Fig. 1
From: Independence estimators for re-randomisation trials in multi-episode settings: a simulation study
Bias and coverage of independence estimators in simulation study 1. PE = per-episode; PP = per-patient; AB = added-benefit; PB = policy-benefit. Error bars are 95% confidence intervals based on Monte Carlo standard errors. Scenario 1: Constant treatment effect. Scenario 2: Treatment effect varies across episode. Scenario 3: Treatment effect varies across patients with different values of Mi. Scenario 4: Treatment effect carries forward. Scenario 5: Treatment becomes less effective on re-use. Scenario 6: Treatment effect varies across episodes, across patients with different values of Mi, carries forward, and becomes less effective on re-use