“…Control arms of studies provided standard medical care without (28, 30 per cent) or with (11, 12 per cent) a resting period, reported no music (13, 14 per cent), reported no intervention (17, 18 per cent), used a device with sham sounds (8, 9 per cent), provided headphones without music (6, 7 per cent) or with noise‐blocking features (2, 2 per cent), used midazolam (1, 1 per cent) or had an unclear description (5, 5 per cent). Eleven RCTs (12 per cent) did not report on quantitative data and could not therefore be included in the quantitative analyses ( Table S1 , supporting information).…”