“How do we measure the quality of our health care?” asked Laura Tellmann, Healthcare Informatics Executive at BJC HealthCare in St. Louis, speaking at the Enterprise Data World 2017 Conference. She shared a photo of a young boy who had survived a brain tumor and post-surgery stroke, as a reminder that the quality of the data in our health care systems can impact the quality of care patients receive. “It’s my belief that with better data, the fate of this young child would have been better.” Tellmann shared three key tactics she used to guide BJC toward better quality data through the creation of a Data Governance plan.
BJC Healthcare is one of largest employers in Missouri, encompassing 15 hospitals, an affiliation with Washington University School of Medicine, and a comprehensive network of care facilities, physician offices, and ancillary services.
Data Governance Defined
Working with a consultant from First San Francisco Partners, BJC created a working definition of Data Governance.
“For us, Data Governance is a formal program run by the business. It’s a set of policies, processes, and standards that ensure consistent nomenclature,” Tellmann said. “We want to improve our quality, reliability, we want to have good privacy and security, all in the goal of being able to use our data for our decision-making,”
Most of 2015 was organized around creating an operating model and setting a vision that arose out of that definition, which is: “Data Governance will guide the transformation of data and the trusted quality information that supports our decision and actions.”
The operating model includes a top level with C-Suite decision makers. The second level includes a group of trustees, “accountable for a grouping of data that we call a domain,” she said. Trustees meet two hours a month, and are charged with building steward communities, writing policies, establishing standards, and sharing best practices, she said. Trustees are chosen from data managers as well as from major business areas. At the bottom of the model, “doing the majority of the work, are the steward communities.”
In 2016, both BJC and Washington University were introducing a brand new electronic health record system, she said. “We wanted to make sure that the foundational pieces of data going into that electronic health record were of the highest quality we could get.” BJC set an aggressive agenda for the lead up to the new system: to identify and train Data Stewards, and then create Data Steward communities that would build business glossaries, standard reference data, and preferred data sources for six Master Data domains.
Tellmann identified three tactics that she used to help them accomplish their agenda.
Tactic 1: Use a Guide
Using the Data Management Book of Knowledge (DMBOK) wheel, The Data Management Maturity Model from the CMMI Institute, the ten steps from the Data Governance Institute Data Governance Framework, and the First San Francisco Partners Data Governance Framework, she built a maturity model. That model was used to create a set of strategic objectives and a matrix to measure the maturity of the data associated with each of those objectives, she said.
“In our planning sessions, we essentially had two activities. The first was to take all the strategic objectives that we believed would exist in 2017, and then figure out what data is associated with those strategic activities. Then we took each one of those pieces of data and placed it on a maturity matrix where we were measuring how important that data was to that strategic objective and how well-managed we felt the data was today.”
Through the planning process they discovered “a nice little grouping of data domains that we felt were not well-governed, and were strategic,” she said. They asked each trustee independently, “Where they thought we were on the maturity curve and where they thought we needed to be in five years,” identifying six of those domains and deciding to “get them from what we call ‘organize’ to ‘operate,’ so that they’re working independent of my group, and can do Data Governance on their own.”
Tactic 2: Getting Executive-Level Support
“[Getting support from C-Suite executives] was something that when I was getting started, I heard over and over and over again, and read it, and read it, and read it,” she said. “OK, I know why they want me to do this, but how am I going to get there?”
Starting with Tellmann’s boss, the Chief Clinical Officer, she started systematically getting buy-in from key corporate leaders. The CCO “was an easy sell because he was currently responsible for all that regulatory reporting, and he saw the landscape, saying that it was going to get worse.”
In healthcare today, she said, most regulatory reporting is transcribed,
“People type it into websites. We’re moving to being dependent on this electronic data, and it was not hard to show that our electronic data was not as accurate as that abstracted information.”
She then used the next six months helping,
“Key members of the organization understand where regulatory requirements were taking us, why our electronic data was bad, and why it wasn’t just a simple matter of telling somebody to get the job done.”
Once the C-Suite was on board, the first project Tellmann brought them involved building a glossary mapped with a Data Dictionary, and establishing reference data standards for two electronic clinical quality measures.
“They thought it was going to end up being a pretty easy project.”
She set up a monthly meeting with the CQO and CIO, both sponsors:
“And what I like to say is, we failed miserably, but because I had that monthly voice with those people, and could actually bring to them very specific examples of why I couldn’t get decisions made because it was nobody’s decision to make, and why the problems that we were having were systemic, so there were no policies or processes, or standards in place for many of this. That very quickly elevated that project to a program.”
