INTRODUCTION — Sepsis is an important cause of morbidity and mortality among newborn infants. Although the incidence of sepsis in term and late preterm neonates is low, the potential for serious adverse outcomes is of such great consequence that caregivers should have a low threshold for evaluation and treatment for possible sepsis in neonates.
The epidemiology, clinical features, diagnosis, and evaluation of sepsis in term and late preterm neonates will be reviewed here. The management and outcome of sepsis in term and late preterm neonates are discussed separately. (See "Management and outcome of sepsis in term and late preterm neonates".)
Other related topics include:
●Sepsis in preterm neonates (see "Clinical features and diagnosis of bacterial sepsis in preterm infants <34 weeks gestation" and "Treatment and prevention of bacterial sepsis in preterm infants <34 weeks gestation")
●Group B streptococcal (GBS) infection in neonates and infants, including management of well-appearing newborns at risk for GBS (see "Group B streptococcal infection in neonates and young infants" and "Management of neonates at risk for early-onset group B streptococcal infection")
●Outpatient evaluation and management of febrile neonates (see "The febrile neonate (28 days of age or younger): Outpatient evaluation and initial management")
●Evaluation and management of ill-appearing infants (see "Approach to the ill-appearing infant (younger than 90 days of age)")
TERMINOLOGY — The following terms will be used throughout this discussion on neonatal sepsis:
●Term neonates are those born at a gestational age of 37 weeks or greater.
●Late preterm neonates (also called near-term neonates) are those born from 34 through 36 completed weeks of gestation [1]. (See "Late preterm infants".)
●Neonatal sepsis is a clinical syndrome in an infant 28 days of life or younger, manifested by systemic signs of infection and isolation of a bacterial pathogen from the bloodstream. A consensus definition for neonatal sepsis is lacking [2]. (See 'Diagnosis' below.)
Sepsis is classified according to the neonate's age at the onset of symptoms.
•Early-onset sepsis is defined as the onset of symptoms before 72 hours of age [3-5], although some experts extend the definition to infections occurring within the first seven days after birth.
•Late-onset sepsis is generally defined as the onset of symptoms at ≥72 hours of age [3]. Similar to early-onset sepsis, there is variability in its definition, ranging from an onset at >72 hours of life to ≥7 days of age.
Newborn infants with early-onset sepsis typically present with symptoms during their birth hospitalization. Term neonates with late-onset sepsis generally present to the outpatient setting or emergency department unless comorbid conditions have prolonged the birth hospitalization. The approach to evaluating febrile neonates in the outpatient setting is discussed separately. (See "The febrile neonate (28 days of age or younger): Outpatient evaluation and initial management", section on 'Neonates 8 to 21 days old'.)
PATHOGENESIS — The pathogenesis differs based on the timing of onset:
●Early-onset infection – Early-onset infection is usually due to vertical transmission by ascending contaminated amniotic fluid or during vaginal delivery from bacteria in the mother's lower genital tract [4]. Maternal chorioamnionitis (or intra-amniotic infection) is a well-recognized risk factor for early-onset neonatal sepsis [6,7]. Maternal group B streptococcal (GBS) colonization is another important risk factor. (See 'Maternal risk factors' below and "Group B streptococcal infection in neonates and young infants", section on 'Risk factors'.)
Use of forceps during delivery and electrodes placed for intrauterine monitoring have been implicated in the pathogenesis of early-onset sepsis because they penetrate the neonatal defensive epithelial barriers [8].
●Late-onset infection – Late-onset infections can be acquired by the following mechanisms:
•Vertical transmission, resulting in initial neonatal colonization that evolves into later infection
•Horizontal transmission from contact with care providers or environmental sources
Disruption of the intact skin or mucosa, which can be due to invasive procedures (eg, intravascular catheter), increases the risk of late-onset infection. Late-onset sepsis is uncommonly associated with maternal obstetrical complications.
Metabolic factors including hypoxia, acidosis, hypothermia, and inherited metabolic disorders (eg, galactosemia) are likely to contribute to risk for and severity of neonatal sepsis (including both early- and late-onset). These factors are thought to disrupt the neonate's host defenses (ie, immunologic response) [8].
EPIDEMIOLOGY — The overall incidence of neonatal sepsis ranges from 1 to 5 cases per 1000 live births [9-11]. Estimated incidence rates vary based on the case definition and population studied. In a systematic review and meta-analysis of population-based studies from around the world, the estimated pooled incidence of neonatal sepsis was 22 per 1000 live births, with an associated mortality rate of 11 to 19 percent [12]. This translates to a global incidence of three million cases of neonatal sepsis per year [12]. Globally, neonatal sepsis and other severe infections were responsible for an estimated 430,000 neonatal deaths in 2013, accounting for approximately 15 percent of all neonatal deaths [13].
