After publication of our article in a recent issue of Critical Care[1], inconsistencies were identified in Figure 6. Four trials - by Fuentes-Orozco and colleagues [2, 3], Hall and colleagues [4], and Goeters and colleagues [5] - were allocated to the wrong subgroup in our meta-analysis. They should have been included in the first subgroup (glutamine <0.3 g/kg per day); this does not change the significance of the results (P = 0.01). The correct Figure 6 is given here in full as Figure 1.

Figure 1
figure 1

A subgroup meta-analysis of the effects of different dosages of glutamine on mortality in critically ill patients (fixed effects models). CI, confidence interval; M-H, Mantel-Haenszel.