Postnatal Bacterial Infections

Epidemiology and Microbiology

Neonatal sepsis is categorized according to the infant's postnatal age at onset of disease. Although the definitions for early-onset and late-onset sepsis vary slightly, most reports define early-onset sepsis as that occurring at or before 72 hours of life and late onset occurring at greater than 72 hours to 7 days. The categories are meant to reflect the different etiologies and pathophysiologic changes associated with timing of disease onset.

In early-onset sepsis, infants acquire infection by vertical transmission either through ascending amniotic fluid infection or through acquisition of bacterial flora from the mother's anogenital tract during vaginal delivery. Examples of bacteria that cross the placenta to cause fetal infection are Treponema pallidum and Listeria monocytogenes . Over the past decade, our understanding of the maternal vaginal and infant microbiome has substantially increased. Exposure of the neonate to maternal vaginal microbiota during delivery is essential for healthy development during the newborn period, providing the primary source for normal gut colonization, initiating host immune maturation, and metabolism. However, these same bacteria are responsible for early-onset sepsis, including group B streptococci (GBS) and Gram-negative enteric bacilli. The development and makeup of the infant microbiome can be altered by cesarean section (C-section), perinatal antibiotics, and formula feeding.

Factors that increase the risk for infection can be divided into those that are intrapartum and those related to the infant after delivery. Maternal intrapartum conditions that increase the risk of infection include maternal GBS colonization, maternal fever, chorioamnionitis, prolonged rupture of membranes (>18 hours), and inadequate intrapartum antibiotic administration before delivery. Newborn conditions that increase the risk of infection include the degree of prematurity and lower birth weight.

The epidemiology of neonatal sepsis has changed significantly over the past several decades and continues to change. Universal antenatal screening for GBS colonization with intrapartum antibiotic prophylaxis for women colonized with GBS has significantly reduced the rate of early-onset GBS sepsis. The overall incidence of early-onset sepsis in the United States is estimated to be 0.77 per 1000 live births, with the highest rates occurring among preterm, low birth weight, and African American infants. For those infants with a birth weight less than 1500 g, the rate is 10.96 per 1000 live births, and for those with a birth weight between 1500 and 2500 g, the rate is 1.38. Mortality rates are also inversely proportional to gestational age. Although the case fatality rate among full-term infants has dramatically decreased over the past several decades to 2.5%-3%, the mortality rates for early preterm infants remain high between 30%-54%. The most common causative bacteria are GBS and Escherichia coli , accounting for 38% and 24% of early-onset cases, respectively. Staphylococcus aureus , viridans group Streptococci, Enterococci, group A Streptococci, Listeria monocytogenes , Haemophilus, and other enteric Gram-negative organisms, including Klebsiella , Enterobacter , Citrobacter , Acinetobacter , and Pseudomonas , are also known pathogens associated with early-onset neonatal sepsis. Among preterm infants, E. coli was the most common infection with the highest case fatality ratio (32.1%).

In late-onset sepsis, acquisition of infection is predominantly through the infant's environment. The infant becomes colonized with pathogenic bacteria that are ubiquitous in their physical environments, including part of the flora of their caregivers. The gut microbiome is often involved in the pathogenesis of late-onset sepsis. While the agent can vary, the causative organism is usually found to be abundant in the infant's gastrointestinal tract. Extreme prematurity is one of the greatest risk factors for late-onset sepsis. A 2010 National Institute of Child Health and Human Development Neonatal Research Network study of morbidity and mortality rates for extremely preterm infants showed that late-onset sepsis is a frequent complication for these patients. Among 9575 infants born between 22 and 28 weeks’ gestation, 36% developed late-onset sepsis. As compared with term infants, premature infants have more impaired innate and adaptive immune function and thus are more susceptible to invasive infections.

