Neonatal Sepsis: An Update

Document Type: Review Article


1 Department of Pediatrics , Mahatma Gandhi Mission Hospital Aurangabad, 431001, Aurangabad, India

2 Department of Pharmacology, Mahatma Gandhi Mission Hospital Aurangabad, 431001, Aurangabad, India


Sepsis is the most common cause of neonatal mortality. As per National Neonatal-Perinatal Database (NNPD), 2002-2003, the incidence of neonatal sepsis in India was 30 per 1000 live births. Signs and symptoms of sepsis are nonspecific; therefore empirical antimicrobial therapy is promptly initiated after obtaining appropriate cultures. The early manifestations of neonatal sepsis are vague and ill-defined. Novel approaches in the diagnosis of neonatal sepsis include heart rate analysis on ECG, and colorimetric analysis of skin color. Although blood culture is the gold standard for the diagnosis of sepsis, culture reports are available only after 48-72 hours. In this era of multidrug resistance, it is mandatory to avoid unnecessary use of antibiotics to treat non-infected infants. Thus, rapid diagnostic test(s) that include Interleukien-6 (IL-6), neutrophil CD64 index, procalcitonin and nucleated RBC count– and differentiate the infected infants from the non-infected, particularly in the early neonatal period– have the potential to make a significant impact on neonatal care. The aim of this review is to specify the diagnostic criteria, treatment guidelines, and a summary of recent diagnostic tests of sepsis, along with the preventive measures.



Neonatal sepsis is a clinical syndrome characterized by systemic signs of infection, and accompanied by bacteremia in the first month of life (1). Sepsis is the most common cause of neonatal mortality, and is responsible for 30-50% of total neonatal deaths, each year in developing countries (2-4). The term neonatal sepsis, refers to the systemic infection of neonates including septicemia, pneumonia, meningitis, arthritis, osteomyelitis, and urinary- tract infection. As per National Neonatal-Perinatal Database (NNPD) (2002-2003), infection is the primary cause of mortality in 18.6% of intramural neonates, among which Klebsiella pneumoniae is the most frequent bacterial isolate (32.5%), followed by Staphyloco-ccus aureus (13.6%). Sepsis is 12 times more common in extramural admissions (39.7%). In extramural admissions, Klebsiella is the most prevalent bacteria responsible (27.5%), preceding S. aureus (14.9%). Sepsis is responsible for deaths in 38.0% of these extramural babies (4, 5).

  Knowledge of all the common pathogens causing septicemia, along with multidrug-resistance organism in neonates, like antibiotic-resistant nosocomial pathogens— such as vancomycin-resistant Enterococcus, carbapen-emas eproducing Enterobacteriaceae, Pseudo-monas aeruginosa, and Acinetobacter baumannii (6) — and their antimicrobial susceptibility is essential in order to select the appropriate antimicrobial treatment, along with a multidrug resistance organism.



National Neonatal Forum of India has defined neonatal sepsis as follows: (4)

  1. 1.    Probable (clinical) sepsis: It is found in an infant having a clinical picture suggestive of septicemia if any one of the following criteria are present:
  • Existence of predisposing factors: Maternal fever, foul smelling liquor, prolonged rupture of membranes (>24 hrs), or gastric polymorphs (>5 per high-power field). The septic screen would be positive due to the presence of two of the four parameters namely, TLC (< 5000/mm), band to total polymorphonuclear cells ratio of >0.2, absolute neutrophil count < 1800/cumm, C-reactive protein (CRP) >1mg/dl and micro ESR > 10 mm-first hour.
  • Radiological evidence of pneumonia.
  1. 2.    Culture Positive Sepsis: In an infant having a clinical picture suggestive of septicemia, pneumonia or meningitis, if either of the following criteria are found:
  • Isolation of pathogens from blood or CSF or urine or abscess (es)
  • Pathological evidence of sepsis in the autopsy.



