Pediatric Annals

Management of Neonatal Sepsis and Meningitis

Hanspeter Gnehm, MD; Jerome O Klein, MD

Abstract

Bacterial sepsis remains a significant cause of morbidity and mortality in the newborn infant. The incidence of sepsis in the neonate (defined as systemic signs of infection accompanied by bacteremia in the first month of life) varies from less than one to eight per 1000 live births. The incidence of meningitis is approximately one third the number of infants with sepsis. The mortality rate of neonatal sepsis varies between 10% and 30% and is higher if meningitis ensues. The continuing high mortality rate and significant sequelae in those infants who survive indicates that we must critically judge present management of sepsis in the newborn and consider new and possibly more effective methods. In this article we review the following features of management of systemic infection in the neonate; "early-onset" and ulate-onset" forms of diseases; bacteriology; diagnosis; use of antimicrobial agents; adjuncts to antimicrobial therapy; and sepsis in the infant recently discharged from the hospital.

Two patterns of disease, "early-onset" and "late-onset," have been associated with neonatal sepsis (Table I).1 "Early-onset" disease presents as a fulminant multisystemic illness during the first few days of life. These infants have a history of one or more significant obstetric complications, including premature rupture of maternal membranes, premature onset of labor, chorioamnionitis or peripartum maternal fever. Many of the infants are premature or of low birthweight. Bacteria responsible for "early-onset" disease are those present in the maternal birth canal. The mortality rate varies in recent series between 15% and 50%.

"Late-onset" disease may occur as early as five days of age but is usually recognized after the first week of life. Infants may have a history of obstetric complications but these are less characteristic than in "early-onset" disease. Bacteria responsible for "late-onset" sepsis and meningitis include those acquired from the maternal genital tract and organisms acquired after birth from human contact or from contaminated equipment or materials. The mortality rate is approximately 10% to 20%. Since different bacteria are responsible for the"early"and "lateonset" forms of neonatal sepsis, the choice of antimicrobial agents may differ.

MICROBIOLOGY

Escherichia coli and group B streptococcus are currently the bacterial pathogens most often responsible for sepsis and meningitis in the newborn infant but many bacterial species including other gram-negative enteric bacilli and other groups of streptococci, staphylococci and anaerobic bacteria are also occasional causes of invasive disease. Figure 1 indicates the experience from 1966 to 1978 at the Yale Medical Center. Group B streptococcus and E. coli were each responsible for 32% of 239 cases of neonatal sepsis.

Table

Appropriate management of the child with sepsis who has recently been discharged from the hospital nursery requires: careful history of infections in the household; careful physical examination for foci of infection such as otitis media; and collection of appropriate materials for culture including urine.

Since fever in this age group may indicate invasive bacterial disease with significant morbidity and mortality, a conservative approach to management, including early hospital iza ti o n and use of appropriate antimicrobial agents is warranted. For suspected sepsis without a focus, use of a penicillin and an aminoglycoside should be considered for term infants up to two weeks of age and for infants who were low birthweight or premature to one month of age. Chloramphenicol may be substituted for the aminoglycoside after these periods as H. influenzae becomes a more likely cause of sepsis than a gram negative enteric bacillus acquired at delivery or in the nursery.

1. Freeman RM, Ingram DL. Gross I, et al: A half century of neonatal sepsis at Yale. Am J Dis Child 1981: 135:140.

2. Bacterial meningitis and meningococcemJa -…

Bacterial sepsis remains a significant cause of morbidity and mortality in the newborn infant. The incidence of sepsis in the neonate (defined as systemic signs of infection accompanied by bacteremia in the first month of life) varies from less than one to eight per 1000 live births. The incidence of meningitis is approximately one third the number of infants with sepsis. The mortality rate of neonatal sepsis varies between 10% and 30% and is higher if meningitis ensues. The continuing high mortality rate and significant sequelae in those infants who survive indicates that we must critically judge present management of sepsis in the newborn and consider new and possibly more effective methods. In this article we review the following features of management of systemic infection in the neonate; "early-onset" and ulate-onset" forms of diseases; bacteriology; diagnosis; use of antimicrobial agents; adjuncts to antimicrobial therapy; and sepsis in the infant recently discharged from the hospital.

