As inbound data is processed, it is first checked for quality standards.  Messages not meeting quality standards are logged and quarantined for review and reprocessing.  Typically, quarantined messages require additional codeset mappings to be added / registered.  Messages that move on for processing next go through codeset mapping, tagging of sensitive data and filtering of bad or incomplete data.  Every transaction that is successfully ingested by the FHIR CDR is logged in an audit trail.  If a transaction causes an error, it can be tracked and reviewed via reports / alerts.

Through the development of these tools, Outcome Healthcare is able to ensure that every message received from any given participant will be carefully processed and ingested with a full audit trail attached.

Data Quality reports are extracted from three sources in our Audit database:

• message_log – each row represents a file processed and includes JSON metrics about the conversion
• message_error_log – each row generally represents when minimum standards weren’t met to generate a distinct FHIR resource
• codeset_error_log – each row generally represents when there was an issue converting or looking up coded concepts.

Data Quality Reports are generated periodically or ad-hoc.  Our reports breakdown problematic sources, source message paths, codes and codeset, to name a few.

The following sections on this page outline several examples of data quality reports, available to our HIE Platform users.