Neonatal infections continue to cause morbidity and mortality in infants. Among approximately 400,000 infants followed nationally, the incidence rates of early-onset sepsis infection within 3 days of life are 0.98 cases per 1000 live births. Newborn infants are at increased risk for infections because they have relative immunodeficiency. This article provides evidence-based practical approaches to the diagnosis, management, and prevention of neonatal infections.
Key points
- •
Neonatal infections continue to cause morbidity and mortality in infants. Group B streptococcus and Escherichia coli are the most common agents of early-onset sepsis, whereas coagulase-negative Staphylococcus is the predominant cause of late-onset sepsis.
- •
Other important agents include Listeria monocytogenes , syphilis, Staphylococcus aureus , herpes simplex virus, cytomegalovirus, and Candida spp.
- •
There is increasing recognition of respiratory viral infections contributing to ruling out sepsis in very young infants whose presentations are similar to bacterial infections.
- •
Initial work up for neonatal infection consists of complete blood count and blood culture, with the option of performing cerebrospinal fluid analyses and culture if clinically indicated. Serial determinations of biomarkers (C-reactive protein, procalcitonin, or neutrophil CD64) may be used adjunctively in the diagnosis and management of neonatal infection.
- •
Ampicillin and gentamicin remains the cornerstone of initial antimicrobial regimen for neonatal infections. Third-generation cephalosporins should be used judiciously.
- •
The use of antiviral (acyclovir, ganciclovir, valganciclovir, and oseltamivir) and antifungal agents (fluconazole, amphotericin B, and voriconazole) may reduce mortality and morbidity due to specific viral and fungal disease.
- •
Different strategies, such as group B streptococcal prophylaxis, hand hygiene, immunization and immunoprophylaxis, antimicrobial stewardship, probiotics, and prebiotics, and care bundles may be used in preventing infections in neonates.
Introduction
Neonatal infections continue to cause morbidity and mortality in infants. Among approximately 400,000 infants followed nationally, the incidence rates of early-onset sepsis (EOS) infection within 3 days of life were 0.98 cases per 1000 live births. More than two-thirds of the frequently isolated organisms were associated with group B streptococcus (GBS) (43%) and Escherichia coli (29%). Although 20% of the term infants were treated in the newborn nursery, 77% of the infected infants required intensive care management. Of those who survived beyond 3 days of life, about 21% had an episode of late-onset sepsis (LOS) infection after 3 days of life. The overall mortality rate of infected infants was 16%.
Newborn infants are at increased risk for infections because they have relative immunodeficiency. This may be due to decreased passage of maternal antibodies in preterm infants and to immaturity of the immune system in general. The innate immune functions in infants are impaired with decreased production of inflammatory markers (interleukin 6 and tumor necrosis factor) and with decreased dendritic and neutrophil functions. The adaptive immune system is less than optimal with decreased cytotoxic functions, decreased cell mediated immunity, and delayed or lack of isotype switching. Complement is important in opsonization and bacterial killing. In term infants, complement levels are approximately half compared with adults. Taken together, these predispose infants to severe, prolonged, or recurrent infections associated with bacterial, viral, or fungal infections.
Suspected sepsis, presumed infection, and ruling out sepsis remain the most common diagnoses in the nursery intensive care unit (NICU). The American Academy of Pediatrics (AAP) Committee on Fetus and Newborn has published a clinical report extensively discussing clinically relevant challenges: identifying newborns with signs of sepsis with high likelihood of EOS requiring antimicrobial regimen and identifying healthy-appearing newborns with high likelihood of EOS requiring antimicrobial regimen. The committee concluded that, although these guidelines are evidence-based, they may be modified by the clinical judgment of the provider. The primary reason is that the clinical presentation of neonatal infection may be subtle and nonspecific, and may overlap with noninfectious causes. Many clinicians empirically start broad spectrum antimicrobial regimen for infants considered at risk for sepsis but antibiotics are occasionally continued despite a negative blood culture. This practice may be detrimental to the infant because it increases the risk of invasive fungal infections, necrotizing enterocolitis (NEC), or death, which increases the pressure for selecting multidrug-resistant organisms and even the risk of LOS.
The purpose of this article is to provide evidence-based practical approaches to the diagnosis, management, and prevention of neonatal infections.
Introduction
Neonatal infections continue to cause morbidity and mortality in infants. Among approximately 400,000 infants followed nationally, the incidence rates of early-onset sepsis (EOS) infection within 3 days of life were 0.98 cases per 1000 live births. More than two-thirds of the frequently isolated organisms were associated with group B streptococcus (GBS) (43%) and Escherichia coli (29%). Although 20% of the term infants were treated in the newborn nursery, 77% of the infected infants required intensive care management. Of those who survived beyond 3 days of life, about 21% had an episode of late-onset sepsis (LOS) infection after 3 days of life. The overall mortality rate of infected infants was 16%.
