De-escalation of antifungal treatment in critically ill patients with suspected invasive Candida infection: incidence, associated factors, and safety

Abstract Background Antifungal treatment is common in critically ill patients, but only a small proportion of patients receiving antifungals have a proven fungal infection. However, antifungal treatment has side effects such as toxicity, emergence of resistance, and high cost. Moreover, empirical antifungal treatment is still a matter for debate in these patients. Our study aimed to determine the incidence, associated factors, and safety of de-escalation of antifungals in critically ill patients. Methods This retrospective study was conducted in a 30-bed mixed ICU, from January 2012 through January 2013. Patients hospitalized for > 5 days and treated with antifungals for first suspected or proven invasive Candida infection were included. Exclusion criteria were prophylactic antifungals, suspected invasive aspergillosis, and neutropenia. De-escalation was defined as switch from initial systemic antifungals (except fluconazole) to triazoles, or stopping initial drugs within the 5 days following their initiation. Results One hundred and ninety patients were included. Antifungal treatment was empirical, preemptive, and targeted in 55, 27, and 24% of study patients, respectively. Caspofungin (53%), fluconazole (43%), voriconazole (4%), and liposomal amphotericin B (0.5%) were the more frequently used antifungals. De-escalation was performed in 38 (20%) patients. Invasive mechanical ventilation was independently associated with lower rates of de-escalation (OR 0.25 [95% CI 0.08–0.85], p = 0.013). Total duration of antifungal treatment was significantly shorter in patients with de-escalation, compared with those with no de-escalation (med [IQR] 6 (5, 18) vs. 13 days (7, 25), p = 0.023). No significant difference was found in duration of mechanical ventilation (22 [5–31] vs. 20 days [10–35], p = 0.43), length of ICU stay (25 [14–40) vs. 25 days [11–40], p = 0.99), ICU mortality (45 vs. 59%, p = 0.13), or 1-year mortality (55 vs. 64%, p = 0.33) between patients with de-escalation and those with no de-escalation, respectively. Conclusions De-escalation was performed in 20% of patients receiving systemic antifungals for suspected or proven invasive Candida infection. Mechanical ventilation was independently associated with lower rates of de-escalation. De-escalation of antifungal treatment seems to be safe in critically ill patients