Creating a Define-XML 2.0 define.xml File from SDTM Source Data

The SAS Clinical Standards Toolkit supports the currently published CDISC Define-XML 2.0 submission standard, which supports representation of the CDISC SDTM and CDISC SEND tabulation data sets and ADaM 2.1 analysis data sets in metadata form.

You have a library of CDISC SDTM SAS data sets (which are not read by this process) from which a set of metadata (in the form of source_study, source_tables, source_columns, source_codelists, source_values, and source_documents data sets) has been created. This metadata must contain the expected, correctly typed columns created for the sample study provided by SAS. The source_study, source_tables, and source_columns data sets are required to create a Define-XML 2.0 define.xml file.

The Define-XML driver programs are located in the sample study library directory/cdisc-definexml-2.0-1.7/programs directory.

The initial task is to extract the available SDTM metadata into Define-XML 2.0 metadata files. The SAS representation of Define-XML 2.0 involves 54 data sets, but only 31 of these are typically used to create a Define-XML 2.0 file. The other 23 data sets contain ODM (Operational Data Model) metadata or Analysis Results metadata that are extensions of the Define-XML 2.0 model. The sample driver program create_sasdefine_from_source.sas, modified to point to your specific SDTM study metadata, must be submitted to create the SAS representation of Define-XML 2.0 metadata. This process builds the 31 core data sets, but it might not populate all of them (depending on the completeness of the metadata of your study).

%define_sourcetodefine(
     _cstOutLib=srcdata,
     _cstSourceStudy=sampdata.source_study,
     _cstSourceTables=sampdata.source_tables,
     _cstSourceColumns=sampdata.source_columns,
     _cstSourceCodeLists=sampdata.source_codelists,
     _cstSourceDocuments=sampdata.source_documents,
     _cstSourceValues=sampdata.source_values,
     _cstFullModel=N,
			_cstCheckLengths=Y,
     _cstLang=en
     );

The macro is located in the global standards library directory/standards/cdisc-definexml-2.0-1.7/macros directory.

In the sample study, the source files for the SDTM 3.1.2 study are in the sample study library directory/cdisc-definexml-2.0-1.7/sascstdemodata/cdisc-sdtm-3.1.2/metadata directory.

At this point, all available content for the define.xml file has been captured in the SAS representation (31 data sets) of the CDISC Define-XML 2.0 standard. The SAS Clinical Standards Toolkit provides a sample driver program, create_definexml.sas. This program builds and validates the define-sdtm-3.1.2.xml file. Submit the create_definexml.sas driver program.

In this driver program, the call to the primary task macro requests that the default style sheet provided by SAS (the source of which is CDISC) be copied to the folder location containing the generated define.xml file. The macro is located in the global standards library directory/standards/cdisc-definexml-2.0-1.7/macros directory.

Here is the macro:

%define_write(_cstCreateDisplayStyleSheet=1,
     _cstHeaderComment=%str(Produced from SAS data using the SAS
     Clinical Standards Toolkit &_cstVersion)); 

The final task in the sample create_definexml.sas driver program is to call the %CSTUTILXMLVALIDATE macro to perform the schema validation. This involves verifying that the Define-XML define.xml file is valid structurally and syntactically according to the XML schema.

The Results data set provides process information and the location of the generated define-sdtm-3.1.2.xml file. The Results data set confirms that no problems were found with the file after validation of the file.