Fig. 2From: Time-varying SMART design and data analysis methods for evaluating adaptive intervention effectsExample of time-varying two-stage SMART design with unequal probability allocation: only non-responders are re-randomized in the second stageBack to article page