“…bo-controlled studies (e.g., in the discussion section of the publication, in 3 trials) (Guay et al, 1999;Guez et al, 1996;Saekeland and Albrecht, 1997), or included an additional placebo or 'minor control' (a treatment assumed to be inferior compared to the standard included in the actual trial) group (Aubier et al, 1999;Day et al, 2000;DuBiner et al, 1996;Dulin et al, 1998;Elfstraem et al, 1999;Farnier, Portal and Maigret, 2000;Lallemant et al, 2000;Matthys et al, 1998;Rehn et al, 1996;Sall, 2000;Singer et al, 2000). Here the median of the 'preserved' effect referred to the point estimate amounts to about 15% (data not shown), including four trials in which the irrelevance margin was even larger than that point estimate.…”