Early Onset Sepsis (EOS)

Presence of foul smelling liquor or three of the aforementioned risk factors should be considered as having EOS, and be treated with antibiotics. Presence of ≥2 risk factors should be investigated with sepsis screens, and be treated, accordingly (9).


Late Onset Sepsis (LOS)

LOS usually appears after 72 hr of age. The source of infection is either nosocomial or community-acquired, and is present in neonates with septicemia, pneumonia or meningitis (10, 11). Risk factors concerning the development of LOS include:

  • NICU admission
  • Poor hygiene
  • Low birth weight (LBW)
  • Poor cord care
  • Prematurity
  • Bottle feeding
  • Invasive procedure
  • Superficial infection (pyoderma, umbilical sepsis)
  • Prelacteal feeding
  • Ventilation
  • Aspiration of feeds


Clinical features

Manifestation of neonatal sepsis is vague and ill-defined. Alteration in the established feeding behavior is an early and common symptom, though nonspecific. Other symptoms are hypothermia or fever (former is more common in LBW babies), lethargy, poor cry, poor perfusion i.e. prolonged capillary refill time (>2 seconds), hypotonic or absent neonatal reflexes, bradycar-dia or tachycardia, respiratory distress i.e. apnea or gasping respiration, hypoglycemia or hyperglycemia, and metabolic acidosis. Specific system-wise features are:


Central nervous system

These are bulging anterior fontanel, blank look, high-pitched cry, excessive irritability, coma, seizures, and neck retraction. Presence of these signs should raise clinical suspicion of meningitis.



The cardiac signs are mainly hypotension and poor perfusion. A recent study emphasized the value of early diagnosis of sepsis by analyzing heart rate characteristics by ECG monitoring. Griffin et al found that abnormal heart rate characteristics such as reduced variability, and transient decelerations occurred 24 hours prior to the onset of symptoms in sepsis and sepsis-like diseases (12). Another study found that sample asymmetry of RR intervals increased in 3-4 days prior to sepsis, with the steepest increase in the last 24 hours. These tests may prove helpful in starting the therapy long before the baby shows signs of deterioration (13).



The symptoms present in this system are feed intolerance, vomiting, diarrhea, abdominal distension paralytic ileus, and necrotising enterocolitis.



The common hepatic signs are hepatomegaly and direct hyperbilirubinemia. (Infants with the onset of jaundice after 8 days of age, or with direct hyperbilirubinemia, were more likely to have urinary tract infection)(14).



There may be acute renal failure.



Hematological signs are bleeding, petechiae, and purpura.


Skin signs

There may be multiple pustules, sclerema, mottling, umbilical redness and discharge. De Felice et al used colorimetric analysis of skin color to analyze the severity of sepsis. Color readings were taken from 10 different body sites using a portable tristimulus (15).



HRC monitoring [heart rate characteristics]

Research about cardiac electrical patterns has revealed that reduced variability and transient decelerations in heart rate may be early indicators of clinical instability, and are hypothesized to be mediated by the cholinergic anti-inflammatory pathway (16). The HRC index is a statistically-derived interpretation of the beat-to-beat variation in a patient (17). A low index indicates normal variation, but as normal variation is lost, the index rises, and so does the risk of clinical deterioration.

A recent randomized controlled trial of >3,000 very-low-birth-weight infants revealed that the use of HRC monitoring significantly decreased the 30-day mortality rate after a septic-like event, without a significant increase in antibiotic days (16). The mechanism by which mortality decreases in the monitored cohort remains unclear. In a separate study, neonates with culture-proven sepsis had a statistically higher HRC during the 24 hours leading to the septic episode, compared with the healthy control group (17). However, neonates with culture-negative, septic-like events had a statistically similar rise in HRC. Therefore, HRC is able to serve as an early warning sign of impending clinical instability. Additional research is needed to determine if it can differentiate between true sepsis and a culture-negative, septic-like event.


