Nine RCTs were included in the review, with samples sizes that ranged from 59 to 581 patients (total of 2,254 patients). Trial quality scores ranged from 5 to 7 (median 6).
Significantly fewer patients were withdrawn due to adverse effects with fluconazole prophylaxis compared with those treated with itraconazole (RR 0.45, 95% CI 0.27 to 0.75; seven RCTs; Ι²=74%).
There were statistically significant differences favouring treatment with itraconazole for fungal infections (RR 1.34, 95% CI 1.08 to 1.67; nine RCTs; Ι²=0%) and invasive fungal infections (RR 1.33, 95% CI 1.02 to 1.73; nine RCTs; Ι²=4%)
There were no statistically significant differences between the two treatments for overall mortality (eight RCTs), fungal-related mortality (seven RCTs), and proven fungal infections (seven RCTs).
Sensitivity analyses yielded similar results.