CDISC Data Automation In Clinical Trials Submission

BackgroundFDA recommends that sponsors submit clinical trial data in CDISC (Clinical Data Interchange Standards Consortium) SDTM (Study Data Tabulation Model) and ADaM (Analysis Data Model) standards, accompanied with metadata (aka Define.xml).CDISC standards are becoming popular in the industry:include data collection data standard CDASH, data interchange standard SDTM, and data analysis standard ADaM.Up to now, the transformation processes from CDASH data, to SDTM data, and to ADaM data, for most companies, are not automatic, and in the manually processing steps. There is no mature commercial software packages, and the no good automatic mechanism.If the collected data in clinical trials uses CDISC CDASH data standard, and converts to SDTM and ADaM standard thereafter, then there are many parts/process can be re-useable:from the design of CRF (Case Report Form), to the database design, to the generation of reports, etc. So, based on CDISC standards, we can1) Develop a set of SAS macros, or other software utilities;2) Reuse our current SAS macros, or other software utilities as many as possible;MethodThis study starts from a simple clinical protocol, and designs a CRF based on this protocol, generates CDASH data based on this CRF, develops the config file and conversion program to generate SDTM and ADaM data automatically, as well as the metadata (define.xml). And if the raw data is not in CDASH standard, but user own standard, it develops a mapping program to convert the raw data to CDASH data. The whole conversion process is very standardized. For example, the variables’ names and value lists are all in CDISC standards. It can be used in different types of studies.The conversion process is highly optimized for general use: modules like demography, medical history, physical examination, adverse event, concomitant medication, vital signs, and study summary or disposition, etc, because the data structure are very similar and can be standardized, which makes it very suitable for general use.The standardized process then can improve the data quality, reduce the time and cost on the data collection and derivation. The saved time can be used on the protocol design and data analysis.ResultsA clinical protocol on the common indexes, mean modified whole mouth gingival index and mean plaque index in oral care area, was developed. Based on this protocol, we designed the case report forms (CRF), and simulated the source data (including RAW format and CDASH format). A series of SAS macros and programs were developed, in combination of excel configuration files and mapping specifications.ConclusionsThis automation system can convert the source/raw data (both CDASH and non-CDASH format) to SDTM+ format and then SDTM+ to ADaM format very quickly by user configuration and mapping, thereby resulting in high quality statistical analysis, and reducing study time and cost.The main features include:1) Based on the source data (CDASH standard, or non-CDASH standard);2) Generate SDTM and ADaM format data automatically;3) The data format complies with FDA CDER/CBER requirements.4) User interface is in Microsoft Excel format, user friendly, easy to use. Users without CDISC knowledge can get started very quickly.