balakotu
★    

India,
2011-03-29 17:25
(5563 d 18:02 ago)

Posting: # 6827
Views: 7,421
 

 Multivariate test procedure [Design Issues]

Dear All,

In FDA guidance for dissolution profile comparison we could find other method than similarity factor. If within batch variation is more than 15%CV we could use multivariate model independent procedure for dissolution profile comparison. Could anyone give the explanation about?
  1. How to calculate the similarity limits in terms of multivariate statistical distance (MSD) based on interbatch differences in dissolution from reference batches?
  2. How to estimate 90% confidence interval of true MSD between test and reference batches.
Any possible softwares are used to do this calculation plz post it soon.


Thanks.


Edit: Category and subject line changed. [Jaime]
d_labes
★★★

Berlin, Germany,
2011-03-29 18:11
(5563 d 17:15 ago)

@ balakotu
Posting: # 6829
Views: 6,469
 

 Multivariate profile similarity

Dear balakotu,

did you ever considered searching this forum? :no:

We had this recently. See this thread
Look especially at the EXCEL add-in named DDSolver.
In the accompanying material to this add-in (which can be accessed online) there is a description of some of the alternative methods to the similarity factor f2. Among them Tsongs method using the Mahalanobis distance, which is nearly identical to the description in the FDA guidance, also no reference is given in the guidance.

But beware of using EXCEL in such a difficult analysis. Remember Martin's mantra: "never never never never use excel for any statistical calculations". And beware of the results of the 90% CI's of the Mahalanobis distance.

I personally don't belief in that results of DDSolver.
But unfortunately I can't check or verify because the Visual basic code behind that Add-in is password protected.

Unfortunately I'm not aware of any other software doing that calculations.

Moreover the theory behind Tsongs method is criticised fundamentally here.

Regards,

Detlew
UA Flag
Activity
 Admin contact
23,655 posts in 4,993 threads, 1,572 registered users;
143 visitors (0 registered, 143 guests [including 25 identified bots]).
Forum time: 11:27 CEST (Europe/Vienna)

Scientists cannot simply hang their subjectivities
up on a hook outside the laboratory door.    Ruth Bleier

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