Epidemiology and outcome of Candida bloodstream infection in an intensive care unit in Hong Kong

Abstract

OBJECTIVE. To study the epidemiology of Candida bloodstream infection in the Intensive Care Unit.

DESIGN. Retrospective study.

SETTING. A 22-bed, mixed medical and surgical Intensive Care Unit of a 1400-bed university teaching hospital in Hong Kong.

PATIENTS. All adult patients (>18 years) who had at least one blood culture positive for Candida.

RESULTS. During the 9 years of the study period, there were 128 patients with episodes of candidaemia (point prevalence, 9.6 per 1000 Intensive Care Unit admissions), 72 entailed albicans candidaemia and 56 non-albicans candidaemia. Albicans was still the predominant species, but the incidence of tropicalis was increasing. The median lengths of hospital and Intensive Care Unit stays prior to taking of the culture revealing candidaemia were 15 and 6 days, respectively. In all, 61% of patients did not have Candida colonisation within 2 weeks of their candidaemia. The main anti-fungal agents used were fluconazole and amphotericin B, but only 89 (70%) of the patients received appropriate anti-fungal treatment. Intensive Care Unit and hospital mortalities were 70% and 78%, respectively. Patients who did not receive appropriate treatment within 3 days had a worse outcome than those who did.

CONCLUSIONS. Our data showed a high point prevalence of candidaemia in the Intensive Care Unit. Albicans was still the predominant species. Candidaemia occurred early during Intensive Care Unit stay, and a significant proportion of patients did not have prior fungal colonisation. Candidaemia in the Intensive Care Unit was associated with high morbidity and mortality. Many patients did not receive appropriately early anti-fungal therapy, and endured higher mortality than in the remainder.