MII Releases 'Document' Core Dataset Module for Standardized Use of Clinical Texts

Details

The Medizininformatik-Initiative (MII) announced the 'Document' core dataset module, designed to structure clinical free texts (e.g., discharge summaries or operation reports), making them available for medical research across different institutions. The foundation of this module is the international standard HL7 FHIR, specifically the DocumentReference resource. The module was designed with compatibility in mind regarding existing German standards (such as those from the Kassenärztliche Bundesvereinigung and Gematik's ISiK specifications), but it specifically targets secondary use for research. For uniform document classification, the Clinical Document Class List (KDL) is recommended. Furthermore, an integrated NLP status extension tracks the processing status of a document, indicating if it has been de-identified or annotated. The primary driver for this development was the GeMTeX project, which aims to build a nationwide clinical text corpus in Germany. The technical formalization utilized FHIR Shorthand (FSH), and the Implementation Guide is published on the Simplifier platform. For Data Integration Centers (DIZ), this module forms the basis for providing text data via the research data portal Gesundheit (FDPG). This allows researchers to query specific patient cohorts that contain certain document types in a defined processing state.