Bioequivalence and Bioavailability Forum

Hello VSL,

❝ In calculating descriptive statistics why Tmax is calculated as Median instead of Mean ?

Since Tmax does not follow a normal distribution it is often evaluated with non-parametric statistics. These are almost always based on medians rather than means. It has to do with tradition as well as with maximum likelihood.

---

❝ 2. The time to maximum drug plasma concentration (Tmax) is examined but is not required to be within any specified limits. Why doesn't it require to prove Bioeuivalence on Tmax?

We are often interested in rate and extent of absorption or rate and extent of drug becoming available at the site of action.
It turns out that Cmax is often (often, not necessarily meaning always, see below) a good enough indicator of rate while AUC is a good enough indicator of extent. Since both can be handled with parametric statistics after logs these metrics have some clear practical advantages.

❝ 3. If there is larger difference in Tmax between Test and Reference Formulation but not in Cmax,it would be rational comparison between T & R as for some therapeutic class drugs 'on set time of action' is more important?

I think difference in Tmax but not in Cmax would often imply difference in AUC, so this would be trash anyway. I get your point in onset of action. And yes, in such cases Tmax could become important. You could possibly face that with certain bronchodilators, smoke cessation remedies or cardiac nitrates.

---

Edit: Merged the second part from a follow-up post. [Helmut]