Control Terminology vs. Warnings from Open CDISC Validation: A Submission (SBLA) Preparation Experience

• Activity Number: 270
• Type: Invited
• Date/Time: Tuesday, July 31, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Section for Statistical Programmers and Analysts
• Author(s): Jian Wang*+
• Companies: Regeneron Pharmaceuticals, Inc.
• Keywords: CDISC ; ADAM ; SDTMIG ; eCRF

Abstract

Study Data Tabulation Model (SDTM), developed by the Submission Data Standard working group of the Clinical Data Interchange Standard Consortium (CDISC), provides a valuable representation that may facilitate FDA review. Generation and validation are two main tasks in SDTM data preparation and define.xml preparation, and require appropriate usage of controlled terminology (CDISC defined, external dictionary, sponsor defined). This presentation will emphasize (1) controlled terminology preparation for SDTM generation by domain (2) openCDISC validation steps to find deficiencies in controlled terminology by giving warnings. Examples wih details including steps implemented will be shared. It is noted that for SDTM mapping, about half of time and resources were spent on mapping control terminology including addressing warnings from open CDISC validation. Fewer warnings represent better CDISC compliance. A set of preparation steps are therefore proposed to be implemented during early study set up stage: including preparing code list based on SDTMIG; implementing code list as much as possible in eCRF/database design; external lab data standardized format preparation; revisiting code list after completing database setup and coming up with the list of non-implemented code (most time are non CDISC vocabularies), and getting programming ready for mapping.