Febrile Infant/Encephalitis

Educational Objectives

The goal of this program is to improve management of the febrile infant and the child with encephalitis. After hearing and assimilating this program, the clinician will be better able to:

Stratify febrile infants according to modifications to the Rochester criteria.

List the 4 most important risk stratification criteria for febrile infants, according to the modified Rochester criteria.

Choose the appropriate inpatient management strategy for a febrile infant, based on his or her risk classification.

Differentiate clinical features of encephalitis from those of encephalopathy.

Recognize the most common causes of encephalitis in infants and children, including herpes simplex virus, en­terovirus, congenital infections, and metabolic disorders.

Faculty Disclosure

In adherence to ACCME Standards for Commercial Support, Audio-Digest requires all faculty and members of the planning committee to disclose relevant financial relationships within the past 12 months that might create any per­sonal conflicts of interest. Any identified conflicts were resolved to ensure that this educational activity promotes quality in health care and not a proprietary business or commercial interest. For this program, the faculty and plan­ning committee reported nothing to disclose.

Acknowledgements

Dr. Byington was recorded at the 11th Annual Frontiers in Pediatrics, held December 5-7, 2008, in Charleston, SC, and sponsored by the Medical University of South Carolina. Dr. Gonzalez spoke at Masters of Pediatrics Leadership Conferences: Exploring Contemporary and Future Pediatrics, held January 28 to February 2, 2009, in Miami Beach, FL, and sponsored by the University of Miami Miller School of Medicine. The Audio-Digest Foundation thanks the speakers and the sponsors for their cooperation in the production of this program.

Evidence-based Model for Evaluation of the Febrile Infant

Carrie L. Byington, MD, Professor of Pediatrics and Associate Chair for Clinical Research, University of Utah School of Medicine, Salt Lake City

Fever in infant <90 days of age: occurs in 10% to 12% of infants; temperature >38^°C associated with serious bacte­rial infection (SBI; ie, meningitis, bacteremia, or urinary tract infection [UTI]); UTI most common SBI in this age group

Studies conducted by speaker and colleagues: prospective enrollment of 1779 infants (between years 1996 and 2002); since 2002 (when electronic medical records used to track infants), 2000 infants enrolled annually

Prospectively enrolled infants: average age 31 days; by Rochester criteria, 33% considered low-risk for SBI and 67% considered high-risk; 10% had SBI (consistent with international rates); 35% had respiratory or enterovi­rus infection; sensitivity and negative predictive value lower than those cited in Rochester study; in Salt Lake City study, sensitivity 90% (ie, 10% of babies misclassified as low-risk)

Criteria associated with SBI (modified Rochester criteria [MRC]): 1) urinalysis (UA) with >10 white blood cells (WBCs) per high-power field (odds ratio of SBI, 38.8); 2) absolute band count >1500/mL; 3) WBCs <5000/mL or >15 000/mL; 4) chronic illness (major congenital anomaly, known immunocompromise-associated condi­tion, or chromosomal defect) and prematurity (delivery at <37 wk); speaker and colleagues now stratify infants according to these criteria

Age as risk factor: infants 1 to 28 days of age treated more conservatively than those between 29 and 90 days of age; at speaker’s hospital, 99% of younger infants admitted, compared to 50% of older infants; older high-risk infants »40% less likely to have SBI than younger  high-risk infants; any infants younger than 28 days consid­ered high-risk

Viral infection: infants with confirmed viral illness »70% less likely to have SBI than infants without viral illness

Impact on sensitivity and negative predictive value: combination of MRC, infant’s age, and viral status has greater impact than or similar impact to original Rochester criteria

“Stop-light” evaluation: green light (low risk)  —  infants with no MRC; if age >29 days and no viral infection pres­ent, risk for SBI 0.7%; yellow light (moderate risk)  —  patient has MRC, but viral infection also present; risk for SBI, 4% to 6%; red light (high risk)  —high risk according to MRC, with no virus; risk for SBI, 20% to 33%

