Eleven RCTs (n=886) were included.
The median quality score was 8.5 (range: 5 to 12). Of the 11 studies included in the review, five had adequate allocation concealment, six were blinded and nine used an intention-to-treat analysis. There was no evidence of publication bias for the mortality outcome.
Mortality. Antioxidants (either alone or in combination) were associated with a statistically significant reduction in overall mortality compared with placebo (RR 0.65, 95% CI: 0.44, 0.97; based on 866 patients from 11 trials). There was no evidence of heterogeneity between the study results (I-squared 0%). Subgroup analyses showed similar significant reductions in risk for single antioxidants (RR 0.52, 95% CI: 0.27, 0.98; based on 188 patients from 5 trials) and antioxidants given parenterally (RR 0.56, 95% CI: 0.34, 0.92; based on 236 patients from 7 trials), but no evidence of any significant effect for combined antioxidants, enteral administration, selenium, non-selenium antioxidants, or different doses of selenium.
There was no evidence of a difference between antioxidants and placebo for infectious complications (RR 0.90, 95% CI: 0.65, 1.21; based on 728 patients from 5 trials). There was a small amount of heterogeneity (I-squared 23.4%). Subgroup analyses showed no evidence of a difference between treatment and placebo for combined antioxidants, enteral or parenteral administration, selenium or non-selenium antioxidants.
There were insufficient data to pool for length of stay.