Outcome assessment by central adjudicators in randomised stroke trials: Simulation of differential and non-differential misclassification.

For trials without adequate blinding, central adjudication is vital to control for differential misclassification. However, for large blinded trials, adjudication is of less importance and may not be necessary.