Fluconazole vs. amphotericin B for the management of candidaemia in adults: a meta-analysis.

The incidence of bloodstream infections caused by Candida species is rising. Few published studies have compared the efficacy of fluconazole with that of amphotericin B. We performed a meta-analysis of the prospective studies that compared fluconazole and amphotericin B for the treatment of candidaemia in adults. Data on total mortality, candidaemia-attributable mortality, efficacy, microbiological failure, and toxicity were extracted from eligible studies. All studies appeared homogeneous with respect to the outcome measures. Most patients were at relatively low risk for death as evidenced by the low average physiologic score and the lack of intense immunosuppression. The odds ratios (OR) of treatment with amphotericin B versus fluconazole and 95% confidence intervals (CI) were as follows: total mortality (OR, 1.06; CI, 0.89-1.25), candidaemia-attributable mortality (OR, 1.0; CI, 0.70-1.45), clinical response (OR, 1.14; CI, 0.93-1.39) and microbiological failure according to all Candida species (OR, 0.99; CI, 0.78-1.26). A trend favouring amphotericin B was seen in mycological eradication of non-albicans Candida species (OR, 0.70; CI, 0.47-1.06). Finally, amphotericin B was more toxic than fluconazole (OR, 2.94; CI, 2.14-4.4). In conclusion, fluconazole is as efficacious and less toxic than amphotericin B in stable, not severely immunosuppressed candidaemic patients at low risk for death. However, fluconazole may be less effective than amphotericin B in candidaemias caused by some non-albicans Candida species.