Bioequivalence and Bioavailability Forum 14:36 CEST

Main page Policy/Terms of Use Abbreviations Latest Posts

 Log in |  Register |  Search

roro2907
Junior

Ireland,
2018-03-14 10:43

Posting: # 18530
Views: 3,207
 

 Error in Number of Subjects [R for BE/BA]

Dear All,

I am using "bear" to carry out a BE study. I have noticed that it will only allow a maximum of 14 subjects. My study contains data for 27 subjects. Would it be possible to change this?
yjlee168
Senior
avatar
Homepage
Kaohsiung, Taiwan,
2018-03-14 11:07

@ roro2907
Posting: # 18531
Views: 3,076
 

 Error in Number of Subjects

Dear roro2907,

It sounds unlikely. I never release any trial version, or restrict subject number. However, I cannot figure out why you have such an error or error messages. Are you running with the demo dataset? If yes, please turn off (uncheck) the demo run and try again.

All the best,
---Yung-jin Lee
bear v2.8.3-2:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan http://pkpd.kmu.edu.tw/bear
Download link (updated) -> here
roro2907
Junior

Ireland,
2018-03-14 11:28

@ yjlee168
Posting: # 18532
Views: 3,061
 

 Error in Number of Subjects

Dear yjlee168,

No I have not selected the demo dataset. I have tried with my own data: I included 14 subjects and it worked, then when i tried with 27 subjects it gave me this message:

Error in lm.fit(x, y, offset = offset, singular.ok = singular.ok, ...) :
NA/NaN/Inf in 'y'

:confused:
yjlee168
Senior
avatar
Homepage
Kaohsiung, Taiwan,
2018-03-14 11:54

@ roro2907
Posting: # 18534
Views: 3,045
 

 Error in Number of Subjects

Dear roro2907,

The error implies that your dataset has errors. Please check your dataset carefully. But I have tried to avoid the error long time ago, by screening the dataset first before analyzing it. What is the version of bear that you are using? The current version is v2.8.3. You can find the download site in my signature panel. With the current version, it will show you which subject' data is not correctly input.

» Error in lm.fit(x, y, offset = offset, singular.ok = singular.ok, ...) :
» NA/NaN/Inf in 'y'

All the best,
---Yung-jin Lee
bear v2.8.3-2:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan http://pkpd.kmu.edu.tw/bear
Download link (updated) -> here
roro2907
Junior

Ireland,
2018-03-14 12:54

@ yjlee168
Posting: # 18538
Views: 3,043
 

 Error in Number of Subjects

Dear yjlee168,

As far as I am aware there are no errors in my dataset. I am using the most recent version of "bear" with the most recent version of R. It does not specify which subject that the errors are included. Like I said before, it works fine if i select 13 or 14 subjects, but when I include any more subjects i get the above error message. I am selecting the NCA-Statistical analysis option in "bear".
yjlee168
Senior
avatar
Homepage
Kaohsiung, Taiwan,
2018-03-14 13:07

@ roro2907
Posting: # 18540
Views: 3,037
 

 Error in Number of Subjects

Dear roro2907,

interesting. Could you send me your dataset (with 27 subjects)? Thanks.

» As far as I am aware there are no errors in my dataset...

All the best,
---Yung-jin Lee
bear v2.8.3-2:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan http://pkpd.kmu.edu.tw/bear
Download link (updated) -> here
roro2907
Junior

Ireland,
2018-03-14 13:08

@ yjlee168
Posting: # 18541
Views: 3,033
 

 Error in Number of Subjects

Dear yjlee168,

I have found the reason why it was not working, the dataset sent to me had a 0 for conc. when it should have been NA. All is resolved. Thank you for your help.
yjlee168
Senior
avatar
Homepage
Kaohsiung, Taiwan,
2018-03-14 13:32

@ roro2907
Posting: # 18542
Views: 3,030
 

 Error in Number of Subjects

Dear roro2907,

I am very happy to hear that the problem has been solved. But I still wonder why the dataset can pass the screening or data check. :ponder: Could you please give me a example here? I may need to fix it later. Thanks.

