“…The control group of a four-arm trial 4 was triplicated (ie, data for patients in the control group were copied twice to have three control arms to compare with each experimental arm), a 2 × 2 trial (EORTC 22962 45 ) included three relevant comparisons for the meta-analysis, and three three-arm trials 36,37,55 included two relevant comparisons. The 33 trials included in the analysis of fractionation schedules (comparison 1) were divided into four predefined subgroups, depending on the type of radiotherapy: hyperfractionation (eight comparisons, including the unpublished EORTC 22962 trial), 4,33,44,45,49,50,56 moderately accelerated radiotherapy (19 comparisons), 2,4,30,32,34–39,41,42,46,54,55,57–59 very accelerated radiotherapy (seven comparisons, including the unpublished CHARTWEL trial), 3,47,51–53,60 and moderately hypofractionated (dose per fraction between 2–2·5 Gy [two comparisons]; 31,40
appendix pp 17–18).…”