As overall care and survival rates for premature and low birth weight infants continue to improve, long hospital stays and indwelling vascular catheters provide additional infectious risk factors for these infants. Common causes of late-onset infection in this population of infants include coagulase-negative staphylococci and S. aureus , as well as invasive candidiasis. Group B streptococcus and E. coli are also commonly implicated in late-onset sepsis. Escherichia coli is frequently a cause of urosepsis in young infants. Although screening for maternal GBS colonization and intrapartum antibiotic prophylaxis have markedly decreased the incidence of early-onset GBS, it has not reduced the incidence of late-onset GBS disease in either term or preterm infants. Similarly, improved surgical care and survival for neonates with congenital heart disease have led to prolonged hospitalizations with associated risk for sepsis. Coagulase-negative staphylococci, S. aureus , E. coli , and candidiasis are common causes of bloodstream infection in this population, which has been associated with an increased mortality risk compared with uninfected infants with congenital heart disease. The use of histamine-2 blockers and proton pump inhibitors, as well as gastrointestinal tract pathology, have been associated with an increased risk for late-onset Gram-negative bloodstream infections. Numerous efforts are under way to decrease the incidence of late-onset sepsis, including comprehensive catheter-care bundles and aggressive enteral feeding programs with early line removal.

Clinical Presentation

The clinical signs of sepsis in a neonate are varied and often nonspecific. Detection requires that clinicians maintain a high index of suspicion. Familiarity with epidemiologic risk factors is crucial to determining the threshold index of suspicion. For early-onset infection, consider any perinatal risk factors that may be present, including maternal GBS status, chorioamnionitis, prolonged rupture of membranes, and gestational age. For late-onset infection, consider whether the patient has indwelling foreign bodies such as a central venous catheter or endotracheal tube, is dependent on parenteral nutrition, or receives proton-pump inhibitor or histamine-2 blocking therapy.

Clinical signs and symptoms are variable and nonspecific and can reflect noninfectious etiologies. These include, but are not limited to, findings of hyper- or hypothermia, lethargy or irritability, hypotonia, respiratory distress, cyanosis, apnea, feeding difficulties, poor perfusion, bleeding problems, and abdominal distention. Lethargy or poor feeding may be the only symptoms initially. Metabolic changes may include hyperglycemia or hypoglycemia, acidosis, and jaundice. Meningismus is uncommon in neonates with central nervous system (CNS) infection. Full fontanelle, irritability, lethargy, and seizures may occur. Noninfectious etiologies that can mimic newborn sepsis are numerous and include respiratory distress syndrome, cardiogenic pulmonary edema, metabolic acidosis, and meconium aspiration syndrome. Lethargy, irritability, and seizures can also occur secondary to electrolyte, endocrine, or metabolic disturbances.

Diagnosis

Because signs and symptoms are nonspecific, the clinical diagnosis of neonatal sepsis is extremely challenging. A definitive diagnosis requires the isolation of a pathogen from a normally sterile body site, including blood, cerebrospinal fluid (CSF), and urine. Isolation of bacteria from blood is considered the gold standard for the diagnosis of sepsis. A blood culture should be drawn in any infant with suspected sepsis. A collection volume of at least 1 mL is recommended for improved recovery of microorganisms in culture, particularly for those patients with low colony count bacteremia. A lumbar puncture should also be considered in any infant with suspected sepsis. (See Table 48.1 for normal values for CSF cell counts, protein, and glucose according to gestational age, postnatal age, and birth weight.) If the infant is critically ill with respiratory or hemodynamic instability, the procedure can be deferred until the patient is more stable. Studies have shown poor correlation between results of blood and CSF cultures, so blood culture results alone should not be used to determine which patients should receive a lumbar puncture. In infants with bacteremia, the incidence of meningitis has been shown to be as high as 23%. A study among very low birth weight infants with meningitis showed that one-third of them had negative blood cultures.

For suspected early-onset sepsis, a urine culture is not part of the recommended work-up. In newborns, urinary tract infections are primarily caused by renal seeding during bacteremia, and thus urine cultures are of low yield in early-onset sepsis. In older infants, urinary tract infections increasingly result from ascending infection, so urine cultures should be part of the evaluation of late-onset sepsis. Suprapubic aspiration or sterile catheterization are the preferred procedures to obtain cultures.

