A method for ensuring that an equal number of participants with a characteristic that may influence the prognosis or the response to the intervention will be allocated to each of the compared groups, to reduce the potential for confusion bias.