Early Discontinuation of Empirical Antifungal Therapy and Biomarkers

Empirical antifungal therapy (EAT) is frequently prescribed to septic critically ill patients with risk factors for invasive Candida infections (ICI). However, among patients without subsequent proven ICI, antifungal discontinuation is rarely performed, resulting in unnecessary antifungal overuse. The investigators postulate that the use of fungal biomarkers could increase the percentage of early discontinuation of EAT among critically ill patients suspected of ICI, as compared with a standard strategy, without negative impact on day 28-mortality. To test this hypothesis, the investigators designed a randomized controlled open-label parallel-group study.

Patients requiring EAT will be randomly assigned to: * intervention group: a strategy in which EAT duration is determined by (1,3)-B-Dglucan and mannan serum assays, performed on day 0 (day of EAT initiation) and day 3. Early stop recommendation, provided before day 7, will be determined using an algorithm based on the results of biomarkers. * control group: a routine care strategy, based on international guidelines, which recommend 14 days of treatment for patients without subsequent proven ICI, and who improve under antifungal treatment, or less in other situations.