. Author manuscript; available in PMC: 2011 May 10.

Published in final edited form as: Stat Med. 2010 May 10;29(10):1061–1071. doi: 10.1002/sim.3676

Table 3.

Operating characteristics of a 90 endpoint product Phase 2b trial vs a 300 endpoint Phase 3 trial

(a) Decision guideline for Phase 2b design (n = 90 infections)
	Harm	Useless	Plaus Eff	Eff 1 Ph 3	Eff 2 Ph 3

	Probability of obtaining an estimated V E
True V E	≤ −.42	> −.42 and ≤ .24	> .24 and ≤ .55	> .55 and ≤ 1.0	> 1.0
−.5	.60	.40	<.001	<.001	<.001
0	.05	.85	.10	<.001	<.001
.3	<.001	.35	.63	.02	<.001
.5	<.001	.02	.66	.32	<.001
.7	<.001	<.001	.04	.96	<.001
(b) Decision guideline for Phase 3 design (n = 300 infections)
	Harm	Useless	Plaus Eff	Eff 1 Ph 3	Eff 2 Ph 3

	Probability of obtaining an estimated V E
True V E	≤ −.26	> −.26 and ≤ .20	> .20 and ≤ .45	> .45 and ≤ .53	> .53

−.5	.94	.06	<.001	<.001	<.001
0	.02	.95	.02	<.001	<.001
.3	.<.001	.13	.84	.02	<.001
.5	<.001	<.001	.20	.49	.31
.7	<.001	<.001	<.001	<.001	1.0

Bolded entries are correct decisions