We hope you'll join us for our 4/23 webinar on using data tables to apply reference ranges and AE codes in OC4. For more information and to register, visit https://register.gotowebinar.com/register/2882170018956684555

Delete CRF version in OC 3.7

toskriptoskrip Posts: 273 ✭✭✭
Hi all,

I have currently noticed one strange behaviour in my testing OC 3.7 instance. In previous version I was used to delete old unused CRF versions e.g. during CRF development process. The workflow was that for successful CRF version deletion it was necessary not to use it in any event definition as default CRF version and also having all the subjects that have used the old CRF version migrated to the new CRF version.

I have noticed that in OC 3.7 this is no longer working (or I am doing something wrong). Basically I have CRF version that is not used anywhere. It had data but the data was deleted from study subject and study subject is now using different CRF version. There is just one study subject in a study so it is definitely not used anywhere else. If you try to delete CRF I get a message that CRF has associated item data and cannot be deleted. Showing the associated data as the even of the one study subject that I have.

My question is, whether anybody also run in this sort of troubles and more importantly what have been changed in OC that now it behaves this way. Basically I would like to know whether this is a bug or feature.

best

Tomas
«1

Comments

  • Hello Thomas,

    You say you have deleted the data of the subjects. This implies that you have used this CRF version to enter data. As you cannot delete data but only mark it as removed I guess that OpenClinica does not allow you to delete or replace the CRF version as it already has been used for data entry.

    Best
    Marco
  • toskriptoskrip Posts: 273 ✭✭✭
    Yes I was using it for data entry, than I have deleted collected data and migrated subject to the new CRF version. I am asking because in older versions 3.3 and 3.4 it was possible to delete CRF version if that version was not the default one for any of events nor it was used within any of study subjects.
  • ccollinsccollins Posts: 380 admin
    Hi Tomas,

    I think you're seeing an effect of the changes we made to strengthen the audit trail upon deletion events (https://docs.openclinica.com/release-notes/release-notes-openclinica-3.6, look under 'Changes & Fixes'). These changes were important for compliance purposes as the audit information in earlier versions was obscured when an event CRF was deleted by an administrator.

    Does it greatly complicate the CRF build/test process for you? Have you been able to come up with any workarounds, and if not what behavior would you like to see?

    Cal
  • toskriptoskrip Posts: 273 ✭✭✭
    Hey Cal,

    I think I will be looking from some workaround (I have somewhere SQL scripts for deleting removed CRFs, hope this still works).

    > Regarding the changes in audit trail in 3.6

    Does it mean that when I delete data from CRF it actually creates new data with empty string as a content? Because this is the other change that I have noticed. When I delete data from CRF in 3.3 and 3.4 the CRF reset status to not started enabling the possibility to select the CRF version and start data entry. However when I delete data from CRF in 3.7 the status is set to data entry started the CRF version is the last used one and further more it still displays Delete button (red) as if the data was actually there (which I think is, because now it consider empty strings as data).

    So now I do not see the possibility to reset the CRF into the state as it was before data entry. This can complicate some processes mainly because there is no possibility to select the version of CRF for low privileged user now (it goes with the version which was entered and deleted before). If we have wrongly entered data that need to be deleted and multiple versions of CRFs, the data entry person cannot choose the CRF version and it have to be done with higher role.

    Is this behaviour also feature rather than bug?

    Tomas

    P.S.: now I understand why you offer bridge training for OC :)
  • ccollinsccollins Posts: 380 admin
    Hi Tomas,
    Pardon the slow reply. I spoke with Alicia, she's going to follow up as she is much more knowledgeable on this topic than I am.

    Cal
  • mvirtosumvirtosu Posts: 276 ✭✭
    Hi All,

    I would like to chime in and answer Cal's question: "what behavior would you like to see?"

    I would personally would like to see the deletion appearing in the audit trail and the event CRF being actually deleted instead of being replaced with empty values.

    Looking forward to Alicia's response.

    Mihai
  • agoodwinagoodwin Posts: 131 admin
    via Email
    Hello Everyone,

    Apologies for being late to the thread.

    In summary we made the change because we felt it was what we MUST do to
    meet regulatory requirements. We felt that with the architecture that we
    have and user requirements that this was the best solution. There was
    actually quite a long discussion about this change which you can read
    about here:

    http://forums.openclinica.com/discussion/comment/17249/#Comment_17249

    While I don't think that we can hard delete records, one change that might
    be worth considering is who is able to "reassign" a version (lessening the
    burden on the Data Manager), but this should be a deeper discussion. Please
    let me know your thoughts (both on our rationale of changing delete and
    opinions on "reassigning")

    Best,
    Alicia



    On Thu, Nov 5, 2015 at 3:52 PM, mvirtosu
    wrote:

    > Hi All,
    >
    > I would like to chime in and answer Cal's question: "what behavior would
    > you like to see?"
    >
    > I would personally would like to see the deletion appearing in the audit
    > trail and the event CRF being actually deleted instead of being replaced
    > with empty values.
    >
    > Looking forward to Alicia's response.
    >
    > Mihai
    >
    > --
    > To manage your email notifications, please visit:
    > https://www.openclinica.com/forums#/profile/preferences
    >
    > Reply to this email directly or follow the link below to check it out:
    > https://forums.openclinica.com/discussion/comment/17474#Comment_17474
    >
    >
  • mvirtosumvirtosu Posts: 276 ✭✭
    Alicia,

    I did not suggest hard delete. I meant the pre-3.6 delete (which enabled deleting versions as well) and also populating the audit trail about the delete.

    Mihai
  • toskriptoskrip Posts: 273 ✭✭✭
    Actually, I was already expressing my worries about keeping deleted data in audit trail in the previous thread and I was also proposing the solution of having this configurable on a study lvl and if not than there should be a possibility to anonymise subject which means generation of random ssid (if used also personid secondariid) without auditing this ID changes (you could also have such anonymization feature configurable parameter once setting up a study, so it will not be default for all studies).

    However back to the issue of CRF data delete. If I do not need to delete data for whole subject (in case of consent withdrawal), but I only need to delete one CRF data (because of mistake during data entry, that could be related e.g. with wrongly assigned CRF version). In this situation it is OK to do data deletion with empty values replacement (OK also for audit) but you should clean up the GUI. E.g. the delete button should no longer be accessible (there is nothing more to delete).

    Regarding the possibility to choose the CRF version, I have one very specific proposal that could be considered, because I also did not like how older (3.3) version of OpenClinica have been dealing with this. If the CRF

    1 . is scheduled (data entry not started) or
    2. the CRF was opened in edit mode (get data entry started status) but no data have been stored (edit mode closed without save) or
    3. the CRF was deleted (all data items are set to empty string)

    The data entry person should always be able to reassign such empty CRF to a new version. It is actually very common source of error, because data entry often schedule CRF and than go to edit mode but forget to change the default CRF version and afterwards they end up with wrong CRF version (data entry started) and are unable to correct it (need to ask data manager to do it for them).

    Tomas

  • mvirtosumvirtosu Posts: 276 ✭✭
    toskrip said:

    ...you should clean up the GUI. E.g. the delete button should no longer be accessible (there is nothing more to delete).

    I do agree with Tomas on this one. It does not make sense to have a delete button and "data entry started" status for deleted event CRFs.

    Mihai

Sign In or Register to comment.