Mar 9, 2026: Improvements to Reciprocity Data via Healthie Integration
We fixed a set of issues affecting reciprocity flows for our Healthie EHR Integration focused on CCDA quality, timely availability, and source attribution. The release includes both partnership work with Healthie to improve CCDAs and a set of Zus changes: improved ingestion, stricter attribution, corrupted-data cleanup, parsing improvements, and new monitoring to detect and prevent future failures. Together these changes improve the fidelity of shared clinical records and further strengthen the integration with Healthie's EHR.
Background
Customers and internal Zus support observed data-provenance problems stemming from malformed or delayed CCDAs retrieved from Healthie. Reprocessing those CCDAs inside Zus caused additional data loss or distortion, and incomplete provider attribution created ambiguity for receiving clinicians. The team agreed on a two-track approach: align expectations and fixes with Healthie, and harden Zus ingestion, processing, and monitoring.
What changed
Work with Healthie
- Identified and rectified issues with source system generation of CCDA documents that led to downstream confusion about data quality.
- Coordinated a change to ensure CCDAs are available quickly after encounters so Zus can pull them promptly.
Changes implemented by Zus
- Corrupted data cleanup
- Removed problematic Healthie records from the data lake and downstream relational marts.
- Validated deletion and confirmed no remaining bad rows.
- Targeted CCDA retrieval
- Implemented logic to pull CCDAs more reliably and promptly after encounters, reducing missing or delayed clinical documentation.
- Monitoring & safeguards
- Added reciprocity-quality monitoring to detect malformed documents, delayed ingestion, and attribution issues.
- Created alerts to accelerate detection of rising validation failures or suspension rates.
Impact
- More reliable delivery path for Healthie CCDAs.
- Improved data provenance for clinicians consuming reciprocity records.
- Better detection of future CCDA quality issues through monitoring and alerts.