Because of Tellmann’s efforts, leadership began to see the gravity of the situation and established that the program needed to report directly to the C-Suite, she said, “And now I have that audience.”
Secondly, “Everyone was fearful that we were trying to grow another big organization. That was one of the main reasons that our C-Suite resisted any belief that Data Governance was something that we should do,” so she made it clear that the priority was to train and facilitate Data Governance throughout the organization, not get someone from outside to do the work. “We’re training the organizations, so be clear that the resources that you’re going to be using are going to come from your existing workforce.”
Tactics to Increase Engagement
Tellmann got leadership involved in the process of implementation by ensuring that at each meeting, “We showed them something we had done, we made them make a decision, and we showed them an issue that we were going to come back with that would again need another decision,” she said. “That really helped get their engagement.”
She stressed that it was also important to teach the language used in Data Governance, and encouraged the team to feel comfortable using the terminology and asking questions, a process that took about six months. “My C-Suite now says ‘model.’ They now say, ‘data maturity’ and ‘data as an asset,’” which supports the feeling that they are a full contributor to the process.
Giving the team the responsibility for picking the Master Data Management trustees was another way Tellmann further increased the level of engagement, lowering potential resistance by providing resources and information as the conversation evolved over time.
“We showed them clear roles and responsibilities for the trustees. That got them a little bit; they were like, ‘Okay, I kind of know what you want, but I’m still not really clear.'” She then developed clear selection criteria as well as bringing in data gleaned from interviews with ten similar healthcare systems. This showed the team who might be suitable for those roles, telling them, “This is what other people are doing that are trying to manage their data. These are the types of people that they’re putting in these roles,’ and that got them a little more engaged.” She stressed to the team how important it was to find people who were “good with culture change and who could effectively weigh in on decisions and make sure that they support the transparency that we were trying to build around data,” she said. “And then we made suggestions.”
Executives still assumed that the implementation of the electronic record keeping system would solve many of the problems they faced, so Tellmann presented them with a clear representation of issues with and limitations of the existing record keeping situation.
“We had seven different independent groups trying to manage identities for patients across our organization. That’s really hard – and on top of that, we had put in three different technologies that were all master patient index technologies,” she said. “So they saw, for the first time, that this major investment they had made in this electronic health record was not going to solve this problem, and once they saw this, they realized that we needed to make an organizational change.”
At that point, she said, “The selection of that trustee suddenly became very important, and that’s when I got their buy-in. That’s when I actually heard my CEO go, ‘Wait, so this person’s going to actually have a lot of power,’ and I said, ‘Now, he gets it.’”
Tactic 3: Metadata Management
“It’s important to develop a repeatable process around your Metadata Management,” she said, to ensure that the project is solving an immediate need and is within scope. It’s also important to make a plan to transition that project to the operating stage as soon as possible, “So that you can get the people identified that are going to do that work” functioning independently.
Once you’ve identified the right people, she said, ground them in basic technology and teach them the terminology. “We never start them with a blank slate. Every time [we] create a vocabulary term or start some sort of mapping, we make sure that they clearly understand what the industry already calls things, and what our data dictionaries already call things.”
Be clear about the scope of your projects, she said. Tellmann described a seemingly simple scenario where they were to identify only the source systems that originated data for 27 critical elements, yet the decisions that had to be made to determine which records were included made it a very complex question.\
“It was very helpful for our people to see things in the frame. ‘This is what we are going to do,’ and then out of the frame: ‘This is what we’re not going to do.’ And it’s not that you’re never going to do it, it’s just that it’s not part of what we’re doing right now.”
Tellmann showed another photo of the young boy from the start of her presentation and shared that he is her son, Daniel.
“That young man I showed at the beginning now has seven electronic health records, just for two of his providers. If I want to see what just two of his providers are doing with him, I have to log in to seven different electronic health records and try to piece that all together. So, it’s pretty clear that we’re not getting the right information to the right providers in any sort of timely manner. We know we can do better, and Data Governance is a huge part of that mission.”
Photo Credit: Andy Dean Photography/Shutterstock.com
Here is the video of the Enterprise Data World 2017 Presentation. Her presentation was titled “Data Governance Tactics”:
Check out Enterprise Data World at www.enterprisedataworld.com
