Having problems with message search?
Dear Elaine,
The response by Chris is nearly accurate. One
of the primary principles of ADaM is clear communication about the analysis
that was performed for a particular clinical trial. Therefore, when faced
with questions about how to represent important variables such as treatment and
treatment dates, the first question to ask yourself is how this information
will be used for the analyses. it sounds like you would not need 2
treatment variables in the non-ADSL analysis datasets. You could add a
supportive variable in non-ADSL domains that would indicate which period a
given observation was from if this is something of interest to your analysis.
However I do not agree that a generic
answer for implementing ADaM is ‘however you want to do it’. It
is true that the model is flexible. However, ADaM is trying very hard to
get to a point where ADSL is predictable in its content. Treatment start
and stop dates are required for all designs. Since you have 2 treatments
and 2 periods, you need 2 treatment variables, each with their own start and
stop dates in ADSL.
It is very difficult to respond to
questions like this without knowing more detail about the protocol design and
desired analyses so my response is based on limited knowledge of your specific problem.
Susan Kenny
From:
rtp_cdisc@yahoogroups.com [mailto:rtp_cdisc@yahoogroups.com] On Behalf Of Chris Decker
Sent: Friday, October 31, 2008
10:13 AM
To: rtp_cdisc@yahoogroups.com
Subject: Re: [rtp_cdisc] ADaM
question
Elaine,
We have the ADaM gurus in our user group ;-), so you should be able to get a
clear answer but I will take a shot. This is based on our experience
implementing in the CDISC Pilot. I probably have more questions for you then
answers but here it goes.
The generic answer is 'however you want to do it' as the the model is
flexible. However, the real question is how would you plan to use those
variables. If you have an analysis that needs the treatment for both
periods (even though they are the same) then include both. If you are not
going to use them then I suggest not including it and just having one
treatment. The other question is: Do you plan to also collect the
period date information within ADSL - TRT1STDT, TRT2STDT, etc. If you are
collecting this information within ADSL then I would suggest including TRT2P so
you have consistency across your variables.
My last question is are you using the new (Draft) IG because I believe, from
the conversations in the Pilot, that there are some subtle differences in how
the TRT variables are interpretated. I would follow the IG 2.1 guide.
Disclaimer: This is only from my one experience and I would recommend
waiting until one of the ADaM team members who are part of our group chime in.
Chris
On Thu, Oct 30, 2008 at 1:35 PM, Elaine <elaine.dempsey@
Can anyone answer this
ADaM question for me? This is my first
experience trying to conform to CDISC ADaM guidelines.
I have a study that has two treatment periods. However, it is not a
crossover study. The same treatment remains in effect across the two
periods. So, do I just need to populate TRT1xx variables since there
is only one treatment? Or, do I also need to have the TRT2xx variables
to represent the second study period but set them to the same value as
the TRT1xx variables?
Bear in mind that this will be combined with other studies in an ISS
that have only one period so will not have the TRT2xx variables.
Is there any guideline on something like this?
Thanks.
Elaine Dempsey
Offline 
Send Email
