“…Only five of the relevant studies were randomized, and none of these trials reported adequate concealment of allocation [13,14,16-18], an important factor in limiting selection bias, especially in trials where the participants know their group allocation. Furthermore, blinding of the clinical examinations to establish the presence of atopic disorders was only reported in five studies [14,[16][17][18]20]. Sample sizes were also small (n ¼ 16-60 per group), except in one trial [16], where 288 mother/infant pairs completed the trial.…”