Neonatal Sepsis Workup

Updated: Jun 13, 2019
  • Author: Nathan S Gollehon, MD, FAAP; Chief Editor: Muhammad Aslam, MD  more...
  • Print

Approach Considerations

Laboratory studies used to evaluate for early-onset and late-onset sepsis include a complete blood cell (CBC) count and differential, measurement of levels of C-reactive protein (CRP) and other infection markers. Culture of blood, urine, and cerebrospinal fluid (CSF) samples remains the gold standard. [37] A Gram stain may provide early identification of the gram-negative or gram-positive status of the organism for preliminary identification. DNA-based identification techniques are becoming available and may supplement culture results and provide rapid diagnostic information. [38] Rapid pathogen detection with multiplex polymerase chain reaction (PCR) may facilitate more timely selection of targeted antibiotic therapy while limiting exposure to broad-spectrum antibiotics. [39]

Because of the low incidence of meningitis in the newborn with negative blood culture results, clinicians may elect to culture the cerebrospinal fluid (CSF) of only those infants with documented or presumed sepsis. However, data from large studies show a 38% rate of culture-positive meningitis in neonates with negative blood culture results and suspected sepsis. Accordingly, a lumbar puncture should be considered when evaluating the infant with suspected sepsis.

Imaging studies employed in the workup of neonatal sepsis should target the neonate's symptoms and may include chest radiography to evaluate pulmonary involvement, as well as computed tomography (CT) scanning, magnetic resonance imaging (MRI), and ultrasonography of the head in cases of meningitis.


Laboratory Studies


Aerobic and anaerobic cultures are appropriate for most of the bacterial pathogens associated with neonatal sepsis. Anaerobic cultures are especially important in neonates who have abscesses, processes with bowel involvement, massive hemolysis, or refractory pneumonia. In one study, anaerobic infections were responsible for 16% of early-onset sepsis amongst very low birthweight infants. [40]  

Bacterial culture results should generally reveal the organism of infection within 36-48 hours; subsequent initial identification of the organism occurs within 12-24 hours of the growth. Single-site blood cultures are effective for isolating bacteria in neonates with sepsis, although obtaining two cultures from separate sites has been shown to be useful in determining if commensal species represent true infection or a contaminated sample. [40]  Blood culture from the umbilical cord may be considered. This route is attractive because larger volumes of blood may be drawn without concern, optimizing recovery of the offending organisms. However, contaminants have been reported at high rates in some, studies and specimen handling can be challenging. [41] Urine cultures are most appropriate for the investigation of late-onset sepsis.

Complete blood cell (CBC) count and differential

A CBC and differential may be ordered serially to determine changes associated with the infection (eg, thrombocytopenia or neutropenia) or to monitor the development of a left shift or changes in the ratio of immature to total neutrophils, although the sensitivity and specificity of these markers is low. Serial monitoring of the CBC may be useful in aiding the differentiation of sepsis from nonspecific abnormalities due to the stress of delivery.

Platelet count

The platelet count in the healthy newborn is rarely lower than 100,000/µL in the first 10 days of life (normal, ≥150,000/μL). Thrombocytopenia (platelet count < 100,000/µL) may be a presenting sign of neonatal sepsis and can last as long as 3 weeks; 10%-60% of infants with sepsis have thrombocytopenia. [42]  However, thrombocytopenia is an insensitive and nonspecific finding as well as a late indicator of serious bacterial infection, making its utility in the initial workup of neonatal sepsis questionable.

Because of the appearance of newly formed platelets, mean platelet volume (MPV) and platelet distribution width are significantly higher in neonatal sepsis after 2-3 days of life. These measures may assist in determining the cause of thrombocytopenia. However, owing to the myriad of causes of thrombocytopenia and its late appearance in neonatal sepsis, the presence of thrombocytopenia does not aid the diagnosis of neonatal sepsis.

