sangeeta ☆ India, 2007-06-25 18:04 (6121 d 15:47 ago) Posting: # 834 Views: 12,831 |
|
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 |
|
Helmut ★★★ Vienna, Austria, 2007-07-03 21:18 (6113 d 12:33 ago) @ M.Vasu Posting: # 860 Views: 11,259 |
|
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! Helmut Schütz 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 |
|
sangeeta ☆ India, 2007-07-21 14:05 (6095 d 19:46 ago) @ Helmut Posting: # 920 Views: 11,023 |
|
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 ★★★ Vienna, Austria, 2007-11-16 16:07 (5977 d 16:44 ago) @ sangeeta Posting: # 1303 Views: 11,662 |
|
Dear Sangeea, since I recently had to deal with deficiency letters from the Gulf States (Saudia Arabia, Emirates), I would suggest the following procedure:
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.*
— Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz 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 |
|
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 ★★★ Vienna, Austria, 2007-11-26 12:49 (5967 d 20:02 ago) @ MGR Posting: # 1331 Views: 10,921 |
|
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! Helmut Schütz 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 |
|
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 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 |
|
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 |