Vertical sepsis (VS), also known as early-onset sepsis, continues to be a significant cause of morbidity and mortality in newborn infants today. It results from the colonization of the fetus or neonate by pathogens present in the maternal genital tract or, in rare cases, through transplacental hematogenous spread. The administration of intrapartum antibiotic prophylaxis under certain maternal circumstances and early administration of antibiotics upon suspicion of VS in the neonate have achieved a reduction in the incidence and severity of VS.1 However, this approach is not without risks, which chiefly concern the short- and long-term consequences for the pregnant mother/fetus or the neonate of the exposure to antibiotics.2 Therefore, identifying which neonates require early antibiotherapy and when or under what conditions to stop it is of utmost importance. The aim of this article is to provide updated information on the epidemiology of VS based on nationwide data from the Grupo Castrillo Neonatal Network3 and, based on this information, to update the recommendations for the management and treatment of neonates at risk of or with suspected VS.

Confirmed vertical sepsis is defined by isolation in blood culture of a pathogenic organism combined with the development of clinical manifestations of sepsis (Table 1) within 3 days of birth (≤72 h), which is why it is also known as early-onset neonatal sepsis. Although, based on data from the Grupo Castrillo Neonatal Network, up to 95% of cases have onset in the first 3 days post birth, the window for the onset of symptoms can extend to 7 days post birth. However, confirmation of vertical transmission in such cases requires identification of perinatal risk factors for infection (maternal colonization by Streptococcus agalactiae/Group B Streptococcus [GBS], prolonged rupture of membranes, chorioamnionitis or spontaneous preterm labor), ruling out nosocomial infection, and isolation in culture of a pathogen commonly involved in vertical sepsis (GBS, Escherichia coli).

Suspected vertical sepsis is one of the most common diagnoses in neonatology, and the condition is suspected based on the presence of clinical signs suggestive of infection and/or elevation of biomarkers associated with infection (C-reactive protein [CRP], procalcitonin [PCT], interleukin-6 [IL-6]) in a newborn infant, especially when there are perinatal risk factors for infection. This clinical picture prompts hospital admission of the neonate and initiation of empirical antibiotherapy, which in many cases is maintained for a prolonged period despite negative blood culture results, the low positive predictive value of current biomarkers, and the low incidence of vertical sepsis.4 In this context, the term “clinical vertical sepsis” is used to refer to infants born to mothers with perinatal risk factors for infection and who received intrapartum antibiotherapy, who present with suggestive symptoms and elevation of biomarkers in the first 3 days post birth, and who receive a complete course of antibiotherapy despite negative blood culture results.5

The incidence of VS in Spain has declined significantly since the implementation in 1998 of the strategy for the prevention of perinatal infection by GBS (the most common cause of vertical sepsis at that time), and is currently among the lowest in the world.3 This international strategy, adopted and endorsed by several scientific societies in Spain (Spanish Society of Neonatology [SENeo] and Spanish Society of Gynecology and Obstetrics [SEGO], among others) continues to be the only available effective preventive measure. It is based on the universal screening of GBS in pregnant women with administration of intrapartum antibiotherapy to carriers and establishing the indication of prophylaxis in mothers of unknown carriage status based on the presence of risk factors.6,7

The generalized implementation of this preventive protocol has changed epidemiological trends in vertical sepsis worldwide.8 In Spain (Fig. 1), its incidence has decreased by 60% (from 2.4 cases per 1000 live births in 1996 to 1–1.2 cases per 1000 live birth since 2001) and a significant shift in its etiology, with GBS accounting for 50% of cases of sepsis in 1996 compared to less than 30% at present (from 1.2 cases per 1000 live births to 0.2−0.3 per 1000), as cases caused by Escherichia coli surged, going from 12% to 30% of cases, and becoming the leading cause of vertical sepsis in preterm neonates (5–11 cases per 1000 live births with birth weight ≤ 1500 g) (Fig. 2A and B). Other pathogens involved in vertical sepsis include Enterococcus spp (and other streptococci) and Listeria monocytogenes among the gram-positive bacteria and Klebsiella spp among the gram-negative bacteria3 (Table 2).

Figure 1.

Temporal trends in the incidence of confirmed vertical sepsis per 1000 live births, overall (solid line) and in infants with birth weight ≤ 1500 g (dotted line), in the 1996–2023 period in Spain. Source: Grupo Castrillo Neonatal Network.

Figure 2.

(A) Temporal trends in the incidence of confirmed vertical sepsis caused by GBS per 1000 live births in the 1996–2023 period in Spain. Source: Grupo Castrillo Neonatal Network. (B) Temporal trends in the incidence of confirmed vertical sepsis caused by E coli in infants with birth weight ≤ 1500 g per 1000 live births in the 1996–2023 period in Spain. Source: Grupo Castrillo Neonatal Network.

