The epidemiology of candidaemia and mould infections in Australia

Abstract

Abstract A retrospective review of Candida bloodstream infections (BSI) in Australian hospitals between 1995 and 1998 was performed. Nine tertiary referral hospitals in the States of New South Wales, Victoria, South Australia and Queensland participated. Of all isolates, 56% were Candida albicans, 14% Candida parapsilosis, 13% Candida glabrata, 5% Candida krusei and 4.5% Candida tropicalis. There were significant differences in the distribution of species in different patient groups. Among surgical patients, 69% of candidaemia was due to C. albicans, whereas among medical patients the proportion was 52%, and in haematology patients 43% (P = 0.0012; Pearson χ2). BSI with C. krusei were almost exclusive to haematology patients and were the second leading cause of Candida infection in that group, accounting for 27% of infections. The temporal pattern of Candida isolates also revealed a relationship between the latter years of the study and a lower likelihood of infection with C. albicans. Logistic regression showed year of the study [P = 0.032; odds ratio (OR) 0.81; 95% confidence interval (CI) 0.67–0.98] and surgery (P = 0.005; OR 2.02; 95% CI 1.2–3.1) to be significant variables. The rate of candidaemia in Australian hospitals was similar to that reported for US hospitals at 0.1–0.27 per 1000 discharges. Since April 1998, the clinical database on the internet resource Mycology Online has invited submission of clinical details from cases of invasive mycoses from Australian clinicians. To date, descriptions of 43 patients with proven or presumptive mould infection have been entered on the database. Of these, the leading cause was infection with Aspergillus (n = 16), followed by Zygomycetes (n = 7), Fusarium (n = 5), Scedosporium (n = 5) and Exserohilum (n = 1). Although Aspergillus infections were the most frequent on the database, the variety of mould infections seen in this short time was surprising. A knowledge of local patterns of infection and antifungal susceptibility is useful in selecting empirical therapy and formulating prophylactic and pre-emptive strategies.