Data platforms
Terminology and Patient Identity Services in Healthcare Data Platforms
Every count a data platform publishes rests on two answers no source system can give it: whether two records describe one patient, and whether two codes mean the same thing. This paper is written for the organisation that runs both services. It follows one CMS measure through both failures, sets out the FHIR terminology operations with real exchanges, the US, Canadian and international code systems and their terms, servers and mapping errors, how a release changes a value set by itself, the kinds of patient matching, thresholds and the review queue worked exactly, privacy-preserving linkage and its costs, the identifier and privacy rules in the US, Canada, the EU and England, uneven linkage, and how to compare products and test both services. Written from the specifications, the regulators and the published studies, current to October 2026.
What’s inside
- Two failures in one published measure: CMS’s Hybrid readmission measure, with CIHI’s indicator for comparison
- Seven FHIR terminology operations, each with a dated exchange against HL7’s reference server
- The code systems a US, Canadian or international platform carries, with each steward’s terms
- One value set under two SNOMED CT releases: 133 concepts, then 124, with no edits
- Terminology servers on what can be checked, and where LOINC mappings go wrong
- Duplicates, overlaps, overlays, and how a merge travels
- Thresholds and the review queue, worked exactly for stated inputs
- Privacy-preserving linkage, the rules on tokens, and what hardening costs
- Identifier and privacy rules in the US, Canada, the EU and England
- Uneven linkage, the quoted figures traced, and how to test both services on your own data
Request this resource
Related insights
- Article · Data platforms
Terminology and Patient Identity Services in Healthcare Data Platforms
Every count a platform publishes rests on two answers: whether two records are one patient, and whether two codes mean the same thing. What each service does, and what it leaves to you.
· 19 min read - Article · Data platforms
What a Healthcare Data Platform Is Made Of
The same two words describe an integration engine and a cloud lakehouse. Naming the parts, six components and two shared services, is what makes two very different platforms comparable.
· 14 min read