White blood cell counts and ratios

Although white blood cell (WBC) counts and ratios are more sensitive for determining sepsis than platelet counts are, they remain very nonspecific and have a low positive predictive value. Normal WBC counts may be initially observed in as many as 50% of cases of culture-proven sepsis. Infants who are not infected may also demonstrate abnormal WBC counts related to the stress of delivery or to any of several other factors. A low WBC count (< 5,000/µL) is associated with a higher likelihood ratio for sepsis than an elevated WBC count (>20,000/µL). [43]

A differential may be of use in diagnosing sepsis; however, these counts are largely dependent on the laboratory technician performing them. The total neutrophil count (polymorphonuclear cells [PMNs] and immature forms) is slightly more sensitive for determining sepsis than the total leukocyte count (percent lymphocyte + monocyte/PMNs + bands), although the overall likelihood ratio remains low. 

Abnormal neutrophil counts at the time of symptom onset are observed in only two thirds of infants; therefore, the neutrophil count does not provide adequate confirmation of sepsis. Neutropenia is also observed with maternal hypertension, severe perinatal asphyxia, and periventricular or intraventricular hemorrhage.

Neutrophil ratios may have some limited utility in the diagnosis of neonatal sepsis. The immature-to-total (I/T) ratio is the most sensitive (60%-90%). All immature neutrophil forms are counted. The maximum acceptable I/T ratio for excluding sepsis in the first 24 hours is 0.16. In most newborns, the ratio falls to 0.12 within 60 hours of birth. Because elevated I/T ratios may be observed with other physiologic events, their positive predictive value is limited. In addition, the specificity of the I/T ratio is only 50%-75%, limiting its clinical usefulness. 

The utility of the CBC increases after the first 4 hours of age, with the WBC count findings, I/T ratio, presence of thrombocytopenia, and low absolute neutrophil count (ANC) all having significantly improved likelihood ratios in this later timeframe. Delaying the CBC until 4 hours or later may be prudent if the intent of its use is to make decisions regarding the likelihood of infection. In this context, decisions about antibiotic treatment should largely be based upon clinical findings and maternal risk factors. [43]

C-reactive protein (CRP), procalcitonin, and other markers

Levels of CRP, an acute-phase protein associated with tissue injury, are elevated at some point in 50%-90% of infants with systemic bacterial infections. [44] CRP levels rise secondary to macrophage, T-cell, and adipocyte production of interleukin (IL)–6. This is especially true of infections with abscesses or cellulitis of deep tissue.

CRP levels usually begin to rise within 4-6 hours of the onset of infection, become abnormal within 24 hours of infection, peak within 2-3 days, and remain elevated until the inflammation is resolved. The CRP level is not recommended as a sole indicator of neonatal sepsis but may be used as part of a sepsis workup or as a serial study during infection to assess the response to antibiotics, determine the duration of therapy, or identify a relapse of infection.

Immunoglobulin M (IgM) concentration in serum may be helpful in determining the presence of an intrauterine infection, especially if the infection has been present for some time. Elevated IgM levels in umbilical cord sera suggest intrauterine infection. The clinical availability of such testing and access to timely results limits this assay’s utility.

Evidence on the use of infection markers such as CD11b, soluble CD14 subtype, CD64, IL-6, IL-8, IL-10, and granulocyte-colony stimulating factor (G-CSF) for evaluation of sepsis in neonates shows that they may be helpful as adjunctive tests. [45, 46, 47, 48] Their value may be further enhanced by performing serial measurements and using combinations of tests. At present, however, the consensus is that these tests should not be used alone to determine the need for antibiotic therapy, although in some cases they may prove useful in determining when to stop antibiotic therapy.