Newborn infants are at increased risk for infections because they have relative immunodeficiency. This may be due to decreased passage of maternal antibodies in preterm infants and to immaturity of the immune system in general. The innate immune functions in infants are impaired with decreased production of inflammatory markers (interleukin 6 and tumor necrosis factor) and with decreased dendritic and neutrophil functions. The adaptive immune system is less than optimal with decreased cytotoxic functions, decreased cell mediated immunity, and delayed or lack of isotype switching. Complement is important in opsonization and bacterial killing. In term infants, complement levels are approximately half compared with adults. Taken together, these predispose infants to severe, prolonged, or recurrent infections associated with bacterial, viral, or fungal infections.
Suspected sepsis, presumed infection, and ruling out sepsis remain the most common diagnoses in the nursery intensive care unit (NICU). The American Academy of Pediatrics (AAP) Committee on Fetus and Newborn has published a clinical report extensively discussing clinically relevant challenges: identifying newborns with signs of sepsis with high likelihood of EOS requiring antimicrobial regimen and identifying healthy-appearing newborns with high likelihood of EOS requiring antimicrobial regimen. The committee concluded that, although these guidelines are evidence-based, they may be modified by the clinical judgment of the provider. The primary reason is that the clinical presentation of neonatal infection may be subtle and nonspecific, and may overlap with noninfectious causes. Many clinicians empirically start broad spectrum antimicrobial regimen for infants considered at risk for sepsis but antibiotics are occasionally continued despite a negative blood culture. This practice may be detrimental to the infant because it increases the risk of invasive fungal infections, necrotizing enterocolitis (NEC), or death, which increases the pressure for selecting multidrug-resistant organisms and even the risk of LOS.
The purpose of this article is to provide evidence-based practical approaches to the diagnosis, management, and prevention of neonatal infections.
Microbiologic agents
The timing of transmission is one of the factors contributing to the cause of neonatal infections. Different pathogens may be acquired during pregnancy (prenatal), during delivery (perinatal), or after delivery (postnatal). Table 1 shows the different periods of transmission of various neonatal pathogens.
| Pathogens | During Pregnancy | During Delivery | After Delivery |
|---|---|---|---|
| Bacteria | |||
| Chlamydia trachomatis | — | + a | — |
| GBS | ++ b | ++ | ++ |
| Enterococcus spp | — | + | + |
| Enterobacteriaceae | — | ++ | ++ |
| Listeria monocytogenes | + | + | + |
| Neisseria gonorrhea | — | + | — |
| Staphylococcus spp | — | — | ++ |
| Treponema pallidum | + | — | — |
| Ureaplasma urealyticum | — | + | — |
| Viruses | |||
| Coronavirus | — | — | + |
| Cytomegalovirus | + | + | + |
| Enterovirus | + | + | + |
| Hepatitis B virus | + | + | — |
| Herpes simplex virus | + | — | — |
| Human immunodeficiency virus | + | + | + |
| Human metapneumovirus | — | — | + |
| Influenza | — | — | + |
| Parainfluenza virus | — | — | + |
| Parvovirus B19 | + | — | — |
| Respiratory syncytial virus | — | — | + |
| Rhinovirus | — | — | + |
| Rubella virus | + | — | — |
| Varicella-zoster virus | + | + | + |
| Fungi | |||
| Candida spp | — | + | + |
| Aspergillus spp | — | — | + |
| Protozoa | |||
| Toxoplasma gondii | + | — | — |
The introduction of new molecular-based assays, such as quantitative real-time polymerase chain reaction (PCR), has paved the way for increasing recognition of respiratory viral infections contributing to ruling out sepsis in late-onset infections. Table 1 includes respiratory viral infections (coronavirus, enterovirus, human metapneumovirus, influenza, parainfluenza virus, respiratory syncytial virus [RSV], and rhinovirus) as possible causes of postnatal infections in infants.
Clinical presentations
Early-Onset Infections
EOS is arbitrarily defined as infection within the first 3 days of life. The most common organisms associated with EOS include GBS and E coli . In general, the risk of bacterial infection in a healthy-appearing newborn remains relatively low. The most common clinical findings include hypoglycemia (<40 mg/dL, 22%) and hypothermia (<36.5 ° C, 20%), followed by hyperglycemia (>140 mg/dL, 19%) and apnea (18%).
Edwards and Baker summarized that newborn infants with sepsis manifest similar clinical signs as those with meningitis, including hyperthermia; hypothermia; respiratory distress; anorexia or vomiting; jaundice; and lethargy. Hypotension may be more frequently found in infants with sepsis, whereas irritability, convulsions, and bulging or full fontanel is found in those with meningitis. However, they cautioned that absence of any of the aforementioned signs do not exclude central nervous system involvement. Furthermore, it was suggested to evaluate infants for various foci of infections such as acute otitis media, conjunctivitis, osteomyelitis, pyogenic arthritis, and skin soft-tissue infections.