To date, no single blood cell index has been shown to be sensitive enough to safely exclude sepsis. Numerous indirect markers of infection have been studied. Although none of these tests can definitively confirm or exclude infection, they can be used to help identify infected infants and guide decisions on duration of antimicrobial therapy. A peripheral white blood cell (WBC) count with differential is commonly obtained with sepsis evaluations. It is recommended to wait 6-12 hours after birth before obtaining a WBC with differential, as later counts are more likely to reflect a pathologic inflammatory response compared with those obtained at birth. Studies have shown leukopenia and a high percentage of immature to total white blood cells (>0.2) were associated with early-onset sepsis. Late-onset sepsis has been associated with both high and low WBC, high absolute neutrophil count, and high percentage of immature to total white blood cells. Newer technology allows automated calculation of the immature granulocyte, which accurately identifies sepsis.

C-reactive protein (CRP) and procalcitonin are two acute-phase reactants that have been studied extensively in neonatal sepsis. C-reactive protein levels increase within 6-8 hours after infection and peak after 24 hours. The sensitivity of CRP for neonatal sepsis is lowest early in infection and then increases over the next 10-12 hours after the onset of infection. Serial determinations may be useful for identifying infants who do not have a bacterial infection or in monitoring response to treatment for infected infants. The specificity and positive predictive value of CRP range from 93%-100%. However, it is worth noting that preterm infants have lower CRP values and response which affect its predictive value.

Procalcitonin concentrations peak as early as 6-8 hours following infection, although there is a physiologic increase within the first 24 hours after birth. Also, unlike CRP, the serum procalcitonin concentration is not affected by gestational age. Increased levels can be seen with noninfectious causes such as respiratory distress syndrome. Thus, procalcitonin appears to have better sensitivity but less specificity than CRP for identifying neonatal sepsis. Given that neither CRP nor procalcitonin values have been shown to be entirely sensitive or specific, the most important information guiding clinical decisions continues to be the patient's overall clinical status and culture data.

Specific cytokines have also been evaluated for potential roles in the diagnosis of neonatal sepsis. Serum concentrations of interleukin (IL)-1β, IL-6, IL-8, tumor necrosis factor-α, interleukin-2 soluble receptor (sIL2R), and granulocyte colony-stimulating factor all increase in newborns with bacterial infection. Several have been shown to rise early in the course of infection, often before clinical findings become apparent. Newer technologies using cytometry to accurately identify immature granulocytes have also shown to be as effective, with faster turnaround time and less operator variability than manual differentials. While cytokine and cytometry assays may be helpful in diagnosis or in guiding decisions regarding therapy, their use is not widely employed related to issues of cost, the need for specialized equipment, and the processing time. These tests hold promise for the future.

Real time online/mobile applications are now available that predict the probability of early-onset sepsis based on objective maternal risk factors available at the time of birth. In early studies, these algorithms perform better than those based on risk-factor threshold values. These have allowed improved recognition as well as reduced use of unnecessary antibiotics by 50% and separation of mother and infant.

Management

Ampicillin and an aminoglycoside are recommended as empiric therapy for early-onset sepsis. This combination provides coverage against GBS and E. coli , the most common causative pathogens, as well as Listeria monocytogenes . Empiric use of third-generation cephalosporins is not recommended because of concerns for development of resistance and the increased risk for invasive candidiasis with prolonged administration. However, if Gram-negative meningitis is suspected, then it is recommended to add cefotaxime to the empiric regimen, given its excellent CNS penetration. Cefotaxime is preferred over ceftriaxone for use in neonates because of ceftriaxone's potential to displace bilirubin and increase the risk for kernicterus and its association with biliary sludging. However, recent shortages of cefotaxime have required the use of alternative agents in infants with invasive Gram-negative disease, including ceftriaxone (in term infants with normal bilirubin) or cefipime.

