You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 2 Next »

  1. Applicants may choose which disposition events and milestones to collect for a study. A milestone is a protocol-specified point in time that is not assigned to an epoch. A disposition event describes whether a subject completed the study or portion of a study (epoch), or the reason they did not complete. Applicants may collect 1 disposition event for the trial as a whole, or they may collect dispositions for each epoch of a trial.
  2. Disposition data may be collected on a CRF dedicated to disposition data, or collected across several forms that also contain data that are not DS. In the latter case, the disposition should be mapped to the appropriate DS tabulation variable (e.g., DSCAT, DSSTDTC, DSTERM, DSDECOD). The same disposition data should not be collected both on domain-specific forms and on a DS form.
    1. The collection DS domain does not specify where to collect protocol milestones within the CRF. Protocol milestones may be included in convenient places in the CRF. For example, informed consent date may be collected on the same CRF page as demographics data and mapped to the tabulation DS domain.
      1. The CDASH Model allows for date of death to be collected on any CRF deemed appropriate by the applicant and mapped to the tabulation DS domain; the date of death also may be collected as part of a DS form. However, consideration should be given to designing the CRF to collect the date of death only once in a study. DTHDAT is included in scenario 2.
  3. Controlled Terminology (NCOMPLT) is focused on disposition events, and is used when DSCAT is "DISPOSITION EVENT". Because the complete list of CT may not be appropriate, sponsors may choose to include only subsets of CT on the CRF. The choices that appear for a DS event are dependent upon the event itself, and the contents of the list can vary if data are collected for multiple epochs in a study. For example, "Non-Compliance with Study Drug" and "Lack of Efficacy” are not applicable choices prior to the start of treatment; "Failure to Meet Randomization Criteria" is an applicable choice only for the epoch that immediately follows the date of the subject's randomization.
  4. If DSUNBLND is "Yes", and the unblinding/unmasking resulted in the subject discontinuing the trial prematurely, DSTERM and DSDECOD capture the applicable details. If the unblinding/unmasking occurred due to an adverse event, DSTERM contains the text of the AE, and the AE “Were any other actions taken in response to this adverse event?” (AEACNOTH [optional]) free text may include text of “Treatment Unblinded”. DSUNBLND may also be used to document intentional unblinding at a protocol-defined point in the trial. DSUNBLIND is include in scenario 1. 
  5. DSCONT and DSNEXT data are used to aid in monitoring and data cleaning. Because the questions relate to future plans, the validity of the responses cannot be ascertained until the subject enters the subsequent epoch or new study. DSCONT and DSNEXT are include in scenario 2. 
  6. Associations between DSCAT and some DSDECOD codelist values are described in the DS codetable, available at https://www.cdisc.org/standards/terminology/controlled-terminology.
  • No labels