The scenario is as follows:
We are recruiting participants to a randomised controlled trial of a novel intervention. The recruitment will extend over a period of six months. Participants are randomised individually (no strata, blocks, etc.). Participants in the intervention group are supposed to engage with the intervention for three months, after which we wish to measure whether or not they have changed their behaviour (binary outcome).
We will be recruiting over an extended period, thus we will also end up measuring the outcome over an extended period. An individual that is part of the control group will be assessed at the same time as if he or she had access to the intervention (i.e. after three months).
The question is as follows:
Since the follow up measurements are not all gathered at the same point in time, should we take this into consideration when analysing our data? When collecting data at one point in time we would go ahead and pool all the data (and assume iid) and do a logistic regression analysis, but now we wonder if there are other considerations to take into account.
Thank you in advance.