Empiric therapy for late-onset sepsis usually consists of ampicillin and an aminoglycoside or cefotaxime, providing coverage for common pathogens like group B streptococci and Gram-negative organisms. As with early-onset disease, if Gram-negative meningitis is suspected, consider adding a third-generation cephalosporin. In preterm infants or infants with complex medical problems necessitating surgeries or prolonged indwelling central catheters, vancomycin should be given in lieu of ampicillin if suspected of having coagulase-negative Staphylococci or S. aureus disease. Carbapenems may be considered depending on local resistance patterns or if the patient had previously received therapy with a third-generation cephalosporin. It is also important to consider empiric antifungal coverage for invasive candidiasis for severe sepsis unresponsive to antibiotics, infants on chronic TPN, or those with unexplained cytopenias. Experience with fungal biomarkers including beta-D-glucan assays for the diagnosis of candidemia in infants is limited.

Once a pathogen is identified, therapy should be tailored to the species and antimicrobial susceptibilities. Duration will be determined by the site of infection and the patient's clinical response. There is little evidence from randomized, controlled trials on the appropriate duration of treatment for culture-proven sepsis, especially in preterm and low birth weight infants. Bacteremia without a focus of infection is usually treated for 10 days. It is reasonable to expect that early preterm infants (<32 weeks’ gestational age) may require slightly longer treatment courses of 10-14 days. In addition, Gram-negative bacteremia tends to be treated with longer courses of 10-14 days. In general, uncomplicated GBS meningitis is treated for 14-21 days. Longer courses are needed for other focal complications of GBS infection (see Specific Pathogens ). For Gram-negative bacterial meningitis, treatment is for 21 days, or 2 weeks beyond the first negative CSF culture, whichever is longer. The use of systemic ciproflocaxin is indicated in those infants with multidrug-resistant Gram-negative disease when alternative antibiotics are not available.

It is often difficult to determine an appropriate duration of antibiotic therapy for suspected sepsis when cultures are negative. In well-appearing infants without clinical or hematologic evidence for infection, standard practice is to discontinue antibiotics if cultures have been negative after 36-48 hours. Management decisions are much more challenging for those infants in whom sepsis is highly suspected but cultures are negative, which is often the case for preterm infants. Infants whose mothers received antibiotics during labor may have false negative blood cultures because of antibiotic suppression. Cerebrospinal fluid culture data may be lacking in infants who are not clinically stable enough to tolerate a lumbar puncture. Noninfectious conditions mimicking sepsis can also complicate the clinical picture. Still, studies have shown potential harm associated with longer duration (>5 days) of empiric antibiotics, including increased risk for necrotizing enterocolitis, candidemia, and mortality among premature infants. There is also concern that early antibiotic exposure may permanently alter the microbiome and predispose to obesity. There is little available evidence on when it is safe to discontinue antimicrobials in such cases, and clinicians must consider each patient's clinical course as well as the risks associated with longer courses of antibiotics. The CDC has led efforts to decrease antibiotic exposure in a campaign entitled “Choosing Wisely.” These efforts have championed use of sepsis risk calculators and early stopping rules.

Prevention

As mentioned, universal antenatal screening for GBS colonization and intrapartum antibiotic prophylaxis for the prevention of early onset GBS disease has been highly successful, although it has not had an effect on the incidence of late-onset GBS. Details of the current recommendations for antenatal GBS screening and intrapartum antibiotic prophylaxis are included in a later section. Preventive efforts to reduce the risk of late-onset sepsis have focused on infection control in the postnatal environment. Successful measures include hand hygiene, proper management of central venous catheters, appropriate use of antibiotics, and limited use of histamine-2 blockers and proton pump inhibitors. A Cochrane review on oral lactoferrin prophylaxis in preterm infants showed a reduced incidence of late-onset sepsis in infants weighing less than 1500 g. The roles of exclusive maternal milk feeding and probiotic prophylaxis in prevention of late-onset sepsis require further study.

Meningitis

Bacterial meningitis is more common in the first month of life than at any other age. Incidence of neonatal bacterial meningitis varies across the globe. While the incidence is estimated to be 0.3 per 1000 live births in developed countries, neonatal meningitis is much higher (0.8-6.1 per 1000 live births) in developing nations. As with neonatal sepsis, neonatal meningitis can be categorized into two patterns of disease: early onset and late onset. Infants with early-onset meningitis present within the first week of life, usually within 72 hours of birth. These infections are vertically transmitted and are associated with the complications of labor and delivery. Late-onset disease occurs after the first week of life and reflects community or nosocomial transmission. Among infants with bacteremia, as many as 25% of them will also have meningitis.

