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

2»

Comments

  • agoodwinagoodwin Posts: 131 admin
    Hi All,

    This change had a lot of complexities with it. Some were regulatory, some were usability and some were technical.

    We have spent a lot of time going over regs (including but not limited to FDA and EMA). We have found compelling reasons to NOT allow deletion of records but we can't find anything that says that we SHOULD allow for deletion. @kristiak mentioned that he was going to talk with legal about guidelines. I'd be interested to hear results.

    Regarding usability and the status of the event CRF now being displayed as "Started" - unfortunately this is a deeply rooted issue in OpenClinica in how it has been architected. If item data exists then the status must be started. To change this in the current OpenClinica code base would require "open heart surgery". THis would be extremely time consuming and introduce a lot of risk.

    @mvirtosu , when you say you would like to to work as it did before, we just can't do it that way any more. Previously we recorded that the event CRF was deleted but we did NOT track the items/values that were deleted. At the time this was done intentionally and we felt it met the regulations at the time, but this is no longer true so we had to change this to record the items and values (not just the event CRF being deleted). Previously when you deleted an event CRF, the item data table references were removed and new item data were created for the event CRF. This is why the status went back to "not started" (because there were no associated item data). For now, we keep the item references and everything is retained in the audit, down to the item-level data changes.

    *NOTE: We have never done a hard delete (the item data still exist in the item data table) but in older versions we did remove the references and therefor the event CRF is seen as deleted but the items are not seen in the audit.

    We thought long and hard and tried to come up with different solutions and we felt this was the least risky, met most of the needs and still met all the regulatory requirements. We are still open to feedback and will certainly keep all of this in mind for our new forms engine.

    This has been a long thread and I'm happy to answer any questions I can about the approach, I'm also happy to have a call about it if anyone feels it would be easier to have a telecon :)

    Thanks for all the feedback - we will give it strong consideration in new development and will continue to think about ways to make things more intuitive, easier and more fun to work with.

    Best Regards,
    Alicia
  • kristiakkristiak Posts: 1,338 ✭✭✭
    via Email
    Hi Alicia,

    I thought I did inform you about the outcome of the discussions I had with the authorities. Here is their answer again.
    On aug 25 2015 I got a response from the Data Inspection Agency. Their resonce in summary said. " According to §12 in the data protection law, a patient who decides to withdraw his/her consent to participate in a study will be allowed to do so. The data can be kept up to the date of the withdrawal but no data can be added or modified. No data maybe collected after the withdrawal. The patient cannot require to have old data deleted. However to comply with GCP quality standards the data maybe unusable since it may no longer meet expected quality standards and thus must be completely unidentifiable, meaning there must be no possibility to trace the data back to the individual from whom the data was originally collected"

    This is not a verbatim translation. Would ypu want a verbatim translation of §12?


    Best

    Krister
  • agoodwinagoodwin Posts: 131 admin
    Hi @kristiak ,
    Thanks!! I must have overlooked it before. Based on my interpretation of that it seems that we are doing the correct thing by not deleting (and as discussed in previous threads going forward there may be another solution for making "completely unidentifyable". Although the later case may be better for a different discussion and is its own use case.

    Best,
    Alicia
  • mvirtosumvirtosu Posts: 276 ✭✭
    agoodwin said:

    in older versions we did remove the references and therefor the event CRF is seen as deleted but the items are not seen in the audit.

    Alicia,

    Thank you for your response. One last question: would it be possible to remove the references but also add each deleted data element to the audit trail?

    Mihai

  • agoodwinagoodwin Posts: 131 admin
    via Email
    Hi Mihai,

    This also did not seem possible. We entertained this but this did not seem
    like a viable solution. One thing that could possibly be done is to run a
    service to populate the "deleted" records to the audit trail but this would
    be complex to do every time and would introduce performance considerations
    (the same code would not handle this).

    As I mentioned before, we'll find a much more elegant way to handle this
    with the new forms engine.

    Best Regards,
    Alicia

    On Thu, Nov 19, 2015 at 11:36 AM, mvirtosu
    wrote:

    > in older versions we did remove the references and therefor the event CRF
    > is seen as deleted but the items are not seen in the audit.
    >
    >
    >
    >
    > Alicia,
    >
    > Thank you for your response. One last question: would it be possible to
    > remove the references but also add each deleted data element to the audit
    > trail?
    >
    > 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/17543#Comment_17543
    >
    >
  • kristiakkristiak Posts: 1,338 ✭✭✭
    via Email
    Yes! Alicia, that is the correct interpretation. I have since heard of a case where a patient insisted on having his data deleted but he was not successful.

    Best
    Krister
  • mikewormikewor Posts: 35
    @Alicia. I've seen numerous occasions where 'document state' and 'process state' are confused or simply collapsed into one, thus leading to lots of spaghetti code. Unfortunately I don't know enough of how CROs operate to be certain, but maybe this is a similar situation and introduction of a 'process state' could help. Its even possible that a customisable 'state sequence diagram' is included which would allow different organisations to match their current workflows.
  • Hi,
    I have a problem which I believe may be related to the change detailed above.

    A data manager began entering data from a CRF into the incorrect CRF (entering data from one form into a completely different CRF). They recognized that it was not the correct CRF and stopped data entry without completing the CRF. In the past, if this occurred, we were able to complete the CRF and then completely delete the CRF and it would change back to status "Not Started". However, now, it remains as "Initial Data Entry Started" perpetually. Since this CRF is not used for all cohorts, it should remain completely blank for this subject. Now I cannot set the status of the entire event to "completed" because this CRF is still identified as "initial data entry started".

    We tried removing the CRF and then changing the status of the event to "completed", but the only options available are "stopped" or "data entry started" so the event will never be completed and we keep checking it to determine why.

    Is there any way to remedy this situation? Why isn't the event status changeable to "completed" if we have removed this CRF?

    Thank you,
    LaShaunda
  • lkeitalkeita Posts: 50
    Hi LaShaunda,

    If you set the Event Status to Stopped. Then edit the Event Status again and you'll have the option to set it to Completed.

    I hope this helps.

    Laura
  • Hi Laura,
    Thanks for all of your assistance. I removed the errant CRF which changed the status of the CRF to "invalid". I changed the event status to "stopped" and then changed it to "completed". The event status is now completed.

    When I did this, it also changed the CRF which I removed and which had previously been identified as "invalid" to "completed". I removed the errant CRF again, and this changed the status of the CRF to "invalid" once again.

    Thanks again for your help, I'm sure I owe you lunch so please let me know when you visit Cleveland.

    LaShaunda
Sign In or Register to comment.