Late-Onset Infections
LOS is arbitrarily defined as infection after 3 days of life. The most common organisms isolated with LOS include coagulase-negative staphylococci in more than a third of the cases, which may or may not be associated with a medical device. Yeast or Candida spp infection is another important pathogen. Also, there is increasing recognition of viral respiratory infections as a possible cause in LOS. The most common clinical findings include hypothermia (41%), hyperglycemia (38%), apnea (38%), and bradycardia (30%).
There are several factors that may increase the risk for LOS. There are significantly more infants with LOS who have an indwelling central vascular catheter at the time of infection than those infants with EOS (78% vs 10%, P <.0001). Additionally, there are more infants with LOS who had a surgical procedure before infection (8% vs 1%, P <.0001).
Diagnostic evaluations
The clinical presentations of infections may overlap with noninfectious causes in newborns. It has been previously demonstrated that relying on symptoms alone may not be sufficient in diagnosing neonatal infections. Bacteremia has been reported in infants without clinical signs of sepsis. There are several diagnostic tests and principles that may guide clinicians in evaluating infants with infections.
Algorithm-Based Guideline
The AAP Committee on Fetus and Newborn have published a clinical report on the evaluation of asymptomatic infants (<37 and ≥37 week gestation) with risk factors for sepsis. Evaluation of asymptomatic preterm infants (<37-week) with risk factors for sepsis is shown in Fig. 1 . Similar algorithms for the evaluation of asymptomatic term infants (≥37 week gestation) are available from the AAP Committee on Fetus and Newborn. ( ).
Additional principles in the evaluation of infants with risk factors for sepsis follow:
- •
Major risk factors for neonatal sepsis include chorioamnionitis, prolonged rupture of membrane 18 or more hours, and colonization of GBS with inadequate intrapartum antimicrobial prophylaxis (IAP).
- •
Chorioamnionitis usually presents as maternal fever greater than 38 ° C (100.4 ° F) and its diagnosis should be discussed with the obstetric providers. Maternal fever may be the only abnormal finding in chorioamnionitis.
- •
Adequate IAP means maternal treatment with penicillin, ampicillin, or cefazolin at or earlier than 4 hours before delivery.
- •
At least 1 mL of blood may be sufficient for a single blood culture from a peripheral vein. Blood culture from umbilical artery catheter or umbilical vein may be a reliable alternative following aseptic techniques
- •
Screening blood cultures have not been proven of value and are not recommended.
- •
Complete blood count with differential has poor positive predictive value and it is suggested waiting 6 to 12 hours after birth to avoid falsely normal values at birth.
- •
Platelet counts remain low days to weeks after sepsis; thus this cannot be used in following response to treatment.
- •
The sensitivity of C-reactive protein (CRP) improves if done 6 to 12 hours after birth. Bacterial sepsis is unlikely if CRP remains normal.
- •
Lumbar puncture may be indicated in infants whom sepsis is highly suspected, those infants with bacteremia, and in infants who fail to respond to antimicrobial therapy.
- •
Urinary tract infection in newborns is associated with episodes of bacteremia; thus urine culture should not be part of routine sepsis workup.
- •
Microbiologic evaluation using gastric aspirates, tracheal aspirates, or superficial body sites cultures are of limited value and are not routinely recommended for sepsis.
Biomarkers
Several acute-phase reactants or biomarkers (neutrophil CD64 [nCD64], procalcitonin [PCT], or CRP) may be used adjunctively in the evaluation and management of neonatal infection. The diagnostic usefulness of the various surrogate markers depends on the phases of neonatal sepsis: early phase or 2 to 12 hours (nCD64), mid phase or 12–24 hours (PCT), and late phase or greater than 24 hours (CRP).
nCD64 is a high-affinity Fc receptor that increases with exposure to bacterial or fungal agents. The usefulness of nCD64 is related to its high negative predictive value as well as decreasing concentration on serial determinations on infants undergoing antimicrobial treatment of bacteremia. However, there is a scarcity of medical evidence to recommend nCD64 for routine evaluation of neonatal infection and this may not be readily available.
Procalcitonin released from tissues increases with infection at around 2 hours and peaks at 12 hours. It may also increase with noninfectious causes such as in respiratory distress syndrome and a physiologic increase during the first 24 hours of birth. PCT may not be readily available and the turnaround time varies in different institutions from 20 minutes to 5 hours.
