Summary
Sepsis is a leading cause of death in the intensive care unit (ICU), with Candida spp. in the forefront among the important pathogens. As recent studies have shown, survival outcome is strongly influenced by adequate antifungal therapy at an early stage that is often delayed by the time lag associated with microbiological diagnosis. Risk factor‐based prediction models have a high negative predictive value, but positive prediction of candidaemia in the individual patient remains elusive. New antigen‐ or DNA‐based methods for early diagnosis still await clinical validation. Their routine use is hampered by methodological issues. Species distribution of invasive Candida isolates in the ICU appears to be influenced primarily by age, previous hospitalisation and colonising species. In the context of the importance of adequate first‐line treatment, recent guidelines favour the use of echinocandins in critically ill patients with symptoms evoking high suspicion of invasive candidiasis. This is supported by robust clinical trial data, a few interactions and low toxicity. Fluconazole is characterised by reduced activity against some important Candida species, elevated rates of persistent infection seen in comparative trials. Amphotericin B deoxycholate should be considered obsolete in ICU patients because of its high toxicity. Invasive aspergillosis (IA) is a rare devastating infection in the general ICU population, but some centres have reported elevated incidences and underdiagnosis as determined in autopsy‐controlled studies. Treatment with mould‐active agents such as voriconazole must be initiated early in patients with suspected IA.