Twenty RCTs (5,725 patients) were included in the review. Most studies did not report adequate information on sequence allocation and allocation concealment. Only four out of 20 studies were blinded and six out of 20 (30%) conducted an intention-to-treat analysis. Half of the studies were conducted in multiple centres.
Compared with fluconazole, mould-active prophylaxis significantly reduced the number of proven/probable invasive fungal infections (RR 0.71, 95% CI 0.52 to 0.98; 18 trials). There was evidence of some heterogeneity (Ι²=33%). Risk of invasive aspergillosis was lower in the mould-active group (RR 0.53, 95% CI 0.37 to 0.75; 15 trials) as was risk of invasive fungal infection-related mortality (RR 0.67, 95% CI 0.47 to 0.96; 15 trials).
There was a significantly increased risk of adverse events leading to antifungal discontinuation in the mould-active group (RR 1.95, 95% CI 1.24 to 3.07; 16 trials). There was no statistically significant difference in aspergillosis-related mortality and no evidence of any difference in overall mortality between the groups.
Results of subgroup analyses, meta-regression and sensitivity analyses were reported. There was no evidence of publication bias.