Molecular epidemiology of Candida colonizing critically ill patients in intensive care units.

BACKGROUND AND PURPOSE: Nosocomial Candida infections are an important cause of morbidity and mortality in critically ill patients. Although there is growing evidence that candidemia develops primarily as a consequence of endogenous colonization, hospital outbreaks of Candida infection are not uncommon. To examine the prevalence and consequence of Candida colonization in critically ill patients and to elucidate the contribution of cross-transmission to the high frequency of nosocomial fungal infection in intensive care units (ICUs), a 6-month prospective surveillance study was conducted. PATIENTS AND METHODS: A total of 342 adult patients with an expected ICU stay of 48 hours or more were enrolled in the study. Surveillance cultures were taken from the rectal region, oropharynx, and urine on ICU entry and weekly thereafter. The electrophoretic karyotypes (EKs) of all isolates were characterized using pulsed-field gel electrophoresis. RESULTS: A total of 873 Candida strains were isolated from 208 of the 342 patients (60.8%) during ICU stay. A comparison of the EK patterns generated from Candida strains isolated from different patients demonstrated a variability of karyotypes, and failed to identify predominant clones colonizing or infecting ICU patients. For 62 of 102 patients colonized with multiple isolates of the same Candida spp., the EK patterns of the Candida strains isolated from individual patients were identical or similar, even when isolated from different anatomical sites, and the patterns remained the same for up to 62 days. A total of 57 episodes of Candida infection occurred in 53 (25.5%) of these 208 patients. Thirty-six episodes (63.1%) of Candida infection were preceded by colonization with the same Candida spp. All infecting strains had identical or similar EK patterns to prior colonizing strains. CONCLUSIONS: While Candida colonization was common in ICU patients, karyotyping did not identify cross-transmission among these patients. Further, only 25.5% of patients with Candida colonization subsequently developed Candida infection. These findings suggest that universal prophylaxis is not warranted in critically ill patients with Candida colonization.