Apr 12 2023

Q&A: Banner Health’s Journey Toward a Unified Data Model

Jennifer Brooks, vice president of population health, value-based care and administrative services organization at Banner Health, discusses the group’s data evolution.

Turning healthcare’s vast oceans of data into actionable insights remains a top priority for leaders. A majority of organizations view analytics as critical to achieving broader strategic goals, and about 51 percent of executives say that data integration and interoperability are the most significant barriers, according to a 2021 Sage Growth Partners report.

As one of the largest health systems in the U.S., Phoenix-based Banner Health is wrangling its data to deliver high-quality care across six states and integrate services for members as a payer. That has led the organization on a journey toward a unified data model, connecting patient data across numerous systems and finding efficiencies in IT operations. 

Jennifer Brooks, vice president of population health, value-based care and administrative services organization at Banner Health, spoke with HealthTech ahead of her appearance at HIMSS 2023, where she’ll share lessons learned by the group so far.

MDP Sidebar


HEALTHTECH: How has healthcare data evolved over the past decade, and how are healthcare organizations adapting? Where do challenges remain for turning insights into action?

BROOKS: Healthcare has been reliant on data and information that is captured in our electronic health records, and EHRs were not traditionally built to capture data and information in a way that could be easily analyzed and then used to act. The platforms that we were so entrenched in using really weren’t designed to do the level of on-demand analytics that you need for it to be actionable, but it’s what we had.

I think that’s still the challenge that we as an industry continue to go through. I think that’s why you see other health systems moving to use Amazon’s or Google’s analytics services, because mastering Big Data is not a healthcare core competency. But we’re learning that it’s a core competency that we need, and there are other industries that do a far better job than we are doing. I think we’re in that learning curve of trying to understand how to leverage that information. But also, how are we different? Because we’re not just talking about taking retail analytics to try to prompt a consumer to buy more things on Amazon, for example, which works really well. That’s not exactly what we want to do with healthcare data.

There are elements of consumerism that we have to pay attention to in healthcare. We need to make sure we’re delivering the right things for our consumers, but we can’t just take the retail consumerism platforms and layer them on top of healthcare analytics. So, what can we use from that space? What makes sense for healthcare? How do you protect your data? We still have to understand those boundaries.

Click the banner below to learn how a modern data platform supports decision making.

HEALTHTECH: How did Banner Health began its journey toward a unified data model? How did the health system recognize it as an area that needed improvement?

BROOKS: Banner Health has been on a journey for a while in this space, trying to understand our needs and complexities. We are a delivery system at our core, but we’re also an insurance payer. We’re moving into that vertical integration space of being care and coverage.

In that space, the need for unified data surfaces very fast, because you see where we’re providing a disconnected service between care and coverage for healthcare consumers. In order for us to be successful as a vertically integrated organization, we have to figure out how you take insurance information and translate that into care information and vice versa.

Without a unified data model, we’re basically operating as two different entities, and that’s no different than any other healthcare organization out there. We knew we had to go down this path. We’re definitely still learning, because it takes time. We’re big and complicated, as a lot of organizations our size are, and we’re still learning through it. What we traditionally had isn’t going to work with our vision to truly bring care and coverage together.

READ MORE: Three key considerations for building a more modern, agile data platform.

HEALTHTECH: Why did Banner Health decide to rely on an integrated, comprehensive health cloud? What are some challenges you’ve faced?

BROOKS: We were traditionally operating in a space where we were using platforms designed for care and trying to scale that to be an insurance platform to manage risk and value-based care, which frankly just doesn’t work. After trying that for a while, the realization that it wasn’t scalable was definitely staring us in the face.

What has gotten better is that we’re understanding which core platforms are related to coverage, which are related to care, and how we’re bringing those two things together to really understand how to take that information and make it actionable. Bringing care and coverage together is our North Star. We’re fully committed to this, and we’re figuring out the roadmap.

Previously, it felt like one side or the other was getting sacrificed without this unified data model. There are still times that we’ll hear that the data is not updated enough or that it’s not 100 percent reflective of the care a clinician delivered two hours ago. Those issues still exist, but we have a clearer pathway to success.

Jennifer Brooks

Jennifer Brooks, Banner Health’s Vice President of Population Health, Value-Based Care and Administrative Services Organization, shared more about the health system’s data journey at HIMSS23 in April. Photography by Steve Craft

HEALTHTECH: How do you think this journey toward a unified data model will impact the future of care delivery at your organization? How will emerging solutions such as artificial intelligence and machine learning, for example, be supported?

