A multi-arm multi-stage clinical trial design for binary outcomes with application to tuberculosis

Table 3 Examples of two-arm two-stage phase II/III TB trials

Design	Stage(i)	α _i	ω _i	n _i	N _i	t _i	ESS \|H ₀	ρ\|H ₀	ρ\|H ₁	A _i	Ω _i
(v)	1	0.5	0.90	56	134	0.67	723	0.10	0.08	0.500	0.900
	2	0.025	0.90	1050	1312	3.84				0.015	0.813
(vi)	1	0.5	0.95	94	178	0.89	745	0.12	0.11	0.500	0.950
	2	0.025	0.90	1050	1312	4.00				0.015	0.857
(vii)	1	0.2	0.90	156	252	1.26	464	0.16	0.14	0.200	0.900
	2	0.025	0.90	1050	1312	4.28				0.008	0.815
(viii)	1	0.2	0.95	214	320	1.60	518	0.19	0.16	0.200	0.950
	2	0.025	0.90	1050	1312	4.54				0.009	0.858

Characteristics of two-arm two-stage TB trials where I = culture status observed 14 weeks after randomisation and D = relapse status at 18 months. Key: for stage i, α _i = significance level, ω _i = nominal power, n _i = total sample size required for analysis i, N _i = cumulative number of patients recruited by the end of stage i, t _i = predicted timing (in years) of the end of stage i, ESS |H ₀ = expected sample size under the null hypothesis, ρ|H _h = correlation between stages under hypothesis H _h, A _i = probability of passing stage i under H ₀, Ω _i = probability of passing stage i under H ₁.

ISSN: 1471-2288