RESUMO
AIM: We assessed the diagnostic accuracy of serum (1 â 3)-ß-D-glucan (BDG) for neonatal invasive candidiasis (NIC) using the recommended cut-off usually used in adults for detecting invasive candidiasis and searched for an optimal cut-off for ruling out NIC. METHODS: We conducted a prospective cross-sectional study at Nantes University medical centre from January 2017 to July 2018. All consecutive newborn infants of less than 28 days of corrected age, with clinically suspected NIC, who underwent BDG assay, were included. Sensitivity and specificity were calculated by using the recommended cut-off of 80 pg/mL. Receiver operating characteristic curve analysis was used to identify an optimal cut-off value. RESULTS: We included 55 newborn infants with 61 episodes of suspected NIC. Their median gestational and chronological ages were 28.0 weeks (interquartile range [IQR] 26.4-34.1) and 10.0 days (IQR 6.0-22.0), respectively. Of 61 episodes, seven revealed NIC. Sensitivity and specificity were 85.7% (95% confidence interval [CI] 42.1%-99.6%) and 51.9% (37.8%-65.7%) with the recommended cut-off, respectively. An optimal cut-off of 174 pg/mL offered the same sensitivity but higher specificity 77.8% (64.4%-88.0%). CONCLUSION: The recommended cut-off of 80 pg/mL was probably too low for ruling out NIC. A higher cut-off might have been more appropriate.