Embed Size (px)
Citation preview
Session I: Scaling procedure and adaptive design(s) • Discussion [Schütz and Labes]The Global Bioequivalence Harmonisation Initiative
Type I Error (nominal α 0.05)Since in reference-scaling BE-limits are random variables, the patient’s risk might be compromisedABEL (EMA)• Example: 4-period 2-sequence full replicate design
60%50%40%30%20%0.000.020.040.050.060.080.10
60%50%40%30% CVwR20%0.000.020.040.050.060.080.10
120100 80 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
120100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
1200.020.040.060.080.10
0.05
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
1200.020.040.060.080.10
0.05TIEemp
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
1200.020.040.060.080.10
0.05TIEemp
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
RSABE (FDA)
60%50%40%30%20%0.000.050.100.150.200.25
60%50%40%30% CVwR20%0.000.050.100.150.200.25
120100 80 60 60%50% 40 40%30% CVwR 20 20%0.000.050.100.150.200.25
120100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.050.100.150.200.25
1200.050.100.150.200.25
0.05100 80n 60 60%50% 40 40%30% CVwR 20 20%
0.000.050.100.150.200.25
1200.050.100.150.200.25
0.05TIEemp
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.050.100.150.200.25
1200.050.100.150.200.25
0.05TIEemp
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.050.100.150.200.25
Session I: Scaling procedure and adaptive design(s) • Discussion [Schütz and Labes]The Global Bioequivalence Harmonisation Initiative
Type I Error (iteratively adjusted α)Proposed alternative:Decreasing the level of the test in the critical region (CVwR, n)• Example cont’dABEL (EMA) RSABE (FDA)
60%50%40%30%20%0.000.020.040.050.060.080.10
60%50%40%30% CVwR20%0.000.020.040.050.060.080.10
120100 80 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
120100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
1200.020.040.060.080.10
0.05
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
1200.020.040.060.080.10
0.05TIEemp
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
1200.020.040.060.080.10
0.05TIEemp
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.020.040.050.060.080.10
60%50%40%30%20%0.000.050.100.150.200.25
60%50%40%30% CVwR20%0.000.050.100.150.200.25
120100 80 60 60%50% 40 40%30% CVwR 20 20%0.000.050.100.150.200.25
120100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.050.100.150.200.25
1200.050.100.150.200.25
0.05100 80n 60 60%50% 40 40%30% CVwR 20 20%
0.000.050.100.150.200.25
1200.050.100.150.200.25
0.05TIEemp
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.050.100.150.200.25
1200.050.100.150.200.25
0.05TIEemp
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.050.100.150.200.25
Session I: Scaling procedure and adaptive design(s) • Discussion [Schütz and Labes]The Global Bioequivalence Harmonisation Initiative
Backup slide (iteratively adjusted α)Proposed alternative:Decreasing the level of the test in the critical region (CVwR, n)• Example cont’dABEL (EMA) RSABE (FDA)
60%50%40%30%20%0.000.010.020.030.040.05
60%50%40%30% CVwR20%0.000.010.020.030.040.05
120100 80 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
120100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
1200.010.020.030.040.050.05
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
1200.010.020.030.040.050.05
αadj
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
1200.010.020.030.040.050.05
αadj
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
60%50%40%30%20%0.000.010.020.030.040.05
60%50%40%30% CVwR20%0.000.010.020.030.040.05
120100 80 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
120100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
1200.010.020.030.040.050.05
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
1200.010.020.030.040.050.05
αadj
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
1200.010.020.030.040.050.05
αadj
100 80n 60 60%50% 40 40%30% CVwR 20 20%0.000.010.020.030.040.05
