Fever Without a Focus in the Neonate and Young Infant

Etiology and Epidemiology

Serious bacterial infection (SBI) occurs in 7% to 13% of neonates and young infants with fever. In this group, the most common SBIs are urinary tract infection (UTI; 5–13%), bacteremia (1–2%) and meningitis (0.2–0.5%). Escherichia coli is the most common organism causing SBI, followed by group B streptococcus (GBS). The decrease in GBS infections is related to increased screening of pregnant women and use of intrapartum antibiotic prophylaxis. Other, less common organisms include Klebsiella spp., Enterococcus spp., Streptococcus pneumoniae , Neisseria meningitidis , and Staphylococcus aureus ( Table 202.1 ). Listeria monocytogenes is a rare cause of neonatal infections, potentially related to changes in public health education and improvements in food safety. Additional details about specific bacteria are available in the following chapters: Escherichia coli ( Chapter 227 ), GBS ( Chapter 211 ), Streptococcus pneumoniae ( Chapter 209 ), Neisseria meningitidis ( Chapter 218 ), Staphylococcus aureus ( Chapter 208.1 ), and Listeria monocytogenes ( Chapter 215 ). Specific bacterial infections that can present with fever in this age-group, although often with symptoms other than isolated fever, include pneumonia ( Chapter 428 ), gastroenteritis ( Chapter 366 ), osteomyelitis ( Chapter 704 ), septic arthritis ( Chapter 705 ), omphalitis ( Chapter 125 ), cellulitis, and other skin and soft tissue infections ( Chapter 685 ).

Herpes simplex virus ( HSV ) infections ( Chapter 279 ) should also be considered in febrile neonates <28 days old, particularly given the high rate of mortality and significant morbidity among survivors. Neonatal HSV is rare, with a prevalence of 0.2–0.3% among febrile neonates. Most of these infections are caused by HSV type 2, though HSV type 1 can also cause neonatal infection. Neonates with disseminated disease and skin, eye, and mouth (SEM) disease typically present at 5-12 days of life. Neonates with central nervous system (CNS) disease generally present at 16-19 days. Perinatally acquired HSV may occasionally manifests beyond 28 days of age, although some of these later-onset cases may represent postnatal acquisition.

In febrile infants who appear well, viral illnesses are much more common than bacterial or serious viral infections. The most common viruses include respiratory syncytial virus ( RSV ; Chapter 287 ), enteroviruses ( Chapter 277 ), influenza viruses ( Chapter 285 ), parainfluenza viruses ( Chapter 286 ), human metapneumovirus ( Chapter 288 ), adenovirus ( Chapter 289 ), parechoviruses ( Chapter 277 ), and rhinovirus ( Chapter 290 ).

Clinical Manifestations

In neonates and young infants, bacterial and viral infections can present with isolated fever or nonspecific symptoms, making diagnosis of serious illnesses challenging. Some neonates and young infants will have signs of systemic illness at presentation, including abnormal temperature (hypothermia <36°C [96.8°F], fever ≥38°C [100.4°F), abnormal respiratory examination (tachypnea >60 breaths/min, respiratory distress, apnea), abnormal circulatory examination (tachycardia >180 beats/min, delayed capillary refill >3 sec, weak or bounding pulses), abnormal abdominal examination, abnormal neurologic examination (lethargy, irritability, alterations in tone), or abnormal skin examination (rash, petechiae, cyanosis). Infants with septic arthritis or osteomyelitis may appear well except for signs around the involved joint or bone or may only manifest with pseudoparalysis (disuse) and paradoxical irritability (pain when attempting to comfort the child).

Diagnosis

No consensus exists on the diagnosis and empirical treatment of febrile neonates and young infants. Traditionally, all neonates <60 or <90 days of age were hospitalized; underwent laboratory evaluation of the blood, urine, and cerebrospinal fluid (CSF); and received empirical antibiotics. Additionally, some patients had stool cultures, chest radiographs, HSV evaluation, and/or received empirical antiviral agents. Under this approach, many infants without SBI or serious viral infection received evaluation, treatment, and hospitalization. Protocols were subsequently developed to identify infants at lower risk of SBI, who may be managed outside the hospital setting. The 3 most widely used are the Rochester, Philadelphia, and Boston criteria ( Table 202.2 ). Clinical prediction rules are further discussed later in the Other Diagnostic Studies section. Despite these protocols, substantial variation continues to exist in the approach to and management of the febrile infant. It must be emphasized that these criteria apply to the well-appearing child; those who appear critically ill (septic) require prompt evaluation, resuscitation, and empirical antibiotic therapy (within 1 hr).

Many experts advocate that all neonates ≤28 days old undergo a complete evaluation for serious infection, receive empirical antimicrobials, and be hospitalized. Of the 3 widely used criteria, only the Rochester criteria allow neonates ≤28 days to be designated as “low risk” and managed outside the hospital without antimicrobials. In one study, <1% of low-risk infants ≤28 days old had SBI; however, in another study applying the Boston and Philadelphia criteria to neonates, 3–4% of those classified as low risk had SBI.

Young febrile infants ≥29 days old who appear ill (with signs of systemic illness) require complete evaluation for SBI, including antimicrobials and hospitalization; however, well-appearing infants can be managed safely as outpatients using low-risk criteria as indicated in Table 202.2 . In each of these approaches, infants must have a normal physical examination, must be able to reliably obtain close follow-up, and must meet certain laboratory and/or radiographic criteria. Based on these protocols, all infants following the Boston or Philadelphia criteria would undergo lumbar puncture (LP), whereas low-risk infants following the Rochester criteria would not. There is substantial variation in clinical practice in the performance of LPs in well-appearing infants >28 days. Clinicians should consider multiple factors, including the home situation and ability to contact the family, when deciding about LP in this age-group.

In addition, approximately 35% of infants with bacterial meningitis do not have a positive blood culture.

The protocols discussed in Table 202.2 were initially developed for use in the emergency department (ED). Infants evaluated in the office setting may warrant a different approach when a relationship between the physician and family already exists to facilitate clear communication and timely follow-up. In one large study of febrile infants <3 mo old who were initially evaluated for fever in the office setting, clinicians hospitalized only 36% of infants but initiated antibiotics in 61 of the 63 infants with bacteremia or bacterial meningitis. These findings suggest that, with very close follow-up (including multiple in person visits or frequent contacts by telephone), some febrile infants perceived to be at low risk for invasive bacterial infection ( IBI ; bacteremia and meningitis), on the basis of history, physical examination, and normal but limited laboratory testing, can be managed in an office-based setting. It is important to note that 3% of infants with SBI did not initially receive empirical antibiotics, necessitating careful consideration of risks and benefits of selective rather than universal testing and empirical antibiotic treatment of febrile infants evaluated in the office setting.