All content on this Wiki is non-binding and any individual opinions expressed should not be considered indicative of the policies or positions of CDISC or any other organization.
Page tree
Skip to end of metadata
Go to start of metadata

COVID-19 disruptions may lead to protocol deviations at the site level. The SDTMIG DV domain represents deviations at the subject level, but does not provide a mechanism for representing deviations using a site identifier. In addition, a deviation at the site level might not affect all subjects at that site (e.g., subjects who have completed the study, future subjects). A site-level deviation can be represented in DV in 2 ways, as illustrated by the following examples.

Example

In this example, when a protocol deviation that affected subjects occurred at a site, the sponsor determined which subjects were affected by the deviation and recorded the deviation for each subject at the site. The sponsor used DVCAT = "SITE DEVIATION" to identify site-level deviations.

 dv.xpt
Rows 1, 2, 4, 8:Show subjects who were affected by a site shutdown due to a flooded basement, a problem not related to COVID-19.
Rows 3, 5-7, 9:Show subjects were affected when all transportation in the country was down due to COVID-19.

dv.xpt

Row

STUDYID

DOMAIN

USUBJID

DVSEQ

DVREFID

DVSPID

DVTERM

DVDECOD

DVCAT

DVSCAT

EPOCH

DVSTDTC

DVENDTC


DVREASDVEPRELI
1CVD-3DV3001

Site shut down
SITE DEVIATION
TREATMENT2020-02-172020-02-20
Flooded basement
2CVD-3DV3011

Site shut down
SITE DEVIATION
TREATMENT2020-02-172020-02-20
Flooded basement
3CVD-3DV3012

Site shut down
SITE DEVIATION
TREATMENT2020-04-062020-04-18
All transportation in country down due to COVID-19.Y
4CVD-3DV3031

Site shut down
SITE DEVIATION
TREATMENT2020-02-172020-02-20
Flooded basement
5CVD-3DV3061

Site shut down
SITE DEVIATION
TREATMENT2020-04-062020-04-18
All transportation in country down due to COVID-19.Y
6CVD-3DV3071

Site shut down
SITE DEVIATION
TREATMENT2020-04-062020-04-18
All transportation in country down due to COVID-19.Y
7CVD-3DV3091

Site shut down
SITE DEVIATION
TREATMENT2020-04-062020-04-18
All transportation in country down due to COVID-19.Y
8CVD-3DV3121

Site shut down
SITE DEVIATION
TREATMENT2020-02-172020-02-20
Flooded basement
9CVD-3DV3122

Site shut down
SITE DEVIATION
TREATMENT2020-04-062020-04-18
All transportation in country down due to COVID-19.Y

DV NSV Metadata

Example


In this example, when a protocol deviation that affected subjects occurred at a site, the sponsor determined which subjects were affected by the deviation and used POOLDEF to define a pool for each deviation. The sponsor used DVCAT = "SITE DEVIATION" to identify site-level deviations.

 pooldef.xpt

pooldef.xpt

RowSTUDYIDPOOLIDUSUBJID
1CVD-3SITEDEV1300
2CVD-3SITEDEV1301
3CVD-3SITEDEV1303
4CVD-3SITEDEV1312
 5CVD-3SITEDEV2301
6CVD-3SITEDEV2306
 7CVD-3SITEDEV2307
 8CVD-3SITEDEV2309
9CVD-3SITEDEV2312
The pool identifier (POOLID) was then used in the DV domain.
 dv.xpt
Row 1:Shows a protocol deviation for the pool of subjects given the identifier "SITEDEV1".
Row 2:Shows a protocol deviation for the pool of subjects given the identifier "SITEDEV2".

dv.xpt

Row

STUDYID

DOMAIN

USUBJID

POOLID

DVSEQ

DVREFID

DVSPID

DVTERM

DVDECOD

DVCAT

DVSCAT

EPOCH

DVSTDTC

DVENDTC


DVREASDVEPRELI
1CVD-3DV SITEDEV11

Site shut down
SITE DEVIATION
TREATMENT2020-02-172020-02-20
Flooded basement
2CVD-3DV
SITEDEV21

Site shut down
SITE DEVIATION
TREATMENT2020-04-062020-04-18
All transportation in country down due to COVID-19.Y

DV NSV Metadata

See Missed Visits for missed visits as protocol deviations and an alternative approach using a custom domain.

  • No labels

8 Comments

  1. Can we use the name DVEPIND instead of DVRELEP, similar to the indicator flag in other domains (DS, EC, EX, VE)?  [I also do not like AERELEP because AERELxxx is used for Causality to other study drugs (IG 4.1.2.8.3). The Model says -REL “Records the investigator's opinion as to the causality of the event to the treatment” and sponsor’s have their own codelist such as RELATED/NOT RELATED for AERELxxx variables.]  I prefer a consistent generic indicator flag -EPIND across domains with label Epidemic Related Indicator. 

    1. I had used --RELEP to parallel

      • --REL, Relationship to study drug
      • --REL, Relationship to non-study treatment
      • --RELDEV Relationship to device

      I suspect you are thinking of NSVs for relationships to specific study drugs, for studies with multiple study drugs.  Is that right?

  2. Yes. In AE, for example, one might have AERELACY (Relationship to Acyclovir), AERELERL (Relationship to Erlotinib), etc. where these are other study drugs in the study.

    1. The NSV name was changed from "RELEP" to "EPREL" in all examples where it appeared.

  3. Once examples 2 and 3 have been updated will we be given another opportunity for review?

    1. I'm not sure what update you're expecting – the realistic examples of site-level deviations? I'm depending on others for those examples.

      The guide needs to be completed by , but that is the only constraint on opportunities to review.

  4. It is becoming clear internally that categorization of deviations is going to be required and controlled terminology will be beneficial for this (there was a separate conversation on this but I can't find it).  I will raise at the core team call today and I can share the list we are working on as an example

  5. Is the SUPPVE variable supposed to be the same as DVEPRELI? It seems it has been updated to "Epidemic Related Change Indicator".