The reported mortality of confirmed VS ranges between 8% and 10% overall, increasing to 30% in infants with birth weight of less than 1500 g. Besides gestational age, other factors associated with mortality are the presence of intrauterine infection, the isolated pathogen (E coli) or the presence of multidrug-resistant bacteria.3

The main perinatal risk factors are:

• •

  Maternal colonization by GBS in the absence of adequate intrapartum antibiotic prophylaxis (IAP).

• •

  Spontaneous preterm labor (<37 weeks of gestation).

• •

  Prolonged rupture of membranes ≥ 18 h.

• •

  Suspected or confirmed intrauterine inflammation, infection or both (“Triple I”) based on the presence of maternal fever ≥ 39 °C or fever between 38.0 and 38.9 °C lasting more than 30 min accompanied by at least one of the following: fetal tachycardia (>160 bpm for ≥10 min), maternal leukocytosis with a white blood cell count greater than 15.000/mm3 (without administration of corticosteroids) or purulent cervical discharge. Maternal tachycardia, uterine irritability or contractions and the elevation of inflammatory markers (such as CRP) are not considered diagnostic criteria for Triple I, although their presence support the diagnosis. Confirmation of Triple I requires evidence of bacterial growth in amniotic fluid or histopathologic evidence of infection in the placenta.9

Neonatal sepsis calculator. It is a tool based on a multivariate predictive model that combines perinatal maternal risk factors and variables related to the clinical presentation of the neonate to estimate the individual risk of VS in a particular infant and act according to the calculated risk. Multiple studies attest to its usefulness, reporting a significant reduction in unnecessary antibiotherapy (40%–60%), performance of laboratory tests and neonatal admissions for suspected sepsis, with no associated increase in adverse events (mortality). Some of its limitations are the restriction of its use to the first 12–24 h of life (there are cases of VS with later onset), the need for staff qualified in the clinical assessment of the neonate (detection of signs of infection) and that it cannot be applied to infants born before 34 weeks of gestational age. It is available as a web calculator and as a mobile app.10,11

Although the risk of VS is inversely proportional to gestational age (GA), it is possible to identify preterm infants at low risk of sepsis (cesarean delivery for maternal indications, absence of labor or rupture of membranes and absence of suspicion of intra-amniotic infection) in whom empirical antibiotherapy may be withheld if they show no clinical signs of VS.12

The current prevention strategy is based on the identification of pregnant women colonized by GBS (10%–30% of pregnant women) and administration of IAP to carriers or, in the subset of women of unknown carriage status, those with risk factors (Table 3).6,7 Its implementation, as noted above, has been followed by sharp decline in VS from GBS, and, in fact, most of the cases recently reported tend to be related to false negatives from vaginal-rectal cultures resulting in failure to deliver IAP.

During pregnancy, screening for GBS through a vaginal-rectal culture is indicated in the following cases:

• •

  All pregnant women between 36+0 and 37+6 weeks of gestation.

• •

  Previous culture performed more than 5 weeks before delivery that was negative for GBS.

• •

  Labor before 37+0 weeks, with or without rupture of membranes.

• •

  In the case of preterm or term prelabor premature rupture of membranes.

Vaginal-rectal culture is not necessary in pregnant women who have had a previous child with neonatal GBS infection or if GBS has been detected in urine during the current pregnancy (both cases in which administration of IAP is indicated).

Intrapartum antibiotic prophylaxis is considered complete if at least one dose of penicillin, ampicillin, or cefazolin has been administered at least 4 h before delivery. Any other antibiotic or time window should be considered inadequate or insufficient.

Intrapartum antibiotic prophylaxis is not effective for prevention of late-onset sepsis from GBS.

Signs of infection in neonates are nonspecific and tend to be few. Neonates with bacterial sepsis may exhibit systemic or focal signs and symptoms of infection, including respiratory distress, thermal instability, hypotension, poor perfusion, pallor, abnormal heart rhythms, apnea, irritability, lethargy, and seizures, among others (Table 1).