Management implications: admit high-risk infants and administer antibiotics; send low-risk infants home; man­agement of moderate-risk infants still under study

Conclusions: to determine risk status of febrile infant, minimal evaluation requirements include patients’ age, history of chronic illness or prematurity, complete blood cell count (CBC), and UA; at speaker’s institution, viral testing also recommended

Viral testing: respiratory viral direct fluorescent antibody (DFA) test used year-round, 5 times/day (cost re­couped in savings on hospital and antibiotic expenses); enterovirus polymerase chain reaction (PCR) testing of blood and cerebrospinal fluid (CSF) offered June through November, and for all infants with meningitis; enterovirus most common cause of fever in infants; accounts for 50% of fever in infants detected between June and November, and 25% of cases overall

Electronic medical record cohort: infants considered high-risk if CBC and UA abnormal; otherwise, patients deemed low-risk (if CBC and UA not performed, patient’s status considered “undetermined”); 10 316 infants considered; of those, 5095 (49%) admitted and 5221 (51%) discharged; of discharged infants, fewer than one-half (49%) had both CBC and UA, despite recommendations established in 1971; risk status undetermined in 66%; within 3 days, 63 (1.2%) readmitted, most commonly due to missed SBI (including 5 cases of meningitis, with significant neurologic complications); of readmitted patients, 52% considered high-risk, based on CBC and UA; risk undetermined in 44%; high-risk, infants twice as likely to be readmitted in £3 days for any reason

Conclusions: most infants managed as outpatients did not have CBC or UA; high-risk infants most likely to be re­admitted, with SBI as most common reason; simple laboratory testing, including CBC and UA should be per­formed before deciding on outpatient management

Management of febrile infants (model program): quality improvement (QI) program started at Intermountain Health Care in 2006; electronic care process model introduced on January 1, 2008; available at all facilities; con­sists of algorithm for emergency care of febrile infants 1 to 90 days of age; directs clinicians to obtain “limited labs” (CBC and UA) for all infants, regardless of risk status; clinician must determine infant’s risk status, then follow steps outlined in algorithm; admission recommended for all high-risk infants; lumbar puncture should at least be attempted before initiating intravenous (IV) antibiotic administration in high-risk infants

Inpatient management: assess tests run in emergency department and infant’s risk level; low-risk infants or high-risk infants with viral infection  —  suggested discharge time, 24 hr; laboratory should read CSF and urine plates at least twice daily; among infants testing positive for virus, likelihood of detecting bacteremia or meningitis in CSF or urine after 36 hr <0.04%; most pathogens detected by 16 hr; high-risk virus-negative infants  —recom­mended discharge time, 36 hr; highest-risk group

Goals: appropriate evaluation of all infants, including age, history of chronic illness or prematurity, CBC, UA, viral diagnostic testing, and documented risk stratification; care process model now approved by American Board of Pediatrics as QI initiative acceptable for maintenance of certification and license

Goals for 2009: appropriate admission of high-risk infants; length of stay (LOS) decreased to 36 hr (current aver­age 50 hr)

Questions and answers

C-reactive protein (CRP): speaker does not measure in febrile infants; single reading not helpful; daily measure­ments very helpful when SBI present; infants with adenovirus or enterovirus may have “very abnormal” CRP levels; procalcitonin  —  being studied as rapid test; high levels associated with SBI, but association with viral in­fection still under investigation

Regional epidemiology: important to know; speaker has found association between prophylaxis against group B streptococci with ampicillin administered during delivery, risk of admission for SBI within first 90 days of life, and risk that pathogen will be ampicillin-resistant; only use of penicillin for prophylaxis against group B strepto­cocci strongly recommended (prophylaxis with ampicillin presents risk to infant; use clindamycin if mother al­lergic to penicillin, but test isolate first for clinadmycin resistance; use vancomycin if organism clindamycin-resistant)