Rates of neonatal sepsis increase with decreasing gestational age:
●Term neonates (≥37 weeks gestational age) – The estimated incidence of sepsis (both early- and late-onset) in term neonates is 1 to 2 cases per 1000 live births [9,10]. In a prospective national surveillance study (2006 to 2009), the incidence of early-onset sepsis (defined as positive blood or cerebrospinal fluid cultures) was 0.98 cases per 1000 live births; the rate among neonates with birth weight >2500 g was 0.57 per 1000 [11].
●Late preterm neonates (34 through 36 weeks gestational age) – The incidence is higher in late preterm than term neonates. In an observational cohort study (1996 to 2007), the reported incidences of early- and late-onset sepsis (defined as positive blood culture) in late preterm neonates were 4.4 and 6.3 per 1000, respectively [14].
●Preterm neonates <34 weeks gestational age – The incidence of sepsis in preterm neonates is discussed separately. (See "Clinical features and diagnosis of bacterial sepsis in preterm infants <34 weeks gestation", section on 'Incidence'.)
The incidence of early-onset sepsis in the United States has decreased, primarily due to reduction in group B streptococcal (GBS) infections, owing to the use of intrapartum antibiotic prophylaxis (IAP) [15-19]. Early-onset GBS infection rates in the United States reported through the Centers for Disease Control and Prevention's Active Bacterial Core Surveillance Report have declined from 0.6 per 1000 live births in 2000 to 0.25 per 1000 live births in 2018 [20,21]. Late-onset GBS infection rates have remained relatively stable in the same interval (0.4 per 1000 live births in 2000 and 0.28 per 1000 live births in 2018). (See "Group B streptococcal infection in neonates and young infants", section on 'Epidemiology'.)
ETIOLOGIC AGENTS — Group B Streptococcus (GBS) and Escherichia coli are the most common causes of both early- and late-onset sepsis, accounting for approximately two-thirds of early-onset infections [16,22-24].
Other bacterial agents associated with neonatal sepsis include (table 1):
●Listeria monocytogenes, although a well-recognized cause of early-onset sepsis, only accounts for rare sporadic cases of neonatal sepsis and is more commonly seen during an outbreak of listeriosis [25,26].
●Staphylococcus aureus, including community-acquired methicillin-resistant S. aureus, is a potential pathogen in neonatal sepsis [27]. Bacteremic staphylococcal infections in term neonates usually occur in association with skin, bone, or joint sites of involvement.
●Enterococcus, a commonly encountered pathogen among preterm neonates, is a rare cause of sepsis in otherwise healthy term newborn infants.
●Other gram-negative bacteria (including Klebsiella, Enterobacter, and Citrobacter spp) and Pseudomonas aeruginosa are associated with late-onset infection, especially in neonates admitted to neonatal intensive care units [28].
●Coagulase-negative staphylococci often are a cause of hospital-associated infection in ill neonates (primarily in premature neonates and/or neonates who have indwelling intravascular catheters). Coagulase-negative staphylococci may be considered a contaminant in otherwise healthy term neonates who have not undergone invasive procedures.
The patterns of pathogens associated with neonatal sepsis have changed over time, as reflected by longitudinal databases from single tertiary centers [16,17]. The incidence of early-onset GBS has declined by 80 percent in the United States with the use of intrapartum antibiotic prophylaxis (IAP). GBS prevention efforts have not led to an increasing burden of early-onset E. coli infection [23]. (See "Prevention of early-onset group B streptococcal disease in neonates".)
Common nonbacterial agents associated with neonatal sepsis include (see 'Differential diagnosis' below):
●Herpes simplex virus (see "Neonatal herpes simplex virus infection: Clinical features and diagnosis")
●Enterovirus and parechovirus (see "Enterovirus and parechovirus infections: Clinical features, laboratory diagnosis, treatment, and prevention", section on 'Neonates' and "Nosocomial viral infections in the neonatal intensive care unit", section on 'Sepsis-like illness and meningitis/encephalitis')
●Candida (see "Clinical manifestations and diagnosis of Candida infection in neonates", section on 'Invasive infection')
MATERNAL RISK FACTORS — Maternal factors that are associated with an increased risk of early-onset sepsis in the neonate, particularly group B Streptococcus (GBS) infection, include chorioamnionitis (or intra-amniotic infection), intrapartum maternal fever, maternal GBS colonization, preterm birth, and prolonged rupture of the membranes [4,6,29]. It is unclear whether prolonged rupture of membranes is an independent risk factor for sepsis in the absence of intra-amniotic infection [30]. The approach to identifying pregnancies at risk for neonatal early-onset sepsis and indications for maternal intrapartum antibiotic prophylaxis (IAP) are discussed separately. (See "Prevention of early-onset group B streptococcal disease in neonates", section on 'Identification of pregnancies at increased risk for early-onset neonatal GBS' and "Prevention of early-onset group B streptococcal disease in neonates", section on 'Intrapartum antibiotic prophylaxis'.)
