Fig. 2From: Sequential Multiple Assignment Randomized Trial (SMART) to identify optimal sequences of telemedicine interventions for improving initiation of insulin therapy: A simulation studyThe (a) SMART and (b) RCT design in comparison. Each intervention component (Nurse, App, App + Nurse) is for 6 weeks long. (a) Randomisation happens at baseline for stage 1 intervention and at week 6 for stage 2 intervention, given stage 1 treatment and response. The four AIs are embedded in the design as a result. (b) Randomisation happens at baseline only, so the four arms correspond to the four AIsBack to article page