Table I.
Experimental Set-up for Model A, B, and C Regarding Parameterization, Sample Size, and Sampling Schedule
| Model | Description | Parameterization | Design | W2/LOQ/BQL | ||
|---|---|---|---|---|---|---|
| I | II | III | ||||
| A | One compartment model with oral dosing and absorption transit compartments, BQL data in absorption phase | Absorption rate constant (KA = 1), mean transit time (MTT = 1.5), number of transit compartments (NTC = 10), clearance (CL = 1), distribution volume (VC = 30). Log-normal intra individual variability of 30% for all typical parameters. Log transformed DV | 100 studies of 100 individuals with 7 samples: 1, 2, 3, 4, 6, 12 and 24 h post-dose | W2: 0.3, LOQ: 0.2, BQL: 12% of all; 28% of samples 1–3 | W2: 0.45, LOQ: 0.4, BQL: 15% of all; 35% of samples 1–3 | W2: 0.75, LOQ: 0.8, BQL: 19% of all; 44% of samples 1–3 |
| B | Two compartment model with iv dosing | Clearance (CL = 5), central distribution volume (VC = 20), inter compartment clearance (Q = 5), peripheral distribution volume (VP = 100). Log-normal intra individual variability of 30% for all typical parameters. Log transformed DV | 100 studies of 100 individuals with 5 samples. Sampling intervals: 0–1, 1–2, 2–6, 8–12, and 12–24 h | W2: 0.55, LOQ: 1.25, BQL: 10% of all; 40% of last sample | W2: 0.7 LOQ: 1.75 BQL: 20% of all; 70% of last sample | W2: 0.9, LOQ: 2.5, BQL: 30% of all; 90% of last sample |
| C | Indirect response model with drug concentration effect on K OUT drug concentration following fix one comp mod | Baseline (BASE = 10), first order elimination constantan (K OUT = 0.015), linear drug concentration effect on K OUT (SLOP = 10). Log-normal intra individual variability of 30% for all typical parameters. Log transformed DV | 100 studies of 100 individuals with 5 samples: 0, 10, 20, 30, and 200 h | W2: 0.65, LOQ: 1.6, BQL: 10% of all; 17% of samples 2–4 | W2: 0.75, LOQ: 2, BQL: 18% of all; 30% of samples 2–4 | W2: 0.9, LOQ: 2.5, BQL: 28% of all; 46% of samples 2–4 |
In the right hand end of the table parameter value W 2, limit of quantification (LOQ) and approximate percentage of BQL samples associated with version I, II and III of each model