Maternal GBS screening and IAP reduce the risk of GBS infection but do not eliminate it. In a prospective national surveillance study that included 117 term neonates with early-onset GBS infection, 66 percent were born to mothers who did not receive IAP, because maternal GBS screening culture was either not performed (29 percent of mothers) or was negative (51 percent of mothers) [11].
CLINICAL MANIFESTATIONS — Clinical manifestations range from subtle symptoms to profound septic shock. Signs and symptoms of sepsis are nonspecific and include temperature instability (hypothermia or fever), irritability, lethargy, respiratory symptoms (eg, tachypnea, grunting, hypoxia), poor feeding, tachycardia, poor perfusion, and hypotension (table 2) [8].
Because the signs and symptoms of sepsis can be subtle and nonspecific, it is important to identify neonates with risk factors for sepsis and to have a high index of suspicion for sepsis when a newborn deviates from their usual pattern of activity or feeding [8].
Signs and symptoms of neonatal sepsis include:
●Fetal and delivery room distress – The following signs of fetal and neonatal distress during labor and delivery may be early indicators of neonatal sepsis:
•Intrapartum fetal tachycardia, which may be due to intra-amniotic infection (see "Fetal arrhythmias", section on 'Tachyarrhythmias')
•Meconium-stained amniotic fluid, which is associated with a twofold increased risk of sepsis [6] (see "Meconium aspiration syndrome: Pathophysiology, clinical manifestations, and diagnosis", section on 'Meconium composition and passage')
●Temperature instability – The temperature of an infected neonate can be elevated, depressed, or normal. Term neonates with sepsis are more likely to be febrile than preterm neonates who are more likely to be hypothermic [8]. Temperature elevation in full-term neonates is concerning and, if persistent, is highly indicative of infection [31,32].
●Cardiorespiratory symptoms – Respiratory and cardiocirculatory symptoms are common in infected neonates. Approximately 85 percent of newborns with early-onset sepsis present with respiratory distress (eg, tachypnea, grunting, flaring, use of accessory muscles) [11]. Apnea is less common, occurring in 38 percent of cases, and is more likely in preterm than term neonates. Apnea is a classic presenting symptom in late-onset group B streptococcal (GBS) sepsis. Early-onset disease can be associated with persistent pulmonary hypertension of the newborn. (See "Persistent pulmonary hypertension of the newborn (PPHN): Clinical features and diagnosis".)
Tachycardia is a common finding in neonatal sepsis but is nonspecific. Bradycardia may also occur. Poor perfusion and hypotension are more sensitive indicators of sepsis, but these tend to be late findings. In a prospective national surveillance study, 40 percent of neonates with sepsis required volume expansion, and 29 percent required vasopressor support [11].
●Neurologic symptoms – Neurologic manifestations of sepsis in the neonate include lethargy, poor tone, poor feeding, irritability, and seizures [8]. Seizures are an uncommon presentation of neonatal sepsis but are associated with a high likelihood of infection. In a prospective study in a single neonatal unit, 38 percent of neonates with seizures had sepsis as the etiology [33]. Seizures are a presenting feature in 20 to 50 percent of neonates with neonatal meningitis [34]. (See "Bacterial meningitis in the neonate: Clinical features and diagnosis" and "Etiology and prognosis of neonatal seizures".)
●Other findings – Other findings associated with neonatal sepsis, and their approximate frequencies are listed below (table 2) [8,11]:
•Jaundice – 35 percent
•Hepatomegaly – 33 percent
•Poor feeding – 28 percent
•Vomiting – 25 percent
•Abdominal distension – 17 percent
•Diarrhea – 11 percent
EVALUATION AND INITIAL MANAGEMENT — Neonates with signs and symptoms of sepsis (table 2) require prompt evaluation and initiation of antibiotic therapy [4,8]. Because the signs and symptoms of sepsis are subtle and nonspecific, the threshold for performing laboratory testing is low. Our approach is generally consistent with guidelines published by the American Academy of Pediatrics [4,35].
Early-onset sepsis — The evaluation of a neonate with suspected early-onset sepsis (onset within first 72 hours after birth) includes all of the following:
●Review of the pregnancy, labor, and delivery, including risk factors for sepsis and the use and duration of maternal intrapartum antibiotic prophylaxis (IAP) (see 'Maternal risk factors' above)
●Comprehensive physical examination (see "Assessment of the newborn infant")
●Laboratory testing (see 'Laboratory tests' below)
The extent of the diagnostic evaluation for sepsis is directed by the neonate's symptoms and maternal risk factors.
