Analytics Focus

In our April 2015 Intelligence Report, healthcare leaders cited clinical and business data integration, EHR interoperability, and data accuracy as their top data-related analytics challenges over the next three years. HealthLeaders Media Council members discuss ways to improve analytics performance.

This article first appeared in the October 2015 issue of HealthLeaders magazine.

Linda Butler, MD
Vice President of Medical Affairs
and Chief Medical Officer
Rex Healthcare, Raleigh, NC

In our organization, we've seen many IT system changes over the past 18 months. We changed our materials management system to conform to other providers in the Triangle area, and we also switched both our revenue cycle and clinical EMRs to Epic.

One of our primary goals had been to have everything on one system in hopes that we would be able to run better reports and gain analytic insight in to costs and quality, but we found that this undertaking wasn't quite as easy as just making that transition. We've since realized that we needed a dedicated analytics team to properly extract the data, and then to figure out how to interpret it. Frankly, we're still on that journey.

There's a real challenge around data accuracy. When you're running a report, you need to put the right parameters in to get data that is meaningful. If you don't have the information entered correctly, then everything downstream will also be incorrect. We've found that when workers were inputting certain data fields, it was easy for them to make a mistake. Sometimes, they didn't understand the consequence of that mistake—the incorrect data feeds into other screens, into orders, and creates other data accuracy problems as well.

Marlene Weatherwax
Vice President and CFO
Columbus (IN)
Regional Health

We recently formed an accountable care organization with Schneck Medical Center in Seymour, Indiana, and are trying to clinically integrate all of our facilities. So far, both hospitals have their own separate electronic medical records, and the physician practices within our Columbus organization are also using an assortment of records systems. We've been trying to find a common data warehouse or repository that can track the clinical indicators that we want to focus on.

Our greatest challenge to improved data quality and analytics is that, currently, EMRs don't talk to each other. You have to implement something that interfaces with them and can accept information from disparate systems, and you have to work with various software vendors to get it, which is expensive to do.

I believe that improved analytics will enhance the care continuum. Our organization is starting this process with the employees of our own hospitals as we try to gather this information. As we see improvement among them, we're hoping we can go out and get other insurance companies and employers interested in using this network, as we'll have a proven track record in helping patients to manage most chronic diseases.

We're hoping that analytics will help us identify patients we see more frequently, or who would benefit from being paired up with a health coach to change some of their practices, and that we will be able to encourage them to live a healthier lifestyle.

Bobby Olm-Shipman
Vice President, Planning
and Project Management
Saint Luke's Health System
Kansas City, MS

The endgame is that we're all moving toward becoming value-based providers. I believe all of our analytics activities are really aiming toward various strategies to achieve that overarching goal.

Analytics and related programs fall in to three buckets: First, we're trying to get a better handle on our data across our enterprises. That includes across the continuum; we want data from hospitals, physician practices, other nonacute locations like ambulatory and retail centers. We're looking to get a good sense of data and a snapshot of what's going on clinically, financially, and around utilization from the entire continuum of care.

The second bucket is that of market intelligence. Historically, we have had a lot of good data relative to inpatient hospitals. But more and more, care is not delivered in the inpatient setting. To understand how many patients are seeking, say, diabetic care in an ambulatory setting or a physician office, that's becoming increasingly important as we craft our population health strategies.

Third—from an analytics perspective—is that we're really trying to turn the massive amounts of data into good useable, meaningful information in the right format that allows the execution of this overarching strategy of transformation. The goal is to use all of this information, which is fantastic, and get it to the right people so they can actually do something with it.

Bill Fenske
Chief Financial Officer
Rice Memorial Hospital
Willmar, MN

On interoperability: Organizationally, we're looking at how we can share clinical data with providers who may not be located at our physical location. As we send our patients out to tertiary care providers, we want to know they're ready to take care of the patient immediately upon arrival. I think one of the most important points of this process is to be able to improve the interoperability and accessibility of medical data as the patient moves across systems.

On resources: Interoperability is a significant challenge for a couple different reasons, and both are related to resources. The first is around having the proper tools to extract clinical data. The second is finding human resources with the ability and time to be able to learn to do that—and to be able to translate it into meaningful information.

On population health: I think there will be population health benefits of improved clinical data, but I believe those benefits are years down the road. But we've identified the need to integrate clinical and business data in analytics. What options might there be in terms of getting over further hurdles? How do we start working together to jointly understand clinical and business analytics related to a specific care delivery or population health model? It will be interesting to watch this develop.