Using re-randomisation designs to increase the efficiency and applicability of retention studies within trials: a case study.

The re-randomisation design can increase the efficiency and generalisability of SWATs for trials with multiple follow-up time points.