Table 4

Designs with maximal power of detecting the interaction, according to sensitivity and specificity

Gene-surrogate
Flexible two-phase design options
Expected Phase One counts
Power#
Cost*

Spec
Sens
n0
n1
ρ0
ρ1
N0
N1



70%
80%
800
400
90%
90%
5902
1373
83%
1564
70%
90%
800
400
90%
90%
5882
1325
87%
1560
80%
60%
800
400
90%
90%
8824
1988
87%
1741
80%
70%
800
400
90%
90%
8780
1889
91%
1733
80%
80%
800
400
90%
90%
8738
1800
94%
1727
80%
90%
800
400
90%
90%
8696
1718
96%
1720
90%
60%
900
300
90%
90%
19286
2000
98%
2264
90%
70%
900
300
90%
90%
19104
2000
99%
2255
90%
80%
900
300
90%
90%
18925
1960
99.6%
2244
90%
90%
900
300
90%
90%
18750
1835
99.8%
2229

Wild et al. Epidemiologic Perspectives & Innovations 2008 5:4   doi:10.1186/1742-5573-5-4