Tina Esposito On The Evolution In How Data Is Being Used

Tina Esposito, VP, Center for Health Information Services, Advocate Health Care

As VP of the Center for Health Information Services, Tina Esposito knows very well the value of digging deep into analytics to find answers. But she has found that when it comes to learning what clinicians actually need, the best course is to simply ask them. By going “to the source” instead of making assumptions, leaders can help provide what those on the frontline need, says Esposito. In this interview, she talks about how the Center came about, the change in thinking that was needed to become an ACO, and the interoperability challenges that come with multiple EHRs. Esposito also discusses the concept of patient-centered care, why her team is focused less on HCAHPS and more on building loyalty, and why she’s excited for the future.

Chapter 1

About Advocate
CHIS’ origin in 2005
“We are, first and foremost, a safe clinical enterprise.”
Monthly scorecards with “drill-down analysis” on safety, quality & engagement
Becoming an ACO — “Our data wasn’t structured that way.”
Partnering with Cerner to create an analytics platform
The “evolution” in how data is used
Clinician input — “You need to ask them.”

LISTEN NOW USING THE PLAYER BELOW OR CLICK HERE TO SUBSCRIBE TO OUR iTUNES PODCAST FEED

Podcast: Play in new window | Download (Duration: 13:35 — 12.4MB)

Bold Statements

We always knew we were doing good work. There was just a need to ensure that there was a level of objectivity in understanding the work that was going on, the outcomes that were being delivered, and the opportunities that may exist.

One of the first things we had to do was pull all of our data together. And not just inpatient-focused data using the traditional siloed approach, but rather, how do we pull all of our EMR source systems together? How do we pull and marry that with our claims data and our PBM data and truly create a platform that reflected all aspects of care, not just individual levels like we had in the past?

We’ve evolved as a team, becoming less focused on the technical aspects of bringing data together and more so on ensuring that we’re using it in the way that provides insight to the organization.

What we quickly understood from our clinicians was the need to ensure we were thinking about all readmissions — not just acute myocardial infarction, heart failure, pneumonia — but thinking about how we ensure that the tools we’re creating would help manage any potential readmission, not just one definition of one entity out there.

Gamble: To start off, can you provide an overview of Advocate? I know you’re a large health system, but can you get a little bit more into the details in terms of number of hospitals, things like that?

Esposito: Absolutely. We are, I believe, the largest, integrated health care system in the state of Illinois. In total, we have 11 hospitals and 6,000 physicians that are aligned with us, either formally employed (about 1,400 of those) or through a clinical integration program, which is our Advocate Physician Partners Program. The system itself does about 6.4 billion in annual revenue. We’re a pretty large accountable care organization — at one time, I know we were the largest. I think now we’re the second largest with about 865,000 attributed lives. We also have a very large post-acute presence with a large home health division that also includes hospice, a skilled nursing facility referral network, and a large palliative care program.

Gamble: We’re certainly seeing the trend of more focus on that post-acute presence and bringing that into the fold in a way that hasn’t always been done. Now, in terms of your background, how long have you been with Advocate?

Esposito: I’ve been with Advocate 18 years, so quite a long time.

Gamble: How long has the Center for Health Information Services been in place?

Esposito: Since 2005.

Gamble: And how long have you been in that role?

Esposito: Since the beginning. It’s funny; it was just of two of us back in 2005. Now, we have about 35, almost 40 people.

Gamble: What’s the reporting structure — do you report to Advocate’s CEO or CIO? How does that work?

Esposito: I actually report up to the chief medical officer. My next-up is the senior vice president over clinical operations, who is a physician. And he reports to our chief medical officer.

There’s certainly an ongoing relationship, as you can imagine, with our information technology division and systems. As things continue to evolve in the industry, and particularly in IT, there’s more and more of a realization that there’s just as much focus around the information as systems. So we, absolutely, work together often. I have many peers at my level in IT and clearly, we wouldn’t have been successful without that collaboration.

Gamble: How often would you say you’re in touch with them?

Esposito: Constantly. It would be odd for me not to have some level of interaction, at least two to three times a week, whether it’s a formal meeting or an informal reach-out or conversation, or email.

Gamble: Now, what services does the Center for Health Information Services provide?

Esposito: At the beginning of it, it was really centered around this need in the organization, at a pretty senior level, to ensure that there’s a level of rigor in understanding, measuring, and providing insight into the clinical care being provided. We always knew we were doing good work. There was just a need to ensure that there was a level of objectivity in understanding the work that was going on, the outcomes that were being delivered, and the opportunities that may exist.

When you look at an organization, typically, from a data measurement analysis perspective, there’s always financial rigor. Our CEO will often say, ‘We are first and foremost a safe clinical enterprise.’ Senior leadership really wanted to ensure that was held in the same vein as we thought about health outcomes, quality, safety, and engagement. So some of the early work we did was centered on ensuring that we consistently measured our key outcomes.

We were incredibly transparent in ensuring that everyone was clear where we are in our health outcomes journey as an organization. Were we fulfilling our vision? Were we on our way there? Some of that, to get very detailed, also started with, what should we be measuring? What are our key outcomes that the organization is after? We still support an annual process to identify our key metrics and ensure that there’s alignment from a strategy and business perspective.

