Essential aspects of a Pick a Winner design
| Randomization between a control arm and a number of novel treatments, with comparisons only between patients randomized between a given novel treatment and control |
| Ability to add new treatments (drug X) in a rolling program by protocol amendment |
| Rapid ascertainment of outcome, or a meaningful surrogate |
| Requirement that the minimal clinically relevant difference is relatively large |
| Evaluation after a small proportion (eg, one-fourth and one-half) of the required number of patients or events |
| Suitable continuation criteria |
| Randomization between a control arm and a number of novel treatments, with comparisons only between patients randomized between a given novel treatment and control |
| Ability to add new treatments (drug X) in a rolling program by protocol amendment |
| Rapid ascertainment of outcome, or a meaningful surrogate |
| Requirement that the minimal clinically relevant difference is relatively large |
| Evaluation after a small proportion (eg, one-fourth and one-half) of the required number of patients or events |
| Suitable continuation criteria |