Health Language Blog

How to Overcome Semantic Interoperability Hurdles

Posted on 12/17/14

semantic_interoperability_hurdles_(1)

Semantic interoperability spans the communications gap among divergent health IT (HIT) systems and data sources.

Electronic data and HIT systems have proliferated across the healthcare sector in recent years. Unfortunately, HIT innovation has developed in isolated pockets, with initial development taking place in billing and claims, followed by localized development of ancillary clinical systems such as laboratory, radiology, and pharmacy, and, more recently, electronic health record (EHR) systems.

The federal Meaningful Use initiative provides some relief, offering incentives for providers to adopt terminology standards such as SNOMED CT,  LOINC and RxNorm. Yet many HIT systems in today’s market still use local or proprietary codes to collect and store patient data. The plethora of differing coding and terminology schemes can introduce errors when healthcare organizations attempt to share and leverage data in such a fragmented environment. For example, a June 2014 report published by the Journal of the American Medical Informatics Association noted “615 observations of errors and data expression variation” across the 21 EHR technologies examined.

Errors aren’t the only problem. Interoperability issues hinder a healthcare organization's ability to aggregate and share data, capabilities which are pivotal for the success of emerging care delivery models such as Accountable Care Organizations (ACOs).

Mapping The Way To Interoperability

In short, the lack of semantic interoperability creates significant communications hurdles. The key to overcoming those hurdles is through a data normalization solution. Data normalization provides terminology maps built by informaticists with the clinical domain expertise, which can enable semantic interoperability through a common and shared vocabulary. For instance, a mapping system can take differing codes for the anti-inflammatory drug Naproxen -- which could be described as Naproxen Tab 250 MG or Naproxen 250mg tablet among other labels -- to a single code under the RxNorm standard. Systems that rely on the RxNorm standard can share data more readily if the drug data is properly normalized.

In addition, a structured approach to mapping terminology can help an organization navigate the multitude of content sets present in a variety of clinical applications and HIT systems. Sets in use today include SNOMED CT, ICD-9, ICD-10, LOINC, CPT, RxNorm, HCPCS and DSM-5. Mapping among those content sets -- and proprietary codes as well -- helps support analytics and reporting initiatives by breaking down semantic barriers. 

In terms of use cases, mapping can help healthcare providers obtain a more comprehensive view of patient data for quality reporting purposes. ICD-9 and ICD-10, for example, are frequently used in situations where data aggregation is helpful, such as measuring quality, monitoring resource utilization, or processing claims for reimbursement. SNOMED CT, meanwhile, is geared toward codifying clinical information captured in an EHR during an encounter. However, an ACO that wants to calculate a particular quality measure may need to pull together claims data encoded in ICD-9 and clinical data encoded in SNOMED CT. With terminology mapping, the ACO can semantically translate between the different standards and compute its quality measure.

Supporting Different Models

ACOs aren’t the only entities to benefit from overcoming semantical hurdles. Patient-centered medical homes (PCMHs) also rely on the exchange of data among primary care physicians, specialists, hospitals and other providers. Health information exchanges (HIEs), whether localized or regional in scope, must be able to move clinical data among the participating parties. And population health management initiatives require a robust data analysis infrastructure that can pull in anonymized patient data from myriad healthcare systems. All of those healthcare efforts can take advantage of terminology mapping.

What are your main semantic interoperability challenges? How do you plan to address them? Leave your comments below.

The Single Source of Terminology Truth

 

Topics: semantic interoperability

About the Author

Brian Diaz is the Director of Integrated Solutions with Health Language, part of Wolters Kluwer Health. When not working, Brian is soaking up the Colorado experience with his family but still cheers on the Golden Gophers.