The causative pathogens for neonatal meningitis are similar to those for neonatal sepsis with group B streptococcus and E. coli predominant. Group B streptococcus is the most common cause of neonatal meningitis and occurs in up to 40% of infants with early onset meningitis. Gram-negative enteric bacilli cause 30%-40% of cases of neonatal meningitis, and E. coli accounts for approximately 50% of the Gram-negative isolates. E. coli is now recognized as the most common cause of early-onset meningitis among very low birth weight (<1500 g birth weight) infants. The majority of E. coli strains causing meningitis contain the K1 polysaccharide capsular antigen, which assists the organism in evading host defenses. Other important Gram-negative organisms include Klebsiella , Enterobacter , Citrobacter , and Serratia species. A known complication of Citrobacter koseri and Enterobacter sakazakii meningitis in neonates and young infants is the formation of brain abscesses, so it is important to obtain brain imaging whenever these species are isolated from the CSF. Although Listeria monocytogenes is a relatively uncommon cause of neonatal meningitis, it can lead to significant morbidity and mortality because of its association with rhombencephalitis. It has been estimated to cause 5%-20% of neonatal meningitis cases. Nosocomial pathogens include coagulase-negative Staphylococci , Candida , and resistant Gram-negative organisms, particularly Pseudomonas , are more prominent in late-onset meningitis.

The clinical presentation of neonatal meningitis is often nonspecific and is similar to that of neonatal sepsis. The most common finding is temperature instability, which occurs in approximately 60% of infants with meningitis. Common neurologic signs include irritability, emesis, lethargy, poor tone, and seizures. Most infants present with a full, but not bulging, fontanelle without meningeal signs. Other signs include poor feeding, respiratory distress, apnea, and diarrhea.

Neonates with suspected bacterial meningitis should undergo a complete sepsis evaluation—including blood culture, complete blood count with differential, urine culture if greater than 6 days of age, and lumbar puncture for CSF Gram stain and culture—cell count with differential, glucose, and protein in order of priority. For clinically stable infants, the lumbar puncture should be performed before administration of antibiotics.

Cerebrospinal fluid studies need to be interpreted based on the infant's gestational age, postnatal age, and birth weight (see Table 48.1 ). A CSF WBC count of greater than 20-30 cells/µL is consistent with meningeal inflammation. However, neonatal meningitis has been shown to occur with normal CSF parameters, and there is some overlap in CSF WBC values between neonates with and without meningitis. If CSF findings are not definitive in an infant whose clinical picture is otherwise suspicious for meningitis, a repeat lumbar puncture obtained 24-48 hours later will still show pleocytosis if true meningeal inflammation is present even with use of antibiotics. Cerebrospinal fluid protein and glucose values are highly variable in neonates with and without meningitis. Cerebrospinal fluid Gram stain may be helpful in providing an early presumptive etiologic diagnosis, although a negative Gram stain certainly does not exclude the diagnosis. If the lumbar puncture is traumatic, it is not recommended to adjust the CSF white blood cell count based on the red blood cell counts, as this does not improve the diagnostic utility for neonates. Ultrasound to guide lumbar puncture may reduce the incidence of traumatic taps. Infants who have had a traumatic lumbar puncture may be treated presumptively for meningitis pending CSF culture results. Increasingly, PCR against common newborn pathogens has been utilized as a diagnostic in situations where the CSF is uninterpretable. Automated systems are now commercially available and may reduce antibiotic exposure.

Empiric therapy for early-onset meningitis includes ampicillin and an aminoglycoside. If infection with a Gram-negative organism is suspected, the regimen may be expanded to include cefotaxime or higher generation cephalosporin in addition to ampicillin. Aminoglycosides are not used as monotherapy because of their poor CNS penetration. Once a pathogen is identified, therapy should be tailored according to the causative organism.