Blood culture

Blood culture is the gold standard for diagnosis of septicemia. It should be done in all cases before starting antibiotics. One-ml sample of blood should be adequate for a blood-culture bottle containing 5-10 ml of culture media. Blood culture should be observed for 72 hours before labeling it sterile. Although it is time-consuming, empirical antibiotics are administered during this period.


Sepsis screen (18, 19)

This is a panel of tests consisting of:


Abnormal value


< 5000/mm3


< as per Manroe chart
for term (12) and Mouzinho’s chart for very LBW (VLBW) infants (13)

Immature/total neutrophil

> 0.2


> 15mm in 1st hr


> 1mg/dl

All neonates, suspected to have sepsis, should have a septic screen to corroborate the diagnosis.

However, the decision to start antibiotics need not be conditional to the sepsis screen result if there is a strong clinical suspicion of sepsis. Sepsis screen is considered positive if two of these are positive. If the screen is negative but clinical suspicion persists, it should be repeated within 12 hours. If the screen is still negative, sepsis can be excluded with reasonable certainty. The absolute neutron-phil count varies considerably in the immediate neonatal period, and normal reference ranges are available from Manroe’s chart (20). For very-low-birth-weight infants, the reference ranges are available from Mouzinho’s charts (21). Presence of two abnormal parameters in the screen is associated with 93-100% sensitivity, 83% specifi-city, and positive and negative predictive values of 27% and 100%, respectively in detecting sepsis.

In a recently published paper, the authors have evaluated the SNAP-II score for the assessment of illness severity which consists of 6 physiological parameters, namely lowest mean arterial pressure (MAP), worst ratio of partial pressure of oxygen (PaO2) to fraction of inspired oxygen (FiO2), lowest temperature (in ºF), lowest serum pH, occurrence of multiple seizures, and urine output (<1mL/kg/hr). They found that SNAP-II can predict mortality as well as organ dysfunction in severely-septic neonates; indivi-dual components of the score do not have equal predictive abilities (22).


Lumbar puncture (LP)

The incidence of meningitis in neonatal sepsis varies from 0.3-3%, in various studies (4, 10). In EOS, lumbar puncture is indicated in the presence of a positive blood culture, or when the clinical picture is consistent with septicemia. In case of LOS, LP should be done in all infants, prior to starting antibiotics. LP should not be done in the following cases: (23)

  • Asymptomatic babies being investigated for maternal risk factors; However, LP should be performed in these cases as well, if blood culture becomes positive, subsequently.
  • Premature neonates afflicted with respiratory distress syndrome (RDS); In this case, LP should be postponed in critically-ill and haemodynami-cally-unstable babies; if traumatic, it should be repeated within 12-72 hours. The cerebrospinal fluid characteristics are unique in the neonatal period, and normal values are presented in Table 1 (24).


Urine culture

The rate of positive urine culture in infants with EOS is low. Given the low yield ofpositive

Table 1. Normal values of CSF in newborn period







9 (0-29)

7 (0-32)


57 %

61 %

Polymorphonuclear Cells

115 (65-150)

90 (20-170)

Protein (mg /dl)

50 (24-63)

52 (34-119)

Glucose (mg / dl)

74 (55-105)

81 (44-248)

CSF glucose : blood glucose

urine culture results, and costs of processing the specimens, urine culture should notbe part of the traditional sepsis evaluation in the first 72 hours of life (24). Urine cultures obtained bysuprapubic puncture or bladder catheterization have been recommended in all cases of LOS. However, neonates at risk of fungal sepsis, and very-low-birth-weight infants with poor weight gain, should have a urine examination to exclude urinary tract infection (UTI). UTI may be diagnosed in the presence of one of the following:

(a) >10 WBC/mm in a 10 ml centrifuged sample.

(b) >10 organisms /ml in urine obtained by catheterization, and

(c) any organism in urine obtained by suprapubic aspiration.



Chest X-ray is done in cases of respiratory distress or apnea. Abdominal X-ray should be done for diagnosis of necrotizing enterocolitis.


