Premium
A Bayesian group sequential small n sequential multiple‐assignment randomized trial
Author(s) -
Chao YanCheng,
Braun Thomas M.,
Tamura Roy N.,
Kidwell Kelley M.
Publication year - 2020
Publication title -
journal of the royal statistical society: series c (applied statistics)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.205
H-Index - 72
eISSN - 1467-9876
pISSN - 0035-9254
DOI - 10.1111/rssc.12406
Subject(s) - randomized controlled trial , bayesian probability , interim , posterior probability , interim analysis , sequential analysis , treatment and control groups , sample size determination , statistics , bayes' theorem , computer science , medicine , mathematics , surgery , geography , archaeology
Summary A small n , sequential, multiple‐assignment, randomized trial (called ‘snSMART’) is a small sample multistage design where participants may be rerandomized to treatment on the basis of intermediate end points. This design is motivated by the ‘A randomized multicenter study for isolated skin vasculitis’ trial (NCT02939573): an on‐going snSMART design focusing on the evaluation of three drugs for isolated skin vasculitis. By formulating an interim decision rule for removing one of the treatments, we use a Bayesian model and the resulting posterior distributions to provide sufficient evidence that one treatment is inferior to the other treatments before enrolling more participants. By doing so, we can remove the worst performing treatment at an interim analysis and prevent the subsequent participants from receiving the removed treatment. On the basis of simulation results, we have evidence that the treatment response rates can still be unbiasedly and efficiently estimated in our new design, especially for the treatments with higher response rates. In addition, by adjusting the decision rule criteria for the posterior probabilities, we can control the probability of incorrectly removing an effective treatment.