For the three studies, treatment allocation was done by an external clinical trial company (Kulnigg study), by computer generated random number table35 and by an Internet based method.36 The allocation was not concealed in the study by Schroder et al.