Neonatal sepsis - Clinical Tree

Introduction

Sepsis is the illness resulting from systemic bacterial or (less commonly) fungal infection. It is an ever-present threat in the NICU, accounting for approximately 10% of infant deaths under 1 month of age worldwide ( ) and contributing significantly to long-term neurodevelopmental impairments among NICU survivors.

The clinical manifestations of neonatal sepsis are variable. The onset can be abrupt or insidious, and many of the most common signs and laboratory findings associated with neonatal sepsis are nonspecific. Although certain neonatal populations—such as very low birth weight (VLBW) and chronically instrumented infants—are at increased risk, sepsis can also strike unexpectedly in low-risk patients. Vigilance and a high index of suspicion are key to diagnosing and treating sepsis before the infection spirals out of control.

Neonatal sepsis can be categorized as early onset or late onset, with some sources using 3 days of age as the cutoff for early onset sepsis (EOS) and others using 7 days (the 3-day definition is more common). The two categories exist because early and late-onset sepsis (LOS) result from different modes of acquisition, with EOS most often caused by vertical microbial transmission during the perinatal period and LOS most often from nosocomial or other environmental sources.

The distinct modes of acquisition are reflected in differences in the organisms recovered from EOS and LOS populations, with corresponding implications for appropriate initial empiric antibiotic therapy for the two groups. A precise demarcation between EOS and LOS is therefore less important than the concept that during the first week of life a shift in the pathogenesis and pathogen profile of sepsis occurs.

This chapter addresses EOS and LOS in separate sections. Each begins with a review of pathogenesis, epidemiology, and microbiology before presenting a series of instructive cases drawn from actual patient histories.

Early onset sepsis

Pathogenesis

During fetal life, the innate immune system develops within the highly protected confines of the amniotic sac, defended from microbes by multiple maternal anatomic and immunologic barriers. Additionally, the fetal skin surface secretes a mixture of antimicrobial peptides into the surrounding amniotic fluid, further bolstering protection from infection ( ). Although recent evidence points to the presence of a limited placental microbiome ( )—whose role in fetal immunity is still unclear—there is no doubt that the perinatal period involves sudden exposure to a vast microbiological ecosystem.

The neonate has numerous adaptive responses to this abrupt transition from a nearly sterile environment. Bacterial colonization of the skin, which begins during passage through the birth canal (or immediately upon delivery for babies born by cesarean section without labor), triggers interleukin-1 (IL-1) and IL-6 release from resident macrophages at the bases of hair follicles. (This local cytokine release is clinically manifest as the benign rash of erythema toxicum neonatorum.) IL-1 and IL-6 bind hepatocyte receptors, triggering systemic release of nonspecific acute phase reactants that protect against unchecked bacteremia ( ). At the same time, early intestinal bacterial colonization activates antiinflammatory signaling networks that promote establishment of a stable microbiome ( ).

However, early exposure to a pathogenic bacterial species—often expressing cytotoxic and immune evasion virulence factors—can overwhelm the orderly process of immune adaptation to a commensal microbiome. In the absence of a robust adaptive immune response, which does not develop until later in toddlerhood, the neonate has few defenses against an expanding bloodborne bacterial population.

EOS caused by group B Streptococcus (GBS) exemplifies this chain of events. Perinatal mucosal exposure to maternal GBS colonization—either during passage through the birth canal or in swallowed breastmilk—allows GBS access to a microenvironment relatively free of competing microbiota ( ). As the GBS population expands, expression of the cytotoxin β-hemolysin/cytolysin permits traversal of epithelial surfaces ( ), allowing bloodstream invasion where toxin expression is further amplified ( ).

Epidemiology

Active, population-based surveillance in the United States by the Centers for Disease Control and Prevention has revealed a stable incidence of EOS over the past 10 years, affecting 0.77 to 0.79 per 1000 live births ( ; ). Smaller studies during the same period have shown slightly higher incidence, documenting EOS rates of approximately 1 case per 1000 live births ( ; ).

Table 13.1 lists evidence-based risk factors for EOS. The single greatest risk factor is prematurity. EOS rates are inversely proportional to gestational age, with premature infants younger than 34 weeks’ gestation more than 10 times more likely to develop EOS than term infants ( ). Extremely premature, extremely low birth weight babies can have EOS rates close to 100 times greater than term neonates ( ).

TABLE 13.1

Major Risk Factors for Early Onset Sepsis

Prematurity

Low birth weight

Prolonged rupture of membranes

Preterm rupture of membranes

Chorioamnionitis

5-minute Apgar score <7

Maternal GBS colonization (with no or inadequate intrapartum antibiotic prophylaxis)

Black maternal race

Maternal age ≤20 years

Multiple digital vaginal examinations

Membrane stripping

Internal fetal monitor

Infectious chorioamnionitis, which is usually the result of invasion of the amnioplacental unit by vaginal microbes, can result in fetal infection and induction of labor, culminating in delivery of an acutely ill newborn ( ). Historically, chorioamnionitis has been considered a key risk factor for EOS, with multiple prevention guidelines recommending empiric antibiotic therapy for all affected neonates, regardless of their clinical status ( ; ; ).

