## protocol amendment that may trigger Group effect test [General Sta­tis­tics]

Hi Silke,

» Situation:
» 1st period was dosed in 1 group and completed in CRO 1
» 2nd period must now be performed at a different CRO (CRO 2)
» 2nd period must be performed in 2 groups in CRO 2

That's annoying

» Protocol amendment:
» Switch in CRO including all tasks
» due to the change of dosing in 1 group to 2 groups in period 2 the washout will be different between the two groups.
»
» Does this trigger the evaluation of a group effect, although the groups may only be separated by 1 week?
»
» Is there special statistics we need to consider to address this properly?

In my opinion it is not overly informative to do a group analysis here, at least not within the fit which provides for the evaluation of BE (here it is a kind of "accidental group within period within CRO", where period somehow is indistinguishable from CRO ).

Can you switch the volunteers to the other CRO, or has it become a parallel trial rather than a crossover trial by now?

This week's list of things I absolutely detest: Corona virus, the which function in R, WIA-WIA interfaces for scanning under Windows 10, the Bee Gees, the smell of my fridge.

Best regards,
ElMaestro