Non-parametric analysis of Tmax in a 2-treatment, 2-period, single-sequence study

Phoenix Certara Forums Phoenix WNL basics
1

Hi,

I am wondering if WinNonlin can perform non-parametric analysis to compare the median differences in Tmax between two periods.

The brief study design is as follows.

Period 1 (drug A) —> washout —> period 2 (drug A and B)

I want to calculate the median, interquartile range(IQR), median difference and its 90% CI and 95% CI.

Can I use linear mixed effect model to do this work?

Thank you!

Liaoliao

2

Dear Liaoliao,
I think what you are looking for is the option under toolbox ‘Crossover’. Take a look at it.
You can find details of this non-parametric assessment in the WNL User’s Guide.

Ana

3

Dear Ana,

Thank you for your reply!

As my understanding, the crossover in WNL is only for 2X2 crossover study.

The purpose of my study design is to investigate the effect of drug B on PK of drug A.

Period 1 (drug A) —> washout —> period 2 (drug B and A)

There is only one sequence. How to map the sequence when using the WNL crossover toolbox?

Also, can the bioequivalence toolbox can work for the one sequence study design? I should choose “crossover” as the type of study in Model option? How to map the data?

Thank you!

Liaoliao

4

Liaoliao, you should probably consult with your statistician as an analysis plan should have been in place to analyze your protocol before the study was run.

It is probable that a Wilcoxon signed rank test is non-parametric test that would be appropriate in your case since as you note the Crossover tool is requiring 2x2 design. Unfortunately it is not available as a menu item, we have already noted this as a potential enhancement QC_phx4410.

It may be that someone (Helmut?) has already coded this as a Phoenix workflow, but in the meantime I would suggest that you could use R for this component, potentially launched from Phoenix if you have a Connect licnese

Sorry I can’t be more helpful at this time, Simon

5

Hi Simon,

Thank you for your reply.

I have another question about the bioequivalence toolbox. Did the BE toolbox can be used for my one-sequence crossover study?

Thank you!

LLLi

6

THe Be wizard will take parallel designs so I think it should be OK but of course it isn’t suitable for assesment of Tmax.

Simon

7

Dear LLLi,

I think you need to dig into the R

This Helmut’s post is very useful:

http://forum.bebac.at/mix_entry.php?id=15518#p15519

Please note that CIs are not CIs like in parametric statistics, but Moses CIs (Hauschke D, Steinijans VW, Diletti E. A distribution-free procedure for the statistical analysis of bioequivalence studies. Int J Clin Pharm Ther Toxicol. 1990;28(2):72–8.)