BROOKS: We have to do a better job of meeting our patients and our members where they need us most and then helping to navigate them to care. Without AI, you can’t figure that out. So, that’s really where we’re trying to leverage those components of AI and machine learning. We’ve been doing elements of machine learning within the delivery setting for a while, which I wouldn’t say is necessarily new for healthcare, but using AI to connect the continuum of care together, that’s definitely new, and that’s where we’re trying to bring it together more cohesively. We’re trying to use it to help us predict and understand needs, and then help navigate our patients and members to the right services, instead of the traditional model of sitting back and waiting for them to come to us.

HEALTHTECH: What has feedback been from clinical, IT or patient stakeholders so far?

BROOKS: The feedback that we’ve gotten from some of the clinicians has been, for example, if you delivered care an hour ago, it’s not going to be reflected in the holistic view of the member right now. By tomorrow morning, it will be, but it’s not right now. We’ve gotten feedback that it’s not happening in real time, but it is closer than it has been.

We’re also hearing from clinicians that they’ve learned something by looking at that holistic patient view, for instance, about a medication that someone else prescribed in a different state that their patient never told them they were on, and it was causing adverse reactions, and the only way the clinician knew about it was because they had a holistic, 360-view of the member that they couldn’t see in their own EHR. So, people are seeing the value, and there are pockets of really insightful members stories where we’re directly helping to improve care or catching things before they become an issue because of this integrated data model. But it’s not perfect yet, so we’re still juggling that. How do we make it better, faster and more immediate, more integrated? I think that’s realistically where we’ll be for the next five years.

Click the banner below for more insights on the growing importance of data in healthcare.

HEALTHTECH: Can you talk more about your upcoming HIMSS 2023 session? What do you hope to learn at the conference?

BROOKS: You’ll hear anecdotes about how moving toward a unified data model is and isn’t working, and about the pain points that we’re still working through as a large organization and as an industry as a whole. You won’t hear us say that we’re perfect. You’ll hear real-life issues that we’re still dealing with. And you’ll hear that we’ve tried and failed and kept trying, and I think that’s a really important message for people to hear. This doesn’t happen overnight. It’s OK to try and not hit all the targets that you were hoping to hit, as long as you just keep going back to the drawing board and reiterating, which has definitely been our journey.

I tell people all the time that, to work in this space, you have to love it because you’re going to sometimes fail more than you succeed. You have to just wake up and passionately believe in what you’re trying to achieve. You’re going to fail. Be comfortable with that and keep trying.

What I hope to learn at the conference is the same lessons I’ve learned from others. What is most valuable for me and my colleagues is just hearing from others, what they’ve tried, because we’re all in this together and we’re stronger when we learn from each other. I love hearing from other people. They always give me ideas and make me think of things I’ve never thought about, so that’s what I’m looking forward to the most.

DIVE DEEPER: Understand the importance of having a modern data platform in healthcare.

HEALTHTECH: What advice do you have for healthcare organizations that are still struggling to create a unified data model? What are some steps that they can take now?

BROOKS: Understand what that unified data model will deliver for you. Stay focused on that as your North Star, and then build your processes to support that.

You’ll probably go through a variety of partners to help you achieve that, and that’s OK. Sometimes, we’d like to think that you could pick one partner and get there, but I’m not sure we as an industry are there yet. If I could rewind to myself two years ago, I would tell myself that.

Be OK with trying and failing, and just keep working on it. It’s complex. We had a data architecture that wasn’t designed to support what we’re trying to achieve now. It’s going to take time to build it. It’s not going to happen overnight. I tried to do that; it didn’t work.

As an initial touchpoint, we used what we currently had and tried to make it deliver on some low-hanging fruit, some very clear deliverables. Let yourself fail fast in that space, which is what we did. We looked at what we could achieve with what we had and we quickly figured out what couldn’t work, which allowed us to pivot and go to market.

Sometimes, organizations like to boil the ocean or eat the elephant, or whatever analogy you want to use. My advice: Don’t try to do that. Just pick the one thing that you want to try to achieve that you haven’t been able to, see if your current model and structure can get you there, and if it can’t, have a fast go-to-market and find a solution that can get you there.

venimo/Getty Images

Become an Insider

Unlock white papers, personalized recommendations and other premium content for an in-depth look at evolving IT