Objective  

To determine whether systemic antifungal prophylaxis decreases infectious morbidity and mortality in nonneutropenic, critically ill, trauma and surgical intensive care unit (ICU) adult patients.

Design  

Systematic review and meta-analysis of randomized clinical trials. We used a fixed effect model, with risk ratio (RR) and 95% confidence intervals (CI).

Participants  

Patients admitted to ICU after surgery or trauma, with multiple risk factors for fungal infections.

Interventions  

Nine studies (seven double blind) with a total of 1,226 patients compared ketoconazole (three) or fluconazole (six) to placebo (eight) or no treatment (one).

Results  

Prophylaxis with azole was associated with reduced rates of candidemia (RR 0.30, 95% CI 0.10–0.82), mortality attributable to Candida infection (RR 0.25, 95% CI 0.08–0.80), and overall mortality (RR 0.60, 95% CI 0.45–0.81). Time to event analysis showed a significantly lower probability of fungal infections in treated patients. There was no evidence of statistical heterogeneity between studies, and publication bias assessment gave a negative results. There was, however, wide variability in the definition and reporting of some relevant clinical outcomes (e.g., confirmed or suspected infections, colonization) and pooling of these outcome measures was not feasible.

Conclusions  

Prophylaxis of candidal infection among critically ill ICU patients has beneficial effect on certain outcome measures, but additional data from well designed clinical trials and long-term epidemiological observations are needed to provide firm recommendations for the selection of subgroups of patients who would most benefit from prophylaxis and to determine the effect of prophylaxis on fungal resistance patterns.