Feature of adaptive/innovative design | Considerations for resources |
---|---|
Number of interim analyses | The more interim analyses generally the more additional resources required. Setting up suitable systems or investing in software may reduce this Additional interim analyses may not always provide additional efficiency of the adaptive design [12] |
Adaptive randomisation | Using outcome adaptive randomisation [13] may require more complex randomisation systems than non-adaptive or alternative ADs that drop arms for lack of benefit |
Multi-arm multi-stage (MAMS) designs | Some approaches such as group-sequential MAMS [14] may have a more variable sample size than alternatives such as drop-the-losers [15]. The more interim analyses and arms there are, the higher the variability in the sample size |
Platform trials | A platform trial [16] may have fixed costs for the underlying infrastructure and an additional cost for each arm added in (the latter of which may be lower compared to a new separate trial) |
Dose-ranging trials | Changing which doses are allocated to participants may have associated pharmacy costs and also impact on excess treatment costs |
Population enrichment | Changes to eligibility Changes to randomisation method/approach (e.g. a stratification factor or minimisation factor may be dropped) Training of site staff to understand implications Changes to PIS |