Forums: Define.xml
Dear Laetitia,
unfortunately I could not find any documentation on this rule, at least not on the P21 website.
So, the rule, and its implementation seems not to be very transparent.
What I can however say as one of the developers of the standard is the following:
When you define a standard for the codelists, as you did, you also need to reference this standard on each of the "CodeList" elements, using the "CodeListOID" attribute.
The Define-XML v.2.1 specification (you should read it) is clear: in the context of a regulatory submission (that is what P21 is about I suppose), you either need to provide "def:StandardOID" (when the codelist is taken or derived from CDISC-CT), or the attribute "def:IsNonStandard" with value "Yes" when it is a "sponsor-defined" codelist.
So, did you add "def:StandardOID" or "def:IsNonStandard" on each "CodeList"?
I don't claim adding these will go the error go away, as it is not clear how DD0138 works or implemented, but you could surely have a look, at best in the define.xml itself, e.g. using simple NotePad++.
For the OID of the "Study" and of "MetaDataVersion", one should know that OIDs are just identifiers, NOT meant to have a meaning. Your software (what are you using?) is however trying to generate OIDs with a meaning, and just takes ... something (?) that comes from your input (Excel???).
Better would be that the software generates OIDs e.g. as UUIDs (https://en.wikipedia.org/wiki/Universally_unique_identifier) like "1ef62f72-c9eb-4c4f-9701-13994384d5fd", OR generated from the true information that you provide like the real Study ID.
You also mention "sheet ‘Dataset’ in define specifications" but there essentially is no such thing: the define.xml itself is "the specification". For the (Excel?) sheet one could maybe speak about a "template", but surely not about "specification", i.e. one cannot have a "specification of a specification".
Some people from the CDISC Define-XML development team, so people with the real knowledge, have developed excellent Define-XML validation tools. So you may also have a look at these.
Attached a snapshot of the Define-XML v.2.1 specification. But anyone working with Define-XML should definitely download that specification from the CDISC website.
Dear Jozef,
Thank you for your answer. I have no problem with controlled terminology for DD0138 (I confirm I added "def:StandardOID" or "def:IsNonStandard" on each "CodeList", I’m working with Define-XML v.2.1 specification, and CDISC example is very helpful).
I have the DD0138 message only for SDTMIG 3.4 and SDTMIG-MD 1.1
I confirm that what I’ve called “excel specification” is Pinnacle21 excel template to create the define through Pinnacle. Sorry for confusing !
I continue to investigate on DD0138, and if you have any information later, I would be interested in.
Thank you very much.
I get this error:
DEFINE Standard OID, Standard Type, Standard Name 01, IG, SENDIG-DART DD0138 Referenced Standard is missing Cross-reference
from this section of the define.xml which seems correct for SEND DART studies:
<def:Standards>
<def:Standard OID="STD.01"
Name="SENDIG-DART"
Type="IG"
Status="Final"
Version="1.1"/>
However, if I copy from
C:\Users\<user>\Documents\Pinnacle 21 Community\configs\2204.1\SEND-IG-DART 1.1 (FDA).xml
to
C:\Users\<user>\Documents\Pinnacle 21 Community\configs\2204.1\SENDIG-DART 1.1 (FDA).xml
then this error no longer occurs (I leave both file spellings in the foler).
So seems this a confusion where for DD0137 check, SENDIG-DART is correct name
and then perhaps DD0138 checks for similar spelling in the file name, but spelling of file delivered by Pinnacle21 has dash between SEND and IG in its name.
With the exact same define.xml file, I am finding inconsistent messages fired between P21E and P21C.
I am working and Define-XML v2.1 and P21E fires DD0138 for the CDISC/NCI controlled terminology dated 2023-03-31, which I had specified for OID as 'STD.2'.
I have checked and that:
1. Version is present and correct
2. Name 'CDISC/NCI' is specified for CTs as per Define-XML Specification v2.1
3. In All places where Codelists are defined, there is a def:StandardOID present (with value 'STD.2').
I cannot figure out why P21E kept firing this when this message did not occur for P21C.
File Upload
Hello,
While validating define xml 2.1 with Engine FDA (2204.1)/SDTMIG3.3, I have the ‘DD0138-Referenced Standard is missing’ error.
I have this error for the 2 standards other than SDTMIG3.3 (ie SDTMIG 3.4, SDTMIG-MD 1.1 are used for some datasets).
DD00138 is a cross reference check, but sheet ‘Dataset’ in define specifications is the only place where I mention these standards.
Can you explain me the problem?
Also I wonder why SDTMIG 3.4 is indicated in my define (define generated with Pinnacle community) for Study OID + MetaDataVersion OID
Thank you for you help,
Laetitia