Two patterns of disease, "early-onset" and "late-onset," have been associated with neonatal sepsis (Table I).1 "Early-onset" disease presents as a fulminant multisystemic illness during the first few days of life. These infants have a history of one or more significant obstetric complications, including premature rupture of maternal membranes, premature onset of labor, chorioamnionitis or peripartum maternal fever. Many of the infants are premature or of low birthweight. Bacteria responsible for "early-onset" disease are those present in the maternal birth canal. The mortality rate varies in recent series between 15% and 50%.

"Late-onset" disease may occur as early as five days of age but is usually recognized after the first week of life. Infants may have a history of obstetric complications but these are less characteristic than in "early-onset" disease. Bacteria responsible for "late-onset" sepsis and meningitis include those acquired from the maternal genital tract and organisms acquired after birth from human contact or from contaminated equipment or materials. The mortality rate is approximately 10% to 20%. Since different bacteria are responsible for the"early"and "lateonset" forms of neonatal sepsis, the choice of antimicrobial agents may differ.

MICROBIOLOGY

Escherichia coli and group B streptococcus are currently the bacterial pathogens most often responsible for sepsis and meningitis in the newborn infant but many bacterial species including other gram-negative enteric bacilli and other groups of streptococci, staphylococci and anaerobic bacteria are also occasional causes of invasive disease. Figure 1 indicates the experience from 1966 to 1978 at the Yale Medical Center. Group B streptococcus and E. coli were each responsible for 32% of 239 cases of neonatal sepsis.

Table

TABLE 1CHARACTERISTICS OF EARLY AND LATE ONSET NEONATAL SEPSIS

TABLE 1

CHARACTERISTICS OF EARLY AND LATE ONSET NEONATAL SEPSIS

Organisms responsible for purulent meningitis are similar to those responsible for sepsis. Figure 2 presents data from a survey of I 6 newborn nurseries participating in a collaborative program in 1971 to 1973 under the direction of George McCracken, Jr., M. D. Again, E. coli and Group B streptococcus are the most frequently isolated organisms.

Although Neisseria meningiîidis, Hemophilus influenzas and Streptococcus pneumoniae are the most frequent causes of bacteremia and meningitis in infants and children, they are reJatively uncommon in newborns. However, bacteremia due to S. pneumoniae and H. influenzae appears to be increasing in incidence in newborn infants in recent years. A survey conducted by the Centers for Disease Control, United States Public Health Service2 indicates that meningitis due to each of these organisms is more common in the first month of life than any period after two years of age (Table 2).

DIAGNOSIS

The diagnosis of systemic infection in the newborn infant is difficult to establish on the basis of clinical findings alone. A history of one or more significant risk factors associated with pregnancy and delivery is usually present in "early-onset" disease. Signs of sepsis and meningitis are frequently subtle and nonspecific initially but may rapidly progress to more obvious signs including apnea, significant temperature elevation and seizures. Isolation of microorganisms from a significant source such as the blood, cerebrospinal fluid (CSF), urine or other body fluid or tissue remains the most valid method of diagnosing bacterial sepsis. Infectious agents cultured from the nose, throat, skin umbilicus or stool indicate colonization and may include organisms that cause sepsis, but isolation of a microorganism from these materials does not establish the presence of active systemic infection.

Figure 1. Bacteria causing neonatal sepsis at Yale-New Haven Hospital, 1966 to 1978 - percentage of 239 cases.'

Figure 1. Bacteria causing neonatal sepsis at Yale-New Haven Hospital, 1966 to 1978 - percentage of 239 cases.'

Figure 2. Bacteria associated with neonatal meningitis, 1971 to 1973 - percentage of 131 cases.

Figure 2. Bacteria associated with neonatal meningitis, 1971 to 1973 - percentage of 131 cases.

USE OF ANTIMICROBIAL AGENTS

If the history and clinical signs lead the physician to consider the infant septic, cultures of blood, urine, and CSF should be obtained and treatment withantimicrobial agents begun. Because of the subtlety of clinical manifestations of sepsis, the rapid progression of the disease and the continuing high mortality, treatment must be started even when only minimal indications of sepsis are present. Many infants are treated for sepsis but documentation of bacterial disease by positive cultures of blood, urine or CSF is accomplished in very few.

