lizhao ☆ US, 2014-11-06 23:19 (3430 d 06:56 ago) Posting: # 13835 Views: 5,067 |
|
If a PD endpoint is an endogenous hormone in human body. The PD metric is the change from baseline after taking the medicine. And it is also the PD metric is a concentration, so I should log-transform it. Then should I use ln(post_treatment-baseline) as the bioequivalence metric? Thanks Edit: Please follow the Forum’s policy. [Helmut] |
Dr_Dan ★★ Germany, 2014-11-07 11:58 (3429 d 18:17 ago) @ lizhao Posting: # 13838 Views: 4,110 |
|
Dear lizhao You determine the baseline for each individual separately and you get a value X. This value is subtracted from each concentration of your PD-profile and then you log-transform the data ln(C1-X, C2-X, C3-X.....) and perform the ANOVA. So I do not understand what you mean by ln(post_treatment-baseline) as bioequivalence metric. Kind regards Dr_Dan — Kind regards and have a nice day Dr_Dan |
Helmut ★★★ Vienna, Austria, 2014-11-07 18:27 (3429 d 11:48 ago) @ Dr_Dan Posting: # 13842 Views: 4,063 |
|
Hi Dan & lizhao, ❝ ❝ If a PD endpoint is an endogenous hormone in human body. The PD metric is the change from baseline after taking the medicine. And it is also the PD metric is a concentration, so I should log-transform it. Pharmacokinetics may be simply With hormones you may stir up a hornets’ nest. Depending on the hormone you may see an increase, decrease, or even a nasty feed-back loop (i.e., steady state is far from what you would predict from a single dose). Can you be more specific? Which hormone? Population? Single/multiple dose? ❝ You determine the baseline for each individual separately and you get a value X. This value is subtracted from each concentration of your PD-profile and then you log-transform the data ln(C1-X, C2-X, C3-X.....) and perform the ANOVA. So I do not understand what you mean by ln(post_treatment-baseline) as bioequivalence metric. Subtraction (if we have an increase, of course) is only the first step. I would not log-transform the corrected concentrations. Simply calculate the Cmax / AUC and then log-transform. Important: For AUC don’t use AUCt but include the last triangle if zero is the result of two values >LLOQ.
— 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 |