Group B streptococcus meningitis is treated initially with ampicillin (or penicillin) plus an aminoglycoside. Group B streptococcus has thus far been shown to be uniformly susceptible to penicillins while resistance to clindamycin and erythromycin is reported. Combination therapy is used because synergy is seen in vitro and animal studies have shown improved outcomes with combination therapy versus penicillin alone. When the patient shows clinical improvement and the CSF is sterilized, therapy can be narrowed to ampicillin or penicillin monotherapy to continue for 14 days after the first negative culture. Although not all patients with GBS meningitis require a lumbar puncture at the end of therapy to confirm treatment response, it should be considered in patients with a complicated clinical course, including seizures, abnormal neuroimaging, prolonged positive CSF cultures, or slow clinical response to therapy.

Gram-negative meningitis is treated with a third-generation cephalosporin (cefotaxime) for at least 21 days or for 14 days after the first negative CSF culture. An aminoglycoside is added until CSF sterilization, which usually takes longer for meningitis caused by Gram-negative organisms than for GBS meningitis. A fourth-generation cephalosporin (cefepime) or a carbapenem (meropenem) in combination with an aminoglycoside may be considered for infection with members of the Enterobacteriaceae family with inducible beta-lactamase resistance (e.g., Citrobacter , Enterobacter , Serratia ) and for Pseudomonas. Studies have shown that recommended meningitic doses of meropenem may be toxic at lower gestational ages and may produce seizures. A lumbar puncture may be considered before discontinuation of antibiotics to confirm response to treatment.

Listeria monocytogenes is not susceptible to cephalosporins and should be treated with ampicillin and an aminoglycoside until CSF sterilization, followed by ampicillin monotherapy for 14 days after first negative culture. The same therapy is recommended for infection with Enterococcus. Vancomycin should be avoided, if able, in patients with infection by Listeria monocytogenes as treatment failures are reported. Meningitis with coagulase-negative Staphylococci is seen in preterm infants and is often associated with the presence of a foreign body in the CNS. Most of these organisms are resistant to penicillin, and treatment with vancomycin is often required. Duration is generally 14-21 days after CSF sterilization, with removal of any foreign body if feasible. Adjunctive use of corticosteroids has not shown improvement in the outcome for neonatal bacterial meningitis other than tuberculosis meningitis.

Mortality from neonatal meningitis has decreased dramatically over the past several decades, although substantial neurologic morbidity continues to be seen among affected patients. Mortality rates were estimated at almost 50% in the 1970s and have decreased to current estimates of 10%-15%, with higher mortality rates among preterm infants and those in developing countries. Neurologic sequelae include developmental delay, seizures, hydrocephalus, cerebral palsy, blindness, and hearing loss. Studies estimate that among survivors, 21%-38% will have mild deficits and 24%-29% will have severe neurologic sequelae. Predictors of poor neurologic outcomes include seizures lasting >72 hours, presence of coma, hypotension requiring the use of inotropes, leukopenia (<5.0), and abnormal electroencephalogram findings.

Pneumonia

Pneumonia is a significant cause of morbidity and mortality in neonates, especially in developing countries where it is a leading cause of death for all children under 5 years of age. The incidence of lower respiratory tract infection in developed countries is estimated at less than 1% among full-term infants, but may be as high as 10% in low birth weight infants. In developed countries, the morbidity and mortality from neonatal pneumonia depend largely on the gestational age of the patient, severity of disease, and underlying medical conditions, especially chronic lung disease. Neonatal pneumonia is also categorized into two patterns of disease according to timing and route of acquisition. Early-onset pneumonia is usually acquired within the first 3 days of life via vertical transmission, including aspiration of infected amniotic fluid and transplacental transmission. Late-onset pneumonia occurs after the first week of life, and infection arises from pathogenic organisms in the infant's environment. The risk for late-onset pneumonia is highest among infants who require mechanical ventilation. Other risk factors include extreme prematurity, prolonged hospitalization, and previous bloodstream infection.