In the early days, it was about getting this data, as I mentioned, in a place that was very transparent with a lot of rigor. We created a health outcomes close, which we now call the Advocate Experience Close, and it’s very similar to a financial close. We close the books every month on our health outcomes performance, and that’s done through online reporting through an EDW, as well as a spreadsheet that is delivered across the organization. And now we have measures that cascade all the way down to the associate level and are part of annual performance reviews along with the subjective efforts in ensuring that an employee is being evaluated appropriately. There’s that.

We also support and deliver a monthly scorecard. We measure everything, and our department is responsible to ensure that we’re taking a balanced approach in measuring our performance; but our team always gets deep, if you will, around quality, safety, and engagement. We provide the further drill-down analysis and the understanding. So I would say very traditional business intelligence work would be a function of the team.

Now, as we became an accountable care organization back in 2011/2012, there was a need clearly to rethink how we were assessing performance. It was no longer, how are we doing on our acute care measures, and separate from that, our clinical integration measures, and separate from that, our home care measures? The need very clearly became how did we perform and deliver care across the care continuum to these lives that we’re accountable to?

It was clear at that time that our data was not structured that way. As part of the Center for Health Information Services before our ACO efforts, we had developed an enterprise data warehouse. It had been incredibly successful and it was very clinically-focused; it wasn’t financially-focused. It helped the organization — and certainly our team — to better understand drivers of outcomes and performance and allow that drill-down in a very easy way, but it didn’t offer the insight that was needed to look across care settings. We knew we had to rethink this in a lot of different ways.

We partnered heavily with the Cerner Corporation at that time and we developed what we called the Advocate Cerner Collaborative. The goal of the collaborative was to start thinking through how we create a data platform, which would ultimately the foundation to an analytic platform, to help support these questions around accountable care efforts and outcomes that spanned the care continuum.

That collaborative actually has a team of 11. And we certainly helped guide the efforts on putting the data together, but much of that was really done both within our own IT team as well as Cerner’s IT team. Right now, we’re leveraging the platform, and I would call it a big data platform, but it’s not — it’s a Hadoop cluster. It’s not Google or Amazon certainly; but in healthcare, it’s quite a bit of data and information. We’re leveraging that with the team to provide some additional and advanced analytics to support questions around pop health needs, predicting readmission risk, and predicting patients that have the potential to be a high utilizer, things like that.

Gamble: Right. So it sounds like this has been in the works for a long time and as far as this need to provide really the full picture of data for the clinicians, it’s been an evolution for the center and for this partnership.

Esposito: I think it’s been an evolution for the organization, and we continue to evolve as an organization in terms of the need of information and data, but also our ability to consume it and provide it in a way that helps answer very key questions and helps support decision-making. So you’re absolutely right.

I think the evolution for the team has been this notion that we were very focused initially on the technical need to pull data together, and so there was a component of the team that was very focused on helping to direct that. We always had an analytic core, as you can imagine, but as this has become more mainstream, and we’ve built it out extensively — and it still needs to be maintained, but it’s nowhere near where it was four years ago in terms of identifying more and more data that needs to be incorporated — the team is very much focused on using it analytically. And so we’ve evolved even as a team, becoming less focused on the technical aspects of bringing data together and more so on ensuring that we’re using it in the way that provides insight to the organization.

The team also has two researchers that are clinicians by background, and so when we put together an algorithm or we think through the analytic components of things, what’s front and center is the need to ensure it is something that can be relevant to the front line, leveraged by the front line, and ultimately, understood. Because whatever we create, if it’s not consumed and if it’s not used, then we’ve just created an academic exercise and that’s not our goal. Our goal really is to ensure that what we’re doing makes a difference, both in the lives of the clinicians, and even more so, in the lives of patients.

Gamble: That touches on something else I wanted to ask. Getting that clinical input in there is a huge piece of the puzzle, and I’m sure it’s not easy because there are so many different scenarios and different situations. How do you approach it when there’s such a variety of what clinicians need?

Esposito: I think, first and foremost, you need to ask them. Sometimes there’s an assumption that we know better or perhaps a question is thrown out there and we’ll say, ‘We know what you mean by readmission,’ but you can’t make that assumption. You need to pull the people who will use your analysis or predictive model or whatever it is you’re doing, into a room, and very foundationally and fundamentally, ask, ‘what’s the problem? Help me understand it from your perspective and let’s really understand what’s driving readmissions.’

I use that example because that was our first visioning session. Any time we have a new project, we try to bring folks in the room to talk through it and really facilitate a discussion. Back in 2012 when we were thinking about creating a predictive model to identify readmission risk, we could have very easily taken the definition that CMS will use to penalize us with the handful of conditions they were looking at, but again, we’re a large accountable care organization. What we quickly understood from our clinicians was the need to ensure that we were thinking about all readmissions — not just acute myocardial infarction, heart failure, pneumonia, or whatever they were looking for at that time, and I know it has expanded — but thinking about how we ensure that the tools we’re creating would help manage any potential readmission not just one definition of one entity out there.

So we always go to the source. Much of what we do, absolutely, is thought of and driven by clinical input. We certainly will do various and numerous lit reviews, and so we go in understanding somewhat what the issues are, but that direction is 100 percent driven by our clinicians that we intend to use these tools. I think it’s an important aspect that we need to further understand as we continue to pivot toward a very consumer-facing need or consumer-engagement need — that clearly is an opportunity in population health, and I think in healthcare in general. So to take a page from that, I think we need to start listening to customers and consumers and patients and really having them help us define how we can best engage them in different ways.

Chapter 2

Related Posts:

Share Your Thoughts Cancel reply