d_labes
★★★

Berlin, Germany,
2008-01-23 17:00
(6166 d 21:50 ago)

Posting: # 1547
Views: 11,145
 

 Method for replicate design [Nonparametrics]

Dear all,

can anybody give me a hint about non-parametric evaluation of BE (average) within replicate cross-over design (2-sequence-4-period and/or 4-sequence-4-period)?
This will be needed for Tmax evaluation, for which most guidances recommend a non-parametric evaluation.

Thanks & regards

DLabes

Regards,

Detlew
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2008-01-23 17:50
(6166 d 21:00 ago)

@ d_labes
Posting: # 1548
Views: 9,717
 

 Method for replicate design

Dear DLabes!

❝ can anybody give me a hint about non-parametric evaluation of BE (average) within replicate cross-over design (2-sequence-4-period and/or 4-sequence-4-period)?


First of all congratulations for being the first colleague posting in my personal favorite category “Nonparametrics”
… but such a difficult question! :ponder:
To my knowledge there’s simply no specific method published at all.
Since replicate designs were at least to some degree submitted to the FDA, tmax did not play an important role, because evaluation of early exposure (partial AUC up to median tmax of the reference – rather than a direct comparison of tmax itself) performed by a parametric method is considered the standard over there.

❝ This will be needed for Tmax evaluation, for which most guidances recommend a non-parametric evaluation.


You are right. But if you go beyond the conventiona 2×2 design, nonparametrics will leave you out in the rain. I know of just two publications (for a 2-treatment, 2-period, 4-sequence design1 and a 6×3 Williams’ design2), although
  • both of them don’t deal with confidence intervals,
  • rely on normal approximations (i.e., are not exact), and
  • don’t help you in the case of a 2×4 or 4×4 replicate design…
One possibility comes into my mind: Performing a parametric evaluation of the replicate design, but not on the untransformed raw data of tmax, but on the ranks within each period.
But I must confess, I have neither done such an analysis myself nor seen it anywhere…

  1. Elswick RK, Uthoff VA. A nonparametric approach to the analysis of the two-treatment, two-period, four-sequence crossover model. Biometrics. 1989;45(2):663–7.
  2. Bellavance F, Tardif S. A nonparametric approach to the analysis of three-treatment three-period crossover designs. Biometrika. 1995;82(4):865–75. doi:10.1093/biomet/82.4.865.

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
d_labes
★★★

Berlin, Germany,
2008-01-24 09:36
(6166 d 05:14 ago)

@ Helmut
Posting: # 1550
Views: 9,531
 

 Method for replicate design

Dear HS,

❝ First of all congratulations for being the first colleague posting in my personal favorite category “Nonparametrics”...


thanks.

But non-parametrics is not in the scope of standard statistical education or standard statistical text books.
Moreover there is not a lot of software out there to deal with.
In SAS for instance, which I have to deal with, no standard method is implemented for obtaining CIs. You have to program methods from the scratch. Not an easy task.

❝ One possibility comes into my mind: Performing a parametric evaluation of the replicate design, but not on the untransformed raw data of tmax, but on the ranks within each period.


Ok, that may help for p values but not for confidence intervals, at least not in SAS.

Thanks a lot for the references.

I have some more for you or whom it concerns, dealing with confidence intervals for designs with p>2 periods, but not specifically with all forms of replicate designs.

[1] Putt, ME and VM Chinchilli
    Nonparametric approaches to the analysis of crossover studies
    Statistical science 19(4), 712-719 (2004)
[2] Duchateau L and P Jansen
    Pairwise non-parametric non-inferiority tests in 3x3 cross-over trials: should we adjust for period
    Statist. Med. 24, 1525-1536 (2005)
[3] Willavize SA and EA Morgenthien
    Non-parametric confidence intervals for tmax in sequence-stratified crossover studies
    Pharmaceut. Statist. (in press) published online Wiley Interscience

Regards,

Detlew
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2008-01-24 13:13
(6166 d 01:36 ago)

@ d_labes
Posting: # 1552
Views: 9,425
 

 Method for replicate design

Dear DLabes!

❝ But non-parametrics is not in the scope of standard statistical education or standard statistical text books.

  1. That’s a pity!
  2. … but in the context of BE extensively covered in ‘specialized’ textbooks:
    Chow & Liu (2000), Senn (2002), Jones & Kenward (2003), Patterson & Jones (2006), Hauschke et al. (2007) – for references see this post.

❝ Moreover there is not a lot of software out there to deal with.


True. At least for a 2×2 design EquivTestPK would help, but even the freeware R is an option (references).

❝ In SAS for instance, which I have to deal with, no standard method is implemented for obtaining CIs. You have to program methods from the scratch. Not an easy task.


OK, but you may opt for Cytel’s StatXact PROCs for SAS. Caution: The correct Hodges-Lehmann CI is not available from the pull-down menu but must be obtained from the command line according to Senn (2002 p.121). I’ve had a lot of conversations with Cytel’s support in the past (and finally gave up with v6.1 of 2003 – maybe the issue is resolved in the current version 8).

❝ ❝ One possibility comes into my mind: Performing a parametric evaluation of the replicate design, but not on the untransformed raw data of tmax, but on the ranks within each period.


❝ Ok, that may help for p values but not for confidence intervals, at least not in SAS.


OK, since I’ve got no SAS license, this is beyond my horizon. :-D
Thanks a lot for your references!

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
kimhuang
☆    

China,
2023-06-14 12:43
(546 d 03:07 ago)

@ Helmut
Posting: # 23592
Views: 1,739
 

 Method for replicate design

Dear Helmut,

Is there any non-parametric evaluation of BE (average) within replicate cross-over design (2-sequence-4-period and/or 4-sequence-4-period) now?


Edit: Full quote removed. Please delete everything from the text of the original poster which is not necessary in understanding your answer; see also this post #5[Helmut]
UA Flag
Activity
 Admin contact
23,336 posts in 4,902 threads, 1,666 registered users;
39 visitors (1 registered, 38 guests [including 10 identified bots]).
Forum time: 14:50 CET (Europe/Vienna)

I’m all in favor of the democratic principle
that one idiot is as good as one genius, but I draw the line
when someone takes the next step and concludes
that two idiots are better than one genius.    Leo Szilard

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5