger. In other cases, we’ll have to do it through partnerships. But you’re absolutely right to ask that question because the continuum is becoming much broader than we ever thought of in the past.