RESUMEN
BACKGROUND: Early, accurate diagnosis of invasive fungal disease (IFD) improves clinical outcomes. 1,3-beta-d-glucan (BDG) (Fungitell, Associates of Cape Cod, Inc., Falmouth, MA, USA) detection can improve IFD diagnosis but has been unavailable in Australia. AIMS: To assess performance of serum BDG for IFD diagnosis in a high-risk Australian haematology population. METHODS: We compared the diagnostic value of weekly screening of serum BDG with screening by Aspergillus polymerase chain reaction and Aspergillus galactomannan in 57 at-risk episodes for the diagnosis of IFD (proven, probable, possible IFD). RESULTS: IFD episodes were: proven (n = 4); probable (n = 4); possible (n = 18); and no IFD (n = 31). Using two consecutive BDG results of ≥80 pg/mL to call a result 'positive', the sensitivity, specificity, positive predictive value and negative predictive value was 37.5%, 64.5%, 23.1% and 80.7% respectively. For invasive aspergillosis, test performance increased to 50%, 90.3%, 57.1% and 87.5% respectively if any two of serum BDG/Aspergillus polymerase chain reaction/galactomannan yielded a 'positive' result. In proven/probable IFD, five of eight episodes returned a positive BDG result earlier (mean 6.6 days) than other diagnostic tests. False-negative BDG results occurred in three of eight episodes of proven/probable IFD, and false positive in 10 of 31 patients with no IFD. Erratic patterns of BDG values predicted false positive results (P = 0.03). Using serum BDG results, possible IFD were reassigned to either 'no' or 'probable' IFD in 44% cases. Empiric anti-fungal therapy use may have been optimised by BDG monitoring in 38.5% of courses. CONCLUSIONS: The BDG assay can add diagnostic speed and value but was hampered by low sensitivity and positive predictive value in Australian haematology patients.