CRP increases around 6 hours associated with an inflammatory response with release of interleukin-6 and peaks at 24 hours. CRP has been used in the algorithm-based guideline from the AAP Committee on Fetus and Newborn for the evaluation of asymptomatic term and preterm infants with a risk factor for sepsis. It is best used as part of a group of diagnostic tests together with blood culture and white blood cell with differential in the evaluation of neonatal infection. However, there is not enough medical evidence at this time to recommend serial determinations of CRP in guiding duration of antimicrobial therapy in infants. Further studies are needed to evaluate the usefulness of sequential determination of CRP and biomarkers for an antimicrobial stewardship program (ASP) in the NICU setting.
Molecular-Based Tests
In 2013, the Infectious Disease Society of America, in collaboration with the American Society for Microbiology, affirmed the importance of close collaboration and positive working relationships between clinicians and microbiologists to better serve patients. The most up-to-date edition of the Red Book provides contact information for expert advice and national collaborative study groups that give guidance on diagnostic assays regarding specific agents causing mother-to-child transmission. It is important to know the various microbiologic resources available locally, which include but are not limited to PCR and matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF).
Rapid antigen tests for respiratory viruses may lack sensitivity, which is important in the NICU setting in controlling local outbreaks. There are several nucleic acid amplification test platforms currently available that differ in the number of analytes detected. It is important to obtain adequate specimens and to use suitable viral transport media following manufacturer instructions.
MALDI-TOF is a valuable alternative to the conventional microbiologic assays; however, it may not be a readily available resource for diagnostic testing in most institutions. However, if it is available, it has several practical applications that may benefit clinical management even in the NICU settings:
- •
Earlier and accurate diagnosis of neonatal sepsis due to various bacteria
- •
Rapid identification of highly virulent GBS that causes meningitis and LOS in infants
- •
Identification of maternal-to-child transmissions (chorioamnionitis and neonatal infections) of opportunistic pathogen
- •
Accurate identification of bloodstream infection associated with fungal infections in the NICU
- •
Identification and monitoring the spread of nosocomial outbreak (eg, methicillin-resistant Staphylococcus aureus [MRSA] and Candida parapsilosis in the NICU).
Therapeutic management
When appropriate specimens for diagnostic evaluations are collected in clinically stable patients, then empirical antimicrobial therapy should be initiated for neonatal sepsis. It is recommended to discuss complicated cases, such as multidrug resistant organisms and infants not improving while on therapy or those requiring unconventional dosing regimens and antimicrobial agents, with pediatric infectious disease specialists.
Antibiotic Treatment
Ampicillin and gentamicin remains the cornerstone of initial antimicrobial regimen for early-onset neonatal infections. The combination of such broad-spectrum antibiotic regimens cover the most common cause (GBS and E coli in more than 70%) of EOS and has synergistic activity (against GBS and Listeria monocytogenes ). The dosing regimen for ampicillin may change over time based on the chronologic age of the infant and body weight. For example, an 8-day-old infant weighing greater than 2000 g may need dosing adjustment of ampicillin from 150 mg/kg/d intravenous (IV) divided every 8 hours to 200 mg/kg/d IV divided every 6 hours.
Once-daily dosing of gentamicin (4 mg/kg IV qd) has been used in the term newborn for more than a decade. The pharmacodynamic characteristics of aminoglycosides that allow the use of once-daily dosing include concentration-dependent killing (peak concentration to minimal inhibitory concentration [peak/MIC] ratio), postantibiotic effect with leukocyte enhancement, and prevention of adaptive resistance.
Third-generation cephalosporins should be used judiciously. There is significant association between the use of third-generation cephalosporins and invasive candidiasis in preterm infants. Cefotaxime has excellent penetration to the cerebrospinal fluid and its therapeutic use should be limited to Gram-negative meningitis. Routine use of cefotaxime for EOS may lead to rapid development of drug-resistant organisms. Ceftriaxone is contraindicated in neonates for 2 reasons: (1) it is highly protein bound and may displace bilirubin progressing to hyperbilirubinemia and (2) concurrent administration with calcium-containing solutions may produce insoluble precipitates (ceftriaxone-calcium salts) leading to cardiorespiratory complications.
The AAP periodically updates the dosing regimens and recommended therapy for selected neonatal infections through Nelson’s Pediatric Antimicrobial Therapy . It provides various antimicrobial regimens (antibiotic, antiviral, and antifungal agents) based on body weight of infants and their chronologic age or gestational and postnatal age. Between new editions, a monthly update of short and interesting reports related to pediatric antimicrobial therapy is posted at www.aap.org/en-us/aap-store/Nelsons/Pages/Whats-New.aspx . Suggested durations of antibiotic therapy for EOS adapted from 2014 Nelson’s Pediatric Antimicrobial Therapy and the AAP Committee on Fetus and Newborn are shown in Table 2 .
