The Use of CDISC Data Standards in FDA Electronic Submissions

Email this to someonePrint this pageShare on LinkedIn0Tweet about this on TwitterShare on Google+0

The preferred format for submitting standardized study data to CDER and CBER are the Clinical Data Interchange Standards Consortium (CDISC) Foundational Standards, including the Study Data Tabulation Model (SDTM), Analysis Data Model (ADaM), and Standard for Exchange of Nonclinical Data (SEND). In 2011, approximately 30% of unique NDAs received by CDER and approximately 20% of BLAs received by CBER included CDISC/SDTM data. Implementing CDISC/SDTM tends to be a difficult and expensive undertaking for sponsors. There is room for interpretation in the standard, which leads to some variability in how it is implemented across sponsors and CROs. In addition, there may be data collected for studies that do not fit neatly into the standards set forth in SDTM. This need for study-specific workarounds can lead to variability in how the standard is implemented across studies within the same development program.

In February 2012, FDA distributed for comment a new draft guidance titled “Providing Regulatory Submissions in Electronic Format – Standardized Study Data.” The purpose of the guidance is to promote the use of data standards in electronic submissions to FDA. The guidance encourages sponsors to implement the standards as early as possible in the development program. If sponsors wait to convert data until after the clinical study report (CSR) is written, the submission-ready datasets will not be same as the datasets used to support the analyses in the report. Ideally, for each study, SDTM datasets will be generated first and the ADaM datasets will be created from the SDTM data. The tables, figures, and listings (TFLs) in the CSR will be generated from the ADaM datasets, and those programs used to generate the (TFLs) to be included in the submission as needed.

The draft guidance also states sponsors are encouraged (not required) to submit a Study Data Standardization Plan as a section in the General Investigational Plan in an IND. The purpose of this section is to open the communication channels early between FDA and sponsors about how data standards will be implemented throughout the development program. I believe creating and maintaining a Study Data Standardization Plan may also serve as a useful communication tool internally to the sponsor company as well. Clearly documenting all datasets and formats for a development program should clarify any standardization work that needs to be done. The Data Standardization Plan should also identify dataset deliverables that will be required for submission (i.e., annotated CRF, data definition files, non-executable programs, etc). For sponsors that utilize multiple CROs for study and data management, the Data Standardization Plan should help clarify how data standards are implemented and also ensure consistent delivery of submission-ready datasets across CROs. It may be that this tool will reveal the need for sponsors to create business rules for how SDTM standards should be consistently implemented throughout the development program, rather than accept each CROs interpretation for each study.

Although you can comment on any guidance at any time, the comment period for the draft guidance closed in April 2012. Here are some useful links to provide more information on data standardization in submissions to CDER and CBER.

Author: Emily Onkka

Share This Post On

Submit a Comment