Health Language Blog

Why It's Important to Automate the Terminology Mapping Process

Posted on 12/03/14


Communication among healthcare IT systems has reached a critical stage as providers and payers rollout more applications to improve patient outcomes and boost operational efficiencies.

Healthcare systems, and the data they house, have proliferated across the healthcare community -- and largely in isolation. Local departmental applications were typically not designed to interact with other systems at the semantic level.  This has become a key problem: the proliferation of local terminologies makes it difficult for systems to talk with each other and even harder for healthcare organizations to support the semantic interoperability needs of data warehousing, ACO, HIE, analytics, and population health initiatives.  In some cases, it can even present obstacles when attesting for Meaningful Use.

This terminology fragmentation impedes data sharing and aggregation. Take the case of two laboratory information systems that each use a different, local terminology. A healthcare organization that wants to trend lab results over time will find that task extremely difficult in light of the disparate terminologies.

What’s needed is the ability to map between local terminologies and standard terminologies that other systems can understand and interpret. Data normalization provides a way to map between otherwise incompatible terminologies. While normalization will help rationalize data, some health systems are still using time-consuming manual methods to make that happen. The mapping process, however, may be automated. Read on to find out how.

The Need To Automate

The job of preparing data so organizations can successfully mine it can take a considerable period of time. Advocate Health Care, for instance, took 18 months to “merge, clean and organize” patient data when it pursued a predictive analytics project, according to Modern Healthcare. That data came from a number of sources including insurance claims, demographic records and electronic health records.

Data handling operations take a long time to perform, underscoring the need for time-saving automated mapping. A health system whose terminology mapping effort is linked to a time-sensitive healthcare initiative may not have the luxury of months to get its data ready.

For example, some healthcare providers face an important deadline in 2015 with respect to Stage 2 Meaningful Use compliance. Medicare providers who began the program in 2013 may qualify for $8,000 incentive payments if they demonstrate a full year of compliance in 2015. And to receive those incentive payments, organizations must adopt such terminology standards as RxNorm for drugs and LOINC for labs. Many providers will need to map their data from local terminologies to those standards.

How To Automate

Technology can help health systems overcome the limitations of manual data mapping. Healthcare organizations should look for a data normalization solution that provides automated mapping to speed up translation from local terminologies to standards.

Automation and the resulting efficiency are critical factors given the vast amounts of data healthcare systems will need to exchange each day. Manual mapping and interpretation won’t work in a high-volume environment. Even though maps are always subject to human review, the goal is to automate as much of the mapping process as possible using automated algorithms.

Additionally, these algorithms must be callable in real time via Web services so that they can be seamlessly integrated into enterprise-class health IT systems. Such mapping algorithms can normalize full catalogs or distinct elements from cryptic and poorly maintained source data into standard terminologies.

As automated mapping algorithms are developed for multiple use cases and become more sophisticated (the ability to learn over time, for instance), the level of human review at the individual code level diminishes. As a consequence, workplace roles turn to the development and maintenance -- including quality control -- of maps for a variety of use cases and the development of algorithmic translation rules.

Automatic mapping is poised to play a central role in helping organizations translate local terminologies to standards-based coding schemas.

Is your healthcare system struggling with localized terminologies? Leave your comments below.

enterprise terminology management


Topics: data normalization

About the Author

Brian Diaz is the Senior Director of Strategy, Health Language, part of Wolters Kluwer, Health. Brian has over 17+ years of leading product and marketing teams for SaaS-based healthcare companies focused on interoperability, data quality, and diagnostic imaging. Brian has a computer engineering degree with the University of Minnesota.