The American College of Obstetrics and Gynecology (ACOG) recently issued updated recommendations for diagnosis and management of suspected intraamniotic infection ( ). When chorioamnionitis is diagnosed based on multiple clinical criteria (including maternal fever, leukocytosis, tachycardia, and purulent amniotic fluid), there is good evidence that it is a significant risk factor for EOS ( ). However, recent work has called into question whether the risk extends to newborns without clinical signs of illness ( ; ) and the consistency with which strict clinical criteria are used to diagnose chorioamnionitis. A large nationwide survey of obstetricians revealed that chorioamnionitis is often diagnosed based on fever alone ( ), which significantly lowers its predictive power for EOS.

Recent literature has therefore suggested alternative, less stringent approaches to managing newborns following a diagnosis of chorioamnionitis ( ). Several studies have demonstrated that asymptomatic late-preterm and term infants born to mothers with chorioamnionitis can be safely managed with serial examinations rather than reflexive empiric antibiotic therapy ( ; ; ).

Microbiology

GBS and Escherichia coli are the most commonly isolated pathogens in EOS, accounting for approximately 75% of cases ( ). GBS is the most frequent isolate, but E. coli EOS is more common among VLBW infants ( ) and may be more common overall than GBS in certain geographic regions or single institutions ( ).

From 2005 to 2014, Streptococcus viridans was the next most common isolate in the CDC surveillance network ( ). The remainder of EOS is caused by a variety of mostly enteric bacteria, including Klebsiella spp., Enterococcus spp., Haemophilus spp., Enterobacter spp., and Listeria monocytogenes.

Case 1

A newborn girl, born at 356⁄7 weeks’ gestation, is brought to the transitional nursery at 1 hour of life after the postpartum nurse expressed concerns about her breathing. The mother is a 27-year-old primigravid woman with no significant past medical history. Her antepartum GBS screen was negative, and she received no intrapartum antibiotics. A maximum maternal temperature of 100.1°C was recorded in the setting of epidural anesthesia administration. Vaginal delivery occurred 16 hours after rupture of membranes, and the Apgar scores were 8 and 9 at 1 and 5 minutes, respectively. The baby is pink and active without respiratory support. Her vital signs are normal except for a respiratory rate of 85 breaths per minute. The physical examination is significant for mild nasal flaring, intercostal and subcostal retractions, good bilateral air entry, and expiratory grunting. The Sp o ₂ is 98%.

Exercise 1

Question

How should this baby be managed?

A.
Because she is still undergoing the physiologic transition to the extrauterine environment, her mild respiratory distress is normal. Return her to the postpartum unit for routine nursery care.
B.
She should be closely observed because her respiratory signs place her at increased risk of EOS. In the absence of improvement over the next several hours, she should undergo laboratory and radiographic evaluation for sepsis.
C.
Her prematurity and respiratory distress, combined with maternal signs of chorioamnionitis, place her at high risk of EOS. She should have a blood culture drawn and receive empiric antibiotic therapy.

Answer

Correct answer: B

Sepsis evaluation and treatment decisions are usually straightforward for asymptomatic newborns with no significant EOS risk factors, who can be managed in the newborn nursery, and acutely ill newborns with clear sepsis risk factors, who must receive empiric antibiotics while EOS is ruled out.

The more challenging cases are those—like the example—where mild symptoms and a constellation of modest risk factors don’t point to an obvious course of action.

A major advance in the management of possible EOS was the introduction of an online, publicly available sepsis risk calculator, based on a large clinical data set and Bayesian analysis of maternal risk factors ( ). A subsequent refinement of the sepsis calculator included stratified risk assignments based on the newborn’s clinical status, which could be characterized as well appearing, equivocal, or clinically ill ( ).

The benefit of the sepsis calculator is that it distills multiple variables—both categorical and continuous risk factors—into a single probability that the newborn is infected. Entering features of the example case into the sepsis calculator and assigning the baby to the “equivocal” clinical status reveals a risk of 9.7/1000 live births. The sepsis calculator also provides management recommendations, which are based on risk cutoffs for each clinical illness category ( ).

It is important to note that these online management recommendations are made in the absence of certain key information, such as the baby’s age and overall clinical trajectory. Therefore although the calculator recommendation for the example case is to administer empiric antibiotics, the clinician may make a different decision based on additional evidence. In this instance the late-preterm baby is still within a timeframe—approximately 6 hours after birth—when it is reasonable to expect some mild respiratory distress as part of the physiologic transition. Continued observation may show that her respiratory distress is improving, which would argue strongly against sepsis; a newborn clinical examination that improves without antibiotic administration is incompatible with EOS, which is a progressive condition. If there were any clinical deterioration—or no clear improvement—within the 6-hour window, it would be imperative to draw a blood culture and start empiric antibiotics without further delay.

Case 2

You are called to the newborn nursery to assess a 1-day-old female infant born at 36 weeks’ gestation following prolonged rupture of the membranes (20 hours). There were no signs of chorioamnionitis. Although she was stable without respiratory support and feeding well throughout the day of birth, today she has been taking smaller volumes and has had intermittent tachypnea. The nurse who called you explains that she just observed an apneic episode with color change that required stimulation. The baby is pink and responsive to examination but has a respiratory rate of 70 breaths per minute and somewhat cool extremities. You decide to admit to the NICU for a sepsis evaluation, including sepsis screening laboratories.

Exercise 2

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here