Early-onset sepsis calculator — Multivariate predictive models for risk of early-onset sepsis have been developed and validated in clinical use, including the early-onset sepsis calculator [29,36-38]. The early-onset sepsis calculator is a web-based tool that can be used to estimate the risk of early-onset sepsis in individual patients based on risk factors (eg, newborn clinical condition, highest intrapartum maternal temperature, maternal group B Streptococcus [GBS] status, administration of maternal IAP, gestational age, duration of rupture of membranes). The calculator requires the user to input the local incidence of early-onset sepsis. If the local incidence of early-onset sepsis is unknown, the users should enter "0.5 per 1000." The calculator provides guidance on the diagnostic evaluation and empiric antibiotic treatment. The threshold used to trigger evaluation and empiric treatment varies, depending on the clinical circumstances. The early-onset sepsis calculator is not valid for preterm neonates (<34 weeks gestation) and does not apply to late-onset sepsis.
Symptomatic neonates — Neonates with signs and symptoms of early-onset sepsis (table 2) should undergo a full diagnostic evaluation and should receive empiric antibiotic treatment. (See 'Clinical manifestations' above and 'Empiric antibiotic therapy' below.)
A full diagnostic evaluation includes (see 'Laboratory tests' below):
●Blood culture. (See 'Blood culture' below.)
●Inflammatory markers (eg, complete blood count [CBC] with differential, C-reactive protein [CRP], and/or procalcitonin [PCT]). These tests are not required, but they may be helpful in determining length of therapy if followed serially. (See 'Complete blood count' below and 'Other inflammatory markers' below.)
●Lumbar puncture (if the neonate is clinically stable enough to tolerate the procedure and it will not delay initiation of antibiotic therapy). (See 'Lumbar puncture' below.)
●Chest radiograph (if respiratory symptoms are present).
Well-appearing neonates — Well-appearing newborns with identified risk factors for neonatal sepsis, particularly GBS, should be observed for 36 to 48 hours. Based on the nature of the risk factor(s) and the use and duration of maternal IAP, they may require a limited diagnostic evaluation (ie, blood culture). Accepted approaches to determining the need for laboratory evaluation and empiric antibiotics include categorical risk assessment (algorithm 1), clinical observation (algorithm 2), and the early-onset sepsis calculator. The early-onset sepsis calculator is discussed above (see 'Early-onset sepsis calculator' above). The other two approaches are discussed in detail separately. (See "Management of neonates at risk for early-onset group B streptococcal infection", section on 'Approaches to risk assessment'.)
Late-onset sepsis — Neonates presenting with signs and symptoms at ≥72 hours of age should undergo prompt evaluation and empiric antibiotic treatment. (See "Management and outcome of sepsis in term and late preterm neonates", section on 'Late-onset sepsis'.)
A full diagnostic evaluation should be performed, including all of the following:
●Blood culture (see 'Blood culture' below).
●Inflammatory markers (eg, CBC, CRP, and/or PCT). These tests are not required, but they may be helpful in determining length of therapy if followed serially. (See 'Complete blood count' below and 'Other inflammatory markers' below.)
●Lumbar puncture (if the neonate is clinically stable enough to tolerate the procedure and it will not delay initiation of antibiotic therapy). (See 'Lumbar puncture' below.)
●Chest radiograph (if respiratory symptoms are present).
●Urine culture. (See 'Urine culture' below.)
●Cultures from any other potential foci of infection (eg, tracheal aspirates if intubated, purulent eye drainage, or pustules). (See 'Other cultures' below.)
Neonates with late-onset sepsis generally present to the outpatient or emergency department setting unless comorbid conditions have prolonged the birth hospitalization. Additional details regarding the evaluation of febrile or ill-appearing neonates are provided separately. (See "Approach to the ill-appearing infant (younger than 90 days of age)" and "The febrile neonate (28 days of age or younger): Outpatient evaluation and initial management", section on 'Neonates 8 to 21 days old'.)
Empiric antibiotic therapy — Indications for empiric antibiotic therapy (after obtaining cultures) include:
●Ill appearance (see "Approach to the ill-appearing infant (younger than 90 days of age)")
●Concerning symptoms, including temperature instability or respiratory, cardiocirculatory, or neurologic symptoms (see 'Clinical manifestations' above)
●Cerebrospinal fluid pleocytosis (white blood cell [WBC] count of >20 to 30 cells/microL) (table 3) (see "Bacterial meningitis in the neonate: Clinical features and diagnosis", section on 'Interpretation of cerebrospinal fluid')
●High estimated sepsis risk based on a validated risk calculator (see 'Early-onset sepsis calculator' above)
The empiric antibiotic regimen should include agents active against GBS and other organisms that most commonly cause neonatal sepsis (eg, E. coli and other gram-negative pathogens) (table 1). Our suggested empiric regimens are summarized in the table (table 4) and discussed in greater detail separately. (See "Management and outcome of sepsis in term and late preterm neonates", section on 'Initial empiric therapy'.)
