Fig. 2

Statistical power of the trial. Calculated for the two interventions in the study were assessed using a cross-over design (i.e. isotonic hydration and time on donation chair after donation; red line) and the two interventions in the study were assessed using a stepped-wedge design (i.e. modified AMT (applied muscle tension) and psychosocial intervention; blue line). Each intervention was tested at a Bonferroni corrected level (type I error) of 0.0125. The study design has four possible sequences for interventions assessed using a stepped-wedge design (X-I-I-I, X-X-I-I, X-X-X-I, X-X-X-X), and two possible sequences for interventions assessed using a cross-over design (X-I-X-I, I-X-I-X) with a study duration of 3 years and each period lasting 9 months (X = current practice, I = intervention). Power calculations assume 15,000 whole blood donations per 9-month period in each donation centre, with an overall rate of severe (syncope) VVRs of 0.18% per donation and a between-centre variance in log odds of 0.06 (estimated from 2014/15 NHSBT data)