Fig. 1From: Estimation of treatment preference effects in clinical trials when some participants are indifferent to treatment choiceThe two-stage randomised trial designBack to article page