Performance of 1,3-β-D-glucan for diagnosing invasive fungal diseases in children.

Abstract

Plasma 1,3-β-D-glucan (BDG) is indicated as a tool for early diagnosis of invasive fungal diseases (IFD). However, data on its diagnostic value are scarce in children. Therefore, definition of BDG test performance in paediatrics is needed. BDG was evaluated in children admitted to "Istituto Giannina Gaslini," Genoa, Italy, who developed clinical conditions at risk for IFD. Results were analysed for sensitivity, specificity, predictive values, likelihood ratios, accuracy, informedness and probability of missing one case by a negative test. A total of 1577 BDG determinations were performed on 255 patients (49% males, median age 5.4years). Overall 46 IFD were diagnosed, 72% proven/probable. The test performance was evaluated for 80pg/mL, 120pg/mL, 200pg/mL, 350pg/mL, 400pg/mL cut offs. Sensitivity was always <0.80 and specificity > 0.90 only for cut offs ≥200pg/mL. Negative predictive value was ≥0.90 for all the cut offs evaluated, while positive predictive value resulted barely 0.50 (8% IFD prevalence). Accuracy was never >0.90, and informedness was at best 0.50. The risk of missing one IFD by a negative result was < 10%. Analyses in haemato-oncological or newborn patients did not show major differences. Detection of serum BDG does not appear a valuable adjunctive diagnostic tool for IFD in paediatrics.