LABORATORY TESTS — The goals of the diagnostic evaluation are to identify and treat all neonates with bacterial sepsis and minimize the treatment of patients who are not infected. Laboratory assessment includes cultures of body fluids that confirm the presence or absence of a bacterial pathogen and other studies that are used to evaluate the likelihood of infection.
Blood culture — A definitive diagnosis of neonatal sepsis is established by a positive blood culture. The sensitivity of a single blood culture (with ≥1 mL of blood) to detect neonatal bacteremia approaches 99 percent [39].
●Blood sampling – The following considerations are important when obtaining a blood culture:
•Sampling site – Blood cultures can be obtained by venipuncture or arterial puncture or by sampling from a newly inserted umbilical artery or vascular access catheter. Positive culture results of blood drawn from indwelling umbilical or central venous catheters can be difficult to interpret since they can indicate contamination or catheter colonization rather than a true systemic infection. In such circumstances, obtaining a peripheral blood culture may help determine if there is a true bloodstream infection [8].
•Number of cultures – We obtain at least one culture prior to initiating empiric antibiotic therapy in neonates with a high clinical suspicion for sepsis, although other institutions may routinely obtain two blood cultures.
•Volume of blood – The optimal volume of blood is based on the weight of the neonate. A minimum blood volume of 1 mL is desirable for optimal detection of bacteremia when a single blood culture bottle is used [4]. At the author's institution, the suggested optimal volume is 2 mL for neonates weighing ≤3 kg and 3 mL for those who weigh >3 to 5 kg. Dividing this volume into two aliquots to inoculate an anaerobic as well as the aerobic culture bottle may optimize recovery of rare strict anaerobic species, and most neonatal pathogens grow under anaerobic conditions.
●Time to positivity – Automated systems for continuous monitoring of blood cultures are routinely used in the United States and have shortened the time to identify positive blood cultures. In most cases of neonatal sepsis, blood cultures become positive within 24 to 36 hours [40].
●Distinguishing infection from contamination – A positive blood culture is diagnostic of sepsis when a known bacterial pathogen is isolated (table 1). Isolation of skin flora (eg, diphtheroids) suggests contamination rather than infection. Contamination is also suggested when multiple species grow in culture. Coagulase-negative staphylococci may be pathogenic in patients with indwelling vascular catheters or other invasive devices, whereas a single blood culture positive for coagulase-negative staphylococci is likely to represent a contaminant in full-term neonates without these risk factors [8].
Lumbar puncture — The clinical presentation of neonatal meningitis is indistinguishable from that of neonatal sepsis without meningitis. Specific clinical signs of meningitis (eg, bulging fontanelle, nuchal rigidity) are often lacking in neonates. For this reason, we suggest performing a lumbar puncture as part of the diagnostic evaluation in symptomatic neonates. Lumbar puncture should ideally be performed before the initiation of antibiotics for neonates in whom the clinical suspicion for sepsis is high, particularly those with critical illness. However, lumbar puncture should be deferred if the procedure would compromise the neonate's clinical condition or if the procedure would delay initiation of antibiotics [4].
Lumbar puncture is indicated in neonates with:
●Clinical findings concerning for sepsis (table 2)
●Positive blood culture
●Worsening clinical status while on antibiotic therapy
Cerebrospinal fluid should be sent for Gram stain, routine culture, cell count with differential, and protein and glucose concentrations. The interpretation of cerebrospinal fluid needs to account for variations due to gestational age, chronologic age, and birth weight (table 3).
The clinical features and diagnosis of neonatal bacterial meningitis are discussed separately. (See "Bacterial meningitis in the neonate: Clinical features and diagnosis".)
Urine culture — Urine culture obtained by catheter or bladder tap should be included in the evaluation for late-onset sepsis (ie, onset at ≥72 hours after birth). A urine culture need not be routinely performed in the evaluation of suspected early-onset sepsis (onset within the first 72 hours after birth), because a positive urine culture in this setting is a reflection of high-grade bacteremia rather than an isolated urinary tract infection [4]. (See "Urinary tract infections in neonates".)
Other cultures — In neonates with late-onset infection, cultures should be obtained from any other potential foci of infection (eg, purulent eye drainage or pustules).
Tracheal aspirates can be of value if obtained immediately after first intubation. However, the trachea and proximal bronchi are not sterile, so a positive tracheal culture may represent colonization rather than infection, particularly in a neonate who has been intubated for several days.
Gram stain or culture of other sites (eg, gastric aspirate, body surface sites [eg, axilla, groin, and external ear canal]) add little to the evaluation and should not be performed [4].
Complete blood count — A complete blood count (CBC) may be used to evaluate the likelihood of sepsis in a neonate with risk factors or signs of infection. However, the CBC cannot establish or exclude the diagnosis of sepsis since it has poor sensitivity and specificity. At the author's institution, CBCs are no longer routinely performed in the evaluation of suspected early- or late-onset neonatal sepsis. Other centers may routinely perform CBCs in this setting.