Initial Therapy

The choice of antimicrobial agents for treatment of suspected sepsis is based on knowledge of the prevalent organisms responsible for neonatal sepsis and the patterns of their antimicrobial susceptibility. Initial therapy of the infant who becomes septic during the first few days of life, "early-onset" disease, must include coverage for gram positive cocci, particularly group B streptococci and gram negative enteric bacilli. Treatment of the infant who becomes septic after day four, "lateonset** disease, must include coverage for hospitalacquired organisms. In some hospitals, coverage for Siaphylococcus aureus or gram-negative bacilli, such as Pseudomonas sp, or Serratia sp. must be considered. A combination of a penicillin such as penicillin G or ampicillin plus an aminoglycoside such as gen t a mi cm or tobramycin is now considered most suitable for initial therapy of "early-onset" infection. If there is a concern for staphylococcal disease, the initial therapy should include a penicillinase-resistant penicillin. The choice of antibiotics should be ree valúa ted when results of cultures and susceptibility tests become available.

Duration of Therapy

The duration of therapy depends on the initial response to the appropriate antibiotic but should be ten to U days in most infants with sepsis or minimal focal infection. The minimal duration of therapy for infants with meningitis due to group B streptococcus or gram negative enteric bacilli is 21 days.

Determination of Dosage Schedules

Dosage schedules to achieve concentrations of antibiotic in blood and body fluids that are effective and safe have been determined and are given in Table 3. Because maturation of renal function leads to more efficient excretion of penicillins and aminoglycosides, different dosage schedules are necessary for infants less than seven days old and seven to 28 days old. More drug given at shorter intervals is required Jn the older infants.

In some cases concentrations in serum of newborn infants are variable and unpredictable. Neonates who receive aminoglycosides or chloramphenicol require monitoring because of the wide range of concentrations of drug in serum. Careful monitoring is important, in particular, for very small infants (those under 1 500) or those who are very premature (less than 34 weeks gestation). Blood should be obtained to determine drug concentrations at the expected peak (one to two hours after intramuscular or intravenous administration) and trough (prior to the next dose). Specimens of blood should be obtained early in the course of therapy (within the first three days) to be certain that effective levels in serum are achieved and at subsequent intervals (every three to four days) to determine that the concentration of drug in serum is insufficient to cause toxicity. Effective concentrations that need to be achieved in blood and body fluids are determined by the susceptibility of the organisms to the drug (Tables 4 and 5). The upper limit of drug activity that must not be exceeded to avoid toxicity is uncertain and estimates are based on limited data (most of which were obtained in adults).

The desired concentrations for the aminoglycosides and chloramphenicol are:

Peak: gentamicin and tobramycin - 5 to 10 µg/ ml

kanamycin and amikacin - 15 to 25 µg/ ml

chloramphenicol - 15 to 25 µg/ ml

Trough: gentamicin and tobramycin - less than 2

µg/ ml

kanamycin and amikacin - less than 1 0 µg| ml

Treatment of the Infant Whose Bacterial Cultures are Negative

If bacterial cultures are negative, the infant appears well, and there is reason to believe that infection was an unlikely cause of the early signs, treatment can be discontinued. If the clinical condition of the infant remains precarious and there is still suspicion of an infectious process, therapy should be continued. Significant bacterial infection may be present without bacteremia, or bacteremìa may be present before but not at the time cultures of blood and body fluids were obtained.

Treatment of Meningitis

Since the pathogens responsiblefor neonatal meningitis are the same as those that cause neonatal sepsis, initial therapy is the same. In spite of availability of potent antimicrobial agents, the mortality due to meningitis ih the neonate remains high. Gram negative bacilli persist in CSF even when bactericidal titers of the antimicrobial agent are present. The continuing high mortality and persistence of organisms in CSF in neonatal meningitis stimulated studies of other modes of administration of antimicrobial agents. McCracken and colleagues1 studied the efficacy of parenteral gentamicin therapy compared with parenteral plus intrathecal gentamicin. Infants also received ampicîllin by vein. No significant difference was noted in mortality rates between parenteral therapy alone and intrathecal plus parenteral therapy. A second collaborative study under Dr. McCracken's direction investigated the value of intraventricular therapy for neonatal meningitis.4 Infants were randomly assigned tu receive either parenteral ampicillin plus gentamicin alone or in combination with intraventricular gentamicin. Infants in the group receiving intraventricular gentamicin had a higher mortality rate than did infants receiving systemic therapy only.

