“…Estimates of the standard-placebo difference formed the basis for the choice of the irrelevance margin in only 20 trials (Anonymous 1995(Anonymous , 1997bAbrams et al, 1998;Benkert, Diehm et al, 1996;Emery et al, 2000;Glasser et al, 1995a;Hawkey et al, 2000;Koopman et al, 1996;Lefebvre et al, 1996;Mattson et al, 2000;Moliterno and Topol, 2000;Peuskens et al, 1999;Pfeffer et al, 2000;Philipp, Kohnen and Hiller, 1999;Schnitzer et al, 2000;Sierakowski et al, 1997;Tyring et al, 1998). Further six trials used estimates of the difference between a new (more effective) and an old (less effective) standard treatment as a basis (Anonymous 1997a(Anonymous , 1999Clarke et al, 1998;Garcia et al, 2000;McCarty et al, 1999;Moore et al, 1999).…”