Joseph Mendelsohn

Case Conference, Aug 1, 2005

Fever in the Neonate: Late-Onset Group B Strep

Case: 3-week old infant with a fever.

Introduction:

Fever in the neonate represents a diagnostic dilemma as it may be the only sign of illness. However it must not be neglected as the underlying illness may be significant. A systematic approach to fever would be ideal—and many have been proposed—but a satisfactory one has not yet been agreed on.

Fever in the neonate is traditionally defined as a rectal temperature of 100.4° F or above. These must all be worked up with the full complement of cultures and labs (see below). A caregiver’s report of prior tactile fever in the absence of any other suspicious findings may usually be monitored without further investigation. However the clinician must ensure that the caregiver is capable of monitoring appropriately.

The number one cause of neonatal fever is viral. The most common viruses include HSV, Influenza, Enteroviruses, and RSV. In neonates, the incidence of bacterial etiology is about 7% and is most commonly found to be GBS and Gram negative enterics.

Workup of an infant (3 weeks) with fever

The goal is to identify those at risk for serious illness. Some clues include a “toxic appearing” baby with symptoms such as irritability, decreased perfusion, decreased tone, decreased activity or lethargy although many ill neonates can still lack these findings. The history can also give clues to etiology such as respiratory or GI symptoms, sick contacts, behavioral changes, maternal GBS status, or HSV exposure. At the same time, some symptoms with which we students are familiar will manifest only with advanced stages of disease or at varying ages. For example, a bulging anterior fontanel is not a good screening test for meningitis as it will only appear with significant disease. Similarly, nuchal rigidity in meningitis will only appear in about 27% of patients less than 6 months old.

Once the decision to work up fever in the neonate has been made, the following tests should be ordered:

-  CBC and Blood culture

-  Urinalysis and Urine Culture. Note that the neonatal urine culture should be obtained ideally via straight or suprapubic catheter to avoid contamination. I.e.: Bags are not a good idea as they are only useful if the urine is negative.

-  Stool WBC: If present, suggestive of Salmonella.

-  LP: While the incidence of findings is low, the high morbidity of meningitis is sufficient so as to convince most clinicians to order the test.

-  CXR: The current recommendation is to order one only in the presence of 1 or more symptoms of pulmonary disease. E.g. tachypnea.

Introduction to GBS infections

Colonization among pregnant women ranges from 15-40% according to the 2003 Red Book. About half of those will transmit GBS during pregnancy or delivery. GBS-colonized babies may develop severe illness including meningitis, pneumonia, and sepsis. This is why there are several prevention strategies for detection and treatment of pregnant GBS+ women prior to delivery.

GBS in the newborn is typically classified into early-onset and late-onset. The early-onset form occurs during the first week of life. It is this form that is susceptible to prophylaxis with prenatal screening and care during labor and delivery. Late-onset GBS can occur anywhere from about 1 week to 3 months of age with a median age of about 2 weeks.

Clinical characteristics of late-onset GBS infection

Early-onset GBS most commonly presents with sepsis, meningitis, apnea, or pneumonia. It carries a mortality of about 5%. These patients tend to be severely ill, most often with overwhelming sepsis, at the time of diagnosis. This is not the case with late-onset GBS where patients are not typically as sick at the time of diagnosis and have a lower mortality rate. About 37% have meningitis and about 45% have focal infections such as pneumonia, septic arthritis, adenitis, or cellulitis.

Late onset GBS commonly presents with a temperature greater than 38ºC. Other clinical findings include irritability, lethargy, poor feeding, tachypnea, grunting, and occasionally apnea. These infants are also less likely to present with severe shock than the early-onset patients.

Empiric Treatment of early or late-onset GBS infection

The initial empiric regimen for suspected GBS is IV ampicillin and an aminoglycoside or IV ampicillin and a 3rd generation cephalosporin. Once GBS has been positively identified, high dose penicillin G can be used. If IV access becomes difficult, IM therapy with ceftriaxone is sometimes used. Duration of therapy is generally 2-3 weeks.

Prognosis of late-onset GBS infection

About 25% of patients with late-onset GBS meningitis can suffer long-term problems such as hearing loss, vision loss, or learning disabilities. Factors that are more predictive of future problems include low birth weight, delayed hospitalization, leukopenia, seizures, focal neurological deficits, CSF protein >300mg/dl, and need for ventilation.

Prevention strategies in early-onset GBS infections

There are no current prevention strategies for late-onset GBS infections because the mechanism for transmission is not well understood. However, prevention for the early-onset form has been widely investigated and several strategies have been proposed including:

-  Screening all pregnant women for GBS and treatment if confirmed.

-  Treatment of all high-risk mother’s during labor. In this case, high-risk typically includes women who have not had prenatal care.

In 2002 ACOG (American College of Obstetricians and Gynecologists) and the CDC developed a consensus on prevention of early-onset GBS infection which can be found at http://www.cdc.gov/mmwr/preview/mmwrhtml/rr5111a1.htm. It is preferred to screen and treat all positive women than to treat those at high-risk; the latter strategy is now used only in women without prenatal screening.

Rochester Criteria

The Rochester Criteria is one of several proposed systematic mechanisms for determining which children may be considered as low-risk and can therefore be observed without the need for a full sepsis evaluation and administering empiric antimicrobial agents. Other proposed protocols include the Boston and Philadelphia protocols. This protocol was reviewed in Pediatrics in 1994 and was found to have a 98.9% negative predictive value for low-risk infants. Even though it has been studied in infants as young as 1 week, it is generally not applied unless the infant is >30 days of age. A summary of the protocol is given here:

  1. Criteria: Reassuring if all criteria are present
  2. Well appearing infant
  3. No skeletal, soft tissue, skin or ear infections
  4. Full term birth
  5. No prior illness
  6. No prior hospitalizations
  7. Not hospitalized longer than mother after delivery
  8. No prior antibiotics
  9. No Hyperbilirubinemia
  10. No chronic or underlying illness
  11. Complete Blood Count normal
  12. White Blood Cell count normal (5000 to 15,000/mm3)
  13. Band Neutrophils < 1,500/mm3
  14. Other Lab Findings
  15. If Diarrhea is present, Fecal Leukocytes <5 per hpf
  16. Urine White Blood Cells <10 per hpf