Antifungal prophylaxis in critically ill patients

We congratulate van Till and colleagues on their review showing that selective decontamination of the digestive tract (SDD) is more effective than single-drug prophylaxis (SAP) in reducing yeast colonisation, infection and mortality [1].

The authors claimed that their review differs from our earlier review, which included paediatric or liver transplant patients. In our review a subgroup analysis was performed in both selected and unselected populations, and demonstrated a significant reduction in yeast carriage and infection in unselected critically ill patients [2].

van Till and colleagues assessed yeast colonisation, lumping together 'positive yeast cultures obtained from sputum, stool, urine and/or wound' [1]. The majority of SDD trials reported positive yeast cultures obtained from surveillance cultures of throat and rectal swabs, whilst the SAP studies mainly assessed positive yeast cultures obtained from diagnostic samples including lower airway secretions, urine and wound fluid. Grouping together surveillance and diagnostic cultures may be misleading in interpreting the efficacy of antifungal interventions.

van Till and colleagues' review demonstrated a nonsignificant 41% reduction in candidemia by SDD and a 68% significant reduction by SAP [1]. The authors concluded that SAP prevents candidemia, whilst SDD does not. We believe that van Till and colleagues evaluated two different populations, as the candidemia rates in the control individuals were 3.79% and 1.69% for SAP and SDD, respectively. A larger sample size is almost certainly needed for SDD to demonstrate a significant reduction in candidemia [3].