Announcement

Collapse
No announcement yet.

Sequencing a Dataset or not

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Sequencing a Dataset or not

    Hi Greg,

    We decided to change the dataset structure to sequenced. The only drawback is not being able to display the number of AEs or Meds as a part of the CRF tracking report, so we will put together a separate report for those.

    Joe

    --------------------------------------------------------------------------
    From: Joe Reed
    Sent: Wednesday, January 11, 2006 3:24 PM
    To: Greg Wagner
    Cc: Ronni Rubenstein
    Subject: RE: follow-up question about sequenced datasets


    Thanks Greg. Yes, you understand the situation exactly and the 2 different options. The AE form that we have collects data for only one event at a time and is not necessarily related to a certain follow-up visit. The Meds form, however, does capture multiple records on one form.

    Your response is very helpful in deciding which option to pursue. We are thinking through the information, but leaning toward the sequenced dataset approach.

    Joe


    -----Original Message-----
    From: Greg Wagner [mailto:[email protected]]
    Sent: Wednesday, January 11, 2006 12:08 PM
    To: Joe Reed
    Cc: Ronni Rubenstein
    Subject: RE: follow-up question about sequenced datasets


    Hi Joe,

    This is always an issue on how to handle AE’s and Meds. In general I would say that you should capture the AE’s as sequenced records. Generally, there are multiple records on any given AE (or meds) page. I think what most clients do when AE pages are transmitted at each visit is to create a page definition for each visit for an AE page and enter all the AE for that visit under that one page. They create a display order just for AE’s that contain just the available AE pages/screens for the study. The same goes for Meds.

    If the pages are not collected at each visit, it becomes a bit more of a challenge. In this case, you basically have two choices. You can either create one AE page for the study and enter all the data against that one page as sequenced records – in this manner you loose the tracking of the individual CRF pages. Or, create a group of AE page definitions such as AE1, AE2, etc…, as many as you think may be required, and track a new page each time you receive one for a patient – this is a bit more difficult as you need to know what you have already received for the patient.

    My choice would be to either have an AE page for each visit if possible or create one and sequence the data. If you create only one page, you would only have one tracking record entered.

    Greg-

    --------------------------------------------------------------------------

    From: Joe Reed
    Sent: Wednesday, January 11, 2006 1:51 PM
    To: Greg Wagner
    Subject: follow-up question about sequenced datasets

    Hello Greg,

    I have another question on the CRF example that we sent you, which collects data at multiple timepoints. For the 1st database that I setup in ClinPlus, I defined the page number in the barcode so that it contains the specific visit checked on the CRF. For example, some page numbers for one of the CRFs in the current database look like:

    08Q_BASE (baseline visit)
    08Q_T2 (treatment week 2)
    08Q_FU12 (follow-up week 12)

    Then I created display orders to correspond to the specific sets of CRFs that we expect to receive (i.e., a display order listing all the baseline CRFs, one for the treatment week 2 CRFs, etc). For the above example, these records go into one dataset, Q, which is NOT sequenced. So this all works great and is doing exactly what I want it to!

    My question for you concerns the AE and Con Med CRFs. I used the same approach described above for these datasets, but now that we are entering test data and I am writing up the training instructions for data entry, I'm wondering if I should instead be defining the AE and Meds datasets as sequenced.

    What I currently have are page numbers defined that account for a conservative number of AE and Meds per subject. Then a separate display order for AE and for Meds. While this does allow multiple records to be entered correctly, I realize that it makes it difficult for the situation where we get AE or Med forms for a subject at various points during the study - - difficult because data entry needs to know what number AE or Med it is before entering the data and this would involve pulling the subject binder and counting the AEs or Meds. Plus navigating the display order when doing a mixture of AEs for different subjects could get confusing since after committing a record, the next page in the display order appears - but this page is not necessarily the appropriate page for data being entered for a new subject.

    If I change the AE and Meds datasets to sequenced, will this be accomplishing the same thing as the current design, but without the display order issue? Then would I define just one page number for the AE and for the Meds CRFs (versus now I have several defined to account for a conservative number of forms to be received)? Also, I'm unclear how data gets tracked into the system for a sequenced dataset. Do you just track the first CRF instance (i.e., the first AE received) and from then on update the sequence number to add new AEs?

    Thank you in advance for your help.
    Joe
    Last edited by Ronni Rubenstein; 01-17-2006, 10:38 AM. Reason: Change title
Working...
X