TMS autoclassification, synchronization, and manual processing

1. DMW TMS processing runs the TMS synchronization job, which checks for changes in the TMS dictionary(ies) associated with a study that impact DMW terms in TMS, either coded or not yet coded. It also checks for TMS data that meets the criteria in New and updated data processed in TMS.

2. TMS loops through all records that meet the criteria and calls autoclassification for each one.

   Autoclassification looks for exact matches to source terms in the coding dictionary level. If there is no exact match, it checks if a TMS user has previously classified the same term to a dictionary term (this is called a verbatim term assignment, or VTA), and classifies the current term the same way. It checks first in the current domain and then, if there is no VTA in the domain, globally.

   • For terms with a dictionary term or VTA match, TMS sends derived data back to DMW during the same transformation job.
   • For terms that do not have a dictionary term or VTA match, TMS creates a discrepancy in DMW and, if the lifecycle stage is Production, an omission in TMS. TMS sets the DMW discrepancy category to TMS in Progress as described in System interaction conflicts and discrepancy statuses.

3. For terms without a match, a TMS user does one of the following:

   • Classifies the term manually, creating a new VTA. During the next transformation, autoclassification finds the new VTA and TMS derives data to DMW and answers the discrepancy, setting the DMW discrepancy category to TMS DM Review as described in System interaction conflicts and discrepancy statuses.
   • Applies an action to the omission, requesting review or clarification of the term. This becomes the discrepancy text in DMW and the query text in InForm. TMS sets the DMW discrepancy category to TMS INV, sending the discrepancy to InForm for investigator review.

4. A DMW user can update a discrepancy as follows:

   • Add a comment. The user can choose to send the comment to TMS, which changes the category to TMS Evaluation, triggering autoclassification in TMS and making it impossible to make any other updates to the discrepancy until it is acted on in TMS.

   • Change the DMW discrepancy category directly to any of the TMS-related categories.

   • Send the discrepancy to the source system. After a discrepancy has been sent to InForm, any additional changes to the discrepancy in DMW or TMS are automatically sent to InForm.

   An InForm user can update a discrepancy as follows:

   • Answer the discrepancy by providing text in response to the discrepancy text.
   • Correct the underlying data item.

     In both cases, the category is updated to TMS Evaluation and TMS autoclassification is triggered.

     If the data change results in a successful classification, TMS sends derived data to DMW during the next transformation run. If a data change results in declassifying a term that was previously classified, previously derived data is deleted.

     If a source term is deleted in InForm, and then DMW, it is deleted in TMS as well.

Note:

TMS autoclassification, synchronization, derivation, actions, and DMW discrepancy management are the same in all three lifecycle stages. However, TMS omissions are created only in Production, so TMS users can do manual classification only in Production, and Force rederivation works only in Production.