All the best,
---Yung-jin Lee
bear v2.8.3-2:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan http://pkpd.kmu.edu.tw/bear
Download link (updated) -> here
roro2907
Junior

Ireland,
2018-03-14 13:41

@ yjlee168
Posting: # 18543
Views: 3,018
 

 Error in Number of Subjects

Dear yjlee168,

The reason why it worked for the first 14 subjects was due to the data error mentioned above being in subject15 so it didn't process it after this.
roro2907
Junior

Ireland,
2018-03-14 16:03

@ yjlee168
Posting: # 18545
Views: 2,988
 

 Error in Number of Subjects

Dear yjlee168,

Is it possible to change the Confidence Interval value? It is set at 90% but is it possible to change it for different BE studies?
yjlee168
Senior
avatar
Homepage
Kaohsiung, Taiwan,
2018-03-14 19:14
(edited by yjlee168 on 2018-03-14 21:46)

@ roro2907
Posting: # 18547
Views: 2,985
 

 other than 90% CI in BE study?

Dear roro2907,

Technically, yes it of course can be done. But I think it is not necessary in regulatory consideration of BE study worldwide.

» Is it possible to change the Confidence Interval value? It is set at at 90% but is it possible to change it for different BE studies?

All the best,
---Yung-jin Lee
bear v2.8.3-2:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan http://pkpd.kmu.edu.tw/bear
Download link (updated) -> here
roro2907
Junior

Ireland,
2018-03-15 10:19

@ yjlee168
Posting: # 18548
Views: 2,929
 

 other than 90% CI in BE study?

Dear yjlee168,

I am just curious as I am reviewing a BE study which has used a 94.12 confidence interval. Do you have any ideas as to why this is?

Thanks,

roro2907
yjlee168
Senior
avatar
Homepage
Kaohsiung, Taiwan,
2018-03-15 11:48

@ roro2907
Posting: # 18551
Views: 2,906
 

 Is it a two-stage BE study?

Dear roro2907,

Is it a two-stage BE study? Please have a look on Helmut's presentation slides. If yes, please do not use bear to analyze the data obtained from a two-stage BE study. Also I strongly suggest that you should search this Forum for more info about two-stage BE study. There are tons of discussions here.

» ... I am reviewing a BE study which has used a 94.12 confidence interval. Do you have any ideas as to why this is?

All the best,
---Yung-jin Lee
bear v2.8.3-2:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan http://pkpd.kmu.edu.tw/bear
Download link (updated) -> here
roro2907
Junior

Ireland,
2018-03-15 17:51
(edited by Ohlbe on 2018-03-15 19:16)

@ yjlee168
Posting: # 18555
Views: 2,887
 

 Is it a two-stage BE study?

Dear yjlee168,

Yes it is a two-stage study I have just discovered. I do not know much about this kind of study. Thank you for your help.


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! [Ohlbe]
Helmut
Hero
avatar
Homepage
Vienna, Austria,
2018-03-15 16:18

@ roro2907
Posting: # 18554
Views: 2,881
 

 other than 90% CI in BE study?

Hi roro2907,

» I am just curious as I am reviewing a BE study which has used a 94.12 confidence interval. Do you have any ideas as to why this is?

As Yung-jin assumed it smells of a Two-Stage Design of the Potvin flavour. ;-) It would need a major rewrite of bear because we not only need an option to specify an adjusted α (which is always <0.05) but – if the study proceeded to the second stage – “stage” in the model. BTW, you might face different CIs in the future (see this presentation).

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. ☼
Science Quotes
roro2907
Junior

Ireland,
2018-03-15 17:53
(edited by Ohlbe on 2018-03-15 19:16)

@ Helmut
Posting: # 18556
Views: 2,939
 

 other than 90% CI in BE study?

Dear Helmut,

Yes I have discovered that it is a two-stage design; could this be explained in simple terms?

Thanks


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! [Ohlbe]
Helmut
Hero
avatar
Homepage
Vienna, Austria,
2018-03-16 10:57

@ roro2907
Posting: # 18559
Views: 2,819
 

 Homework

Hi roro2907,

» […] it is a two-stage design; could this be explained in simple terms?

Please read the presentations about TSDs as Yung-jin and I suggested. Furthermore, as of today we have 547 posts in the category Two-Stage/ GS Designs. If you have specific questions, feel free to ask.

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. ☼
Science Quotes
Activity
 Thread view
Bioequivalence and Bioavailability Forum |  Admin contact
18,782 posts in 4,002 threads, 1,258 registered users;
online 26 (2 registered, 24 guests [including 21 identified bots]).

Power Calculation – A guess masquerading
as mathematics.    Stephen Senn

The BIOEQUIVALENCE / BIOAVAILABILITY FORUM is hosted by
BEBAC Ing. Helmut Schütz
HTML5 RSS Feed