d_labes
★★★

Berlin, Germany,
2008-08-22 12:37
(5697 d 08:02 ago)

Posting: # 2227
Views: 9,934
 

 Category now obsolete? EMEA on non-parametrics [Nonparametrics]

Dear all,
dear HS,

it seems the IRON will of the Great Mahatma EMEA that this category must be dropped from the Forum. :crying:

I just had a first look at the draft of the CPMP/EWP/QWP/1401/98 Rev.1 (see HS/Ohlbe's post here)

In chapter 4.1.8 under Statistical analysis it is stated:
A non-parametric analysis is not acceptable. Boing!

More over it seems, that any analysis of tmax, for which until now non-parametrics was a must, has gone.

Condolences to HS (it is his favourite category :-P ) and to all programmers which have taken much effort to implement non-parametrics into BE analysis.

Congratulation to the statistical expertise of the Great Mahatma EMEA. :clap:
(Hint: Read the paragraph Statistical analysis farther and you will find more gems)

Regards,

Detlew
ElMaestro
★★★

Denmark,
2008-08-22 15:23
(5697 d 05:16 ago)

@ d_labes
Posting: # 2229
Views: 8,148
 

 Category now obsolete? EMEA on non-parametrics

Good point, DLabes.
First: "In studies to determine bioequivalence after a single dose, AUCt, AUCoo, Cmax and tmax should be determined."
And then later:
"The assessment of bioequivalence is based upon 90% confidence intervals for the ratio of the population geometric means (test/reference) for the parameters under consideration."
One could get the impression that the pharmacokinetolophystic subgroup is requesting a parametric analysis of tmax including 90% CI's and the works, had it not been for the final punch:
"In studies to determine bioequivalence after a single dose, the parameters to be analysed are AUCt and Cmax."
...so why do they want tmax at all?

EM.
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2008-08-22 15:31
(5697 d 05:08 ago)

@ d_labes
Posting: # 2230
Views: 8,267
 

 Category now obsolete? EMEA on non-parametrics

Dear DLabes,

yes, it’s unbelievable. It took these gurus 14 months (and 11 internal drafts) since the 'Recommendation on the Need for Revision of NfG on BA/BE' to come up with this

The second paragraph of ‘Statistical analysis’ (lines 500-505) with respect to tmax is bullshit anyhow. If I read it as a ‘cookbook’ (as intended by the authors) does it mean I should not only use a parametric method, but also log-transform data from a discrete sampling distribution?

This is a major statistical flaw. :angry:
If engineers in the aerospace industry dealing with rare events would apply statistical methods based on the Gaussian rather the Poisson distribution we would see quite a lot of plane accidents caused by wrong specs of mechanical/electrical/electronic devices.

After a quick look I only liked this one (lines 514-516):

‘A test for carry-over should not be performed and no decisions regarding the analysis (e.g. analysis of the first period, only) should be made on the basis of such a test.’


I’ll be opening a new thread collecting a critical review anyhow.

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,
2008-08-22 15:56
(5697 d 04:43 ago)

@ Helmut
Posting: # 2231
Views: 8,163
 

 EMEA statistical gem

Dear HS,

❝ After a quick look I only liked this one (lines 514-516):

‘A test for carry-over should not be performed and no decisions regarding the analysis (e.g. analysis of the first period, only) should be made on the basis of such a test.’


This is one of the statistical gems I had in mind.
Although this sentence is totally right, read it with the previous lines, requesting tests and confidence intervals (!) for treatment, period and sequence (!) effects and you see the gem shining :cool:.
   The gem was born of Evil's fire...
   The gem shall be his portal...
   He comes to claim,
   he comes to sire,
   the end of all things mortal


(found on Leo: "Teen Titans Season 4", don't know the author)

Regards,

Detlew
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2008-08-22 16:04
(5697 d 04:35 ago)

@ d_labes
Posting: # 2232
Views: 8,289
 

 EMEA statistical gem

Dear DLabes!

❝ ❝ After a quick look I only liked this one (lines 514-516):

‘A test for carry-over should not be performed and no decisions regarding the analysis (e.g. analysis of the first period, only) should be made on the basis of such a test.’



❝ This is one of the statistical gems I had in mind.

❝ Although this sentence is totally right, read it with the previous lines, requesting tests and confidence intervals (!) for treatment, period and sequence (!) effects and you see the gem shining


Yes this is ridiculous too. Period effect! The only country in the past asking me a couple of times for the CI of all these effects was France. I don’t know whether there was a French member in the ‘pharmacokinetolophystic subgroup’ (©2008 by ElMaestro).

One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.


[image]

Ash nazg durbatulûk, ash nazg gimbatul,
ash nazg thrakatulûk, agh burzum-ishi krimpatul. […]


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
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
80 visitors (0 registered, 80 guests [including 8 identified bots]).
Forum time: 19:40 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