There’s a lot of excitement around predictive analytics, and for good reason. Organizations like University of Mississippi MC are seeing positive results in areas like reducing pressure ulcers. But, as with any advance in technology, there are hurdles that must be overcome, most notably resistance from clinicians about how workflow will be affected. In this interview, John Showalter, MD, talks about his team’s approach to change management and how they’re working to quell clinician fears. He also discusses the pros and cons of going big-bang, what he believes will be the next wave of predictive analytics, what it was like to go from the “well-defined” CMIO role to the more nebulous CHIO role, and his advice on how to communicate more effectively with physician leaders.
- The next wave of predictive analytics: hospital-acquired conditions
- Shifting “to a more of a preventive mentality”
- Managing “cultural resistance”
- From CMIO to CHIO
- “We hit the ground on a sprint.”
- Data visualization with Qlik
- Creating Propel Health IT to “disseminate knowledge”
- Best practices for working with physician leaders
LISTEN NOW USING THE PLAYER BELOW OR CLICK HERE TO SUBSCRIBE TO OUR iTUNES PODCAST FEED
I think there’s going to be a big challenge to sell people on the technology. I think it’s going to seem a little bit magic that you could put 200,000 people through an algorithm and come up with 100 people that need your attention the most.
It was going from a very defined role to a very open and ill-defined opportunity to say, we know we should get benefit by doing analytics on the data we’re producing, but we don’t know what that looks like.
It was a very exciting place to be in to cobble together best practices from other organizations into a unified strategy.
There was a whole lot of knowledge we had received in our training that was not really getting disseminated, and was not part of the standard graduate curriculum. Even people coming out with a Master’s degree weren’t getting insights into the differences of psychology between different groups of clinicians and administrators in healthcare, and the information around the fusion of innovation.
You can’t treat the nurses the same way you treat the docs and the administrators. You need to tailor your message to really how they process it, and structure your meetings and your adoption around how the different groups function. If you just try to aim for the middle, you’re going to miss everybody.
Gamble: As far as predictive analytics, what do you think the next wave is going to be for your organization?
Showalter: There’s definitely a wave of predictive analytics around hospital-acquired conditions, things that are part of value-based purchasing, and things that are going to be tracked in MACRA, but I think the next wave is going to be at a population level. It’s going to be looking at the ambulatory population and predicting who’s likely to get sick in the next 12 to 24 months, who’s likely to become a very expensive patient in an ACO, and then trying to intervene before they become expensive. So I think we’re going to shift from this inpatient, acute mentality to a more of a preventive mentality around predictive analytics.
Gamble: And when that wave comes, do you think it is going to be a matter of selling people on the technology the same way you’ve had to do that with analytics?
Showalter: Yes. I think there’s going to be a big challenge to sell people on the technology. I think it’s going to seem a little bit magic that you could put 200,000 people through an algorithm and come up with 100 people that need your attention the most.
I think it’s also going to be very challenging to decide what to do for those patients. We’re not going to have clinical pathways for what to do if you’re predicted to be one of the 100 people at highest risk for having a costly year due to poor health. I don’t know that we know what to do for those patients. So there’s definitely going to be some cultural resistance to acting on a prediction, because there’s not a clear path on what to do.
Gamble: And is that something where it will help having some tangible results on the inpatient side?
Showalter: I think so. In order for that next wave to happen and for us to see adoption, people are going to have to see the success demonstrated on the inpatient side where the turnarounds are much shorted. Pressure ulcers are occurring two to three weeks after admission, and in two or three months get a really nice validation of that data. With 30-day readmissions, within two months you’re going to have some pretty decent data around that validation. Knowing who’s going to get sick in the next year is going to take you a year to validate, but you’re not going to want to take a year to act. So it’s going to be this balance of trusting in the technology because it’s been successful in a different setting.
Gamble: You talked about when you made the transition from CMIO to CHIO and it made sense at that juncture to do that, can you talk about what that adjustment has been like and how this role has been different?
Showalter: It was a transition from a very well-defined role — a role of somebody that was helping clinicians interact with the EHR better, changing it so that they could be faster and so it was clearer, and working with them to change their workflows to get it to work better for them. It was going from a very defined role to a very open and ill-defined opportunity to say, we know we should get benefit by doing analytics on the data we’re producing, but we don’t know what that looks like.
We spent well over a year not really producing any analytics, just looking at priorities, creating a roadmap, creating the data governance structure, developing the right skill sets in-house to do the analytics, and identifying the right partners for the analytics we didn’t have the skill sets to do. So we had an almost a full 18 months in planning before we did our first analysis on anything.
Gamble: Sounds like an interesting spot to be in, almost unchartered territory. Is that something that you were comfortable with? How do you approach that?
Showalter: I was comfortable with it. It’s really been something that I was trained for. I did a combined internal medicine and clinical informatics fellowship, so I trained for five years to sit in that spot and lead those teams. It was a very exciting place to be in to cobble together best practices from other organizations into a unified strategy. I had a really good support team and great senior executive support for what we were doing to put that together.
