Indications for Testing
Most cases of neonatal sepsis (roughly 80-90%) present in the first 2 days of life. Manifestations range from nonspecific symptoms to multiorgan failure. Infants rarely present with fever unless born to a febrile mother, although some acquire fever immediately after delivery. It is more common for an infant to be hypothermic or present with pneumonia, two nonspecific signs of sepsis. Other presenting signs that warrant investigation include lethargy, irritability, poor feeding, respiratory symptoms (eg, apnea, grunting), and abdominal distention.
Neonatal sepsis is defined as a systemic infection occurring in infants ≤28 days old. Early-onset neonatal sepsis is defined as sepsis presenting within 72 hours of birth, and late-onset sepsis as sepsis occurring after 72 hours in infants in the neonatal intensive care unit (NICU) and after 7 days of life in full-term infants.
Initial Workup of Suspected Neonatal Sepsis
Initial, nonspecific testing (eg, urine cultures, white blood cell counts) may prove useful for identifying infants with a low probability of developing sepsis, but not for identifying infants likely to be infected.
A single blood culture is used to determine the presence of bacterial infection. The culture should be obtained before beginning treatment.
Cerebrospinal Fluid Studies
Cerebrospinal fluid (CSF) studies provide optimal diagnostic sensitivity for detecting sepsis. The American Academy of Pediatrics (AAP) recommends that all newborn infants with suspected sepsis have a lumbar puncture if stable enough for the procedure. CSF studies are not necessary in all infants with suspected sepsis; if there are no obvious maternal risk factors or there is suspicion that a noninfectious etiology may be involved, CSF studies can be deferred. However, if blood cultures are positive or infants fail to respond to therapy, a lumbar puncture should be obtained and tested.
CRP is a nonspecific marker of acute inflammation. Evidence supports the use of CRP to help rule out neonatal sepsis in full-term infants, and recent studies suggest similar efficacy in preterm infants. CRP increases within 6-8 hours of infection and peaks at 24 hours. CRP sensitivity is initially low, but improves 6-12 hours after birth. Two normal readings provide strong evidence that bacterial sepsis is unlikely and that any antibiotic treatment can be discontinued. There is no evidence to support repetition of CRP testing to determine duration of treatment in an infant with an elevated CRP (≥1.0 mg/dL).
PCT is another acute phase reactant that is used as a marker for the early detection of sepsis, and although it is more sensitive than CRP for identifying early sepsis, it is less specific. A physiologic increase in PCT occurs within the first 24 hours after birth and may appear in the presence of noninfectious conditions (eg, respiratory distress syndrome). The interpretive cutoffs established for adults may not be transferable to the pediatric population, but a cutoff of 2.0-2.5 ng/mL is moderately accurate in predicting neonatal sepsis.