Log in
Register|
AngusMcLean ★★ USA, 2014-01-24 01:02 (4185 d 09:57 ago) Posting: # 12252 Views: 8,748 |
|
|
We have implemented a 3 way cross-over balanced design (Williams, 6 sequences) for a fed, fasted, sprinkled on food bioequivalence study. It is not a highly variable drug (CV ~15%). The study has gone well and the results are unambiguous. We did lose one subject at PERIOD 3. I am wondering whether I should exclude the other two treatments for this subject prior to BE data analysis in Phoenix. Or should I include all the data I have. Or indeed do the data analysis both ways. This work is a registration study and will be submitted here to the FDA. Angus |
|
Helmut ★★★   Vienna, Austria, 2014-01-24 05:18 (4185 d 05:41 ago) @ AngusMcLean Posting: # 12253 Views: 7,586 |
|
|
Hi Angus, short answer: Perform the analysis exactly as you have specified it in the protocol.  OK, more serious now. I guess the reference treatment was in fasted state. If R was in period 3, bad luck – no test-ratios. For the FDA the standard setup is a mixed-effects model (also the default in PHX). Try to remove the subject – you should get the same results as if keeping him/her. If you administered in period 3 one of the tests, keep him/her. You’ll loose one degree of freedom in the respective pairwise comparison (say T1/R), but still have the full set for the other (T2/R). Nothing to worry about. — 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 |
|
AngusMcLean ★★ USA, 2014-01-24 15:26 (4184 d 19:33 ago) @ Helmut Posting: # 12262 Views: 7,543 |
|
|
Thank you Helmut: yes; I do regard the fasted leg as the reference. The treatment lost was not the reference....it was one of the tests (fed)that did not show up for Period 3. The protocol was written and provided an overview of the data analysis done in Phoenix. It did not discuss dropout subjects and what to do if subjects dropped out. Perhaps it is a good thing if the protocol is vague on this issue............suitably vague? do you agree? Often I think it is mistake to give too much detail in a protocol, since you can create protocol violations. The FDA did not comment on this issue during their protocol review, but commented on other issues. I will try the options you suggest: At an earlier date for a fed, fasted, sprinkled study I have run a 6 sequence type data analysis (just like this one) All of the subjects completed the study. I showed that the result were identical in SAS (GLM) and WinNonlin and Kinetica. Angus |
|
Helmut ★★★ Vienna, Austria, 2014-01-24 16:28 (4184 d 18:31 ago) @ AngusMcLean Posting: # 12263 Views: 7,622 |
|
|
Hi Angus, ❝ I do regard the fasted leg as the reference. The treatment lost was […] one of the tests (fed) that did not show up for Period 3. OK, so no problem with PHX’s REML. Just run the analysis on your entire data set. ❝ The protocol was written and provided an overview of the data analysis done in Phoenix. It did not discuss dropout subjects and what to do if subjects dropped out. Perhaps it is a good thing if the protocol is vague on this issue … suitably vague? do you agree? I don’t agree. IMHO it is better to invest your intellectual horsepower in the protocol, than to state sumfink ambiguous and have to come up with “creative” ideas in the analysis. If the protocol is written by a genius, the analysis can be done by a dummy. I’m not sure whether it always / easily can be done the other way ’round. Creative solutions might be interpreted by assessors as the first step of cherry-picking:  “Oh, the study passed by applying Method X. But there are alternatives. Would the study have failed with Method Y? Let’s ask the applicant for a justification, …”. Such a request can turn out to be a show-stopper. Methods are based on different assumptions, which are difficult – if not impossible – to proof given the limited sample size in BE studies. So you might be forced to come up with a lot of in your response, which might be accepted – or not… To call the statistician after the experiment is done ❝ Often I think it is mistake to give too much detail in a protocol, since you can create protocol violations. The FDA did not comment on this issue during their protocol review […] I would rather call them protocol deviations; violation is a little bit harsh. See also ICH’s Q&A on E3. However, if you stated contingency plans in the protocol (aka “be prepared fo the unexpected”) you rarely have to deviate from the original method(s). ❝ All of the subjects completed the study. I showed that the result were identical in SAS (GLM) and WinNonlin and Kinetica. Note that Proc GLM = WinNonlin only for complete data sets (balanced or imbalanced, but subjects completed all periods). For imbalanced data sets Proc Mixed = WinNonlin; if you want to get the same results as Proc GLM you have to exclude incomplete subjects in WinNonlin.— 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 |