●Early-onset sepsis – For most neonates with suspected early-onset sepsis, we suggest not obtaining a CBC as part of the diagnostic evaluation.
The CBC does not perform well in predicting risk of infection in neonates. If a CBC is obtained, it should not be used as the sole determinant of whether to treat empirically with antibiotics [41-45]. The CBC can be used in combination with risk factors, clinical assessment, and/or other tests. However, it adds little to these other assessments. CBCs obtained 6 to 12 hours after delivery are more predictive of sepsis than those obtained immediately after birth because the white blood cell count (WBC) and absolute neutrophil count (ANC) normally increase during the first six hours after birth [41,42,46,47].
Two large multicenter studies have evaluated the diagnostic value of CBCs in early-onset neonatal sepsis [41,42]. These studies found that low WBC (<5000/microL), absolute neutropenia (ANC <1000 neutrophils/microL), relative neutropenia (ANC <5000 neutrophils/microL), and elevated ratio of immature to total neutrophil counts (I/T ratio) were associated with culture-proven sepsis. However, none of the tests were sufficiently sensitive to reliably predict neonatal sepsis.
•I/T ratio – An elevated I/T ratio (≥0.2) has the best sensitivity (approximately 60 percent) of the neutrophil indices for predicting neonatal sepsis [6,48]. A normal I/T ratio has been used to "rule out" sepsis; however, an elevated value is not highly predictive of sepsis and may be observed in 25 to 50 percent of uninfected neonates (ie, the test has poor specificity [<10 percent]) [6,48,49].
The I/T ratio is limited by the wide range of normal values, which reduces its positive predictive value, especially in asymptomatic patients [50]. Inter-reader differences in band neutrophil identification with manual differential counts is another limitation. In addition, exhaustion of the bone marrow reserves, which may occur during critical illness, will result in low band counts and lead to falsely low ratios. (See "Evaluating diagnostic tests", section on 'How well does the test perform in specific populations?'.)
•ANC – Although both elevated and low neutrophil counts can be associated with neonatal sepsis, neutropenia has greater specificity since few conditions other than sepsis and preeclampsia depress the neutrophil count in neonates.
Neutrophil counts vary with gestational age (counts decrease with decreasing gestational age), type of delivery (counts are lower in neonates born by cesarean delivery), site of sampling (counts are lower in arterial than in venous samples), altitude (counts are higher at elevated altitudes), and timing after delivery (counts increase during the first six hours of life) [41].
The lower limit of a normal neutrophil count for neonates >36 weeks of gestation is 3500/microL at birth and 7500/microL six to eight hours after delivery (figure 1) [51]. For neonates born at 28 through 36 weeks of gestation, the lower limits of normal neutrophil counts at birth and at six to eight hours after birth are 1000/microL and 1500/microL, respectively (figure 2).
●Late-onset sepsis – CBCs are frequently used to support the diagnosis of late-onset sepsis. In this setting, CBCs are less variable than in the first days of life. However, WBC indices still perform poorly in identifying neonates with late-onset sepsis.
In a study of 37,826 neonates (mostly neonates continuously hospitalized from birth) who underwent evaluation for late-onset (defined as 4 to 120 days) sepsis with blood culture and CBC, abnormal WBC (<1000 or >50,000/microL), high ANC (>17,670/microL), elevated I/T ratio (≥0.2), and low platelet count (<50,000/microL) were associated with culture positivity [52]. However, sensitivity was inadequate to reliably predict late-onset sepsis.
Other inflammatory markers — A number of acute phase reactants have been used to identify infected newborns. Many of these tests have acceptable sensitivity; however, they all lack specificity, resulting in poor positive predictive value [53]. No single biomarker or panel of screening tests is sufficiently sensitive to reliably detect neonatal sepsis [54].
●C-reactive protein (CRP) – CRP is increased in inflammatory conditions, including sepsis. A variety of noninfectious inflammatory conditions can also cause elevated CRP, including maternal fever, fetal distress, stressful delivery, perinatal asphyxia, meconium aspiration, and intraventricular hemorrhage [55].
A single measurement of CRP is not a useful aid in the diagnosis of neonatal sepsis, because it lacks sufficient sensitivity and specificity [56]. However, sequential assessment of CRP values may help support a diagnosis of sepsis. If the CRP level remains persistently normal (<1 mg/dL [10 mg/L]), bacterial sepsis is unlikely [4].
CRP levels can be helpful in guiding the duration of antibiotic therapy in suspected neonatal bacterial infection. Neonates with elevated CRP levels that decrease to <1 mg/dL (10 mg/L) 24 to 48 hours after initiation of antibiotic therapy typically are not infected and generally do not require further antibiotic treatment if cultures are negative. However, routine use of serial CRP measurements can be associated with longer length of hospital stay [57].