Gamble: After you were able to lay that groundwork, did you feel like, ‘I’m ready to hit the ground running?’
Showalter: We did, and then we hit the ground on a sprint. We’re using Qlik to do data visualizations. We have a four-person team that’s working on Qlik. They produced 20 applications in the first year. On the reporting side of things, when we restructured how they were doing reports, we did over 1,100 reports with a six-person team — they were averaging almost a report a day each. We established a very efficient system that focused on the institution priorities and aligned with the institution strategy. I’ve been running that for a year now. It’s been a very effective structure, and it’s been very efficient. It’s taken two and a half years to begin to see the dividends, but we’re definitely seeing the dividends now.
Gamble: I would imagine that all of this is exciting to be a part of, but in terms of the Chief Health Information Officer, I’m sure I don’t have to tell you there’s not exactly thousands of you. So I’m sure it’s a fascinating spot to be in.
Showalter: Somebody told me the other day that we’re up to 11. I don’t know if that’s true, but I was at a conference and someone said, ‘You’re one of 11.’ I said okay.
Gamble: Wow, a very small number. And it seems like with the training you had though that this is really kind of the ideal role combining the informatics with the clinical.
Showalter: Yes. And it was the vision of the CIO that I was working for after medical school to really combine these two. I wanted to go down the informatics route, and I worked with him to develop the five-year program and made sure I hit all of his requirements as my mentor for the program, and it’s worked out very well. I got a Master’s in information systems while I was doing my residency and fellowship, so I’m board certified in internal medicine and clinical informatics.
Gamble: You had that while you were doing your residency?
Showalter: Yes. With the combination of my residency and fellowship, I was able to jockey my heavy clinical months to not be when I was taking classes. So it was a five-year program to get my internal medicine piece done, which is normally three.
Gamble: I was curious as to how that actually happened doing all that at once.
Showalter: I did not have five children at that time. That’s how that happened.
Gamble: Okay, there you go. So obviously a lot of cool stuff going on, and a lot of really great stuff in the works with going down the population health route. I wanted to also talk about Propel Health IT. Is that an initiative outside of the organization? How does it work?
Showalter: Yes, that’s outside of UMMC, and it was started with my partner, Leigh Williams, who is now the senior business administrator for the information systems department at UVA. She sits over the Lawson and Kronos systems at UVA on the technical side. What we learned early on in our partnership — she was part of the team that was implementing Epic at UMMC — was that there was a whole lot of knowledge we had received in our training that was not really getting disseminated, and was not really part of the standard graduate curriculums. Even people coming out with a Master’s degree weren’t getting insights into the differences of psychology between different groups of clinicians and administrators in healthcare, and the information around the fusion of innovation and some process improvement pieces. So we put together an educational curriculum and had some seminars to really get feedback and hone the education and messaging, and we’re in the process now of putting together online education in order to get this information out there and get people much more effective at adopting technology and managing change management around technology.
Gamble: So it was something you guys started on your own and you’re starting to see that grow out at this point?
Gamble: I would think from doing the kind of big bang implementation that you were part of, I’m sure there was a lot of lessons learned.
Showalter: Yes, there were.
Gamble: I know you don’t want to give too much away, but I would ask if you have any kind of best practices for CIOs when it comes to working with physician leaders on a project of that scale. Any tips on how to really get that communication going?
Showalter: Sure. I think the first best practice is you can’t treat them all the same. You can’t treat the nurses the same way you treat the docs and the administrators. You need to tailor your message to really how they process it and structure your meetings and your adoption around how the different groups function. If you just try to aim for the middle, you’re going to miss everybody. Physicians really require a much more hands-on approach getting things much closer to what they do on a daily basis, while nurses do much better learning and adopting in groups, and executives do much better with ‘here’s some information, go and think about it.’ Even when you’re doing really big projects, you still need to address your stakeholders and your communities individually.
The second thing I’d say about that is it really needs to be an organized and intentioned approach. We had written out communication plans for 18 months into the future: on this date, we’re going to send this messaging. On this date, we’re going to send that messaging. On this date, this email has to be drafted so that it can be reviewed by this person before it goes to that group. We had full communication plans laid out for 18 months in advance. And everyone got tailored communication.
Gamble: I would imagine there has to be a little bit of room for agility if things change, but the important things to really have that ground work in place for communication.
Showalter: Oh yeah. I’m not sure we actually ever hit a single date in that communication plan on the date we were supposed to have it. But it very drew out very clearly throughout that the nurses were going to get the same information, but in a different way than the physicians and in a different way than the executives.
Gamble: Okay. Well, that covers what I wanted to talk about. It sounds like you’re doing some really interesting work and it will be fascinating to see where predictive analytics goes, especially with the work you guys are doing there.
Showalter: I’m really excited about that. I think in 5 to 10 years, predictive analytics is just going to be a part of how we do medicine.
Gamble: As a patient, I like to hear that. Well, thank you so much for taking the time to speak with us, we really appreciate it.
Showalter: No problem. Thanks, Kate.