The Most Recent Diagnostic Tests of Neonatal Sepsis

Isolation of bacteria from blood is the most specific and standard method used to diagnose neonatal sepsis. The drawback of culture-based diagnosis is the 24–48 hour assay time. Newer diagnostic tests can be grouped into:

  1. Acute phase reactants
  2. Cell surface markers
  3. Granulocyte colony-stimulating factor
  4. Cytokines
  5. Molecular genetics
  6. Mol cell proteomics


Acute phase reactants

These groups of endogenous peptides are produced by the liver as part of an immediate response to the infection or tissue injury. These reactants are C-reactive protein, procalcitonin, fibronectin, haptoglobin, lactoferrin, neopterin and oromucosoid.


C-reactive protein (CRP)

CRP is synthesized within six to eight hours of exposure to an infective process or tissue damage, with a half life of 19 hours, and can increase more than 1000-fold during an acute phase response. In a study, it was concluded that CRP, IL-6 and IgM are helpful in the early diagnosis of Gram-negative neonatal sepsis, although CRP continues to be the best single test. A CRP value of 5 mg/l was the best among the three parameters with 95% sensitivity and 98% negative predictive value. The best combination was CRP ≧ 5 mg/l and/or IgM of > or = 20 mg/dl. The use of both CRP and IgM in combination was the most helpful method in predicting Gram-negative neonatal sepsis which had a significant role in making decisions regarding antibiotic treatments (25).

Another recent study was carried out in order to compare the efficiency of Serum Amyloid a (SAA) with that of C-reactive protein (CRP) and procalcitonin (PCT), in diagnosis and follow-up of neonatal sepsis in pre-term infants. The results showed that SAA is an accurate and reliable marker for the diagnosis and follow-up of neonatal sepsis. It is especially useful at the onset of inflammation for the rapid diagnosis of neonatal sepsis, and can be safely and accurately used in combination with other sepsis markers, such as CRP and PCT in diagnosis and follow-up of neonatal sepsis in pre-term infants (26).



Procalcitonin (PCT) which is produced by monocytes and hepatocytes, begins to rise four hours after exposure to bacterial endotoxin, reaches its peak after six to eight hours, and remains raised for at least 24 hours, with a half life of 25–30 hours. Both procalcitonin (2.3 ng/ml) and CRP (30 mg/l) had high specificity and positive predictive values (97%, 91% and 96%, 87%, respectively), though with low sensitivity (48% and 41%, respectively) for sepsis diagnosis. The conclusion was that procalcitonin >2.3 ng/ml or CRP > 30 mg/l indicates a high likelihood for neonatal sepsis, and antibiotic therapy should be continued even in the presence of sterile cultures. However, it is not a readily available diagnostic assay in most institutions (27).


Cell surface markers

Neutrophil CD11b and CD64 appear to be promising markers for the diagnosis of early- and late-onset infections. For culture-positive sepsis episodes, the CD64 index had the highest area under the curve (0.852) of all hematological variables, with a sensitivity of 80% , a specificity of 79%, and a cutoff value of 4.02. Therefore, neutrophil CD64 is a highly sensitive marker for neonatal sepsis. Prospective studies incorporating CD64 into a sepsis scoring system are warranted (28).

CD11b is a subunit of the β2 integrin adhesion molecule, normally expressed at a very low concentration on the surface of non-activated neutrophils. There is a 2–4 fold increase in neutrophil CD11b expression in infants with blood culture positive sepsis. The sensitivity and specificity of CD11b for diagnosing EOS are 96–100% and 81-100%, respectively. Nevertheless, in pre-term infants with RDS, significant activation of circulating phagocytes occurs within 1 to 3 hours of the onset of mechanical ventilation, independent of surfactant administration, which indicates that mechanical ventilation may be the inducer of this systemic inflammatory response. Therefore, CD11b is not a good marker for neonatal sepsis (29).


