Five meta-analyses and 18 RCTs were considered, including 12 trials studied the effect of IgM in adults [1-12], 5 in newborns [13-17], and one in children 1-24 months old . All interventions were applied with basic therapy (BT). No difference between IgM and albumin was found for adults (Fig.1). However we found significant efficacy of IgM in adults when compared with all comparators, RR 0.69 [0.56; 0.84] (Fig.2), and BT, RR 0.52 [0.39; 0.69] (Fig.3). In newborns mortality is lower in IgM than in all comparators groups, RR 0.47 [0.29; 0.76] (Fig.4), and in BT with or without placebo, RR 0.50 [0.30; 0.84] (Fig.5). Children under 24 months receiving IgM also had lower mortality than in all comparators group, RR 0.48 [0.34; 0.68]. Indirect comparison of IgM and IgG in adults showed no differences, in newborns the difference is in favor of IgM, RR 0.47 [0.29; 0.77].