CDISC标准介绍

CDISC（Clinical Data Interchange Standards Consortium）临床数据交换标准协会就如何收集数据、收集什么类型的数据以及如何将数据提交给负责审批新药的机构建立起了一套标准。涉及临床研究电子数据的获取、交换、存档、提交的全过程。下表是CDISC的细分标准^[1]：

Standard	Description
Study Data Tabulation Model (SDTM)	The content standard for regulatory submission of case report form data tabulations from clinical research studies.
Analysis Data Model (ADaM)	The content standard for regulatory submission of analysis datasets and associated files.
Operational Data Model (ODM)	The XML-based content and format standard for the acquisition, exchange, reporting or submission, and archival of case report form (CRF)-based clinical research data.
Laboratory Data Model (LAB)	The content and format standard for data transfer between clinical laboratories and study sponsors/CROs.
Case Report Tabulation Data Definition Specification (CRTDDS) – (define.xml)	The XML-based content and format standard referenced by the FDA as the specification for the data definitions for CDISC SDTM datasets. This standard, also known as define.xml, is an extension of the ODM.
Standard for Exchange of Nonclinical Data (SEND)	An extension of the SDTM standard for submission of data from pre-clinical studies.
Protocol Representation (PR)	The content and format standard supporting the interchange of clinical trial protocol information. This is a collaborative effort with Health Level Seven (HL7).
Trial Design Model (TDM)	The content standard that defines the structure for representing the planned sequence of events and the treatment plan of a trial. This is a subset of the SDTM and Protocol Representation.
Clinical Data Acquisition Standards Harmonization (CDASH)	A CDISC-led collaborabtive initiative to develop the content standard for a minimum set of data collection fields in case report forms. This standard is based upon the SDTM.
Terminology	The controlled standard vocabulary and code sets for the all CDISC models/standards.
Glossary	The CDISC dictionary of terms and their definitions, related to the electronic acquisition, exchange and reporting of clinical research information. Abbreviations and acronyms are also listed.

CDISC标准涵盖两方面：内容（data，metadata，terminology）和格式（XML-based）标准。从 Introducing the CDISC Standards¹截取张图说明各个标准在临床试验过程中的分布：

ODM-XML

ODM-XML定义了一种与供应商无关、独立于平台的格式，用于交换和归档（临床的和转化而来的）研究数据及其相关元数据、管理数据、参考数据和审计信息。
ODM-XML有助于符合法规的元数据和数据的获取、存档和交换。ODM-XML已经成为许多EDC表示CRF内容的首选格式。
下面两幅图展示了ODM文件的格式：


XML文件文本内容举例：

CDASH

用于CRF基础数据收集字段的内容标准，基于SDTM标准制定。和SDTM对应，CDASH按领域划分为：
Special-purpose domains:

• CO- Comments
• DM- Demographics

Interventions class domains:

• CM- Prior and concomitant medications
• EC- Exposure as Collected and eX- Exposure
• PR- Procedures
• SU- Substance use

Events class domains:

• AE- Adverse events
• CE- Clinical events
• DS-Disposition
• DV- Protocol deviations
• HO- Healthcare encounters
• MH- Medical history

Finding class domains:

• DA- Drug Accountability
• DD- Death details
• EG- ECG Test results
• IE -Inclusion/Exclusion Criteria not met
• LB- Laboratory Test results
• MB- Microbiology Specimen
• MS- Microbiology susceptibility
• MI- Microscopic Findings
• PC-Pharmacokinetics Sampling
• PE- Physical examination
• QRS- Questionnaires, Ratings, and Scales
• RP-Reproductive System Findings
• RS-Disease response and Clin classification
• SC- Subject characteristics
• TU- Tumor /Lesion identification domain
• TR- Lesion results
• VS- Vital signs
• FA- Findings about events or interventions
• SR-Skin response

Associated Persons Domain

• AP- Associated Persons

SDTM

通过不同途径收集的临床研究数据，汇总成SDTM数据集。

SDTM定义domains如下图所示：

ADaM

有SDTM了，为什么还要ADaM？看看两者的区别^[1]：

SDTM	ADaM
Contains the raw source data and limited derived data	Contains the derived data, both variables and observations, and even entire datasets, that are derived from the raw data. May also contain rawsource data to facilitate review and sensitivity analysis.
The variables allowed in each domain are restricted by the SDTM model	ADaM specifies core variables to be present in analysis data sets yet additional variables and observations can be added if they are needed for traceability and/or facilitation of analyses, provided that rules about when to derive data as rows and when to derive data as columns are followed.
There is no redundancy of variables within SDTM in that a specific variable, such as treatment, is found in only one SDTM domain.	There is redundancy of variables across different analysis datasets to support analysis. For instance, treatment variable(s) would be expected to be in every analysis dataset.
Generally uses character variables, such as character date strings.	Numeric values are included when needed for analysis purposes, such as numeric date values.
Each domain has a specific topic and contains only those variables related to that topic.	Contains variables and observations that may be obtained from a variety of domains. For example, a responder analysis may be based on information obtained from laboratory, imaging, and /or questionnaire data.

Define-XML

Ref:

1. Introducing the CDISC Standards - New Efficiencies for Medical Research
2. www.cdisc.org