Peregrin ☆ Russia, 2015-07-24 16:24 (3169 d 20:39 ago) Posting: # 15142 Views: 4,545 |
|
Dear all, According to section 4.1.1, first paragraph, of the EMA guideline on bioequivalence, 'The study should be designed in such a way that the formulation effect can be distinguished from other effects'. Moreover, in section 4.1.8, subsection ‘Subject accountability’, second paragraph, of the same guideline, it is stipulated that ‘It is not acceptable to have a protocol which specifies that ‘spare’ subjects will be included in the analysis only if needed as replacements for other subjects who have been excluded.’ Does this imply that the study should be conducted simultaneously in the whole study population? If so, can you please cite the source where this requirement is provided explicitly, perhaps in the scientific guidelines of other regulatory bodies. I would be very thankful for providing any response. Kind regards, Peregrin |
Helmut ★★★ Vienna, Austria, 2015-07-24 20:34 (3169 d 16:28 ago) @ Peregrin Posting: # 15144 Views: 3,995 |
|
Hi Peregrin, ❝ Does this imply that the study should be conducted simultaneously in the whole study population? Not at all. It means that you shouldn’t replace dropouts. You estimate the sample size (based one the expected T/R-ratio & CV, desired power) and take the expected dropout-rate into account. See this post. If the actual dropout-rate is lower than expected, fine. You get a reward (lower producer’s risk). If it is higher, you will loose some power. Example for 2×2 crossover, CV 25%, T/R 0.95, desired power 80%, expected dropout-rate 10%: You estimate the desired sample size with 28 (power 80.7%), and dose 32 (uprounded):
The GL is talking about this idea: You desire 28, dose 32, have just 2 dropouts (say #27/28), and analyze #1–#26/#29/#30 (you don’t analyze #31/#32). Though that would safe some money in bioanalytics, there were regulatory concerns about this approach in the past:
See also FDA’s 2001 guidance, Section VII.A. — 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 |
Peregrin ☆ Russia, 2015-07-25 16:31 (3168 d 20:31 ago) @ Helmut Posting: # 15145 Views: 3,842 |
|
❝ Whether or not you have to modify the statistical model is a gray zone. Search the forum for “group effect”. ❝ See also FDA’s 2001 guidance, Section VII.A. Dear Helmut, Thank you so much! This is the exact thing I've been looking for! Several more questions have arised when I was considering your answer: Should the sponsor explicitly state in the protocol that the study for logistical reasons will be carried out in two or more groups? If the study has been conducted in several groups without statistical arrangements, should the sponsor at least provide some reasons for not doing so? Kind regards, Peregrin |
Helmut ★★★ Vienna, Austria, 2015-07-25 22:53 (3168 d 14:09 ago) @ Peregrin Posting: # 15146 Views: 3,743 |
|
Hi Peregrin, ❝ Should the sponsor explicitly state in the protocol that the study for logistical reasons will be carried out in two or more groups? ❝ If the study has been conducted in several groups without statistical arrangements, should the sponsor at least provide some reasons for not doing so? Two times yes. Always state in the protocol what you plan to do and why. — 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 |
Peregrin ☆ Russia, 2015-07-26 11:36 (3168 d 01:27 ago) @ Helmut Posting: # 15147 Views: 3,639 |
|
❝ Two times yes. Thank you! |
sciguy ☆ Canada, 2015-08-11 06:03 (3152 d 07:00 ago) @ Helmut Posting: # 15219 Views: 3,222 |
|
Hi everyone, Here is a slightly different but related scenario for an EMA study: Regulatory has imposed some recruiting restrictions on our BE study which has forced us to dose in 2 groups due to an inability to recruit a sufficient number in a single group within the current timeframe. Say we estimate we need to enroll 60 to complete 50 (~17% dropout rate) so we say in the protocol that Group 1 and Group 2 will have 30 subjects each. Can we do the following:
Thanks for your valuable opinions! |
Helmut ★★★ Vienna, Austria, 2015-08-11 16:06 (3151 d 20:57 ago) @ sciguy Posting: # 15229 Views: 3,148 |
|
Hi sciguy, I would avoid amendments if ever possible (time consuming, nasty to read, …). You already know beforehand that
BTW, I developed R-code in order to keep the pooled dataset in two-stage designs as balanced a possible (see here). Should be easy to modify it to your needs.
— 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 |