Leaders at One Community Hospital-Based System Leverage Analytics to Improve Outcomes

Leaders at the two-community-hospital, 320-bed Southeast Georgia Health System, based in Brunswick, and with a facility in Brunswick and one in nearby St. Marys, have been moving forward in leveraging business intelligence tools to fuel readmissions reduction and quality improvement work. They have been partnering with the Greenville, South Carolina-based Quantros in their ongoing work. Among the leaders of the current initiative are Christopher Pavlo and Jerry Roe, who are both data analysts at the SGHS health system.

Recently, Pavlo and Roe spoke with Healthcare Innovation Editor-in-Chief Mark Hagland regarding their current initiative. Below are excerpts from that interview.

Tell me about the overall quality initiative that you’re leveraging data analytics for?

Jerry Roe: There are certain measures we have to comply with, for readmissions, mortality, AMI, etc., for CMS [the federal Centers for Medicare & Medicaid Services]. Quantros does an excellent job when we run these reports. We’ve built multiple reports for each of those measures. One example is around sepsis, which of course continues to be an issue across the U.S. healthcare industry. Yet it’s a measure that it’s hard to get a handle on. With Quantros, we’re able to get the data. Chris and I worked together, and went from zero to 100, and said, what can we do with this data to show where we’re at? We’ve developed our own database, and have created our own dashboards, for nine measures.

What are the nine measures involved?

The nine measures are all-cause readmissions; all-cause mortality; AMI; CHF [congestive heart failure]; COPD [chronic obstructive pulmonary disease]; pneumonia; sepsis; stroke; hip and knee replacement; and diabetes. We’ve been analyzing readmissions and morality around those areas. Quantros gives us the ability to look at every single inpatient coming to Southeast Georgia. We look at data across all patients, and also stratified to Medicare only.

What has the process been like so far?

Nobody in the organization had been using the tool for what it could be used for. It was a lot of manual work, as far as looking at the data, to making sure they’re getting the right numbers to share with CMS or for internal reporting. When Chris and I started running reports, we asked, what can we do to really prepare useful, meaningful dashboards, ones that show how we’re really performing?

When did you begin the process of reengineering that?

Christopher Pavlo: We started about a year ago. Previously, the process had been very rudimentary and non-standardized, with crude readmissions reports. Now, we can write these reports, consistent data. And we can risk-adjust for secondary diagnoses. For example, if a COPD patient comes in with a history of diabetes, etc., a readmission wouldn’t be unexpected. Quantros allows us to get a better understanding of what our patient population looks like.

Roe: Also, we took this a step further, with Chris’s experience on the quality side and mine on the IT side—we decided to make this more visual, as far as more bar graphs. Now, it’s easier for senior executives and leaders to look at these statistics. We can look at sepsis, for example, to help us examine what’s behind these measures, in order to try to drill down on issues.

You’re able to better drill down to see some of the contributing factors to readmissions and mortality?

Pavlo: That’s correct. And, in addition to the value of the data, we can look at individual physicians, individual days of the week, any number of factors. And now, these reports can be run in a matter of minutes.

Roe: I’m running our dashboards, and today, within 30 seconds to a minute and a half, each one of the reports was done. And I’ve come across many, many software packages in both manufacturing and healthcare. This is one of the easiest products to run reports on. We know that this data is coming out of our IT department, and going to the vendor, for them to help display the data for us.

How are you applying some of the data?

Pavlo: What a lot of hospitals are struggling with today is, how is CMS judging us? We started with mortality and readmissions, because they’re key elements, are high-dollar, high-impact programs within CMS. The benefit of working with our vendor partner is that if numbers don’t match exactly, we can see what’s happening with regard to readmissions back to our facility, while CMS looks at readmissions to all facilities. We’ve validated that our vendor partner is doing a very good job of predicting readmissions rates to our facility, where CMS is judging us. And Hospital Compare runs 2014 through 2017 data; it’s very old. But now, we’re looking at readmissions that are a couple of months old, versus the reports that CMS provides us, that are two to three years old.

Roe: Yes, so we can drill down and figure out what is causing patients to be readmitted. By having much newer data, it allows us to be much more proactive, to eliminate some of those readmissions.

Are you able to share some of this data with physicians at the individual level?

Roe: Absolutely. The nice thing about our vendor partner—it’s not just these six measures—is that if the orthopedic department or cardiology department wanted us to run data, we could run data for them on length of stay or along any other dimension. We can help them with issues in their departments.

Pavlo: yes, and we’re starting to get traction with certain specialties. We’ve gained credibility with this product, and with physicians wanting to dig further down into the data.

Roe: Being able to run reports on specific physicians and specific service lines, really helps the physicians to talk among themselves, around what they might be able to do, to help their patients. So this is great for not only the hospital and the facility, but for the individual physicians as well.

What have been the biggest lessons learned so far, in leveraging data for these purposes?

Pavlo: I think one of the biggest lessons learned was just getting confidence… We had to show the data to the key players in the organization, and gain their confidence that what we were presenting was valid. These reports have a ton more of credibility. And by showing the overall trend of our Hospital Compare CMS data and overlaying that with the Quantros data, and things matched up, and we can trend further out by 16-18 months. So, getting confidence with the c-suite and physicians, these reports gain credibility.

Roe: Chris is absolutely dead-on with that. The thing that was really hard to do with a lot of companies, was to trust in the data. And by being able to validate the data and project it—that’s where all the visualization helps, not just to see physical numbers, but to see bar graphs, by pulling all that in together, we were able to get the executive staff and quality department, and multiple departments on board with the data, because they know and trust the data, and now, when we run the data, it’s trusted.