Background Candida spp. are frequently cultured from the respiratory tract in critically ill patients. Most intensivists start amphotericin-B deoxycholate (ABDC) inhalation therapy to eradicate Candida spp. from the respiratory tract. However, the safety and efficacy of this treatment are not well established. The purpose of this study was to assess the safety and efficacy of ABDC inhalation for the treatment of respiratory Candida spp. colonization in critically ill patients.MethodsAll non-neutropenic patients admitted into the intensive care unit (ICU) of a university hospital from December 2010¿2011, who had positive Candida spp. cultures of the respiratory tract for more than 1 day and required mechanical ventilation >48 h were retrospectively included. The decision to start ABDC inhalation had been made by attending intensivists on clinical grounds in the context of selective decontamination of the digestive tract. Infection characteristics and patient courses were assessed.ResultsHundred and thirteen consecutive patients were studied. Fifty-one of them received ABDC inhalation and their characteristics at baseline and day 1 of respiratory colonization did not differ from those of colonized patients not receiving treatment (n¿=¿62). The ABDC-treated group had a similar Candida spp. load but did not decolonize more rapidly as compared to untreated patients. The clinical pulmonary infection and lung injury scores did not decrease as in the untreated group. In a Cox proportional hazard model, the duration of mechanical ventilation was increased (P¿<¿0.003) by ABDC treatment independently of other potential determinants and Candida spp. colonization. No differences in ventilator-associated pneumonia or in overall mortality (up to day 90) were observed.ConclusionTreatment of respiratory Candida spp. colonization in non-neutropenic critically ill patients by inhaled ABDC may not affect respiratory colonization but may increase duration of mechanical ventilation, because of direct toxicity of the drug on the lung.