FHIR Database Schema: Designing for Interoperability and Search
FHIR database schema - Anpheros Platform — Dashboard
Summary
A technical guide on designing databases for FHIR resources. The article compares three storage methods—normalized tables, JSON+indexes, and graph databases—and details the robust, production-grade schema used by Anpheros in PostgreSQL.
Details
FHIR resources are deep, optional, and highly extensible, making full normalization into relational tables impractical due to required schema migrations. The article recommends the mainstream approach: storing the full resource as JSON (JSONB) and creating separate, typed index tables for common search parameters. The Anpheros schema utilizes dedicated tables for current resources, history, and specific indexes (like token, date, reference), along with projection tables for simple REST views. This design provides best practices for maintaining fast search performance while ensuring complete data history and flexible extensibility.
Original content copyright by respective publishers