Bioequivalence and Bioavailability Forum 15:12 CEST

Main page Policy/Terms of Use Abbreviations Latest Posts

 Log in |  Register |  Search

Pandu
Regular

India,
2009-02-23 14:00

Posting: # 3281
Views: 5,170
 

 Selection of chromatograms and its programme [Bioanalytics]

Dear All,

Please give me the information about the chromatograms and how to select the chromatograms (20%) for submission of regulatory studies.

Is there any SAS programme for selection of chromatograms. If aviable give me the information.

Thanks and Regards,

Pandu
Helmut
Hero
avatar
Homepage
Vienna, Austria,
2009-02-23 15:36

@ Pandu
Posting: # 3282
Views: 4,798
 

 20% of randomly chosen chromatograms

Dear Pandu,

some regulators want to see the subjects already stated in the protocol. I don’t think that’s a good idea because analysts will try to produce especially ‘nice ones’. OK, subjects should be selected in a random manner – but IMHO after the analyses.
You can set up anything in
[image], but why don’t you go for ‘true’ random numbers?

Cheers,
Helmut Schütz
[image]

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

India,
2009-02-24 06:08
(edited by Ohlbe on 2009-02-24 11:14)

@ Helmut
Posting: # 3283
Views: 4,706
 

 20% of randomly chosen chromatograms

Thanks for your reply.
Protocol says that simply submission of 20% chromatograms. at present we given first 20% subjects data.

As per your mention random.org. in that I want some clarifications. For example if sample size is 40. The 20% chromatograms are given that 8 subjects. As per random.com we select 8 time and taken that numbers. The clarifications are:

1. If subject number repeat.
2. If subject withdrawn form the study (Clinics).
3. If that subject samples asked for PK repeats.

Is there is any program in SAS for selection of this 20% chromatograms.

Thanks and Regards,

Pandu
--
Edit: Full quote removed. Please see this post! [Ohlbe]
Obinoscopy
Junior

Nigeria,
2018-06-03 18:14

@ Helmut
Posting: # 18846
Views: 1,887
 

 20% of randomly chosen chromatograms

Hi,

Sorry for ressurecting an already thoroughly discussed topic.

» OK, subjects should be selected in a random manner – but IMHO after the analyses.
» You can set up anything in
» [image], but why don’t you go for ‘true’ random numbers?

My question is this, which part of the guideline was it stated that the 20% Chromatogram has to be selected in a random fashion?

Is the requirement for randomly selecting 20% Chromatogram in any Guideline?

Regards,


Edit: Merged with a later (deleted) post. Hint: You can edit your original post for 24 hours. [Helmut]

Scopy
ElMaestro
Hero

Denmark,
2018-06-03 19:16

@ Obinoscopy
Posting: # 18848
Views: 1,872
 

 20% of randomly chosen chromatograms - In any Guideline?

Hi Obi,

» Is the requirement for randomly selecting 20% Chromatogram in any Guideline?

It is in the guidelines to the same extent as the requirement for cherry-picking the 20%.

Make sure to have an SOP which gives an objective way to select them.

if (3) 4

Best regards,
ElMaestro

"(...) targeted cancer therapies will benefit fewer than 2 percent of the cancer patients they’re aimed at. That reality is often lost on consumers, who are being fed a steady diet of winning anecdotes about miracle cures." New York Times (ed.), June 9, 2018.
Obinoscopy
Junior

Nigeria,
2018-06-03 22:50

@ ElMaestro
Posting: # 18850
Views: 1,857
 

 20% of randomly chosen chromatograms - In any Guideline?

Hi ElMaestro,

Thanks for your response.

» It is in the guidelines to the same extent as the requirement for cherry-picking the 20%.

Lol @ cherry-picking. But yeah, that should be a concern for the Sponsor or even the Regulator as the CRO site could cherry-pick runs of samples with wonderfully looking Chromatogram. However I noticed the FDA Guideline did state that chromatogram for NDAs and BLAs be randomly selected while chromatogram for ANDAs be serially selected. Which is great. I hope EMA considers this in subsequent reviews.

» Make sure to have an SOP which gives an objective way to select them.

Yea, the draft FDA Guideline did state that. But it was not stated in the final FDA guideline.....strange.

Regards,

Scopy
ElMaestro
Hero

Denmark,
2018-06-04 10:51

@ Obinoscopy
Posting: # 18852
Views: 1,840
 

 20% of randomly chosen chromatograms - In any Guideline?

Hi Obi,

» » Make sure to have an SOP which gives an objective way to select them.
»
» Yea, the draft FDA Guideline did state that. But it was not stated in the final FDA guideline.....strange.

Isn't this to some extent to the wrong philosophy? Those chromatograms are an essential part of your submission, so of course the matter should be covered by an SOP, whether or not the need for an SOP is stated explicitly in the guidance. I see it that way.:-)

if (3) 4

Best regards,
ElMaestro

"(...) targeted cancer therapies will benefit fewer than 2 percent of the cancer patients they’re aimed at. That reality is often lost on consumers, who are being fed a steady diet of winning anecdotes about miracle cures." New York Times (ed.), June 9, 2018.
chiragkhatri
Regular

India,
2009-02-24 09:17

@ Pandu
Posting: # 3284
Views: 4,730
 

 Selection of chromatograms and its programme

Dear Pandu,

Selection of 20% chromatograms is done by writing a simple code in SAS by input of the total number of volunteers in the study.

This should be done prior to start of analysis to avoid bias. In this case bioanalytical dept can request the biostatistician to generate a random number, after the randomization is generated it will remain with biostatistician and shared with bioanalytical after the analysis is completed.This will also avoid bias.

For eg the random number generated is 12, and you require 20% of 40 subjects, ideally you select 12-19 subjects serially, but

if any subject is dropout inbetween then select next number e.g if 13 is dropout then select 12 and 14-20.

