Most trials we encounter use a parallel design, where the intervention and control groups are followed simultaneously from the time of allocation to one or the other. The next most common type of trial we might encounter is the cross-over trial.

In a cross-over trail, each participant is used at least twice – once in the control group and once in the intervention group-though the sequence may vary. This creates challenges for ADaM programming, so how do we tackle this? 

ADSL: 

ADSL is the ADaM to start with. ADaMIG requires a variable  called TRTSEQP (Planned Sequence of Treatments) to be added. For example, TRTSEQP can be populated with “Placebo – Drug A” or  “Drug A – Placebo” to distinguish between the two treatment sequences. A participant might receive a sequence of treatments different from what was planned, so TRTSEQA (Actual Sequence of Treatments) might be included as well.

The next question naturally arises regarding the timing variables. As each participant is used at least twice, there will be multiple treatment periods. In ADSL, we can add TR01SDT, TR02SDT, TR01EDT and TR02EDT to denote the date of first/last exposure in period 1 and 2, respectively. Note that the start of the second period (TR02SDT) is not equal to the end of the first period (TR01EDT) because there will be a washout period between the two periods to ensure that the effect of the treatment in the first period doesn’t carryover to the second period.

If you want to cite web pages and programs from the Statistical Research Work, please use:

Li, G. (2024). Statistical Research Work. Retrieved from https://statisticalresearchwork.com.