Pediatric emergency medicine physician Douglas Baker recalled a recent case of a 44-day-old infant who was not feeding well and felt warm to her mother. The mom reported a rectal temperature of 101.0, which concerned her pediatrician. But when Baker saw her in the ED, the infant’s temperature had returned to normal levels, her vital signs were good and she looked well, too. Also, there was nothing remarkable in her medical history. So, Baker asked residents at a recent Hopkins Children’s Grand Rounds, “Now what do you do?”
The laughter that followed a moment of silence spoke volumes about the subject. Do you observe and send the patient home? Test and treat? Why is there such a huge variability in the management of fever in young infants, or those under 2 months of age? As Baker asked, “Why has this issue generated decades of investigation?”
Fever in young infants is more of a concern because they’re less immune-competent – their immune system to fight infectious pathogens is not fully developed – and less socially responsive than older infants. It’s not uncommon to see group B strep or gram negative bacteria in this age range, Baker noted, as well as the occasional Listeria infection and herpes simplex virus.
“There are plenty of pathogens that affect young infants,” Baker said, “and about 1 in 10 have some kind of bacterial disease.”
So, what is fever and how is it managed? Most pediatricians, Baker said, consider any rectal temperature – the most reliable method of measuring body temperature – above 38.0 centigrade a fever. In 1993 – whether you were a pediatrician, family practitioner or emergency medicine physician – fever tended to be managed very conservatively in young infants with full workups, hospital admission and antibiotics. By 1998, however, these infants were increasingly being treated as outpatients. “The trend was more home management,” Baker said.
Speed dial to 2004, Baker added, and some pediatricians were practicing a more liberal approach to management of fever in young infants, even those under the age of one month.
“In one study of pediatricians practicing in non-urban settings, fewer than half of those infants were getting a full workup and admission for fever and treatment with antibiotics,” Baker said, citing data from a prospective cohort study by the Pediatric Research in Office Settings, or PROS, network of the American Academy of Pediatrics (JAMA 2004;291:1203-1212).
Does well appearance of an infant with fever rule out a serious bacterial illness? Citing his own experience and data from his controlled 5-year fever-management study (New England Journal of Medicine 1993;329:1437-1441), Baker said “The unequivocal answer to that is ‘no.’”
Can you rely on the physical exam to screen for bacterial illness in young infants with fever? Baker noted that in his 1993 controlled study, serious bacterial illness was diagnosed in 65 infants, and of those 64 had been identified by the screening criteria for inpatient care and antibiotic treatment. Of the 287 infants assigned to observation and no antibiotics, 286 did not have serious bacterial illness. The “Philadelphia screening tool” that Baker developed has been used as a standard of care in many institutions since it was published almost two decades ago.
“We had one failure in this system, an infant with bacteremia in the in-patient group not initially treated with antibiotics,” Baker said. “I believe in this system and practice with this tool today. I know that it works for 1- to 2-month-old infants with fever.”
But not necessarily for infants under 1 month of age. Of 109 of these very young infants categorized as low-risk for bacterial disease in one study, Baker noted, five were found to have bacterial illnesses (Archives of Pediatrics & Adolescent Medicine 1999;153:508-511).
“Up to 4 percent of these infants considered to be at low risk for bacterial disease were misclassified as safe for management as outpatients without antibiotics,” said Baker. “These children under one month of age are not reliable, and they will take every opportunity to outsmart you no matter how smart you think you are.”
Is a chest X-ray required? Studies indicate X-ray is not needed if there are no signs or symptoms of lower respiratory disease in an infant under 2 months of age with fever. But, what does it mean to be free of signs of lower airway disease? For infants under 12 weeks of age, Baker said, it means a respiratory rate within a normally acceptable range, no crackles, wheezes, rhonchi, retractions, grunting or nasal flaring, and no cough or runny nose.
“But does anyone know an infant who looks like that?” Baker said. “They’re out there but they’re few and far in between. If the infant has none of those symptoms, there is very low risk of a positive chest X-ray.”
Baker also raised the issue of lumbar puncture (LP), which he called “anxiety provoking” in the eyes of many parents. Is LP required? “If you look at the numbers, the answer is ‘no,’” Baker said. “The rate of bacterial infection or bacterial meningitis in this cadre of patients is really quite low, less than 1 percent or half of one percent if you look at some large studies.” However, the single reliable means of proving absence of bacterial meningitis is the spinal fluid analysis.
Urinalysis has been found to be an unreliable indicator of urinary tract infection in young infants with fever, with a sensitivity of only 80 percent. Other laboratory markers – like C-Reactive Protein (CRP) and Procalcitonin (PCT) – have been shown to help indicate the likelihood of bacterial disease. Baker pointed to one study that reported a tenfold increase in CRP values among those young infants with bacterial disease (Pediatrics 2001;108;1275-1279). Also, in a study of children under 36 months of age with fever, the Spanish Society for Pediatric Emergency Medicine found a sizeable difference in PCT values between those with viral and bacterial disease (Pediatric Infectious Disease Journal 2003:22;895-904).
Despite these encouraging data, neither the CRP nor the PCT are single reliable indicators of presence of bacterial disease; 10-20 percent of infants with bacterial disease have CRP or PCT values well within normal ranges. On the other hand, Baker said, “PCT may be somewhat helpful in figuring out who might have deep tissue infection vs. other types of viral or bacterial disease.”
Other biological markers of bacterial disease include sPLA2, which is generated in response to inflammation. In a 2009 study of children 2 weeks to 14 years of age, the viral group had very low sPLA2 values, while contained infections were five-fold higher, and bacteremia 20-fold higher.
Does the presence of viral infection preclude co-existence of bacterial disease? Baker cited four studies showing an “almost zero incidence” of bacterial disease in infants with clinical bronchiolitis (infants with fever and wheezing). On the other hand, several large prospective studies have shown a 6-to-7 percent rate of concurrent bacterial disease in infants with positive viral testing.
Are there treatable conditions other than bacterial that should be considered in evaluating fever in the first month of life? Yes, Baker said, pointing to herpes simplex virus as a chief culprit in infants younger than 3 weeks of age. “Often children with herpes simplex virus do not have fever and we need to recognize other signs and symptoms to identify them,” Baker said. “It’s important to make the diagnosis because there is treatment for herpes virus, and you can ward off evil in a good number of these kids if they’re treated early.”
Summing up, Baker said, “If you fully evaluate the source of fever in 1- to 2-month-old infants, and provide reliable and consistent follow up with these children, you can manage those with negative work-ups at home. If you adhere closely to the Philadelphia criteria, low risk infants are safe to manage without antibiotics.
The patient Baker saw in 2001?
Based on a spinal fluid Gram stain that was the single test result that came back positive, the infant was admitted and treated with antibiotics. That was a good thing, Baker said, because 8 hours later the blood culture and CSF culture came back positive for group B strep. With antibiotics, the child recovered uneventfully.
“This case was maybe one out of 5,000, but they’re out there,” Baker said. “The point is, you cannot trust the physical-exam appearance of children when they present with fever in this age range. They can look well and still be ill.”