Can there be multiple mITTs in a single study? For example, researchers in below paper defined and used 3 mITTs, which they have pulled from a singlt ITT.
http://www.ncbi.nlm.nih.gov/pubmed/19705345
Yes, there could be! you can have two sets of mITT population for efficacy parameters and safety parameters.
It is true that you may consider different sets of populations, depending on the objectives that you plan to evaluate, and as Mohd Kashif has mentioned ther could be one mITT for safety and another for efficacy, and even a tird one for pharmacokinetics. However I would say that this is very unsual, and in fact I have never seen published more than one mITT set for efficacy, which I presume this would be your issue. Nevertheless what it is possible, is to mix mITT, ITT and PP approaches for different analyses. For instance, in a meta-analysis published in 2010 in BMJ (http://www.bmj.com/content/bmj/340/bmj.c2697.full.pdf) it is mentioned that:
"In 12 trials (0.2%) the modified intention to treat approach was used as a secondary analysis. In these studies the primary analysis was by intention to treat (11 trials) and by per protocol (one trial)."
In any case, if you have a good reason for using more than one mITT set you should mention it in advance in your study protocol and SAP and include the rational supporting it as well as evaluate the potential bias on your planned analyses.
Thanks Javir. Yes, it is very clear from your answer that though unusual, if needed, a sub-population can be defined as mITT provided it should be mentioned in advance in the protocol with proper justification. This answered my question.
Good to know that you got your answer! I couldn't get time to go through your publication to understand how author had defined the population. Even for different sets of efficacy outcomes there could be more than two mITT populations provided it has been prior defined as indicated by Mr. Javier
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts