ID Consult

Evaluating fever in the first 90 days of life


Fever in the youngest of infants creates a challenge for the pediatric clinician. Fever is a common presentation for serious bacterial infection (SBI) although most fevers are due to viral infection. However, the clinical presentation does not necessarily differ, and the risk for a poor outcome in this age group is substantial.

In the early stages of my pediatric career, most febrile infants less than 90 days of age were evaluated for sepsis, admitted, and treated with antibiotics pending culture results. Group B streptococcal sepsis or Escherichia coli sepsis were common in the first month of life, and Haemophilus influenza type B or Streptococcus pneumoniae in the second and third months of life. The approach to fever in the first 90 days has changed following both the introduction of haemophilus and pneumococcal conjugate vaccines, the experience with risk stratification criteria for identifying infants at low risk for SBI, and the recognition of urinary tract infection (UTI) as a common source of infection in this age group as well as development of criteria for diagnosis.

Fig. 1. Distribution of clinical syndromes in febrile infants less than 90 days of age (Greenhow) Tara Greenhow, Yun-Yi Hung, Arnd Herz, et al; Pediatric Infectious Disease Journal; The Changing Epidemiology of Serious Bacterial Infections in Young Infants; Volume 33, Issue 6; Jun 1, 2014

Fig. 1. Distribution of clinical syndromes in febrile infants less than 90 days of age (Greenhow)

Over the last 4 decades, the pediatric community has recognized that not all febrile infants have the same risk for SBI; low-risk infants have a 1%-2% incidence while high-risk infants may have rates as high as approximately 20%. The strategies for assigning risk include criteria developed in multiple centers (Rochester, Boston, and Philadelphia, to name a few) that focus on age (30-90 days), history, physical exam, and laboratory parameters. Term infants, with no prior antibiotic administration or comorbid illness, discharged home with mother following birth, with normal physical exam and feeding, and white blood cells (WBC) counts between 5,000 and 15,000/mm3, urine analysis less than 10 wbc/hpf, cerebrospinal fluid (CSF) with less than 8 wbc/mm3, and a normal chest x-ray (if signs/symptoms of respiratory disease) have been found to be at low risk for SBI. Therefore, those older than 30 days can be managed as outpatients with close observation.

A further nuance was subsequently added with the introduction of rapid diagnostics for viral infection. Byington et al. found that the majority of febrile infants less than 90 days of age had viral infection with enterovirus, respiratory syncytial virus (RSV), influenza or rotavirus.1 Using the Rochester risk stratification and the presence or absence of viral infection, she demonstrated that the risk of SBI was reduced in both high- and low-risk infants in the presence of viral infection; in low risk infants with viral infection, SBI was identified in 1.8%, compared with 3.1% in those without viral infection, and in high-risk infants. 5.5% has SBI when viral infection was found, compared to 16.7% in the absence of viral infection. She also proposed risk features to identify those infected with herpes simplex virus; age less than 42 days, vesicular rash, elevated alanine transaminase (ALT) and aspartate aminotransferase (AST), CSF pleocytosis, and seizure or twitching.

Greenhow et al. reported on the experience with “serious” bacterial infection in infants less than 90 days of age receiving care at Northern California Kaiser Permanente during the period 2005-2011.2 As pictured, the majority of children have UTI, and smaller numbers have bacteremia or meningitis. A small group of children with UTI have urosepsis as well; those with urosepsis can be differentiated from those with only UTI by age (less than 21 days), clinical exam (ill appearing), and elevated C reactive protein (greater than 20 mg/L) or elevated procalcitonin (greater than 0.5 ng/mL).3 Further evaluation of procalcitonin by other groups appears to validate its role in identifying children at low risk of SBI (procalcitonin less than 0.3 ng/mL).4

Currently, studies of febrile infants less than 90 days of age demonstrate that E. coli dominates in bacteremia, UTI, and meningitis, with Group B streptococcus as the next most frequent pathogen identified.2 Increasingly ampicillin resistance has been reported among E. coli isolates from both early- and late-onset disease as well as rare isolates that are resistant to third generation cephalosporins or gentamicin. Surveillance to identify changes in antimicrobial susceptibility will need to be ongoing to ensure that current approaches for initial therapy in high-risk infants aligns with current susceptibility patterns.

Dr. Stephen I. Pelton, chief of the section of pediatric infectious diseases and coordinator of the maternal-child HIV program at Boston Medical Center.

Dr. Stephen I. Pelton

In summary, the current approach to the febrile infant less than 90 days of age is based on risk stratification. Using a combination of established criteria based on history, physical exam, and laboratory testing, low-risk infants, at least those over 30 days of age, can be identified and managed as outpatients. Laboratory studies such as urine analysis and culture of a catheter-obtained specimen should be routine. Complete blood count with differential, blood culture, and AST and ALT for those less than 42 days of age, and C reactive protein and/or procalcitonin testing all will be of added value in distinguishing high- and low-risk infants. Enteroviral polymerase chain reaction on blood and CSF (when obtained), RSV detection from nasopharyngeal secretions during RSV season, and herpes simplex virus testing on blood and CSF (when obtained) in selective infants will further enable a specific diagnosis to be established. For the child less than 30 days with fever, especially those under 21 days, there still is concern about relying on these stratification strategies, and hospitalization and presumptive antimicrobial treatment remains the current recommendation. Updated guidelines from the American Academy of Pediatrics are expected.5 The proposed guideline is expected to stratify management by ages (7-28 days, 29-60 days, and 61-90 days), to include a role for inflammatory markers, and allow for a “kinder, gentler” approach, including withholding certain treatments and procedures if infants are at low risk of infection. Regardless of the initial strategy implemented, close follow-up until resolution is critical for optimizing outcomes.

Dr. Pelton is chief of the section of pediatric infectious diseases and coordinator of the maternal-child HIV program at Boston Medical Center. He said he had no relevant financial disclosures. Email him at

Next Article:

   Comments ()

Recommended for You

News & Commentary

Quizzes from MD-IQ

Research Summaries from ClinicalEdge