Granulocyte colony-stimulating factor

Granulocyte colony-stimulating factor (GCSF), a mediator produced by bone marrow, facilitates proliferation and differentiation of neutrophils, and has been proposed to be a reliable infection marker for early diagnosis of neonatal sepsis. A concentration ≥ 200 pg/ml has a high sensitivity (95%), and negative predictive value (99%) for predicting early onset neonatal bacterial and fungal infections (30).



As antigen-specific immunity develops later in infant’s life, e.g. at 2 years of age in case of encapsulated bacteria, neonates initially depend on their natural (innate) immunity. This includes phagocytes (by monocytes, tissue macrophages, and neutrophils), natural killer cells, and humoral mediators (CRP, complements, and transp-lacentally- acquired maternal antibodies). In response to antigens such as bacterial endotoxins, activated tissue macrophages produce tumor necrosis factors (TNF) and interleukins (IL). These proinflammatory cytokines stimulate endothelial cells to express receptors for intercellular adh-esion molecules on white blood cells. This initiates the cytokine cascade towards the increased production of IL6, IL8, and chemokines. Newborn infants display a higher percentage of IL6 and IL8 positive cells than adults do. There is a sharp rise in IL6 concentration on exposure to bacterial products, which precedes the increase in CRP. Umbilical cord blood IL6 has consistently been shown to be a sensitive marker for diagnosing early-onset neonatal sepsis at the onset of infection, compared with other biochemical markers, including CRP, IL1ß, TNFα, and Eselectin, although sensitivity is reduced at 24 and 48 hours, since IL6 concentrations fall rapidly and become undetectable after 24 hours. The measurement of IL6 (early and sensitive) along with CRP (late and specific) in the first 48 hours of presumed septic episodes, improves the sensitivity compared with either of them alone (31).

IL-6 levels may be useful in the initiation, as well as early termination of antibiotic therapy in late-onset neonatal sepsis (32). IL8 is a proinflam-matory cytokine that is predominantly produced by monocytes, macrophages, and endothelial cells, with similar kinetics to IL6. IL8 is considered to be a highly accurate marker with its sensitivity ranging from 80% to 91%, and specificity from 76% to 100%.

TNF-α is a proinflammatory cytokine that stimulates IL6 production and has a broad spectrum of biological actions on several types of target cells, both immune and non-immune. Newborns developing early-onset infections are born with higher TNF-α concentrations than non-infected infants. Other markers studied over the last few years include adhesion molecules (intercellular adhesion molecule-1, vascular cell adhesion molecule-1, Eselectin, L-selectin), complement activation products (C3a-desArg, C3bBbP, SC5b-9), and IL-1alpha, IL-1beta, and IL-receptor antagonist (IL1RA), which have been found to significantly increase during sepsis, though these findings require further evaluation for clinical application in the diagnosis of newborns’ infections. It has been demonstrated that median IL-6 and TNF-α levels were significantly higher in groups of patients with a diagnosis of clinical sepsis than in the controls. The optimal cutoff point was 32 pg /ml for IL-6 and 12 pg /ml for TNF-α. The combination of TNF-α and IL-6 provided a sensitivity of 98.5%, and it is a highly sensitive marker of sepsis in the immediate postnatal period (33).

In a recent study, it was demonstrated that the cytokines released in sepsis have an important role in stimulating nucleated RBC (NRBC) production, independent of hypoxia. In this study, significantly elevated NRBC was observed in EOS infants (no Early Onset Neonatal Sepsis (EONS) (n=49)) 1330 cells/ cmm (665-2630), EONS (n=19)3020 cells/cmm (1388-4558), p=0.011), along with significantly elevated IL-6 in EONS; although no increase in the level of umbilical cortisol or erythropoietin was noticed. Increased NRBC count immediately after the birth could be an interesting marker of EONS in the absence of hypoxia and it awaits further evaluation (34).


Molecular genetics (35-38)

Polymerase chain reaction (PCR) analysis relies on the fact that the bacteria specific 16S rRNA gene is highly cons