Health Language Blog

5 Trends Driving Data Normalization

Posted on 10/13/14


Data normalization addresses a key problem in healthcare informatics: the lack of a single, universally accepted standard terminology that defines the meaning of every type of healthcare data. Solutions supporting data normalization use automated mapping to quickly match incompatible terminologies to a shared vocabulary. This technology helps healthcare organizations overcome semantic uncertainty and the ambiguity that arises when multiple terms describe the same concept.

Against this backdrop, data normalization seeks to map data coming from different sources -- and encoded in different reference terminologies and classification systems --  into standard terminologies. But what makes normalization a critical tool for the healthcare industry? Here are five trends currently driving the demand for semantic mapping:

1. Electronic Information Overload

Healthcare was among the last industries to move from paper-based business processes to digital technology and workflow. But the sector has migrated to electronic information systems in a big way in recent years. Initiatives such as the federal government’s Meaningful Use program have triggered a wave of electronic health record (EHR) technology adoption. The Centers for Disease Control and Prevention reports that the EHR penetration rate has climbed from 18 percent of office-based physicians in 2001 to 78 percent in 2013.

Unfortunately, electronic information has proliferated across incompatible healthcare IT (HIT) systems, each with their own way of representing clinical terms. Normalization has become a necessary tool for liberating data locked in localized and proprietary terminologies.

2. Interoperability is Required to Support New Delivery Models

Emerging healthcare business models such as Patient Centered Medical Homes (PCMHs) and Accountable Care Organizations (ACOs) aim to coordinate patient care across multiple providers. Those models rely on the ability to share patient data among participants in a PCMH, ACO or other groups collaborating on patient care. However, the use of disparate HIT systems among healthcare providers creates communication barriers and hinders the ability to coordinate care. The absence of a common clinical vocabulary is the root cause of this problem. To facilitate care coordination, disparate terminology lexicons must be normalized into standard terminologies.

3. Clinical Data Repositories

The “Big Data” trend impacts many industries and healthcare is no different. In healthcare organizations, big data takes the form of clinical data repositories (CDRs). CDRs serve as data warehouses that pull together patient-oriented health data from a variety of IT systems -- everything from lab reports to ICD-9 codes to patient demographic data from an electronic health record. Data fragmentation stemming from the diversity of data types and formats hinders the healthcare sector’s ability to leverage big data. Data normalization facilitates the data analytics task, semantically mapping between, for example, localized diagnosis codes and standardized terminologies such as ICD-9-CM.

4. Data Integration

Healthcare organizations need to integrate claims and clinical data to obtain a more accurate picture of health across various patient populations. This integration also supports models such as ACOs. The traditional separation of clinical and claims data makes bringing the two data sets together a daunting challenge. Clinical and claims systems have operated as distinct entities for decades, making their data holdings difficult to reconcile. Normalization can rationalize the differences between claims and clinical data, enabling providers and payers to match different terminologies and create a system of shared meaning.

5. Medication Error Reduction

Clinical decision support systems automate the process of checking for medication issues: drug-drug interactions, drug-allergy interactions, and contradictions among other pitfalls. Meaningful Use has made screening for adverse drug combinations one of its compliance criteria. But the same medication may be described in numerous ways,  complicating efforts to run checks on drugs. Medication terminologies in use include RxNorm, NDC, NDF-RT and MediSpan. So, organizations need to implement data normalization to run a thorough drug interaction check and avoid medication errors.

Looking for Help?

As you can see, normalization impacts a variety of healthcare business models, initiatives and technologies. How would a normalization solution help your healthcare operation? What are your key use cases? Leave your comments below.

data normalization

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.