Fig. 4From: A comparison of methods to adjust for continuous covariates in the analysis of randomised trialsSimulation results for binary outcomes (bias)Back to article page