Log in
Register|
PTM139 ☆ India, 2012-05-03 08:01 (4760 d 08:08 ago) Posting: # 8496 Views: 8,999 |
|
|
Dr_Dan ★★ Germany, 2012-05-03 17:14 (4759 d 22:55 ago) @ PTM139 Posting: # 8501 Views: 8,441 |
|
|
Dear PTM139 If you keep the same number of subjects not only your chances of success will increase but also your costs (two periods vs. four). So the answer to your question is not that simple..... Kind regards Dan — Kind regards and have a nice day Dr_Dan |
|
PTM139 ☆ India, 2012-05-03 18:35 (4759 d 21:34 ago) @ Dr_Dan Posting: # 8505 Views: 7,952 |
|
|
Helmut ★★★   Vienna, Austria, 2012-05-03 18:58 (4759 d 21:11 ago) @ PTM139 Posting: # 8507 Views: 8,041 |
|
|
Dear PTM139! ❝ But irrespective of the cost chances of success are more in case of HVD? You didn’t tell us that you are dealing with a HVD/HVDP. As Dan already stated it is not that simple. Design of studies depend on the regulation:
— 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 |
|
PTM139 ☆ India, 2012-05-05 09:16 (4758 d 06:53 ago) @ Helmut Posting: # 8517 Views: 7,834 |
|
|
drsinghs ☆ India, 2012-05-09 14:33 (4754 d 01:36 ago) @ Helmut Posting: # 8542 Views: 7,782 |
|
|
Dear HS and Dr. Dan, My query is similar to the ongoing discussion. Below situation may not be practically feasible but still I am asking this to gain some knowledge. Irrespective of regulatory requirements, cost and other issues; I request valuable opinion from both of you on following. This is regarding a HVDP. If I check same batch of test and reference by both 2 way crossover and replicate design, what kind of differences in results can I expect? To be more precise I will share a hypothetical case - if 2 waycrossver study is already completed and observed T/R ratio is 115 for Cmax and 125 for AUCt with > 30% intraCV for Cmax and > 45% for AUCt, Can I expect successful results if I go for replicate design? Thanks & regards, drsinghs |
|
Helmut ★★★ Vienna, Austria, 2012-05-09 15:05 (4754 d 01:04 ago) @ drsinghs Posting: # 8543 Views: 7,820 |
|
|
Dear drsinghs! ❝ To be more precise I will share a hypothetical case - if 2 waycrossver study is already completed and observed T/R ratio is 115 for Cmax and 125 for AUCt with > 30% intraCV for Cmax and > 45% for AUCt, Can I expect successful results if I go for replicate design? Setting different regulatory requirements aside (EMA scaling allowed for Cmax only, different statistical models for FDA and EMA) in all regulations the T/R ratio has to be within 80.00–125.00%. If your ratio for AUC in the 2×2 cross-over was already 125% your chance to get a ratio ≤125% is only 50% at any given sample size. — 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 |
|
drsinghs ☆ India, 2012-05-09 15:41 (4754 d 00:28 ago) (edited on 2012-05-09 16:31) @ Helmut Posting: # 8545 Views: 7,681 |
|
|
❝ Setting different regulatory requirements aside (EMA scaling allowed for Cmax only, different statistical models for FDA and EMA) in all regulations the T/R ratio has to be within 80.00–125.00%. If your ratio for AUC in the 2×2 cross-over was already 125% your chance to get a ratio ≤125% is only 50% at any given sample size. Thanks HS for your opinion, Ok if my study is for EMA and observed ratio for AUCt is 125% with 90% CI 110 lower and 140 upper, Can I expect 90% CI to be within 80.00-125.00% in replicate design study. Also would really appreciate your comment on my generic query in prevoius post about differences in study results by these 2 designs. Thanks & regards, drsinghs Edit: duplicate post deleted. You can edit your post within 24 h if needed [Ohlbe] |
|
Ohlbe ★★★ France, 2012-05-09 18:40 (4753 d 21:29 ago) @ drsinghs Posting: # 8547 Views: 7,714 |
|
|
Dear drsinghs, ❝ Ok if my study is for EMA and observed ratio for AUCt is 125% with 90% CI 110 lower and 140 upper, Can I expect 90% CI to be within 80.00-125.00% in replicate design study. If your observed T/R ratio is 125 %, your chances to get a 90 % CI within 80.00 - 125.00 % in the next study, even with a replicate design, are close to nil. Regards Ohlbe — Regards Ohlbe |
|
drsinghs ☆ India, 2012-05-10 07:49 (4753 d 08:20 ago) @ Ohlbe Posting: # 8549 Views: 7,653 |
|
|
❝ If your observed T/R ratio is 125 %, your chances to get a 90 % CI within 80.00 - 125.00 % in the next study, even with a replicate design, are close to nil. Thanks ohlbe, Can I once again request you to please comment on my other query as below? This is regarding a Modified release HVDP. If I check same batch of test and reference by both 2 way crossover and replicate design, what kind of differences in results can I expect? Basically I want to understand the impact of change in study design on BE results. Thanks & regards, drsinghs |
|
ElMaestro ★★★ Denmark, 2012-05-09 17:00 (4753 d 23:09 ago) @ drsinghs Posting: # 8546 Views: 7,741 |
|
|
Dear drsinghs, ❝ To be more precise I will share a hypothetical case - if 2 waycrossver study is already completed and observed T/R ratio is 115 for Cmax and 125 for AUCt with > 30% intraCV for Cmax and > 45% for AUCt, Can I expect successful results if I go for replicate design? Big risk. If the CV for AUC is bigger than CV for Cmax then there is a risk the study will be flagged for inspection by some authorities, especially FDA. It is often a sign of something having gone badly wrong, and is generally not lending the data much credibility. Second, I know of IECs/IRBs now beginnng to understand that power and statistics and sample size are not just academic issues but need to be dealt with practically. It may be a seriously uphill struggle to dimension a study with a T/R of 125% (AUC) - try and calculate the sample size for 80% power, 125% T/R and 45% CV. Your only practical chance is to assume T/R being much closer to 100%. Then you can at least arrive at fairly reasonable sample size, but on the other hand that implies that you disregard the data from the first study. You generally need a very good and justified reason to do so. I would either say it is game over with the current formulation or suggest to try very hard with audits and internal reviews in order to figure out if something went wrong before proceeding. The consideration above is regardless of replicate design or no repliacte design. — Pass or fail! ElMaestro |
|
drsinghs ☆ India, 2012-05-10 07:43 (4753 d 08:26 ago) @ ElMaestro Posting: # 8548 Views: 7,691 |
|
|
❝ Big risk. ❝ ❝ If the CV for AUC is bigger than CV for Cmax then there is a risk the study will be flagged for inspection by some authorities, especially FDA. It is often a sign of something having gone badly wrong, and is generally not lending the data much credibility. Dear ElMaestro, My formulation is Modified release do you still feel it will be flagged for inspection just because CV for AUC is bigger than CV for Cmax. Thanks & regards, drsinghs |
Ing. Helmut Schütz