An elevated CRP level alone does not justify continuation of empiric antibiotics for more than 36 to 48 hours in well-appearing neonates with negative culture results [58]. Additional evaluation may be warranted to investigate alternative explanations for persistently elevated CRP values.
●Procalcitonin (PCT) – PCT is the peptide precursor of calcitonin. It is released by parenchymal cells in response to bacterial toxins, leading to elevated serum levels in patients with bacterial infections. Several observational studies have suggested that PCT may be a useful marker to identify neonates who are infected [59-61]. In a 2015 systematic review of 18 studies, the sensitivity of PCT for detection of neonatal sepsis ranged from 72 to 79 percent, and the specificity ranged from 72 to 90 percent [61]. Thus, although PCT is a promising marker, it does not appear to be reliable as the sole or main diagnostic indicator for neonatal sepsis.
PCT may have utility in guiding the duration of antibiotic therapy in neonates with suspected sepsis. In a multicenter randomized trial involving 1710 neonates with suspected early-onset sepsis, neonates assigned to a treatment protocol in which duration of antibiotic therapy was guided by serial PCT measurements in addition to standard criteria (ie, clinical examination, blood culture results, WBC, and CRP) had shorter duration of antibiotic therapy compared with neonates managed according to standard criteria without PCT measurements (55.1 versus 65 hours, respectively) [62]. The rate of reinfection was low in both groups (0.7 percent for the PCT group and 0.6 percent for the control group). However, the study protocol's suggested duration of antibiotic therapy was overruled by the treating clinician in 45 percent of neonates in the PCT group and 41 percent of those in the control group. The findings reinforce the concept that when serial PCT levels are obtained, they should be used in conjunction with other clinical indicators of sepsis and should not be the sole basis of decision-making.
●Cytokines, chemokines, and other biomarkers – Both proinflammatory cytokines, such as interleukin-6 and tumor necrosis factor-alpha, and antiinflammatory cytokines (interleukin-4 and interleukin-10) are increased in infected neonates compared with those without infections [63-66]. Elevations of serum amyloid A and the cell surface antigen CD64 also have high sensitivity for identifying neonates with sepsis [61,67]. However, these biomarkers are generally not used in clinical practice.
Further research aimed at better understanding the neonatal inflammatory response to sepsis may result in the identification of sensitive and specific markers of inflammation or the development of pathogen-specific rapid diagnostic tests for early detection of neonatal sepsis. With a sensitive and specific marker for systemic bacterial infection, the management of neonatal sepsis would be significantly altered so that antimicrobial therapy could be safely withheld in neonates for whom sepsis is unlikely.
DIAGNOSIS — The diagnosis of sepsis is based upon isolating a pathogenic organism in culture. Few neonates who undergo sepsis evaluation are ultimately diagnosed with sepsis.
This was demonstrated in a retrospective study of 2785 newborns who underwent evaluation for sepsis based on clinical symptoms (54 percent) or risk factors (46 percent) [6]. Culture-proven sepsis was identified in 22 neonates (0.8 percent) and culture-negative clinical sepsis ("probable sepsis") was diagnosed in 40 neonates (1.4 percent).
Culture-proven sepsis — The isolation of pathogenic bacteria from a blood culture is the gold standard to confirm the diagnosis of neonatal sepsis. A positive blood culture is diagnostic of sepsis when a pathogenic organism is isolated (table 1). Isolation of skin flora (eg, diphtheroids and coagulase-negative staphylococci) in culture suggests contamination rather than infection, although coagulase-negative staphylococci can be pathogenic in patients with indwelling vascular catheters or other devices [8]. (See 'Blood culture' above and "Management and outcome of sepsis in term and late preterm neonates", section on 'Culture-proven sepsis'.)
Probable sepsis — In some cases, a pathogen may not be isolated in culture, yet the neonate has a clinical course that is concerning for sepsis (eg, ongoing temperature instability; ongoing respiratory, cardiocirculatory, or neurologic symptoms not explained by other conditions; or laboratory abnormalities suggestive of sepsis [cerebrospinal fluid pleocytosis, persistently elevated C-reactive protein]).
A composite of observational assessment and serial laboratory testing is typically used to make a diagnosis of probable sepsis [68]. The criteria used are usually broad in an attempt to ensure that all infected neonates are identified, but at the cost of testing and treating a number of uninfected neonates. There is no consensus definition for the diagnosis of probable sepsis in neonates [2].
Alternative diagnoses (table 5) should also be entertained when a neonate with suspected sepsis has negative cultures. (See "Management and outcome of sepsis in term and late preterm neonates", section on 'Probable but unproven sepsis' and 'Differential diagnosis' below.)
Infection unlikely — Neonates with mild and/or transient symptoms (ie, fever alone or other symptoms that quickly resolve) who remain well-appearing with negative cultures at 36 to 48 hours are unlikely to have sepsis. Empiric antibiotic therapy should be discontinued after 36 to 48 hours in these neonates [4,69].
