“…However, these studies are imperfect for several reasons including: heterogeneous study population (Madhusoodanan et al, 1995;Joshi and Joshi, 1996;Davidson et al, 2000;Madhusoodanan et al, 2000), retrospective design (Kiraly et al, 1998;Madhusoodanan et al, 1999a;Solomon and Geiger, 2000) single arm (Madhusoodanan et al, 1995;Berman et al, 1996;Madhusoodanan et al, 1999b;Davidson et al, 2000;Madhusoodanan et al, 2000;Solomons and Geiger, 2000) small sample size (Madhusoodanan et al, 1995;Berman et al, 1996;Barak et al, 2001;Madhusoodanan et al, 2000) nonspecific or limited outcome measures (Kiraly et al, 1998) and short follow up period (Madhusoodanan et al, 1999b;Maguire et al, 2001;Jeste et al, 2001). Despite these limitations they have consistently indicated efficacy in improving core schizophrenic symptoms, reducing extra pyramidal side effects (EPSE) and being well tolerated.…”