Table

TABLE 2AGE SPECIFIC INCIDENCE OF BACTERIAL MENINGITIS IN THE UNITED STATES, 1978*

TABLE 2

AGE SPECIFIC INCIDENCE OF BACTERIAL MENINGITIS IN THE UNITED STATES, 1978*

The optimal therapy at present for neonatal meningitis caused by gram negative enteric bacilli remains uncertain. U se of chloramphenicol may be of value if the organism is sensitive since high concentrations of the drug are achieved in the CSF. E. coli and selected gram negative enteric bacilli are very susceptible to the new cephalosporins, cefoperazone, cefotaxime and moxalactam (Table 5). Very high concentrations of these drugs are achieved in serum (more than 100 /xg/ml) and approximately 1 0% of the drug diffuses into the CS F. Therefore, more than enough drug would be expected at the site of infection in the CSF. Initial studies of treatment of infants with gram negative meningitis with moxalactam5 or cefotaxime6 suggest that the promise of the in vitro results will be borne out in clinical efficacy. At present, because of the failure of prior regimens and the preliminary data which suggest efficacy, the Committee on Infectious Diseases of the American Academy of Pediatrics suggests use of moxalactam for infants who have gram negative meningitis.7 The suggested dosage schedule is: infants less than seven days: 100 mg/kg/day in two doses; infants seven to 28 days: 150 mg/kg/day in three doses; infants more than 28 days: 200 mg/kg/day in four doses. An initial loading dose of 100 mg/kg is recommended.

ADJUNCTIVE THERAPIES FOR TREATMENT OF NEONATAL SEPSIS

Despite appropriate antimicrobial and optimal supportive therapy, mortality resulting from neonatal sepsis remains very high. With the hope of improving survival and decreasing the severity of sequelae in survivors, recent investigations have considered various methods of increasing the immunologie defenses of the newborn infant. These methods include granulocyte transfusion, exchange transfusion and the use of specific immune globulins. At present, these methods appear to promise substantial improvement over use of antimicrobial agents alone but must be considered investigational until studies of appropriate randomized control design are performed.

SEPSIS IN THE NEONATE RECENTLY DISCHARGED FROM THE HOSPITAL

When fever or other signs of systemic infection occur in the first weeks and after the newborn is discharged from the hospital, appropriate management requires consideration of the various sources of infection.

Infection acquired from a household contaci is the most likely event. Respiratory and gastrointestinal viruses may be transmitted from children or adults in the househoíd. Focal infections may occur: by three months of age almost 1 0% of infants have had at least one episode of otitis media (Klein J.O., Teele D.W., unpublished data). Invasive disease due to S. pneumoniae or H. influenzae occurs with significant frequency even during this early period of life (Table 2).

"Late-onset" disease due to an organism acquired at the time of delivery may present weeks to months later as sepsis or meningitis. Infection from group Bsïreptococcus is now the most frequent cause of "late-onset" sepsis in the neonate (Table 2). Neonatal disease due to Listeria monocytogenes resembles disease due to the streptococcus including early and Jate onset forms. The most common infection in this category is conjunctivitis and pneumonitis due to Chlamydia irachomatis. The organism is usually acquired at the time of delivery during passage through the birth canal. The syndrome includes a relatively well appearing child without fever, but with clinical and radiologie signs of pneumonitis, a pertussislike cough, and the presence of or history of conjunctivitis.

"Late-onset" disease may also be caused by organisms acquired in the nursery. Lesions of the skin and soft tissues or other focal infection due to S. aureus may occur in the second week or later. The organism was acquired in the nursery from hospital personnel or other infants but the disease is not evident for several weeks. In most areas of the United States, staphylococcal disease in newborn infants is not as significant now as it was during the pandemic of the late 1950s and 1960s but wemust remain alert for the return of virulent organisms that would cause disease in neonates.

Finally, infection in the neonate appearing after discharge from the nursery may be associated with an underlying anatomic or physiologic defect. The infant who fails to thrive or presents with fever may have infection of the urinary tract that is the first indication of a physiologic or anatomic abnormality. Other examples include infants with atresia of the lacrimal duct or choanae who may develop focal infections of the obstructed passages.

