Pharma_88 ☆ India, 2020-12-11 06:22 (1451 d 14:25 ago) Posting: # 22138 Views: 3,300 |
|
Dear all, Please provide your suggestion for below concern. for the bioequivalence studies, Analyst is always blind despite of regulatory requirement. By keeping different sampling time points i.e. 20 time-points for test and 22 time points for reference, whether analyst will blind in this condition? Edit: Category changed; see also this post #1. [Helmut] — Regards, Pharma_88 |
Helmut ★★★ Vienna, Austria, 2020-12-11 11:38 (1451 d 09:09 ago) @ Pharma_88 Posting: # 22141 Views: 2,729 |
|
Hi Pharma_88, ❝ Analyst is always blind … Analytics would be a tough job for a blind analyst. You mean blinded for treatment. ❝ By keeping different sampling time points i.e. 20 time-points for test and 22 time points for reference, whether analyst will blind in this condition? Of course, not (assuming that he/she is not stupid). Sometimes you aim at different sampling times (say when comparing MR to IR, DR to IR, IR to IV or an solution) in order to have an optimal schedule describing the profiles. Keep the number of time points equal and give the sequence of samples (1, 2, …, n) and not the time points themselves (t1, t2, …, tn) on the vials’ label. — 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 |
ElMaestro ★★★ Denmark, 2020-12-11 12:07 (1451 d 08:40 ago) @ Helmut Posting: # 22142 Views: 2,713 |
|
Hi both, ❝ ❝ Analyst is always blind … ❝ ❝ Analytics would be a tough job for a blind analyst. ❝ You mean blinded for treatment. Ah, the necessary continuation of last week's game of words. ❝ ❝ By keeping different sampling time points i.e. 20 time-points for test and 22 time points for reference, whether analyst will blind in this condition? ❝ ❝ Of course, not (assuming that he/she is not stupid). Sometimes you aim at different sampling times (say when comparing MR to IR, DR to IR, IR to IV or an solution) in order to have an optimal schedule describing the profiles. ❝ Keep the number of time points equal and give the sequence of samples (1, 2, …, n) and not the time points themselves (t1, t2, …, tn) on the vials’ label. It is an interesting situation though. There is quite some focus on blinding at the moment. I guess one could pad the series holding only 20 samples with a sample 21 and 22 based on extracted un-spiked plasma. That would kind of win the olympic gold, wouldn't it? The analyst will presumably be able to tell nothing about the nature of samples due to just the number of samples per series (or to say it diffrently: the CRO has done what it can to avoid situations when she/he can). I would love to see this done in practice. — Pass or fail! ElMaestro |
Helmut ★★★ Vienna, Austria, 2020-12-11 18:44 (1451 d 02:03 ago) @ ElMaestro Posting: # 22145 Views: 2,694 |
|
Hi ElMaestro, ❝ ❝ Keep the number of time points equal and give the sequence of samples (1, 2, …, n) and not the time points themselves (t1, t2, …, tn) on the vials’ label. ❝ ❝ […] I guess one could pad the series holding only 20 samples with a sample 21 and 22 based on extracted un-spiked plasma. That would kind of win the olympic gold, wouldn't it? Perhaps silver. ❝ The analyst will presumably be able to tell nothing about the nature of samples due to just the number of samples per series […]. Evidently, my dear Dr Watson. But: Sooner or later a smart analyst would start to wonder why in one period the last two samples are always BLQ. — 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 |
ElMaestro ★★★ Denmark, 2020-12-11 22:57 (1450 d 21:51 ago) @ Helmut Posting: # 22146 Views: 2,670 |
|
Hi Helmut, ❝ Evidently, my dear Dr Watson. But: Sooner or later a smart analyst would start to wonder why in one period the last two samples are always BLQ. Fantastic... you are right. We could put anything into the last two samples like sample 21 = "sample 20 diluted x.95" and sample 22 ="sample 20 diluted x.90" or something. The smart analyst you refer to may then look at the two profiles to check if the Cmax occurs on sample X in one series and on sample Y in another and try to work out something from the design. Perhaps one should make an effort ascertain that the sample number that corresponds to Cmax occurs at the same spot in the two series. But speculation and cleverness aside, how could one rock-solidly assure analyst-blindedness in those trials, in your opinion? — Pass or fail! ElMaestro |
Helmut ★★★ Vienna, Austria, 2020-12-12 13:06 (1450 d 07:41 ago) @ ElMaestro Posting: # 22147 Views: 2,625 |
|
Hi ElMaestro, ❝ We could put anything into the last two samples like sample 21 = "sample 20 diluted x.95" and sample 22 ="sample 20 diluted x.90" or something. Who is “we”? The clinical staff, right? Then they would have to harvest 3times the usual volume of sample 20 and have some blank matrix at hand. ❝ The smart analyst you refer to may then look at the two profiles to check if the Cmax occurs on sample X in one series and on sample Y in another and try to work out something from the design. Correct. ❝ Perhaps one should make an effort ascertain that the sample number that corresponds to Cmax occurs at the same spot in the two series. Oh dear! ❝ […] how could one rock-solidly assure analyst-blindedness in those trials, in your opinion? No idea. BTW, we had never fun with FIM trials. Verum generally a bolus or fast infusion. We always crossed fingers to have no concentrations >LLOQ in some of the profiles (guess, which ones). Blinded for treatment? Gimme a break. — 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 |