The most common cause of early-onset pneumonia in developed countries is GBS. Other common causes include S. pneumoniae , nontypable H. influenzae , S. aureus , E. coli , Klebsiella , and atypical organisms. Ureaplasma urealyticum has been potentially linked to the development of chronic lung disease in colonized infants. However, the significance of this association is unknown, and the efficacy of antimicrobial therapy for those colonized infants is also uncertain. Chlamydia trachomatis pneumonia can occur in the first week of life but more typically presents between 2 and 4 weeks of age, given its long incubation. Chlamydia trachomatis pneumonia and ophthalmia neonatorum are discussed in more detail elsewhere. While syphilis, Listeria monocytogenes, and Mycobacterium tuberculosis can be transmitted across the placenta, these pathogens are uncommon causes of neonatal pneumonia.

Definitive culture data are often lacking for cases of late-onset, community-acquired neonatal pneumonia. Streptococcus pneumoniae is considered a predominant causative pathogen in this population. Other important pathogens include S. aureus , S. pyogenes , nontypable H. influenza , and Gram-negative enteric organisms. Staphylococcus aureus , Streptococci , Klebsiella pneumoniae , Citrobacter , Enterobacter , Serratia , and Pseudomonas have all been shown to have the potential to cause extensive lung injury, including abscess formation, empyema, and pneumatoceles. Although it generally causes self-limiting illness in older children and adults, Bordetella pertussis infection in young infants can lead to respiratory failure and death. The Centers for Disease Control and Prevention (CDC) updated their vaccination guidelines in January 2013 to include a recommendation that all pregnant women get a dose of the tetanus, diphtheria, and pertussis (Tdap) vaccine during each pregnancy. Receipt of Tdap vaccine during pregnancy allows for increased maternal pertussis antibody transfer to the neonate. This provides added protection for infants from birth to the time of infant pertussis vaccination at 2 months.

The clinical presentation of early-onset pneumonia often includes respiratory distress within the first few hours of life, with tachypnea, retractions, nasal flaring, or grunting. Other associated signs are apnea, lethargy, poor feeding, temperature instability, abdominal distention, poor perfusion, and metabolic acidosis. None of these signs is specific for pneumonia, and any infant with these symptoms should undergo a complete sepsis evaluation. Copious or purulent tracheal secretions may also be present.

The diagnosis of neonatal pneumonia can be difficult. The differential diagnosis is broad and includes many noninfectious causes such as transient tachypnea of the newborn, respiratory distress syndrome, meconium aspiration, pulmonary hemorrhage, pneumothorax, hypoglycemia, and metabolic acidosis. Anatomic abnormalities that can cause respiratory distress include primary pulmonary hypoplasia, tracheoesophageal fistula, choanal atresia, congenital diaphragmatic hernia, and congenital heart disease, among many others. Causative pathogens are rarely obtained and identified from the lower respiratory tract. Blood and CSF cultures should be obtained from any neonate with suspected pneumonia. Chest radiographs can be helpful in making the clinical diagnosis, although there is significant radiologic overlap between pneumonia and other respiratory disorders of the newborn. In pneumonia, radiographs often reveal bilateral alveolar densities with air bronchograms, or irregular, patchy pulmonary infiltrates. Pleural effusions are seen in up to two-thirds of pneumonia cases and are almost never seen in uncomplicated respiratory distress syndrome. The chest radiograph may appear normal in up to 15% of pneumonia cases. Tracheal aspirate Gram stain and culture results must be interpreted with caution. These specimens may be of some value if obtained immediately following intubation. If an infant has been intubated for several days, the endotracheal tube will invariably become colonized and will be of little value in the evaluation for sepsis.

Like most early-onset newborn infections, empiric antibiotic treatment includes ampicillin and an aminoglycoside. Therapy can then be tailored to susceptibility results if a causative organism is identified. Empiric antibiotic therapy for late-onset pneumonia will depend on local bacterial resistance patterns in both the hospital and the community. Vancomycin and a cephalosporin, often cefotaxime, are commonly used as empiric therapy to provide coverage against newborn pathogens as well as coagulase-negative Staphylococci and methicillin-resistant S. aureus (MRSA) . If Pseudomonas is suspected, an aminoglycoside plus an anti-pseudomonal beta-lactam, such as ceftazidime, cefipime, or piperacillin-tazobactam, should be given. Recommended duration of therapy for uncomplicated pneumonia is 7-10 days.