Table

TABLE 3ANTIBACTERIAL DRUGS FOR NEWBORN INFANTS

TABLE 3

ANTIBACTERIAL DRUGS FOR NEWBORN INFANTS

Table

TABLE 4ANTIMICROBIAL SUSCEPTIBILITY OF GRAM POSITIVE BACTERIA OF IMPORTANCE IN NEONATAL SEPSIS MEDIAN MIC (Mg/ml)

TABLE 4

ANTIMICROBIAL SUSCEPTIBILITY OF GRAM POSITIVE BACTERIA OF IMPORTANCE IN NEONATAL SEPSIS MEDIAN MIC (Mg/ml)

Table

TABLE 5ANTIMICROBIAL SUSCEPTIBILITY OF GRAM NEGATIVE ENTERIC BACILLI OF IMPORTANCE IN NEONATAL SEPSIS MEDIAN MIC (/ig/ml)

TABLE 5

ANTIMICROBIAL SUSCEPTIBILITY OF GRAM NEGATIVE ENTERIC BACILLI OF IMPORTANCE IN NEONATAL SEPSIS MEDIAN MIC (/ig/ml)

Appropriate management of the child with sepsis who has recently been discharged from the hospital nursery requires: careful history of infections in the household; careful physical examination for foci of infection such as otitis media; and collection of appropriate materials for culture including urine.

Since fever in this age group may indicate invasive bacterial disease with significant morbidity and mortality, a conservative approach to management, including early hospital iza ti o n and use of appropriate antimicrobial agents is warranted. For suspected sepsis without a focus, use of a penicillin and an aminoglycoside should be considered for term infants up to two weeks of age and for infants who were low birthweight or premature to one month of age. Chloramphenicol may be substituted for the aminoglycoside after these periods as H. influenzae becomes a more likely cause of sepsis than a gram negative enteric bacillus acquired at delivery or in the nursery.

REFERENCES

1. Freeman RM, Ingram DL. Gross I, et al: A half century of neonatal sepsis at Yale. Am J Dis Child 1981: 135:140.

2. Bacterial meningitis and meningococcemJa - United States. 1978. MMWR 1979; 28:277.

3. McC racken GH Jr. Mize SG: A controlled study of intrathecal antibiotic therapy in gram-negative enteric meningitis of infancy. Report of ? he Neonatal Meningitis Cooperative Study Group. J Pediatr 1976; 89:66.

4. McCracken GH Jr. Mize SG. ThrelkeM N: Imraventricular gentamicin therapy in gram-negative ba ciliary meningitis of infancy. Lancet 1980: 1:787.

5. Schaad UB. McCracken G H Jr. Threlkeld N. el al: Clinical evaluation of a new broad -spectrum o xa- be t a- lac t a m antibiotic, mo ? a lac tarn, in neon al es and infants. J Pediatr 1981:98:129.

6. Kafet/is DA. Brater DC. Kapiki AN. et at: Treatment of severe neonatal infections with cefotaxime. Efficacy and pharmacokineiics. J Pediair 1982; 100:483.

7. Report of the Committee on Infectious Diseases, ed 19. Evansion. Illinois. American Academy of Pediatrics. 1982, p 78.

TABLE 1

CHARACTERISTICS OF EARLY AND LATE ONSET NEONATAL SEPSIS

TABLE 2

AGE SPECIFIC INCIDENCE OF BACTERIAL MENINGITIS IN THE UNITED STATES, 1978*

TABLE 3

ANTIBACTERIAL DRUGS FOR NEWBORN INFANTS

TABLE 4

ANTIMICROBIAL SUSCEPTIBILITY OF GRAM POSITIVE BACTERIA OF IMPORTANCE IN NEONATAL SEPSIS MEDIAN MIC (Mg/ml)

TABLE 5

ANTIMICROBIAL SUSCEPTIBILITY OF GRAM NEGATIVE ENTERIC BACILLI OF IMPORTANCE IN NEONATAL SEPSIS MEDIAN MIC (/ig/ml)

10.3928/0090-4481-19830301-03

Sign up to receive

Journal E-contents