BE-proff
●    

2016-04-06 23:39
(2913 d 00:37 ago)

Posting: # 16167
Views: 9,261
 

 PowerTOST for steady state [🇷 for BE/BA]

Hi All,

Is it possible to calculate number of subjects for steady state study with PowerTOST? :confused:

What parameters are required?


Edit: Category changed. [Helmut]
d_labes
★★★

Berlin, Germany,
2016-04-07 12:54
(2912 d 11:22 ago)

@ BE-proff
Posting: # 16168
Views: 8,339
 

 PowerTOST for steady state

Dear BE-proff,

❝ Is it possible to calculate number of subjects for steady state study with PowerTOST? :confused:


Of course, it's the Gold Standard :-D!

❝ What parameters are required?


The same as for sample size estimation for other types of BE studies.
  • An educated guess of the variability of your primary targets (usually AUCtau and Cmax,ss in case of steady state studies)
  • target power (>=80% usually), alpha (0.05 usually)
  • 'True' GMR for which power should be calculated (usually set to 0.95 if no other information is available)
  • design of the study
Thats all. More arguments of the sample size estimation functions in PowerTOST you may find in the man pages. These are more or less only for special purposes.

Regards,

Detlew
BE-proff
●    

2016-04-07 17:42
(2912 d 06:34 ago)

@ d_labes
Posting: # 16171
Views: 8,237
 

 PowerTOST for steady state

Hi d_labes,

Could you share a sample of script for calculation?

How to write design type in it? :confused:
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2016-04-07 18:09
(2912 d 06:06 ago)

@ BE-proff
Posting: # 16172
Views: 8,229
 

 RTFM

H BE-proff,

❝ Could you share a sample of script for calculation?


As Detlew wrote above

❝ ❝ More arguments of the sample size estimation functions in PowerTOST you may find in the man pages.


Type help(package=PowerTOST) in the R-console or use the online-version. Most interesting functions are sampleN.TOST() for ABE, sampleN.scABEL() for the EMA’s ABEL, and sampleN.RSABE() for the FDA’s RSABE.
At the end of all man pages you will find examples.

❝ How to write design type in it? :confused:


RTFM (argument design) of the sampleN-functions:
  • sampleN.TOST
    Character string describing the study design.
    See known.designs() for designs covered in this package.

  • sampleN.scABEL and sampleN.RSABE
    Design of the study to be planned.
    2x3x3 is the partial replicate design (TRR|RTR|RRT).
    2x2x3 is the 3-period replicate design (TRT|RTR).
    2x2x4 is the full replicate design with 2 sequences and 4 periods.
    Defaults to design="2x3x3"
Get others from known.designs(). The column design shows the argument to specify in the functions (e.g., design="2x2x2" or design="parallel").

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
BE-proff
●    

2016-04-09 22:15
(2910 d 02:00 ago)

@ Helmut
Posting: # 16176
Views: 8,098
 

 RTFM

Hi Helmut,

I have tried to enter known.designs() and it returned the following list:
1            2 parallel groups
2                2x2 crossover
3              2x2x2 crossover
4                3x3 crossover
5              3x6x3 crossover
6                4x4 crossover
7    2x2x3 replicate crossover
8    2x2x4 replicate crossover
9    2x4x4 replicate crossover
10   partial replicate (2x3x3)
11            Balaam's (2x4x2)
12 Liu's 2x2x2 repeated x-over
13                paired means


Where is my so wanted steady state design? :-)
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2016-04-10 02:50
(2909 d 21:26 ago)

@ BE-proff
Posting: # 16177
Views: 7,983
 

 RTFM‼

BE-proff,

❝ Where is my so wanted steady state design? :-)


Are you a troll? You estimate the sample sample for the PK metrics of interest (see this post) and the values Detlew stated already above. Only you know which design you will use! Plug the argument in one of PowerTOST’s functions. What’s your problem‽

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
BE-proff
●    

2016-04-10 15:42
(2909 d 08:34 ago)

@ Helmut
Posting: # 16178
Views: 7,945
 

 RTFM‼

Hi Helmut,

I am not a troll :)
Below is a TOST example for calculation of parallel design study.

sampleN.TOST(CV=0.23, design="parallel")

+++++++++++ Equivalence test - TOST +++++++++++
            Sample size estimation
-----------------------------------------------
Study design:  2 parallel groups
log-transformed data (multiplicative model)

alpha = 0.05, target power = 0.8
BE margins        = 0.8 ... 1.25
Null (true) ratio = 0.95,  CV = 0.23

Sample size (total)
 n     power
46   0.802730


What word must be written instead of "parallel" for calculate sample for steady state study?
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2016-04-10 15:48
(2909 d 08:27 ago)

@ BE-proff
Posting: # 16179
Views: 7,958
 

 RTFM‼

[image]Hi BE-proff,

how shall I know in which design you plan to perform the steady state study?

❝ What word must be written instead of "parallel" for calculate sample for steady state study?


Heck, a steady state study can be done in any of the designs given by known.designs()!
If it’s not a parallel but a simple crossover use design="2x2x2". If it’s a four-period full replicate use design="2x2x4", et cetera, ad libitum, ad nauseam.

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
BE-proff
●    

2016-04-12 15:36
(2907 d 08:40 ago)

@ Helmut
Posting: # 16186
Views: 7,800
 

 RTFM‼

Hi Helmut,

Yes! Now it is clear to me!
Many thanks for your help and patience.
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
84 visitors (0 registered, 84 guests [including 7 identified bots]).
Forum time: 23:16 CET (Europe/Vienna)

Nothing shows a lack of mathematical education more
than an overly precise calculation.    Carl Friedrich Gauß

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