03-02-2014 05:46 PM
Hello, I am a newbie here and needed your help. Can anyone please let me know the difference between "intent to treat" (itt) and safety populations?
I was in the process of flagging some data which I was working on, and got stuck not knowing the difference.
Thanks for your help.
03-18-2014 08:19 AM
it is not defined somewhere? i think the itt may be larger, the safety are those people from which safety measurements are known like vital signs etc, thus actually treated. itt may contain drop-outs, or may be smaller than safety when it is decided that some primary objectives are relevant for a subgroup.
03-18-2014 09:45 AM
This is not really a SAS specific discussion. It should be defined in your specs/protocol somewhere what these are. As for actual definitions - Wikipedia:
So ITT is those randomized based on randomized treatment not what they were actually given.
Safety is all randomized with any treatment.
Again, might be slightly different depending on your specs.
03-19-2014 06:32 AM
to add some more definition
Includes all randomized patients in the groups to which they were randomly assigned, regardless of their adherence with the entry criteria, regardless of the treatment they actually received,
and regardless of subsequent withdrawal from treatment or deviation from the protocol. it simply means that any subject who is randomized is considered in ITT, even if the subject is not 100% compliant with the study, withdrew from study. So if the subject is once randomized is still considered for analysis.
subjects randomized and who took atleast one dose of study medication are considered as safety population.
01-02-2015 04:02 PM
A references are papesr by John Lachin "Statistical Considerations in the Intent to Treat Principle"
21 Controlld Clinical Trials 167 (2000); and Lewis and Machin "Intention to treat -- who should use
ITT?" 68 Br. J. Cancer 647 (1993). Pure ITT is irrespective of events occurring post-randomization
and classifies subjects wtih respect to the randomized treatment assignment. ICH E9 is also pertinent.
Note that ICH E9 refers to the "Full Analysis Set", which appears to me to be something less than pure
ITT -- calling for the analysis set to be as close as possible to the "ITT ideal". And in practice, sometimes
one will see trials using MITT (modified ITT) which requires the receipt of at least one dose of a study