Taking health care data analytics to the next level
A holistic view from an enterprise perspective
In this interview, Christine Santos, chief of strategic business analytics for Providence Health Services and Chris DeBeer, principal at Deloitte Consulting LLP explain the value of enterprise data analytics.
The evolution of health care analytics
There’s been a lot of buzz around “analytics” in health care in recent years; how would you describe the current state?
Chris DeBeer: As the industry shifts its focus from volume to value, there is increasing emphasis on creating long-term views of patient populations. This is especially evident in the bundled payment programs and regulations coming from the Centers for Medicare & Medicaid Services (CMS), such as those related to total joint replacements and other disease states. Robust analytics are key to defining populations, understanding the care that’s provided to these groups, and determining variations across practitioners, physicians, facilities, or regions.
Despite the need for strong analytics, many organizations often struggle in this area, in part due to the sheer volume of available data. Since hospitals and health systems have adopted electronic health records (EHRs), there has been a data explosion. Unfortunately, as volume has gone up, access to meaningful information on which to make sound decisions has dropped because organizations are unable to fully appreciate the data they have.
Christine Santos: Although EHRs provide an opportunity to capture more data at the point of service, this doesn’t necessarily translate into more information and greater collaboration. In addition, the shift from fee for service to value requires additional data sources and capabilities. The result is that most organizations are data rich but information poor.
What effect do you think industry consolidation has had on analytics?
Santos: Aggressive merger and acquisition activity has led to a significant number of large health systems having multiple EHR and financial solutions, yielding disparate data that takes time and resources to normalize and translate into useable information. At Providence Health Services, we often spent 75 percent of our time collecting, manipulating and normalizing data to get to a point where we could analyze the data to transform to insight and then ultimately actions. Eventually, we began consolidating our various systems into enterprise solutions, starting with our EHR system.
The next evolution of our analytics roadmap is to work on the integration of these new “enterprise” standards and ensure that data and insights are leveraged across multiple cohorts (finance, quality, supply chain, etc.) to truly understand the levers that affect our enterprise performance. We also want to shift from descriptive analytics (what happened, where and why) to predictive and prescriptive analytics (to model what will happen next and what’s the best that can happen).
DeBeer: Health systems pursuing mergers and acquisitions tend to underinvest in analytics capabilities—focusing more on the semantics of combining operations. However, it’s becoming increasingly apparent that health systems have to go beyond just keeping the lights on—they must be able to get the value out of an acquisition. In health care, bigger has not necessarily led to better; it’s just led to bigger. With the growth of data that comes from mergers and acquisitions, there are so many versions of the truth that it’s hard to make informed decisions.
What is the role of the finance department in the evolution of analytics?
Santos: The role of the finance department is shifting from managing the organization’s day-to-day finance activities (i.e. accounting, payroll, budgeting, etc.) to becoming a strategic thought partner in providing the right insights at the right time to support organizational decision making. Finance is expected to understand the rhythm of the business and have a current pulse on the organization. At Providence, we have recently deployed our first wave of an enterprise cost accounting and decision support system, which allows us to bring together multiple data sources and generate a single source of truth regarding the cost and margin of our health care services. We have been able to deploy this solution on a large scale—implementing it in 33 hospitals across five states. Before this initiative, we had multiple data sources, and it was primarily the finance department that could access the information.
Now, we have democratized the data and delivered a more self-service analytics tool. As a result, people across functions (finance, quality, operations, supply chain, and strategy) can immediately get access to information and gain their own insights rather than relying on finance as the gatekeeper to information. As such, finance can shift their decision support resources from providing data extracts and reports to becoming analytical consultants.
DeBeer: Although generating an accurate per capita cost was the key financial driver behind this project, the indirect benefits that have surfaced are pretty amazing. For instance, clinical leaders can look at care pattern variation and quantify the cost differences across various facilities and between different providers. Also, Providence can review service line profitability and understand where best practices exist and where it can make improvements. So, even though finance drove the initiative, the level of data granularity has benefited many stakeholder groups and allowed finance to become a business partner rather than a scorekeeper.
For most organizations, analytics investments have tended to be siloed with clinical, operational, and financial leaders only concentrating on their pieces of the pie. Getting a consolidated view that lets leaders leverage information in all three domains is critically important. These types of projects are the ones that are getting funded because they lead organizations toward a single version of the truth that yields great insight.
What is driving investments in analytics?
DeBeer: Instead of arguing over whose answer is right or whose truth is more accurate, health care organizations are becoming more interested in having systems with a high degree of transparency across multiple functions. They want to be able to trust in the information and receive it in a timely fashion without having to spend critical resources sorting through disparate systems, data, and semantics.
Organizations are also considering their missions and goals when prioritizing analytics investments. For example, academic medical centers may emphasize investing in research and development analytics because R&D is fundamental to their success. However, large, regional or national faith-based health systems may choose to prioritize other areas because their missions are different than those of academic medical centers. Perhaps a faith-based organization is most interested in managing its facilities to subsidize ones that are serving underserved communities. In the end, there is no one right answer on what should be the top priority—it should be a customized decision.
What advice would you offer to organizations launching analytics initiatives?
Santos: Look at it holistically from an enterprise perspective, and establish a realistic vision for analytics. Then, it is important to prioritize the analytic capabilities that will provide the most value to your organization both from a margin and mission perspective. The key differentiator for our analytics initiative was that although finance focused on cost accounting, it was an enterprise initiative. To drive organizational change and adoption, it was critical to have leaders across the organization see the value in having an integrated, coordinated, and collaborative approach to analytics. Buy-in not only from the top down but among multiple cross-functional areas (quality, finance, operations, and strategy) was the key to success. The executive committee included our CFO and CIO, leaders from our hospitals and medical groups, clinicians, and population health. We made sure that all leaders that affect the Triple Aim were a part of the dialogue, so we could implement a functional and accessible analytics solution that would address everyone’s needs, vertically and horizontally.