None of the biomarkers that are currently available can be used to diagnose neonatal sepsis with sufficient sensitivity and specificity. The most commonly used diagnostic methods are:

Blood culture. It is the gold standard for diagnosis of VS. Advances in blood culture processing, such as automation and the use of mass spectrometry techniques, have reduced the time required to detect bacterial growth, facilitating the early discontinuation of empirical antibiotherapy. Blood culture media contain antibiotic-binding resins that limit the antimicrobial activity of any antibiotics present in the blood sample, allowing bacteremia to be identified at a level of 1–10 colony-forming units (CFU) per mL. Blood culture volumes of at least 1 mL maximize the probability of growth in cases where the bacterial load in the bloodstream is low. Smaller volumes may reduce the sensitivity and increase the risk of contamination and the time required for any microorganism to grow.13,14

Molecular techniques based on the identification of bacterial DNA/RNA in biological samples. These novel techniques are increasingly used to identify the presence of microorganisms in blood and other biological fluids. Their advantages include faster identification (less than 12 h), requiring smaller blood volumes (they can even be performed on the same sample used for blood culture), greater sensitivity and specificity, and the ability to simultaneously detect multiple pathogens and even antibiotic resistance genes (nested multiplex polymerase chain reaction). The disadvantages are that they cannot differentiate between infection and contamination (so the clinical evaluation continues to be a priority), the cost is still high, and they are not available in every hospital.15,16

Cerebrospinal fluid (CSF) analysis. Identifying meningeal involvement is essential, as it affects decisions regarding the dose, type and duration of antibiotic treatment, and it is associated with a higher morbidity, mortality, and incidence of long-term sequelae. In the experience of the Grupo Castrillo network, the incidence of vertical meningitis is 0.2 cases per 1000 live births (10%–20% of vertical sepsis cases are associated with meningitis). Given its incidence, lumbar puncture should be performed when there is a strong suspicion of sepsis and/or the clinical presentation is suggestive of meningitis. Preferably, it should be performed before initiation of antibiotherapy, although it can be delayed if the patient is unstable, without delaying antibiotherapy to await its performance.17

White blood cell differential. Abnormal findings in this test should not be used to make the diagnosis of sepsis, as the white blood cell count can be elevated in conditions involving inflammation without infection, such as chorioamnionitis, complicated delivery, pneumothorax, seizures, perinatal asphyxia, meconium aspiration syndrome, or maternal fever, and neutropenia in cases of preeclampsia, maternal use of corticosteroids, and even cesarean delivery.15,18 Counts higher than 20 000/mm3 or below 5000/mm3 have been associated with the presence of VS, but offer a very low sensitivity. The white blood cell count changes in the first days of life, so serial counts could be useful. As for the immature/total neutrophil ratio (I/T), the value range associated with the presence of neonatal sepsis is ≥0.2 up to 72 h post birth and ≥0.12 thereafter.

Management of neonates with gestational age ≥ 35 weeks22 (Fig. 3).

• •

  Asymptomatic neonate with risk factors. Close monitoring of the infant along with the mother is recommended for the first 36–48 hpost birth (especially the first 24 h) to rule out the presence of symptoms suggestive of infection. Risk factors include maternal fever ≥ 38 °C and/or inadequate IAP in a pregnant woman in whom it was indicated.

• •

  Neonate with symptoms suggestive of vs. the recommended approach is collection of a blood sample for culture and initiation of standard empirical antibiotherapy (ampicillin + aminoglycoside), which should be stopped after 24–48 h if the cultures are negative and the infant is stable. If symptoms persist and no infectious cause is detected, alternative etiologies for the clinical presentation should be explored.

• •

  Neonate with respiratory symptoms. It has been customary practice to treat all infants with respiratory distress at birth with antibiotics due to suspected sepsis/pneumonia, independently of the presence or absence of infection risk factors, even though it is known that most of these neonates have transient tachypnea. In this subset of patients, it is important to carefully consider infection risk factors and, above all, their clinical course in the first hours of life before choosing to initiate antibiotherapy.

Figure 3.

Algorithm for the management of neonates born at or after 35 weeks of gestation at risk of vertical sepsis.

Abbreviations: AST, antimicrobial susceptibility testing; CSF, cerebrospinal fluid; GA, gestational age; h, hours; NB, newborn.

Management of neonates with gestational age < 35 weeks23 (Fig. 4).

Figure 4.

Algorithm for the management of neonates born before 35 weeks of gestation at risk of vertical sepsis.

Abbreviations: BC, blood culture; CSF, cerebrospinal fluid; GA, gestational age; h, hours; NB, newborn; PROM, premature rupture of membranes; VS, vertical sepsis.

Since the risk of VS is inversely proportional to gestational age, many preterm infants receive antibiotics from birth and often for prolonged periods. However, it is possible to identify preterm deliveries carrying a low risk of sepsis in which initiation of empiric antibiotherapy can be avoided as well as preterm infants whom, despite the presence of risk factors, are not going to develop vertical sepsis, in whom it is possible to stop the course of empirical antibiotherapy early.