If the random number itself is dropout then select next number ie from 13-20.

if the random number is generated is last subject number then select for e.g 40 and 1-7.

I hope its clear.

Regards

Chirag
Pandu
Regular

India,
2009-02-24 10:16

@ chiragkhatri
Posting: # 3285
Views: 4,686
 

 Selection of chromatograms and its programme

Dear Chirag

Thanks for your information. As per your information, statistican will given one code (that is subject number) after that number we will take the subjects continuesly.

But my clarificaion is that after that code we take continues subets (1-7 or 13-20). As per our bioanalytical department requires that total 20% subjects select at random (ex: 1,6,15,22,23,36,40,8 etc).

Please give me the code how to generate this in SAS.

Thanks and Regards

Pandu
krish3108
Junior

2009-03-03 09:10
(edited by Ohlbe on 2009-03-03 09:32)

@ Pandu
Posting: # 3305
Views: 4,565
 

 Selection of chromatograms and its programme

Dear Pandu,
Normally the trend in industry is to submit those subjects in same QC run, and 20% of the number of subjects that have completed the study.
Regards
Krishnan
--
Edit: Full quote removed. Please see this post! [Ohlbe]
Pandu
Regular

India,
2009-03-04 08:31

@ krish3108
Posting: # 3312
Views: 4,570
 

 Selection of chromatograms and its programme

Dear Krishnan

Thanks for your information. As per your information 20% of the number of subjects is selected and given in the report. My clarification is that how to select that 20% of subjects. (Please see my previous entry)

Thanks and Regards,

Pandu
d_labes
Hero

Berlin, Germany,
2009-03-04 12:15

@ Pandu
Posting: # 3313
Views: 4,556
 

 Equal chance to every one

Dear Pandu,

one very simple procedure:
  • Make lots with the numbers of your volunteers on it
  • Take a hat and put the lots into it
  • Mix them substantial (very important!)
  • Now draw 20% of the lots (or better leave this job to a nice looking young woman)
and voilà - you have the numbers of subjects to report.

No terrible software like "The power to know" needed, no criticism of bad 'pseudo' random number generators :-D .

If you cannot avoid SAS I suggest you Gooooogle:
SAS Simple Random Sample Without Replacement

Regards,

Detlew
Helmut
Hero
avatar
Homepage
Vienna, Austria,
2009-03-04 12:51

@ d_labes
Posting: # 3315
Views: 4,579
 

 Equal chance to every one

Hi D. Labes!

Great procedure, justified, well described, pointing out the crucial points, etc.

But:

» […] leave this job to a nice looking young woman

Although I like the idea and I’ve seen it in the recent past – when you have to defend your procedure I would suggest to stick to the ‘old fashioned way’!
IMHO the drawing of lots should be done by an orphan (out from a bowl, not a hat). If you want to be very serious, you may use an urn instead of the hat.

Cheers,
Helmut Schütz
[image]

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

Berlin, Germany,
2009-03-04 13:20

@ Helmut
Posting: # 3316
Views: 4,559
 

 Orphans, hats, bowls and urns

Dear Helmut!

» IMHO the drawing of lots should be done by an orphan (out from a bowl, not a hat). If you want to be very serious, you may use an urn instead of the hat.

Point 1: Orphan is a very good idea. But should be mandatory womanly. I have never seen a male "Lottofee" in Germany.
But it may be that we run in an availability problem by requesting Orphans (see here for a percentage of available orphans on world).

Point 2: I think a hat may be more available than a "Bowle", because alcohol is not allowed during work, at least here with me :-D .

Point 3: Urns are only good for the deads.

Regards,

Detlew
ElMaestro
Hero

Denmark,
2009-03-05 20:43
(edited by ElMaestro on 2009-03-05 21:20)

@ d_labes
Posting: # 3327
Views: 4,557
 

 Equal chance to every one

Dear d_labes,

» - Now drew 20% of the lots (or better leave this job to a nice looking young woman)

I am terribly sorry but I personally have very bad experience with your suggested random number generator, and it stems in part from the fact that a typical NLYW is generally very much more unpredictable than Mersenne Twister* and her allies. In fact I have tried several times to interact with NLYWs but the result often is a general syntax error, occasionally also an access failure, and debugging does not seem to be an option.
Therefore, I would strongly advise against your recommendation.

Anyone who believes unpredictability is a desired property of a random number generator should thus think twice.

EM.

d_labes
Hero

Berlin, Germany,
2009-03-06 10:02

@ ElMaestro
Posting: # 3330
Views: 4,514
 

 NLYWs

Dear ElMaestro,

» [...] I have tried several times to interact with NLYWs but the result often is a general syntax error, occasionally also an access failure, and debugging does not seem to be an option.

Yes, yes :yes:

"Ein unergründlich Rätsel ist das Weib"
(from the Opera "Mona Lisa" by Max von Schillings
rough translation: woman is a inscrutable mystery)

In case of syntax difficulties eventually the "Language of flowers" :flower: is an option.

But I think interacting with a NLYW is not an real option for Old Dogs (like me :-D ).

[image]In the context of the described random number generator the NLYW has only to interact with a hat and this is one of the best known interactions since long times (6.780.000!!! entries in Gooooogle picture search: "woman with hat").

Nofretete, nicest looking ancient pharao woman, of course with a hat.

Regards,

Detlew
Back to the forum Activity
 Thread view
Bioequivalence and Bioavailability Forum |  Admin contact
18,700 posts in 3,984 threads, 1,243 registered users;
online 16 (0 registered, 16 guests [including 11 identified bots]).

I suppose it is tempting, if the only tool you have is a hammer,
to treat everything as if it were a nail.    Abraham Maslow

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