Intérêt de la procalcitonine dans les infections néonatales

Abstract

The value of procalcitonin (PCT) measurement is not presently completely assessed for the diagnosis of neonatal infections. This parameter was assessed in a prospective study in the neonatal intensive care unit of Clermont-Ferrand Hospital (France) in comparison to C-reactive protein. All newborn infants admitted before 24 h of life (day 0) in the neonatal intensive care unit were included in the study. Newborns (102) were assigned to one of four groups: group 1: non-infected newborns (n = 41); group 2: possibly infected newborns (n = 33); group 3: probably infected newborns (n = 10); group 4: confirmed infections (n = 18 bacterial or fungal infections). C-reactive protein and PCT were determined in the sera at D0, D1, D3 and D8. We determined the optimal cutoff value of PCT using the Receiver Operating Characteristic curves (R.O.C.). The cutoff value is 1.5 ng/mL at D0 and 10 ng/mL at D1. PCT cutoff value is significantly higher at D1 because of a significant PCT peak on the first day of life independent of any infectious stimulus. Our study shows that at D0 and D1 infected newborn infants had significantly higher mean PCT and C-reactive protein values than non infected newborn infants. C-reactive protein has a better specificity but PCT has better sensitivity and negative predictive value. PCT seems to be an interesting marker of neonatal infections especially during the first 24 h of life even though the mechanism of PCT synthesis remains unclear.