“…Power calculations were only reported for two studies, one at 85 % [10] and the other at 90 % [20]. Power was not relevant in the trial by Brisbois et al, as it was a proof-of-principle pilot trial [9], and power was not discussed in the remaining studies [12,13,[16][17][18][19]. Second, generalizability was reduced by the study of specific populations; four studies included only individuals with head and neck cancer [12,15,16,20], and another study limited recruitment to patients receiving docetaxel and paclitaxel [17].…”