文献速递医学人工智能与信息学医学教育与科研方法

临床研究中数据对象的元数据模式

A metadata schema for data objects in clinical research.
PubMed 文献 · Trials · 2016-11-24

为解决临床研究数据对象存储库多样且分散的问题,作者提出利益相关方需就简单一致的元数据方案达成一致。该方案基于DataCite标准扩展,涵盖研究标识(含临床试验注册链接)、数据对象特征与标识符、以及位置、所有权和访问信息,以填补现有临床研究相关标准未涉及的空白。

要点

  • 提出基于DataCite标准扩展的元数据方案,用于临床研究数据对象的存储与共享。
  • 方案涵盖研究标识(含注册链接)、数据对象特征与标识符、位置、所有权与访问信息。
  • 该方案可与其他数据结构化举措整合,但不依赖它们。

核心结论

  • 该元数据方案被提出作为广泛认可标准的自然扩展,填补了其他临床研究相关标准(如临床数据交换标准协会、生物医学研究集成领域组)未涉及的空白。

研究设计与样本

  • 本文描述了相关数据对象及其可能的存储方式,并识别了支持临床研究数据共享的元数据需求,探讨了研究与数据对象的持久标识符问题。

主要发现

  • 提出了一种基于DataCite标准的方案,并扩展以满足临床研究者的需求,具体提供研究标识数据(包括临床试验注册链接)、数据对象特征与标识符、以及涵盖位置、所有权和数据对象访问的数据。

临床怎么用

  • 作者提出该元数据方案可作为广泛认可标准的自然扩展,并可与其他改善临床研究数据结构的举措整合,但不依赖它们。

局限性

  • 原文未报告局限性

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: A large number of stakeholders have accepted the need for greater transparency in clinical research and, in the context of various initiatives and systems, have developed a diverse and expanding number of repositories for storing the data and documents created by clinical studies (collectively known as data objects). To make the best use of such resources, we assert that it is also necessary for stakeholders to agree and deploy a simple, consistent metadata scheme. METHODS: The relevant data objects and their likely storage are described, and the requirements for metadata to support data sharing in clinical research are identified. Issues concerning persistent identifiers, for both studies and data objects, are explored. RESULTS: A scheme is proposed that is based on the DataCite standard, with extensions to cover the needs of clinical researchers, specifically to provide (a) study identification data, including links to clinical trial registries; (b) data object characteristics and identifiers; and (c) data covering location, ownership and access to the data object. The components of the metadata scheme are described. CONCLUSIONS: The metadata schema is proposed as a natural extension of a widely agreed standard to fill a gap not tackled by other standards related to clinical research (e.g., Clinical Data Interchange Standards Consortium, Biomedical Research Integrated Domain Group). The proposal could be integrated with, but is not dependent on, other moves to better structure data in clinical research.

来源:Trials,查看原文。版权归原作者所有。

引用

  1. Canham S, Ohmann C. A metadata schema for data objects in clinical research. Trials. 2016. PMID: 27881150

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。