Reference Data Management: The Cornerstone of Reliable Analytics

brianlevy-200By Brian Levy, M.D., Vice President of Global Clinical Operations, Wolters KluwerHealth Language
Twitter: @Wolters_Kluwer

There is a lot of discussion about “clean” data today in healthcare. Recent movements with health IT, interoperability and data exchange hold great promise in fueling the high-level analytics needed to drive performance improvement. Yet, healthcare organizations, by and large, struggle in their quest to deploy infrastructures that support complete, accurate capture of data and ongoing management of those assets.

Simply put, data remains one of healthcare’s greatest opportunities and challenges. As such, healthcare organizations are increasingly turning to reference data management (RDM) as a best-practice strategy for managing the growing volumes of clinical and claims data needed to successfully position within an evolving regulatory environment. Increased momentum with value-based care through such initiatives as the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA) and 21st Century Cures Act notably increase the urgency to act.

A function of Master Data Management, RDM provides the framework for organizing data around a central service. It enables the industry’s wide array of disparate healthcare terminologies and codes to be represented via a single source of truth. For instance, analyzing a population health cohort for heart failure requires that all representations of the condition be normalized across a multitude of IT systems within an integrated delivery network. These representations must then be mapped to an appropriate industry standard for clean, accurate capture of data—and reliable, meaningful analytics.

A comprehensive approach to RDM encompasses managing local and standard content, then mapping and grouping that content. Some practical examples include:

  • Centralized management of standard and local codes

Health IT infrastructures are filled with the need for standard terminologies—from clinical to billing to administrative. Healthcare stakeholders rely on code sets such as SNOMED, RxNorm, LOINC and ICD-10 to comply with regulatory requirements and improve information exchange. Because provider data warehouses assemble information from multiple sources, data must be normalized to these standards.

Providers and payers are also increasingly turning to RDM as a best practice for managing local or custom codes. For instance, a health system aggregating data from hundreds of clinics within its network may need to manage local concepts outside of industry standards for internal use. A payer may use the same strategy to run claims adjudication systems that rely on dozens of code sets.

  • Mapping of code sets

Healthcare organizations managing local codes that originate from disparate clinical information systems need to map these codes to standards when appropriate. For instance, a health information exchange receiving lab data from various member organizations will need to normalize all incoming codes to LOINC before information can be exchanged and analyzed in a meaningful way.

  • Grouping of code sets

Providers, payers and vendors are all turning to the creation of code groups to effectively address analytics and reporting needs. These custom groupings of codes help answer specific use cases unique to an organization’s goals. For instance, a provider may need to establish and update value sets for Clinical Quality Measures to support MIPS. On a more system-specific level, a healthcare vendor may leverage specific code groupings to drive application user interfaces.

Increasingly, health system IT departments find that manual efforts to mature RDM exhaust and consume resources in such a way that it is impossible to create a sustainable strategy. The good news is that advanced solutions exist in the industry that can automate and streamline the complexities of RDM by addressing the following:

  • Content—establish a single source of truth for all terminology-related maps, value sets, and code sets by accessing all updates from a web-based content portal.
  • Applications—enable interoperability and increase the quality of your analytics with support for mapping clinical and claims data, modeling custom content, managing value sets, and searching terminologies and code sets.
  • Web-based APIs—integrate reference data into existing data warehouses or analytics platforms by utilizing a suite of cloud-based APIs, which complements current IT and infrastructure investments.
  • Consulting Services—leverage a team of terminology experts consisting of medical informaticists, certified coders, pharmacists, and PhDs to elevate the quality of your analytics and data management programs.

To thrive in value-based care, healthcare organizations must gain a better mastery of big data and analytics. In particular, health IT professionals need to improve data exchange with industry stakeholders, acquire more complete and accurate patient data and perform advanced analytics. Without an REM strategy in place that addresses each of these key areas, it is difficult for organizations to achieve their overall population health or financial goals.