RESUMEN
BACKGROUND: Candida is a leading cause of infection in the intensive care unit. Colonization versus infection remains a challenge. A Candida Score (CS) of 3 or greater has been used to target antifungal therapy in surgical patients at risk of candidemia but has not been well evaluated in medical patients with sepsis. OBJECTIVES: The aim of this study was to assess utility of the CS in detecting candidemia early in patients with sepsis. METHODS: This was a secondary analysis of patients with sepsis (n = 77) who were followed up for development of new infections. Patients with known fungal infection at admission were excluded. Candida colonization was defined as Candida cultured from any baseline culture, except blood, as a part of routine clinical care. RESULTS: Candidemia was detected in 8 of 77 participants (10.4%; 4 [15.4%] with a CS ≥3 and 4 [7.8%] with a CS <3). Demographic variables (age, race, sex) were similar among those who did and did not develop candidemia. Using the recommended CS of 3 or greater, sensitivity was (4/8) 50%, specificity was (47/69) 68.1%, positive predictive value was (4/26) 15.4%, and negative predictive value was (47/51) 92.2%. Baseline colonization was significantly higher among those who developed candidemia (50% vs 11.6%; P = .02), but no significant differences were observed among CS components or total scores. CONCLUSIONS: Despite a relatively poor sensitivity, a reasonable specificity with a strong negative predictive value makes this tool a viable option for screening medically ill patients who may require antifungal agents. The CS should be evaluated in a larger, more inclusive, medical population.