sangeeta
☆    

India,
2007-06-25 18:04
(6121 d 15:47 ago)

Posting: # 834
Views: 12,831
 

 Statistical Analysis Plan [Design Issues]

Hi to all,

Can anybody suggest me the statistical analysis plan, if the subjects are divided into groups in two-way cross-over design.

Waiting for reply, :-|

Thanks in advance

Sangeeta.
M.Vasu
★    

India,
2007-07-02 10:29
(6114 d 23:22 ago)

@ sangeeta
Posting: # 856
Views: 11,130
 

 Statistical Analysis Plan

hi madam,

As per my knowledge, the subjects are divided into groups to know the "group effect", plz go through this guide - "FDA Statistical Approaches to Establishing Bioequivalence".

M.Vasu
Apotex,QA.

--
Edit: Full quote removed. Please see this post! [Helmut]
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2007-07-03 21:18
(6113 d 12:33 ago)

@ M.Vasu
Posting: # 860
Views: 11,259
 

 Two groups (e.g. logistics)

Dear M.Vasu,

❝ As per my knowledge, the subjects are divided into groups to know the "group effect", plz go through this guide - "FDA Statistical Approaches to Establishing Bioequivalence".


You are right, Section VII.A states:

If a crossover study is carried out in two or more groups of subjects (e.g., if for logistical reasons only a limited number of subjects can be studied at one time), the statistical model should be modified to reflect the multigroup nature of the study. In particular, the model should reflect the fact that the periods for the first group are different from the periods for the second group. This applies to all of the approaches (average, population, and individual BE) described in this guidance.

But also:

If the study is carried out in two or more groups and those groups are studied at different clinical sites, or at the same site but greatly separated in time (months apart, for example), questions may arise as to whether the results from the several groups should be combined in a single analysis. Such cases should be discussed with the appropriate CDER review division. (my emphasis)


Although such a procedure is required for submission in the USA (and Canada?!), I haven't seen a single study in the EU where a group effect was included in the statistical model if groups were administered in a timely manner (let's say, not more than one week apart).

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
M.Vasu
★    

India,
2007-07-05 10:40
(6111 d 23:11 ago)

@ Helmut
Posting: # 866
Views: 11,031
 

 Two groups (e.g. logistics)

Dear HS

i am also verified this one, you are right, right now i am not seen EU submissions to know group effect studys.


Edit: Full quote removed - like in all your previous posts! Please see this post! [Helmut]
sangeeta
☆    

India,
2007-07-21 14:05
(6095 d 19:46 ago)

@ Helmut
Posting: # 920
Views: 11,023
 

 Two groups (e.g. logistics)

Thanks for all who gave me a great reply.,

The analysis model which i have taken:

Group, sequence, sequence*group, period*group, formulation and formulation*group interaction as fixed effects and subject nested within group*sequence as a random effect. Sequence, sequence*group and group were tested using the Type III mean-square term for subject nested within sequence*group as the error term, at a 10% level of significance.

The formulation, period*group and formulation*group effects were tested against the residual mean square error. Probability (p) values were derived from Type III sums of squares with 5% level of significance.

In this model which significant effects i should consider…

And If i get formulation*group interaction was statistically significant for one PK parameter so, what is the next step i have to take in this…? Will it effect on the BE criteria...?


Thanks in advance…

Sangeea
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2007-11-16 16:07
(5977 d 16:44 ago)

@ sangeeta
Posting: # 1303
Views: 11,662
 

 Two groups (e.g. logistics)

Dear Sangeea,

since I recently had to deal with deficiency letters from the Gulf States (Saudia Arabia, Emirates), I would suggest the following procedure:

Fixed:  Sequence + Treatment + Period + Group + Treatment × Group
Random: Subject(Sequence)

Tests for Group and Treatment × Group.

If the ‘Treatment × Group Interaction’ term is statistically not significant (p>0.05), the simple model (i.e., naïve pooling of data without both Group-terms) can be applied.*


  • S Bolton S, Bon C. Pharmaceutical Statistics. New York: Marcel Dekker; 4th ed. 2004. p. 344–5.

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
MGR
★    

India,
2007-11-26 07:55
(5968 d 00:56 ago)

@ Helmut
Posting: # 1330
Views: 10,750
 

 Two groups (e.g. logistics)

Dear HS,

A small doubt regarding the above reply you told about the model, in that can we introduce the period*group effect term. Is it necessary or not? Please clarify this as iam in a confusion.

Thanks in advance.

Regards,
MGR
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2007-11-26 12:49
(5967 d 20:02 ago)

@ MGR
Posting: # 1331
Views: 10,921
 

 Two groups (e.g. logistics)

Dear MGR!

❝ […] can we introduce the period*group effect term. Is it necessary or not?


IMHO the answer is a straight ‘no’.
If you use the same protocol for both groups (the washout between periods in particular), I don’t see any reason which may cause such a type of interaction.

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
arl_stat
★    

India,
2013-03-26 10:42
(4020 d 22:09 ago)

@ Helmut
Posting: # 10284
Views: 7,611
 

 Two groups (e.g. logistics)

Dear HS,

❝ since I recently had to deal with deficiency letters from the Gulf States (Saudia Arabia, Emirates), I would suggest the following procedure:

   Fixed:  Sequence + Treatment + Period + Group + Treatment × Group
   Random: Subject(Sequence)

Tests for Group and Treatment × Group.


❝ If the ‘Treatment × Group Interaction’ term is statistically not significant (p>0.05), the simple model (i.e., naïve pooling of data without both Group-terms) can be applied.


In additional to above model, Is it necessary to add factor Period(group). If yes then what it would imply. This is For a 2 way crossover study where subject has been divided into two equal groups with equal washout period.
MGR
★    

India,
2007-07-02 15:27
(6114 d 18:24 ago)

@ sangeeta
Posting: # 857
Views: 11,201
 

 Statistical Analysis Plan

Hi madam,

If it is mentioned in protocol (sentence: in two groups) then the same statistical analysis plan is used. That Anova GLM model is used for the statistical analysis and Pharmacokinetic analysis is same. May be this helps.


Edit: Full quote removed. Please see this post! [Helmut]

Regards,
MGR
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
98 visitors (0 registered, 98 guests [including 3 identified bots]).
Forum time: 08:52 CET (Europe/Vienna)

With four parameters I can fit an elephant,
and with five I can make him wiggle his trunk.    John von Neumann

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