Special Topic: Performance Considerations

SAS Clinical Standards Toolkit validation should first be run on a subset of source data to identify general process problems, missing or inconsistent process control metadata, and common and perhaps correctable data errors.

The SAS Clinical Standards Toolkit standard-specific Validation Master data set should be subsetted to remove duplicate checks. For example, CDISC SDTM 3.1.1 Janus checks are generally duplicates of WebSDM checks with occasionally different resultseverity values.

The _cstDebug option should be toggled off except for when you want to debug specific program errors to avoid exceeding the SAS log-size limitations or to avoid generating large SAS log files.

A SAS Clinical Standards Toolkit validation process that involves a large number of checks, should be run in batch or using PROC PRINTTO. This is also true for a SAS Clinical Standards Toolkit validation process that is run with the _cstDebug option toggled on. Doing so avoids exceeding the SAS log-size limitations.