|
AngusMcLean ★★ USA, 2014-01-24 17:28 (4184 d 17:31 ago) @ Helmut Posting: # 12265 Views: 7,601 |
|
|
Thanks Helmut: the comparison work was done> 5 years ago before Phoenix Nonlin was introduced so I used the version of WinNonlin current at that time. The SAS was done using GLM model. The work in Kinetica was done just along the same lines as the other programs. It only works with the full data sets (i.e. no dropouts). I have since recently repeated the work using Phoenix WinNonlin and the results is exactly the same as previous work in WinNonlin. ❝ ❝ All of the subjects completed the study. I showed that the result were identical in SAS (GLM) and WinNonlin and Kinetica. current Helmut: I take it your remarks below for WinNonlin also apples to current Phoenix Nonlin program yes? ❝ Note that Angus |
|
Helmut ★★★ Vienna, Austria, 2014-01-24 17:59 (4184 d 17:00 ago) @ AngusMcLean Posting: # 12266 Views: 7,503 |
|
|
Hi Angus, ❝ I take it your remarks below for WinNonlin also apples to current Phoenix Nonlin program yes? If you mean Phoenix WinNonlin, correct.  — 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 |
|
AngusMcLean ★★ USA, 2014-01-26 16:58 (4182 d 18:01 ago) @ Helmut Posting: # 12270 Views: 7,438 |
|
|
❝ I don’t agree. IMHO it is better to invest your intellectual horsepower in the protocol, than to state sumfink ambiguous and have to come up with “creative” ideas in the analysis. If the protocol is written by a genius, the analysis can be done by a dummy. I’m not sure whether it always / easily can be done the other way ’round. Creative solutions might be interpreted by assessors as the first step of cherry-picking: “Oh, the study passed by applying Method X. But there are alternatives. Would the study have failed with Method Y? Let’s ask the applicant for a justification, …”. Such a request can turn out to be a show-stopper. Methods are based on different assumptions, which are difficult – if not impossible – to proof given the limited sample size in BE studies. So you might be forced to come up with a lot of in your response, which might be accepted – or not…Helmut: Then we need a structured data analysis plan for our clinical protocol. Well let us assume that we can make a general statement in the clinical protocol that covers the eventuality of “drop out” subjects when using Phoenix WinNonlin. “if a subject drops out during the reference period of the study then the corresponding test pharmacokinetic data (Cmax and AUC) from that subject can be omitted from the data set prior to BE data analysis irrespective of whether the study is a two or three way cross-over study. On the other hand if one subject drops out during a test period then the following applies: For a two way cross-over study the corresponding reference period can be omitted from the data set. For a three way cross over study the one remaining test data set available is included with the reference in the study. So how does that sound for a start? Angus Edit: Full quote removed. Please delete everything from the text of the original poster which is not necessary in understanding your answer; see also this post! [Helmut] |
|
Helmut ★★★ Vienna, Austria, 2014-01-26 17:25 (4182 d 17:34 ago) @ AngusMcLean Posting: # 12271 Views: 7,578 |
|
|
Hi Angus, ❝ Then we need a structured data analysis plan for our clinical protocol. Exactly. For my studies I give an overview in the clinical protocol and the details in the SAP. ❝ Well let us assume that we can make a general statement in the clinical protocol that covers the eventuality of “drop out” subjects when using Phoenix WinNonlin. […] ❝ So how does that sound for a start? Good. For the two-way I would rather write “[…] the corresponding reference period will be omitted from the data set.” PS: In the future please don’t full quote (see also this post). — 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 |
Ing. Helmut Schütz