Empiric use of acyclovir: test liver function in every infant younger than 28 days as part of sepsis evaluation; if in­fant has vesicles on skin, eyes, or mucous membranes, consider herpes simplex virus (HSV) and administer acy­clovir; if CSF abnormal, consider central nervous system (CNS) infection with HSV and administer acyclovir, especially during first 28 days; disseminated herpes easy to miss (no vesicles and CSF normal); occurs in approx­imately one-third of infants with herpes; abnormal liver function tests (>1000 mg/L) key to detection

Encephalitis

Ivan A. Gonzalez, MD, Associate Professor of Clinical Pediatrics, Pediatric Infectious Diseases, and Immunol­ogy, University of Miami Miller School of Medicine, Miami, FL

Definition: any inflammation of CNS; typically diagnosed clinically; must be distinguished from encephalopathy, which appears clinically similar (CNS inflammation only histologic difference)

Clinical presentation: seizures, lethargy, ataxia, movement disorders, neurologic deficits, personality changes, and altered sensorium

Features that distinguish encephalitis from encephalopathy: fever, headache, weakness, and irritability

Questions to ask: history of previous encephalitis (recurrent encephalitis sometimes associated with immunodefi­ciency; 5% of individuals with HSV encephalitis can relapse posttreatment); maternal history of fever (if patient neonate); maternal HSV (however, only 25% of patients have positive history; in many cases, patient unaware of infection)

Etiology: identified in only one-third of cases; pathogens often isolated from other sites (eg, nasopharynx, stool, urine, serum, CSF); single antibody titer sufficient for diagnosis in highly fatal cases, such as rabies

Most common pathogens: neonates  —  determine whether illness meningitis or meningoencephalitis (requires faster action); HSV most common cause of encephalitis in this age group; other pathogens include enterovirus, adeno­virus, group B streptococci, Listeria, Citrobacter, and other gram-negative organisms; congenital infections such as toxoplasmosis, rubella, cytomegalovirus (CMV), and HSV (TORCH), or syphilis; metabolic disorders; neuro­logic disorders; infants and children  —  enterovirus main cause in this age group; HSV responsible for most se­vere cases; arthropod-borne viruses (West Nile, St Louis, and Eastern Equine); Epstein-Barr virus (EBV); adenovirus; controversial issues include mycoplasma infection (test result may be false positive; may indicate postinfectious process), human herpes virus-6, and mumps; rule out inborn errors of metabolism, systemic lupus erythematosus vasculitis, and tumors (all can mimic encephalitis); consider toxic encephalitis associated with Bordatella, Shigella, Campylobacter, or Haemophilus influenzae; also consider acute ingestion of toxin

Epidemiology: infectious encephalitis typically occurs in summer; rate, 0.3 to 0.5 per 100  000 individuals; affects children, elderly, and immunocompromised; rate declines over time, with highest incidence during first year of life; severity increases with age; postinfectious disease rare in children <1 yr of age; fewer epidemics today, but mortality rates unchanged; HSV leading cause of severe cases in all age groups; if HSV encephalitis suspected in neonate, begin immediate treatment with acyclovir, even if suspicion later disproved; enterovirus and adenovirus usually cause systemic illness rather than isolated encephalitis; other possible causes include human parechovi­ruses (HPe; HPeV1 [formerly echovirus 22] and HPeV2 [formerly echovirus 23]), and congenital infections

HIV encephalitis: typically occurs during seroconversion; degenerative encephalopathy late complication

Arboviruses: West Nile and California encephalitis most common; <1% of patients have neurologic symptoms

Pathogenesis: pathogen can invade any cell in CNS; leads to cell lysis and inflammation; depending on organism, may remain focal or cause generalized neuronal lysis; severe lysis causes cortical gray matter inflammation, which may lead to vasculitis, hemorrhage, and necrosis; spread may be hematogenous, through neuronal tract, or from la­tency site

