Log in
Register|
RK ☆ India, 2013-09-14 14:17 (4260 d 03:33 ago) Posting: # 11497 Views: 9,242 |
|
|
Hi all, Can anybody tell me the pros and cons for using 6 sequences than the normal two sequence in fully replicate design. The sequences are AABB BBAA ABAB BABA ABBA BAAB regards rk Edit: Category changed. [Helmut] |
|
Helmut ★★★   Vienna, Austria, 2013-09-15 15:47 (4259 d 02:04 ago) @ RK Posting: # 11498 Views: 7,950 |
|
|
Hi RK, counter-questions: What are you trying to achieve? Why do you think such a design would have advantages over RTRT|TRTR? — Dif-tor heh smusma 🖖🏼 Довге життя Україна! ![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png) Helmut Schütz ![[image]](https://static.bebac.at/img/CC by.png) The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
jag009 ★★★ NJ, 2013-09-16 19:24 (4257 d 22:26 ago) @ Helmut Posting: # 11500 Views: 7,819 |
|
|
Hi Helmut, ❝ counter-questions: What are you trying to achieve? Why do you think such a design would have advantages over RTRT|TRTR? Can I answer?  (Love to do this when i was in school, jumping to answer first hehe)I don't like AABB, BBAA, ABBA, BAAB because of giving the same treatment (A or B) in consecutive period. I have no choice with FDA's partial rep - ABB, BBA, ABA... John |
|
RK ☆ India, 2013-09-17 15:35 (4257 d 02:15 ago) @ Helmut Posting: # 11505 Views: 7,916 |
|
|
❝ What are you trying to achieve? Why do you think such a design would have advantages over RTRT|TRTR? Hi all, We already have conducted a fully replicated study with two sequences (ABAB and BABA) but the study did not meet bioequivalence, now the client is planning for a repeat fully replicated study with six sequences (AABB, BBAA, ABAB, BABA, ABBA, BAAB). As the discussions are in its initial stage would like to know the regulatory acceptance (USFDA) and the implication on going ahead with this study design and statistical analysis. Regards RK. |
|
Helmut ★★★ Vienna, Austria, 2013-09-17 15:51 (4257 d 01:59 ago) @ RK Posting: # 11506 Views: 7,811 |
|
|
Hi RK, ❝ ❝ What are you trying to achieve? Why do you think such a design would have advantages over RTRT|TRTR? ❝ We already have conducted a fully replicated study with two sequences (ABAB and BABA) but the study did not meet bioequivalence, […] Why did the study fail? Ratio too far away from one? Variability higher than expected? If the answer to both is no, pure chance? 80% power = 20% chance to fail to show BE. ❝ […] now the client is planning for a repeat fully replicated study with six sequences (AABB, BBAA, ABAB, BABA, ABBA, BAAB). Try to answer my questions above. “The client wants it” is no valid justification. ❝ […] the regulatory acceptance (USFDA) and the implication on going ahead with this study design and statistical analysis. FDA argued even against the four-period four-sequence design TRRT|RTTR|TTRR|RRTT in the Guidance’s Appendix B. — 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 |
|
jag009 ★★★ NJ, 2013-09-17 17:49 (4257 d 00:02 ago) @ Helmut Posting: # 11507 Views: 7,809 |
|
|
Hi Helmut, One question from the guidance B: If the replicated crossover design has only two sequences, use of only the three main effects (sequence, period, and treatment) in the fixed effects model or use of a more saturated model makes little difference to the results of the analysis, provided there are no missing observations and the study is carried out in one group of subjects… Missing observations ⇒ Does that mean we would encounter problems if not all subjects have data for each treatment? i.e, what if we miss some AUCinf values when kel cannot be determined? Thanks John |
|
Helmut ★★★ Vienna, Austria, 2013-09-18 19:32 (4255 d 22:19 ago) @ jag009 Posting: # 11520 Views: 7,801 |
|
|
Hi John, ❝ Missing observations ⇒ Does that mean we would encounter problems if not all subjects have data for each treatment? i.e, what if we miss some AUCinf values when kel cannot be determined? There might be a small bias. However, sequences would have to be extremely imbalanced/incomplete to start hurting. Look at FDA’s replicate example data sets. nAUC∞ ≠ nAUCt is rather a rule than an exception.  — 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 |
|
RK ☆ India, 2013-09-20 15:56 (4254 d 01:54 ago) @ Helmut Posting: # 11533 Views: 7,748 |
|
|
Hi Helmut, On reviewing the data of four period two sequence (ABAB and BABA) fully replicated study retrospectively, there is a statistically significant period effect reported for Cmax. So would like to know whether we can use four period four sequence fully replicated study to nullify this period effect which is described in the FDA guideline (Appendix B- Reasons Unrelated to Carryover Effects). regards RK |
|
d_labes ★★★ Berlin, Germany, 2013-09-23 10:44 (4251 d 07:06 ago) @ RK Posting: # 11541 Views: 8,516 |
|
|
Dear RK, ❝ On reviewing the data of four period two sequence (ABAB and BABA) fully replicated study retrospectively, there is a statistically significant period effect reported for Cmax. So would like to know whether we can use four period four sequence fully replicated study to nullify this period effect ... If there is a period effect in a crossover study (replicate or not) it means that the subjects behave different in the study periods, irrespective of their treatment in that periods. That implies in turn irrespective of their sequence of treatments. Thus I can't imagine any reason why such an effect should vanish by using different (or more) sequences. ❝ ... this period effect which is described in the FDA guideline (Appendix B- Reasons Unrelated to Carryover Effects). Can you elaborate where you found hints about significant period effects in 'Appendix B'  ?— Regards, Detlew |
|
RK ☆ India, 2013-09-24 15:16 (4250 d 02:35 ago) @ d_labes Posting: # 11549 Views: 7,586 |
|
|
Dear Detlew, could you please explain me if this not the one (4 sequence) which will nullify the period effect, then in which condition we can use this additional sequence information??? Here is the section which describes about the four sequence-reasons Unrelated to Carryover Effects Deleted Appendix B from the beginning to the first paragraph of Section 2. We all know the guidance. If you really consider it necessary to paste in such a large amount of text from a PDF, please use the Preview before posting. It came out terrible. [Helmut] Regards RK. |
Ing. Helmut Schütz