The strategies for stratifying the risk of infection in preterm neonates born before 35 weeks are:

• •

  Low risk vs. Preterm infants born by cesarean section solely for maternal/fetal indications (preeclampsia, placental insufficiency, intrauterine growth restriction, etc), in absence of onset of labor or premature rupture of membranes (PROM). In these cases, the recommended approach is avoiding empirical antibiotherapy or, when it is initiated, withdrawing it within 24–48 h if the blood culture turns out negative.

• •

  High risk of vs. Preterm birth due to cervical insufficiency, spontaneous preterm labor, PROM, suspected intrauterine infection and/or nonreassuring fetal status of unknown etiology. In these cases, performance of blood tests and blood cultures should be considered, and empirical antibiotherapy initiated; performance of lumbar puncture should be contemplated as indicated earlier in the text. If the blood culture is negative after 24–48 h, the discontinuation of antibiotic treatment will be considered based on the clinical course of the neonate.

Antibiotics are the drugs most frequently used inappropriately or unnecessarily in neonatal units. The lack of specific symptoms and diagnostic criteria for infection, together with the low positive predictive value of available biomarkers and the fear of sepsis, contribute to the excessive and unnecessary use of antibiotics.24 The use of antibiotics in neonates contributes to the separation of the infant from their parents, delays breastfeeding initiation, alters the intestinal microbiota, and increases health care costs and antibiotic resistance. It has also been linked to a higher incidence of atopy, allergic disorders, asthma, and childhood obesity. In preterm infants, it can also lead to complications such as chronic lung disease, nosocomial sepsis, necrotizing enterocolitis, periventricular leukomalacia, severe retinopathy of prematurity, and poor long-term neurodevelopmental outcomes.25

Empirical antibiotherapy should be initiated as soon as possible if sepsis is strongly suspected. A combination of ampicillin and an aminoglycoside (gentamicin, amikacin) is recommended to cover 83% of the pathogenic bacteria associated with vertical sepsis (GBS, E coli, Enterococcus spp and L monocytogenes).22,23,26 It should be noted that GBS currently remains sensitive to penicillin and ampicillin (100% of isolates according to data from the Castrillo Group), but E coli exhibits a high rate of resistance to ampicillin (75% of isolates), so an aminoglycoside must be added while awaiting the results of blood culture. The empirical use of third-generation cephalosporins (cefotaxime) or carbapenems is not recommended, as there is evidence that it promotes the development of antibiotic resistance and is associated with an increased risk of candidiasis, enterocolitis, and mortality in preterm infants.26,27 The empirical use of cefotaxime could be considered in cases of abnormal CSF analysis results and would be justified in the event of gram-negative bacterial growth in blood or CSF cultures (especially in very low birth weight infants), while awaiting susceptibility test results or in critically ill infants who are not responding to first-line treatment (in this case, if the mother is colonized by a gram-negative beta-lactamase-producing bacterium, the use of meropenem would be recommended).

If active herpes simplex virus infection is suspected based on the obstetric history, the addition of intravenous acyclovir to the empirical antibiotic regimen should be considered.

Definitive targeted antibiotherapy will be based on antimicrobial susceptibility test results, always in monotherapy and using the narrowest-spectrum antibiotic to which the organism is susceptible (for example, penicillin in cases of sepsis not complicated by GBS).

There is no high-quality scientific evidence to establish the ideal duration of antibiotic treatment, but we offer the following broad recommendations based on the experience of the Grupo Castrillo Neonatal Network and the previous literature:

• •

  Duration of treatment for suspected vs. To date, there is no unanimous agreement on the number of doses or days of empirical treatment in neonates with negative blood cultures and clinical improvement or symptoms attributable to another cause, partly due to differences in the processing of blood cultures. Most sources suggest discontinuation after 24–36 h if the blood culture is negative at that time.24

• •

  Duration of treatment for clinical vs. Despite negative blood culture results, if the clinical suspicion of infection persists and is accompanied by risk factors and abnormal blood test results, treatment may be prolonged for five days, and daily reevaluation of its necessity is recommended.12,28

• •

  Duration of treatment for confirmed vs. A maximum of seven days is recommended (for both gram-positive and gram-negative bacteria) as long as the response is favorable, adjusting treatment according to the susceptibility profile and always choosing the narrowest-spectrum option. In the case of sepsis caused by S aureus or Listeria spp, a minimum of 10 days is recommended. For sepsis associated with meningitis, the duration of antibiotherapy may extend to 14 days in the case of uncomplicated meningitis involving a gram-positive agent or to up to 21 total days or to 14 days after CSF cultures become negative in the case of uncomplicated meningitis involving a gram-negative agent.12,29

This research did not receive any external funding.