Using Trial Sequential Analysis for estimating the sample sizes of further trials: example using smoking cessation intervention

摘要

I. Trial Sequential Analysis output of both MiQuit trials using 90% power, 5% significance, to detect a 3.4% absolute difference. Diversity = 0%. Points A and B on the z-curve represent each trial added to the sequential analysis. A – Feasibility trial [7], B – Pilot trial [8]. Monitoring boundaries in Fig. 1 use the Lan-DeMets O’Brien-Fleming alpha-spending approach. II. Point C represents a theoretical trial with a sample size of 630 women, where an absolute difference of 3.17% was observed, in favour of the MiQuit group versus the control group. Diversity = 0%. III. Point D represents a theoretical trial with a sample size of 630 women, with an absolute difference of − 0.63% in favour of the control group. Diversity = 33%