Can someone explain the note of TI.TIRL, "Rule that expresses the criterion in computer-executable form", or provide an example for TIRL mapping? Thanks.
Forums: General Discussion
Oh yes, we use machine-readable/executable rules all the time when selecting or searching for patients as candidates for a study.
It is however not feasible at all to use these in TI(RL) as long as FDA sticks to SAS Transport 5: usually, such machine code is mostly considerably more than 200 characters and may contain non-ASCII characters.
So, when TIRL was added to SDTM, this was more a "wishful thinking" of the SDTM developers than a real possibility.
The only thing that can currently be done is to add "pseudo-code".
Essentially, trial design information does not belong in tables, and surely not in SAS Transport 5 (nor 8) tables with all its limitations.
It is a tragedy that FDA still refuses to accept information in modern formats such as XML or JSON. For example, all trial design information can already easily go into ODM study design files, which are anyway used to set up the EDC and other systems. The same applies to the annotated CRF: why a PDF when all the mapping information can go into the ODM file which has all form definitions with for each field/question has the SDTM mapping information present? A simple stylesheet can then present that information in a user-friendly way.
But yes, ... if anything you know is SAS Transport, everything is a table (in analogy with hammer and nail ...).
Thank you, Kristin. "However, such a method has not been developed" answered my question!!
Thank you, Jozef. As you mentioned, "wishful thinking" of the SDTM developers than a real possibility” is exactly same as the information that Kristin posted, “However, such a method has not been developed”. I can totally ignore this variable for now!!
* Required Field
Example for TIRL
I have never seen TIRL rule used in the TI domain. In the SDTMIG, Assumption #4 in the TI section states the following:
'IETEST contains the text of the inclusion/exclusion criterion. However, because entry criteria are rules, the
variable TIRL has been included in anticipation of the development of computer-executable rules.'
I'm not sure why I/E criteria would ever need to be computer-executable and it seems that CDISC agrees. In addition, there are no examples that use this variable. TIRL may be deprecated in future versions of the SDTMIG. SDTMIG v3.4 added the following in Section 7.4.1 regarding the TIRL variable:
'The variable TIRL was included in the Trial Inclusion/Exclusion Criteria (TI) domain in anticipation of developing
a way to represent eligibility criteria in a computer-executable manner. However, such a method has not been
developed, and it is not clear that an SDTM dataset would be the best place to represent such a computer-executable
So it seems that there was never a use case for it. Hope this helps.