Levels of other acute-phase reactants (eg, procalcitonin and serum amyloid) are often elevated with the onset of sepsis. Procalcitonin, a pro-peptide of calcitonin produced in monocytes and in the liver, may be more sensitive than CRP. It is more specific to bacterial infection than viral infection. Levels of procalcitonin can be elevated in infants with respiratory distress syndrome and in infants of diabetic mothers, and it should be used in conjunction with the entire clinical situation and not as a single determinant of treatment initiation or duration. Procalcitonin may be used in combination with other acute-phase reactants, such as CRP. [19, 49, 50, 51]

Coagulation studies

Disseminated intravascular coagulation (DIC) can occur in infected infants. Predicting which infants will be affected at the onset of sepsis is difficult. [52]

Infants with DIC show abnormalities in the prothrombin time (PT), the partial thromboplastin time (PTT), and fibrinogen and D-dimer levels, and they may need blood products, including fresh frozen plasma (FFP) and cryoprecipitate, to replace coagulation factors consumed in association with DIC. If infants show signs consistent with impaired coagulation (eg, gastric blood, bleeding from intravenous or laboratory puncture sites, or other bleeding), coagulation should be evaluated by checking these values.


Lumbar Puncture and CSF Analysis

Lumbar puncture may be warranted in the workup of early- and late-onset sepsis, although clinicians may be unsuccessful in obtaining sufficient or clear cerebrospinal fluid (CSF) for all the studies. If positive culture results are obtained, a follow-up lumbar puncture is often performed within 24-36 hours after the initiation of antibiotic therapy to document CSF sterility. If organisms are still present, modification of the drug type or dosage may be required for adequate treatment of the meningitis. An additional lumbar puncture within 24-36 hours of the change in therapy is necessary if organisms remain present.

CSF analysis

CSF findings in infective neonatal meningitis are as follows:

  • Elevated white blood cell (WBC) count (predominantly polymorphonuclear neutrophils [PMNs])

  • Elevated protein level

  • Decreased glucose concentration

  • Positive culture results

  • Positive polymerase chain reaction (PCR)-based results

The CSF WBC count is within the reference range in 29% of group B Streptococcus (GBS) meningitis infections but in only 4% of gram-negative meningitis infections. Reference-range CSF protein and glucose concentrations are found in about 50% of patients with GBS meningitis but in only 15%-20% of patients with gram-negative meningitis. Obtaining a CSF culture is critical, in that neonatal meningitis is commonly present in neonates without bacteremia and with normal CSF findings. [53]

The decrease in CSF glucose concentration does not necessarily reflect serum hypoglycemia. Glucose concentration abnormalities are more severe in late-onset disease and with gram-negative infections.

Herpes simplex virus (HSV) PCR testing

No consensus has been reached regarding the inclusion of HSV PCR testing of CSF as part of a routine sepsis workup in the neonate. Practice in this area is variable, with some centers reserving HSV PCR for infants with CSF abnormalities or for infants with clinical symptoms but who have negative cultures and do not respond to antibiotics. [54] However, vesicles are not present in as many as one third of cases of CNS HSV and disseminated HSV. Further research in this area is needed to provide clear practice recommendations.


Imaging Studies

Chest radiography

Chest radiography of infants with concomitant congenital pneumonia may reveal segmental or lobar infiltrate, but it more commonly reveals a diffuse, fine, reticulogranular pattern, much like that seen in respiratory distress syndrome (RDS). Pleural effusions may also be observed.

Head computed tomography (CT) scanning or magnetic resonance imaging (MRI)

CT scanning or MRI may be needed late in the course of complex neonatal meningitis to document obstructive hydrocephalus, the site where the obstruction is occurring, and the occurrence of major infarctions or abscesses. Signs of chronic disease (eg, ventricular dilatation, multicystic encephalomalacia, and atrophy) may also be demonstrated on CT scan or MRI.

Head ultrasonography

Head ultrasonography in neonates with meningitis may reveal evidence of ventriculitis, abnormal parenchymal echogenicities, extracellular fluid, and chronic changes. Serial head ultrasonography can reveal the progression of complications.