Diagnostic accuracy of fungal PCR and β-d-glucan for detection of candidaemia: a preliminary evaluation.

AIMS: Although treatment for candidaemia is time critical, culture-based tests prolong turnaround times and may promote underdiagnosis. Non-culture-based tests have the potential to overcome these difficulties but are in limited clinical use. The aim of this work was to undertake an initial evaluation of two non-culture-based tests for diagnosis of candidaemia.

METHODS: Patients with candidaemia were identified prospectively over a 4-month period. Sera drawn from case (candidaemic) and control (non-candidaemic) patients on the same day as the positive blood culture were tested with both the Renishaw RenDx Fungiplex test and a commercial β-d-glucan (BDG) assay (Fungitell, Associates of Cape Cod). Sensitivity and specificity were calculated independently and in combination, using paired blood culture as the reference standard.

RESULTS: There were 10 eligible case patients and 39 negative controls. PCR sensitivity and specificity were found to be 44.4% (95% CI 18.9% to 73.3%) and 87.2% (72.8% to 94.8%), respectively. BDG sensitivity and specificity were 80% (47.9% to 95.4%) and 89.7% (75.9% to 96.5%), respectively. When combining PCR and BDG, sensitivity was 90% (95% CI 57.4% to 100%) and specificity was 79.5% (64.2% to 89.5%). When two sequential specimens were tested, PCR sensitivity increased to 60% (95% CI 31.2% to 83.3%) and BDG sensitivity to 90% (54.7% to 100%).

CONCLUSION: A combination of tests, or a single test at multiple time points, may be preferable to relying on one test at a single time point. This should be accounted for in design of future diagnostic accuracy studies of tests for invasive candidosis.