Sample size calculation crossover design

[ANKH1]

Hello, 

We are designing a cross-over repeated measures design. Participants will have all 4 treatments at four different sessions. During each session, samples will be collected over several time points for 2 hrs. 

Is there a SAS procedure for such calculation?

Thank you!

[Patrick]

Are you asking for a SAS procedure creating your design or how to analyse the data you gather.

For creating the design: If you've got SAS/Stat then Proc Plan could eventually be useful for you.

https://go.documentation.sas.com/?docsetId=statug&docsetVersion=14.3&docsetTarget=statug_plan_exampl... 

[ANKH1]

Hello, 

Thank you for your response. This is helpful. But I have two questions:

1) Is this proc necessary when there is a washout period? Meaning that a carry-over effect is null or minimized?

2) Is there a proc to calculate sample size? 

Thanks, 

[Patrick]

@ANKH1 

I'm a bit out of my depth here. May be someone else can answer your question.

@Reeza ?

[Reeza]

PROC POWER will calculate sample sizes, not sure if your crossover design can be accommodated, but there's an example here:

https://documentation.sas.com/?docsetId=statug&docsetTarget=statug_power_examples03.htm&docsetVersio...

I'll also move this to the statistical forum. 

IME, whenever you start getting into customized trial designs, simulations are the best method to determine your sample size using a bootstrap approach. 

---

@ANKH1 wrote:

Hello, 

We are designing a cross-over repeated measures design. Participants will have all 4 treatments at four different sessions. During each session, samples will be collected over several time points for 2 hrs. 

Is there a SAS procedure for such calculation?

Thank you!

---

[ANKH1]

Thank you! The example is helpful.