Mapping one data set to another--such as SNOMED CT to ICD-10--is almost always a resource-intensive project requiring hands-on review and considerable knowledge about the source and target, according to a new report on how to maintain data integrity during the process. A lot can go wrong, the American Health Information Management Association (AHIMA) notes in its paper.
For example, SNOMED CT is a comprehensive clinical terminology that contains content for both human and veterinary medicine, and it's vital for maps to use the correct reference set to exclude non-human terms.
Managers of health information should ask vendors to identify the maps in all existing applications, with a keen eye to understanding and monitoring the content and how it is used.
Using maps for healthcare billing can lead to compliance issues or allegations of fraud if the map results in incorrect code submissions, the report points out.
Maps can create errors in drop-down lists by creating inappropriate or less-specific selections and computer-assisted encoding software can misrepresent the facts. Updates in one code set might not be translated to the other.
- Document the business rules employed in development of each map, including the applications in which it will be used, how the rules were developed and deployed in the workflow.
- Develop a process for testing the validity of the map and its ability to be repurposed. Integrity may be lost in a secondary use of the data.
- Create and implement a maintenance program. Source and target code sets may be updated, discontinued or undergo major version changes, which can affect integrity on either end.
LOINC, or Logical Observation Identifiers Names and Codes, a standard for identifying clinical information in electronic reports, and SNOMED are to be aligned in a 10-year collaboration between their founding standards organizations.
Brian Dixon, assistant professor of health informatics at Indiana University and research scientist with the Regenstrief Institute recently wrote that healthcare is moving beyond the days of merely hunting and gathering data to one in which troves of data can be mined using these code sets for new insights.
To learn more:
- find the report