Objectives: To describe the epidemiology of nosocomial candidemia in a 400-bed teaching hospital. Methods: Prospective follow of all patients with positive blood cultures for Candida species was done between August-2003 and October-2004. Searching for underlying diseases, presence of central venous lines, use of antibiotics and antifungal drugs prior to candidemia, and treatment were performed. Outcomes were survival thirty days after candidemia or death. Results: Seventy-three patients with candidemia were followed (65.8% male), mean of age was 47.3 years (range 0 to 87). Previous episodes of bacteremia occurred in 31.5%. Isolation of Candida in other sites before candidemia was present in 15.1% of patients. The most frequent underlying disease was cancer, diagnosed in 35.6%. Diabetes mellitus, cardiopathy, nephropathy, pulmonary or neurological diseases were presents in less than 20% of cases. Almost half of patients (49.3%) were submitted to surgery in the last three months. Antifungal prophylaxis with fluconazol was received by 13.7%. Antibiotics for more than 24 hours were previously used in 97.2% of cases. At the moment of candidemia, mechanical ventilation was present in 30.1% of patients, parenteral nutrition in 57.5% and central venous lines in 83.6%. Hemodyalisis in the 72 hours before candidemia were performed in 13.7% of cases. From all Candida isolates, 53.4% were C. albicans and 46.6% C. non-albicans (23.3% C. tropicalis, 15.1% C. parapsilosis and 8.2% C. glabrata). None risk factor or underlying disease was statistically related with any species of Candida. Antifungal therapy for more than 72 hours was given to 64.4% of patients. Overall mortality in thirty days was 47.9%. Antifungal treatment for more than 14 days was the only variable associated with lower mortality. No difference in the mortality rate was detected between different Candida species infections. Conclusion: Candidemia is a frequent and serious nosocomial infection, with high mortality. Frequency of non-albicans species of Candida is high, but identification of species with lower susceptibility to fluconazol was not common. Antifungal treatment for more than 14 days was associated with lower mortality. Death in the first 72 hours of treatment was not rare and reinforces the importance of suspicion of candidemia in risk patients and early start of therapy.
Full conference title:
16th European Congress of Clinical Microbiology and Infectious Diseases
- ECCMID 16th (2006)