“…However, methodology was often weak, with reporting of method of randomisation, concealment of allocation and blinding lacking from many papers (Brown et al, 1980;Foster et al, 1980;Kinsella et al, 1981;Moriarty et al, 1981;Stryker and Bartholomew, 1986;Salminen et al, 1988;Karlson et al, 1989;Bye et al, 1992;Chowdhury et al, 2002;Delia et al, 2002). The choice of randomisation in papers that reported their methodology was adequate in two studies (Urbancsek et al, 2001;Martin et al, 2002). Intention-to-treat analyses were described in two papers (Urbancsek et al, 2001;Martin et al, 2002).…”