DIFFERENTIAL DIAGNOSIS — The differential diagnosis of neonatal sepsis includes systemic viral, fungal, and parasitic infections and noninfectious causes of temperature instability and respiratory, cardiocirculatory, and neurologic symptoms (table 5). The clinical history, disease course, and laboratory findings may help to distinguish neonatal sepsis from other infectious and noninfectious disorders. Ultimately, appropriate microbiologic testing is required to confirm the diagnosis.
It is often difficult to differentiate neonatal sepsis from other conditions. However, given the morbidity and mortality of neonatal sepsis, empiric antibiotic therapy should be provided (after obtaining cultures) to infants with suspected sepsis pending definitive culture-based diagnosis.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sepsis in neonates" and "Society guideline links: Group B streptococcal infection in pregnant women and neonates".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topics (see "Patient education: Sepsis in newborn babies (The Basics)")
SUMMARY AND RECOMMENDATIONS
●Epidemiology and microbiology – Sepsis is an important cause of morbidity and mortality in newborn infants.
•Incidence – Although the incidence of sepsis in term and late preterm infants is low (approximately 1 to 6 cases per 1000 births), the potential for serious adverse outcomes, including death, is of such great consequence that caregivers should have a low threshold for evaluation and treatment for possible sepsis. (See 'Epidemiology' above.)
•Common pathogens – Group B Streptococcus (GBS) and Escherichia coli are the most common bacteria causing neonatal sepsis (table 1). (See 'Etiologic agents' above.)
•Maternal risk factors – Maternal risk factors for neonatal sepsis include chorioamnionitis (intra-amniotic infection), intrapartum fever, preterm delivery, maternal GBS colonization, and prolonged rupture of membranes. (See "Prevention of early-onset group B streptococcal disease in neonates", section on 'Identification of pregnancies at increased risk for early-onset neonatal GBS'.)
●Clinical manifestations – Clinical manifestations are nonspecific and include (table 2) (see 'Clinical manifestations' above):
•Fetal distress
•Temperature instability (fever or hypothermia)
•Respiratory and cardiocirculatory symptoms (most commonly respiratory distress and tachycardia)
•Neurologic symptoms (irritability, lethargy, poor tone, and seizures)
•Gastrointestinal abnormalities (poor feeding, vomiting, and abdominal distension)
●Evaluation – Evaluation and initial management of neonates with suspected sepsis should include a review of the pregnancy, labor, and delivery; complete physical examination; laboratory evaluation; and prompt initiation of empiric antibiotics (after obtaining cultures). Our suggested empiric antibiotic regimens are summarized in the table (table 4) and are discussed in detail separately. (See 'Evaluation and initial management' above and "Management and outcome of sepsis in term and late preterm neonates", section on 'Initial empiric therapy'.)
•Early-onset – In neonates presenting with clinical signs of sepsis within 72 hours after birth, the diagnostic evaluation includes (see 'Symptomatic neonates' above and 'Laboratory tests' above):
-Blood culture. (See 'Blood culture' above.)
-Inflammatory markers (eg, complete blood count [CBC] with differential, C-reactive protein [CRP], and/or procalcitonin [PCT]). These tests are not required but they may be helpful in determining length of therapy if followed serially. (See 'Complete blood count' above and 'Other inflammatory markers' above.)
-Lumbar puncture (if the infant is clinically stable enough to tolerate the procedure and it will not delay initiation of antibiotic therapy). (See 'Lumbar puncture' above.)
-Chest radiograph (if respiratory symptoms are present).
Empiric antibiotics should be provided to these neonates pending culture results (table 4). (See "Management and outcome of sepsis in term and late preterm neonates", section on 'Early-onset sepsis'.)
The approach to evaluating well-appearing newborns who have risk factors for early-onset sepsis is discussed separately. (See "Management of neonates at risk for early-onset group B streptococcal infection".)
•Late-onset – Neonates presenting with clinical signs of sepsis at ≥72 hours after birth should undergo a diagnostic evaluation similar to that described above for early-onset sepsis but also including a urine culture and cultures from potential foci of infection. Empiric antibiotic treatment should be initiated in these neonates pending culture results (table 4). (See 'Late-onset sepsis' above and "Management and outcome of sepsis in term and late preterm neonates", section on 'Late-onset sepsis'.)
●Diagnosis – Isolation of a pathogen from a blood culture confirms the diagnosis of neonatal sepsis. (See 'Diagnosis' above.)
The differential diagnosis of neonatal sepsis includes other systemic infections and noninfectious conditions including respiratory diseases (eg, transient tachypnea of the newborn and respiratory distress syndrome), cardiac diseases (eg, congenital heart disease and supraventricular tachycardia), neurologic injury (eg, from anoxia or hemorrhage), inborn errors of metabolism, and neonatal abstinence syndrome (table 5). (See 'Differential diagnosis' above.)
