Validation Process

Starting with the SAS Clinical Standards Toolkit 1.7, SAS has removed OpenCDISC checks for SDTM and ADaM. Supplemental checks that are not addressed by OpenCDISC remain in the Validation Master data set provided for each standard and standardversion.

The most current information about checks required for FDA submissions can be found at the OpenCDISC (http://www.opencdisc.org/projects/validator/cdisc-validation-rules-repository) and FDA Study Data Standards Resources (http://www.fda.gov/ForIndustry/DataStandards/StudyDataStandards/default.htm) websites.

No. The full set of validation checks supported by the SAS Clinical Standards Toolkit is provided in the global standards library data set named validation_master. For SDTM 3.1.2, this data set is located n the global standards library directory/standards/cdisc-sdtm-3.1.2-1.7/validation/control directory.

Some checks assess structural (metadata) compliance, and other checks evaluate specific data sets, columns, and data values at the “cell” level.

For any validation process, the SAS Clinical Standards Toolkit expects that you have defined and referenced (in SASReferences, with type/subtype set to control/validation) a run-time set of checks that might be some subset of the validation_master checks. You have the option of both creating this subset and ordering the checks to be run.

The run-time set of checks might change over time. For example, once the metadata structure of the domains is confirmed, those checks do not need to be run again.

See “Building a Validation Process” in Chapter 7, “Compliance Assessment Against a Reference Standard,” in the SAS Clinical Standards Toolkit: User's Guide.

Yes. The SAS Clinical Standards Toolkit offers several distinct methodologies that support validating multiple studies. However, these methodologies depend on your data accrual processes and workflow. It is important that you maintain relationships between the source study data and the metadata that the SAS Clinical Standards Toolkit requires. Each column in each data set in each study must be uniquely identifiable.

See “Case Study 7: Validation of Multiple Studies” in Chapter 7, “Compliance Assessment Against a Reference Standard,” in the SAS Clinical Standards Toolkit: User's Guide.

Create your own copy of the validation_master data set for a specific standard (which might be the run-time validation_control data set). Then, make the changes to the copy’s validation check metadata, including codelogic.

If you believe that the codelogic provided by SAS is in error and you want SAS to correct the error, report the problem through normal SAS Technical Support channels. Approved updates become available with a subsequent hot fix or release of the SAS Clinical Standards Toolkit.