CDISC Achieves Two Significant Milestones in the Development of Models for Data Interchange
AUSTIN, Texas. November 28, 2000.
The Clinical Data Interchange Standards Consortium has achieved two significant milestones towards its goal of standard data models to streamline drug development and regulatory review processes.
CDISC participants have completed metadata models for the 12 safety domains listed in the FDA Guidance regarding Electronic Submissions and have produced a revised XML-based data model to support data acquisition and archive.
The Submissions Data Standards team has been working since early 1999 to define a metadata model that is designed to 1) provide regulatory submission reviewers with clear descriptions of the usage, structure, contents and attributes of all submitted datasets and variables; 2) allow reviewers to replicate analyses, tables, graphs and listings with minimal or no transformations; 3) enable reviewers to easily view and subset the data used to generate any analysis, table, graph or listing without complex programming.
This team, under the leadership of Wayne Kubick of Lincoln Technologies, and Dave Christiansen of Genentech, presented their metadata models to a group of representatives at the FDA on Oct. 10, and discussed future cooperative efforts with Agency reviewers.
"The lack of cross-industry standards for efficient research data transfers has confused clinical trials information sharing and increased errors in this industry for decades. Developing models for data standards will enable automated tools for quality clinical trials analysis and review. The Submissions Data Model team made the first real breakthrough when they established the metadata model, which is an enabler for further developments of models," stated Kaye Fendt, who was the founding FDA Liaison to CDISC and is currently a director on the CDISC Board.
"The hard work of the CDISC teams and the support of the many participants in the regulatory health information arena have moved the standards development efforts many steps closer to an industry vision of efficient research data management and transfer. It is so gratifying to see so many people recognizing the value of the innovative CDISC approach."
The CDISC Operational Data Modeling (ODM) Working Group released their Version 1.0 model for data acquisition, interchange and archive. A small, interdisciplinary team was formed in September of 1999 to examine two different XML-based data interchange models (which had been separately put forward by PHT/Lincoln Technologies and by Phase Forward); specifically to assess the feasibility of developing an integrated, CDISC standard data and metadata model to support data acquisition.
The resulting CDISC model is based on the eXtensible Markup Language (XML), which is gaining wide acceptance as a general data interchange framework, and has been determined to be an effective approach to clinical data interchange. The goal of the CDISC XML Document Type Definition (DTD) Version 1.0 is to make available a first release of the definition of this CDISC model, in order to support sponsors, vendors and CROs in the design of systems and processes around a standard interchange format.
"The release of the CDISC Version 1.0 DTD provides the industry with a foundation of standards that will support unprecedented improvements in the quality and efficiency of future data interchange," said Ken Harter, senior systems analyst, Amgen Inc.
Both CDISC models can be reviewed at http://www.cdisc.org/publications.html.
Comments are requested by Jan. 31, 2001 and should be posted using the CDISC Web site Discussions option.
CDISC is a non-profit organization with a mission to lead the development of standard, vendor-neutral, platform-independent data models that improve process efficiency while supporting the scientific nature of clinical research in the biopharmaceutical and healthcare industries. For more information, visit the CDISC Web site at www.cdisc.org, or contact Shirley McCluskey, director of operations, CDISC, at firstname.lastname@example.org or Rebecca Kush, president, email@example.com.
Shirley McCluskey or Rebecca Kush