<b>Common Data Model for Rare Diseases</b> based on the ERDRI-CDS, HL7 FHIR®, and the GA4GH Phenopackets Schema© (v2.0)

ART-DECOR Project: https://art-decor.org/ad/#/erker-/project/overviewWhile rare diseases (RDs) affect over 260 million individuals worldwide, low data quality and scarcity challenge effective RD care and research severely. A significant knowledge gap exists between specialised RD data and data from routine clinical care. This study aims to harmonise the data standards necessary for RD care and research into a novel RD Common Data Model (CDM). The proposed RD CDM integrates the ERDRI-CDS for international registry use and extends it with data elements based on HL7 FHIR for reliable data transactions and the GA4GH Phenopacket Schema for precise bioinformatic analyses. This work does not present an implementation guide but lays the foundation for developing and defining international RD CDMs aligned with these data standards. An ontology-driven approach was selected, encoding data elements and value sets with SNOMED CT and LOINC to find a common denominator between the data standards. The RD CDM was implemented in multiple German university hospitals capturing real RD patient data for registry or analysis purposes, fixing errors, and ensuring semantic consistency. Our RD CDM version 2.0 comprises 66 data elements, extending the ERDRI-CDS by 50 elements. We evaluated our CDM based on (1) Medical Data Granularity, (2) Clinical Reasoning and Medical Relevance, and (3) Interoperability and Harmonisation. Six layers of harmonisation were identified, ranging from data element alignment and terminology binding to value sets. Over 95% of data elements, 80% of data types, and, due to the ontology-based approach, less than 41% of value sets align with either HL7 FHIR or the GA4GH Phenopacket Schema. The novel RD CDM can serve as a basis for developing and implementing RD CDMs in various healthcare information systems, adhering to HL7 FHIR, GA4GH Phenopacket Schema, and ERDRI-CDS requirements for effective RD research and care. While not an implementation guide, recommendations on cardinalities are given, and this template invites further refinement and international collaboration. This work represents a significant step for clinicians to capture precise RD data based on international interoperability standards, regardless of the healthcare information system used.<br>ART-DECOR项目：https://art-decor.org/ad/#/erker-/project/overview 罕见病（Rare Diseases, RDs）在全球范围内影响超2.6亿人群，但数据质量低下与数据稀缺严重制约了罕见病的有效诊疗与研究。专业罕见病数据集与常规临床诊疗数据集之间存在显著的知识鸿沟。本研究旨在将罕见病诊疗与研究所需的数据标准进行统一，构建新型罕见病通用数据模型（Rare Disease Common Data Model, RD CDM）。 所提出的RD CDM整合了适用于国际注册的ERDRI-CDS，并基于HL7 FHIR（Health Level Seven Fast Healthcare Interoperability Resources）扩展了数据元素以实现可靠的数据交互，同时基于GA4GH Phenopacket Schema（全球基因组与健康联盟表型数据包架构）以支持精准的生物信息学分析。本工作未提供实施指南，但为符合上述数据标准的国际罕见病通用数据模型的开发与定义奠定了基础。 研究选用本体驱动的方法，使用SNOMED CT（Systematized Nomenclature of Medicine -- Clinical Terms）与LOINC（Logical Observation Identifiers Names and Codes）对数据元素与价值集进行编码，以在各类数据标准之间找到通用契合点。 该RD CDM已在多家德国大学医院落地实施，采集真实的罕见病患者数据用于注册或分析工作，同时修正数据错误并确保语义一致性。我们的RD CDM 2.0版本共包含66个数据元素，较ERDRI-CDS新增50个元素。 我们从三个维度对该CDM进行评估：(1) 医疗数据粒度，(2) 临床推理与医学相关性，(3) 互操作性与标准化统一。研究确定了六层标准化统一路径，覆盖从数据元素对齐、术语绑定到价值集的全流程。超过95%的数据元素、80%的数据类型，以及由于采用基于本体的方法，仅有不到41%的价值集符合HL7 FHIR或GA4GH Phenopacket Schema标准。 这款新型RD CDM可作为在各类医疗信息系统中开发与部署罕见病通用数据模型的基础，契合HL7 FHIR、GA4GH Phenopacket Schema与ERDRI-CDS的相关要求，助力实现高效的罕见病研究与诊疗。尽管本工作未提供实施指南，但针对数据基数约束给出了相关建议，且该模板欢迎进一步优化与国际协作。本工作为临床医生基于国际互操作性标准采集精准的罕见病数据迈出了重要一步，不受所使用的医疗信息系统类型限制。