Log in
Register|
Loky do ★ Egypt, 2020-05-22 20:16 (1576 d 03:08 ago) Posting: # 21457 Views: 5,545 |
|
|
Dears In BE studies, in case the participants are both males and females, should the ratio between them be 1:1 or any ratio is accepted, also for randomization procedure is there any special requirements in case both included? Thanks Edit: Category changed; see also this post #1. [Helmut] |
|
Helmut ★★★   Vienna, Austria, 2020-05-23 19:12 (1575 d 04:12 ago) @ Loky do Posting: # 21458 Views: 4,448 |
|
|
Salam Loky do, ❝ […] participants are both males and females, should the ratio between them be 1:1 … The FDA requires „subjects from the general population”; hence ~1:1. Though I have seen studies in males only as well. Possibly the ANVISA requires that as well. AFAIK, in other jurisdictions there are no rules. ❝ … or any ratio is accepted, … See above. I once saw a study were the protocol stated “females and males” and the CRO recruited one female and 15 males. The study was accepted by European agencies but it looked stupid. ❝ … also for randomization procedure is there any special requirements in case both included? See this thread and R-code for stratification there. IMHO, in crossover designs it does not make sense. Do we have “sex” as a factor in the model? No. Do we want to demonstrate BE separate for females and males? Generally not. If yes, we would have to (at least) double the sample size. Parallel designs are another story, of course. Females/males should be evenly assigned to treatment groups. — 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, 2020-05-25 09:43 (1573 d 13:41 ago) @ Loky do Posting: # 21460 Views: 4,524 |
|
|
Hi, ❝ In BE studies, in case the participants are both males and females, should the ratio between them be 1:1 or any ratio is accepted, also for randomization procedure is there any special requirements in case both included? You should aim for 1:1 in the protocol but say at least 60:40 split (you can try 70:30 if you are desperate). I do that on my protocols for many yrs and no problem w FDA. You should randomize w gender as well. I had one CRO who put the first 10 as males and then last 10 as females, lol. No I did not find that funny when I spoke w them. J |
|
Helmut ★★★ Vienna, Austria, 2020-05-25 14:48 (1573 d 08:37 ago) @ jag009 Posting: # 21461 Views: 4,471 |
|
|
Hi John, ❝ You should randomize w gender as well. If it’s a cross-over, why? Sex is not part of the model. ❝ I had one CRO who put the first 10 as males and then last 10 as females, lol. Reduced chance to “grab her by the p••••y” whilst lining up for phlebotomy. ❝ No I did not find that funny when I spoke w them. I would. — 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, 2020-05-25 23:10 (1573 d 00:14 ago) @ Helmut Posting: # 21465 Views: 4,399 |
|
|
Hi Helmut, I was referring to generating the randomization list (you know, unlike the CRO who put first 10 males and next 10 females..) ❝ Reduced chance to “grab her by the p••••y” whilst lining up for phlebotomy. ❝ ❝ No I did not find that funny when I spoke w them. ❝ I would. Well if I had good impression of their overall capability, yes... But on not that day... J Edit: two successive posts merged [Ohlbe] |
|
ElMaestro ★★★ Denmark, 2020-05-25 16:39 (1573 d 06:45 ago) @ Loky do Posting: # 21463 Views: 4,372 |
|
|
Yah, I agree with everything. Just wish to mention that Gender is a between-subject factor (at least for most practical purposes ex-Thailand), so there is no particular gain in including Gender as a factor when you have Subject in the model unless there is a specific regulatory requirement. While PK may be different between Genders, like it would be between old and young subjects, or smokers and non-smokers, African-Americans versus Eskimos etc. But in BE we are interested in the ratio of T/R and this ratio has never in any particularly good study proven to vary between sub-populations. Hence the relaxed attitude towards e.g. studies done in e.g. India for approval in US and so forth. [Crystal ball] When the day comes, when someone is able to prove a case of real and true population difference in BE conclusion between populations, then BE as we know it today might be history and the innovator industry will jump on it like you would not believe possible with lawsuits left and right to protect their brands.[/Crystal ball] — Pass or fail! ElMaestro |
|
Loky do ★ Egypt, 2020-06-10 15:19 (1557 d 08:05 ago) @ ElMaestro Posting: # 21521 Views: 4,044 |
|
|
Thanks dears  |
|
BEQool ★ 2023-05-05 12:42 (498 d 10:42 ago) (edited by on 2023-05-05 12:52) @ ElMaestro Posting: # 23543 Views: 2,249 |
|
|
❝ Just wish to mention that Gender is a between-subject factor (at least for most practical purposes ex-Thailand), so there is no particular gain in including Gender as a factor when you have Subject in the model unless there is a specific regulatory requirement. Hello, maybe a stupid question but is Gender*Treatment also a between-subject factor (like Sequence and Gender here)? In the article Evaluation of sex-by-formulation interaction in bioequivalence studies of efavirenz tablets by González-Rojano et al. it doesnt say anything about which factor they used (within- or between-subject) to test Gender*Treatment interaction. Additionally, how do you know which factor is a within-subject and which is a between-subject? If anyone has an explanation in simple words |
|
Helmut ★★★ Vienna, Austria, 2023-05-05 14:44 (498 d 08:41 ago) @ BEQool Posting: # 23544 Views: 2,223 |
|
|
Hi BEQool, ❝ […] is Gender*Treatment also a between-subject factor (like Sequence and Gender here)? If you replace gender by sex, yes. Muse about it.  BTW, you could even drop sequence from the model…❝ In the article Evaluation of sex-by-formulation interaction in bioequivalence studies of efavirenz tablets by González-Rojano et al. it doesnt say anything about which factor they used (within- or between-subject) to test Gender*Treatment interaction. See p. 1731, right column, top. That’s the model proposed by the FDA earlier:*$$\eqalign{Y&|\;\text{sequence},\,\text{sex},\,\text{sequence}\,\times\,\text{sex},\\ &\phantom{|}\;\text{subject}(\text{sequence}\,\times\text{sex}),\;\text{period},\\ &\phantom{|}\;\text{treatment},\,\text{sex}\,\times\,\text{treatment}}$$ I employed it also in my meta-analysis. The ‘purpose’ of this model is only to assess the interaction. Since treatment appears twice in the model, it is not possible to get an unbiased estimate of it (for a similar story see this article). ❝ Additionally, how do you know which factor is a within-subject and which is a between-subject? If anyone has an explanation in simple words In simple words? I’ll try. For a given subject anything which remains constant throughout a study leads to a between-subject factor (e.g., sex, sequence, stage, group, site). Everything else leads to a within-subject factor (e.g., treatment, period).
— 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 |
|
BEQool ★ 2023-05-11 11:12 (492 d 12:12 ago) @ Helmut Posting: # 23550 Views: 2,172 |
|
|
Thank you for all thorough explanations, everything is clear and understandable now! BEQool |
Ing. Helmut Schütz