Skip to main content

Table 1 Bayesian adaptive designs explored for the Collaborative Ankle Support Trial

From: Bayesian adaptive designs for multi-arm trials: an orthopaedic case study

DesignInterim analysis scheduleaArm allocationControl allocationEarly stopping
1None1:1:1:1Equal to other armsNone
2Every 200 patients1:1:1:1Equal to other armsEfficacy or futility every 200 patients
3Every 50 patientsArm dropping assessed at each interim analysisEqual to other armsEfficacy or futility every 200 patients
4Every 50 patientsRAR at each analysisMatched to best intervention armEfficacy or futility every 200 patients
5Every 50 patientsRAR at each analysisFixed at 40%Efficacy or futility every 200 patients
6Every 50 patientsRAR at each analysisNo designated control; tubular bandage is treated as an intervention armEfficacy or futility every 200 patients
  1. RAR Response adaptive randomisation
  2. aAt number of patients due for primary outcome follow up (at 12 weeks post-randomisation)