Invasive Candida infections in patients of a medical intensive care unit: attempt of improving diagnosis by quantifying the colonization.

Abstract

During the past decades, a steady increase in the incidence of invasive Candida infections in patients requiring intensive care has been reported. Nevertheless, the diagnosis of invasive Candida infections in the intensive care unit setting remains still difficult and is often made late in the course of disease. Additionally, prognosis worsens rapidly with delayed initiation of antifungal therapy. Clinical scoring systems such as the Candida colonization index or the four risk factor-based Candida score could be very useful tools to select patients at high risk of developing invasive Candida infections, who would benefit from the administration of systemic antifungal therapy. The main objective was to determine the incidence of invasive Candida infection among patients of a medical intensive care unit and to investigate its association with the Candida colonization index and the Candida score. A prospective, observational, single-center study was performed from December 2010 to December 2011. All patients over the age of 18 years who were admitted for at least 7 days in a medical intensive care unit were included. Demographic characteristics, underlying diseases, reasons for intensive care unit admission and presence and duration of risk factors for Candida species colonization and infection were collected for each patient. The Candida colonization index and the Candida score were calculated weekly until discharge or death. A total of 65 patients were included. Proven invasive Candida infections were diagnosed in five patients (7.7 %), i.e. two developed candidemia, two peritonitis, and one pneumonia. All of these patients were characterized by a Candida colonization index ≥ 0.5 (incidence rate: 16.7 % [5/30]) and a Candida score ≥ 2.5 (incidence rate: 29.4 % [5/17]). Mechanical ventilation (p = 0.013) and both Candida scoring systems (p = 0.013 versus p < 0.001) were statistically and significantly associated with invasive Candida infections. The mortality rate was high (80 %). Interestingly, treatment with antifungal drugs was not statistically and significantly associated with invasive candidiasis (p = 0.077) and patients outcome (p = 0.057). Both the Candida colonization index and the Candida score with cut-off values ≥ 0.5 and ≥ 2.5 are very useful tools to select patients at high risk of developing invasive Candida infections in the medical intensive care unit setting, who would benefit from early antifungal treatment.