Metadata Verification 


Overview 
An evaluation of the metadata data defined in the Excel spreadsheet found on the portal at:

Home Portal > Teams/Project Portals > SDTM > Metadata Subteam > Working Documents > Metadata - Define  
Tang 2009-04-13.zip
Tang 2009-04-13  4/21/2009 11:14 AMUse SHIFT+ENTER to open the menu (new window).

Within this file, an evaluation was applied to metadata 2008-12-12.xls and the findings are listed as follows:

  1. SE Variables - The SE Variables are not fully defined.  As highlighted below, the variable names appear to be labels and will be excluded from this evaluation.

    Domain Variable Label Type
    SE Study Identifier Study Identifier CHAR
    SE DOMAIN Domain Abbreviation CHAR
    SE Unique Subject Identifier Unique Subject Identifier CHAR
    SE Sequence Number Unique Subject Identifier CHAR
    SE Element Code  Element Code CHAR
    SE Description of Element Description of Element CHAR
    SE Start Date/Time of Element Start Date/Time of Element CHAR
    SE End Date/Time of Element End Date/Time of Element CHAR
    SE Planned Order of Elements within Arm Planned Order of Elements within Arm NUM
    SE Epoch Epoch CHAR
    SE Description of Unplanned Element Description of Unplanned Element CHAR


     

  2. Internal Attribute Consistency - An evaluation was applied to all the datasets from the set of metadata among themselves for consistency.  This is not yet compared with CDISC standards.  Some findings include:

    • Variable Label Inconsistencies 

      Data name Variable Type Length Label
      CO idvar C 100 Identifier Variable Name*
      SQ idvar C 100 Identifier Variable*

      

    SV visit* Visit Number
    VS visitnum* Visit Number
    • Variable Type Inconsistencies

      DA visitnum N* 8* Visit Number*
      SV visitnum C* 100* Visit Name*

     

  3. Attribute Consistent with 3.1.1 - An evaluation was applied to compare against guidelines 3.1.1.  

    • Variable Label Inconsistencies

      Data Source Domain Variable Type Length Label
      Metadata CM cmstdy N 8 Study Day of Start of Event*
      CDISC 3.1.1 CM cmstdy N 8 Study Day of Start of Medication*


  4. PROC CDISC - An example PROC CDISC was applied AE and MH as an example with findings:
     

    • AE
      [16] ERROR: Variable AEPRESP is not defined in this domain (AE) and must be dropped/deleted.
      [17] ERROR: SDTM-compliant datasets may contain only variables defined in the domain (AE).
       

    • MH
      [16] ERROR: Variable MHPRESP is not defined in this domain (MH) and must be dropped/deleted.
      [17] ERROR: SDTM-compliant datasets may contain only variables defined in the domain (MH).


    last updated on June 9.2009