Log in
Register|
martin ★★ Austria, 2010-12-17 15:32 (5316 d 23:50 ago) Posting: # 6317 Views: 9,060 |
|
|
Dear All ! I have a question regarding interpretation of dose proportionality based on the power-law model. According to Gough et al. (1995), the power-law model has the form AUC = a × doseb. The parameter a is the constant of proportionality, where dose proportionality requires that b=1. I evaluated dose proportionality after IV bolus in two subgroups which gave nearly identically estimates for the parameter b but the parameter a differed substantially between the two subgroups considered. According to my understanding this may indicate a higher bioavailability with one subgroup compared to the other subgroup which may be due to a difference in clearance. Any suggestions for interpretation are highly appreciated! best regards martin Gough K, Hutchison M, Keene O, Byrom B, Ellis S, Lacey L, McKellar J (1995). Assessment of dose proportionality: Report from the statisticians in the pharmaceutical industry/pharmacokinetics UK joint working party. Drug Information Journal, 29:1039-1048. |
|
Helmut ★★★   Vienna, Austria, 2010-12-17 16:30 (5316 d 22:53 ago) @ martin Posting: # 6318 Views: 7,768 |
|
|
Dear Martin! ❝ [...] the power-law model has the form AUC = a × doseb. The parameter a is the constant of proportionality, where dose proportionality requires that b=1. Bingo. ❝ I evaluated dose proportionality after IV bolus in two subgroups which gave nearly identically estimates for the parameter b but the parameter a differed substantially between the two subgroups considered. According to my understanding this may indicate a higher bioavailability with one subgroup compared to the other subgroup which may be due to a difference in clearance. It may help to transform the model to log(AUC) = log(a) + b × log(dose) If the estimates are b1 ≈ b2 and a1 # a2, you have a shift in intercepts (parallel straight lines). Since for i.v. CL=dose/AUC and in your case doses in the subgroups are the same, bingo again. P.S.: From the transformation it is more clear that the power model is empiric, not mechanistic. Any (significant) a >0 would mean an AUC for a zero dose, and what the hell would a <0 mean? — 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 |
|
ElMaestro ★★★ Denmark, 2010-12-17 20:54 (5316 d 18:29 ago) @ Helmut Posting: # 6319 Views: 7,669 |
|
|
Hi HS, ❝ P.S.: From the transformation it is more clear that the power model is empiric, not mechanistic. Any (significant) a >0 would mean an AUC for a zero dose, and what the hell would a <0 mean? a<0 sometimes happens in practice. It would mean:
Best regards, EM  - moored at West Palm Beach  |
|
Helmut ★★★ Vienna, Austria, 2010-12-18 19:52 (5315 d 19:30 ago) @ ElMaestro Posting: # 6320 Views: 7,709 |
|
|
Ahoy my capt'n! ❝ ❝ P.S.: From the transformation it is more clear that the power model is empiric, not mechanistic. Any (significant) a >0 would mean an AUC for a zero dose, and what the hell would a <0 mean? ❝ ❝ a<0 sometimes happens in practice. Right. ❝ It would mean: ❝ 1. That your data is drowning in scatter. Hhm, I said significant >0. ❝ 2. That you might want to seriously consider constrained (a>=0) fitting. ❝ and/or Oh wow. Might end up with a=0 and the software ringing the bell that the estimate is at the boundary.  ❝ and/or ❝ 3. That the empirical model is useless. Yes. ❝ EM Lucky you!  — 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 |
Ing. Helmut Schütz