Correlation with immunodeficiency: humoral  —  chronic enterovirus meningoencephalitis; cell-mediated  —pro­gressive, multifocal leukoencephalopathy (PMLE) with JC virus; subacute HSV infection or measles encephalitis, and progressive rubella encephalitis; recurrent encephalitis with CMV, varicella zoster, or toxoplasma; acquired  —toxoplasma, CMV, Cryptococcus neoformans, or PMLE

Diagnosis: consider meningitis first, especially in neonate; start treatment promptly; inquire about prodromes; if child has respiratory infection, consider influenza; with exanthems, consider measles; history of vaccination may help include or exclude certain conditions; ask about exposure to cats, ticks, and mosquitoes; in teenager with pharyngitis and fatique, consider EBV

Work-up: lumbar puncture (brain biopsy no longer necessary); blood culture and serum antibody testing; viral test­ing; rule out metabolic disorders; urine viral culture; toxicology screen; imaging (magnetic resonance imaging [MRI] first choice); electroencephalography (helps diagnose HSV encephalitis)

Laboratory findings: brain biopsy definitive but rarely performed; CSF cell count, <200 cells/mL; protein, 50 to 200 mg/dL; blood glucose should be normal; perform CSF PCR when possible, but titers usually not seen until 4 to 5 days into illness; MRI can detect edema and inflammation; may be normal initially; gadolinium may help; demyelination may occur in postinfectious cases; computed tomography better if TORCH suspected

Management: antibiotics or antivirals, initiated as soon as possible; other than bacterial meningoencephalitis or HSV, many cases usually self-limited; watch for syndrome of inappropriate secretion of antidiuretic hormone

Prognosis: mortality  —  3% to 4% overall; age-dependent (40%-80% of infants die); 14% among neonates with HSV; morbidity  —  7% to 10% of patients develop neurologic sequelae; 50% of neonates who survive HSV de­velop neurologic sequelae

Postinfectious encephalomyelitis (acute disseminated encephalomyelitis): clinically diagnosed by evidence of neu­rologic deficit and positive MRI after infection; typically affects white matter; acute multifocal inflammatory de­myelinating disorder; incidence decreasing with use of vaccines; autoimmune process that leads to perivenulitis and demyelination; T cells penetrate CNS and attack neurons; treat with corticosteroids or IV immunoglobulin; mortality »5%; 70% to 90% of patients recover completely

Suggested Reading

Byington CL et al: Human herpesvirus 6 infection in febrile infants ninety days of age and younger. Pediatri Infect Dis J 21:996, 2002; Byington CL et al: Serious bacterial infections in febrile infants 1 to 9 days old with and without viral infections. Pediatrics 113:1662, 2004; Cherry JD: Recognition and management of encephalitis in children. Adv Exp Med Biol 634:53, 2009; De Tiège X et al: The spectrum of herpes simplex encephalitis in children. Eur J Paediatr Neurol 12:72, 2008; Doja A et al: Pediat­ric Epstein-Barr virus-associated encephalitis: 10-year review. J Child Neurol 21:384, 2006; Garra G et al: Reappraisal of criteria used to predict serious bacterial illness in febrile infants less than 8 weeks of age. Acad Emerg Med 12:921, 2005; Hsieh WB et al: Outcome of herpes simplex encephalitis in children. J Microbiol Immunol Infect 40:34, 2007; Luxmore B et al: Absolute band counts in febrile infants: know your laboratory. Pediatrics 110:e12, 2002; Malm G: Neonatal herpes virus infection. Semin Fetal Neonatal Med Feb 19, 2009 [Epub ahead of print]; Rittichier KR et al: Diagnosis and outcomes of enterovirus infections in young infants. Pediatr Infect Dis J 24:546, 2005; Schultze D et al: Benefit of detecting tick-borne encephalitis viremia in the first phase of illness. J Clin Virol 38:172, 2007; Tavakoli NP et al: Detection and typing of enterovirus from CSF specimens in pa­tients diagnosed with meningitis/encephalitis. J Clin Virol 43:207, 2008.