Executive summary. Data-quality work moves faster when the problem is described by record, meaning, timing, source, and the decision it affected.
Definition
Data quality describes whether information is sufficiently accurate, complete, timely, consistent, and understandable for its intended use.
Describe the problem precisely
Instead of saying the feed is bad, identify the record, the field or concept, the expected meaning, the observed meaning, the date range, and the decision affected. That gives the other party something they can investigate.
Include examples with the minimum necessary information. One representative, properly handled example beats a broad complaint with no context.
Separate immediate workarounds from lasting fixes
A team may need a temporary review step to protect a workflow while the source issue is investigated. Document that workaround and its owner so it does not become invisible permanent work.
When the issue is resolved, test the outcome with the people who use the information. A technically corrected field can still be confusing in the real workflow.
Frequently asked questions
Who owns a data-quality issue?
Ownership is often shared. The source owner, the receiving workflow owner, and subject-matter experts each have a part, and naming one coordinator keeps the issue from being lost.
Referenced standards and further reading
- HL7 FHIR: Validating Resources ↗HL7 International
Related articles
What to expect in a health information exchange →