I. General issues:

  1. Can have sepsis without fever--proceed with w/u if looks sick, even if afebrile!
  2. 50% of infants under 18mos have nl. rectal temps of up to 37.8C
  3. Bundling (in most cases) and teething do not cause fever
  4. Axillary temps tend to be lower than rectal temps; the degree of variation is itself highly variable (BMJ 320:1174, 2000--AFP)
  5. Keep in mind Kawasaki's Disease as a non-infectious causes of fever in infancy
  6. Consider chest x-ray in any child < 5yo with T > 39.0 and WBC > 20k
    1. 146 children < 5yo with T > 39.0 and WBC > 20k but no clinical signs of pneumonia or other clinically evident source of infection were underwent CXR; 26% had occult pneumonia ("definite" infiltrates on CXR); 74% of these were described as "well-appearing" and all of these had SaO2 > 94% (Ann. Emerg. Med. 33:166, 1999--JW)
  7. Predictors of occult bacterial infections:
    1. Absolute peripheral blood Band counts, % bands, and Band-to-Neutrophil ratio were not found to be sig. diff in bacterial vs. viral infection, in a retrospective study of 100 generally healthy children < 2yo presenting with T > 39'C (or > 38'C if < 3mo), no clear source on exam, and eventual dx based on bacterial cx (of blood or urine or CSF, if done) or viral cx of nasopharyngeal swabs. However, temp, total WBC, and absolute neutrophil count were sig. higher in pts with bacterial than those with viral infections (Arch. Ped. Adol. Med. 153:261, 1999--AFP)
    2. In a study looking at WBC, PMN count, and CRP, the best single predictor of occult bacterial infection was an ANC > 10.6 x 10E9, in a study of 256 children 3-36mo presenting to an ED with temps 39.0-41.3'C. Combining PMN results w/WBC or CRP did not improve the detection rate (Arch. Ped. Adol. Med. 156:855, 2002--JW)

II. Newborn (<28d), T > 38.0

  1. 20% will have life-threatening infections.
  2. Need CBC w/diff, Cx of blood, urine, and CSF; stool sx if with diarrhea. Also check CSF gm stain, prot, glu, cell ct (if RBC?s in CSF, spin it; xanthochromia suggests hemorrhage). CXR if respiratory compromise.
  3. Hospitalize & give IV ABX pending Cx
  4. In a study in 99 well-appearing pts 7-28d old presenting < 12h after onset of fever without an obvious source, CRP > 20mg/L was more sensitive for eventual diagnosis of severe bacterial infection than absolute neutrophil count or WBC count.  Normal CRP on samples obtained > 12h after fever onset had 100% negative predictive value for eventual diagnosis of severe bacterial infection (Pediat. Inf. Dis. e-publication at http://dx.doi.org/10.1097/INF.0b013e3181b9a086-JW)

III. 4wks-3mos

  1. If toxic or not meeting low-risk criteria below, treat as if newborn (see above)
  2. MAY (but don't need to) treat as outpt with daily Ceftriaxone 50mg/kg IM to max of 1g and cx of blood, urine, & CSF, with re-eval in 24h (or with ur. cx and obs.) if low-risk:
  1. Previously healthy
  2. No focal infection other than OM
  3. WBC 5-15k with <1.5k bands
  4. Nl u/a
  5. If with diarrhea, <5WBC/hpf in stool
  1. If manage as in B, and blood cx pos, admit for r/o sepsis and IV abx pending all cx
  2. If manage as in B, and urine cx pos:
  1. If afebrile and well, outpt PO abx
  2. If still febrile, admit for r/o sepsis and IV abx pending all cx

IV. 3-36mos with T>39.0--guidelines per on Pediatrics 92:1, 1993

  1. If looks toxic, admit for sepsis w/u and IV abx
  2. If obvious source, treat that
  3. If have no obvious source, sig. incidence of bateremia or UTI
  1. Do CBC w/diff, u/a, blood and urine Cx
  2. CXR and LP prn focal signs
  1. Note that in the "post-HIB" era, fewer pts in this group may actually have occult bacetermia
    1. In one series of 9,500 such kids, blood cx were positive for pathogenic bacteria in 1.6% (mostly pneumococcus); risk of bacteremia ass'd with higher temp, WBC, neutrophil count, and band count (Arch. Ped. Adol. Med. 152:624, 1998--JW)
    2. In a retrospective review of 2641 immunocompetent pts 2-36mos presenting with fever (39.0-40.9'C) without an apparent source, none of whom received abx, among those who had blood cx done, only 3% were positive; only 2 pts developed serious bacterial infections, and both recovered completely (Arch. Ped. Adol. Med. 156:512, 2002--JW)

V. Antipyresis

  1. In a study in 464 children 6-36mos old with temp > 38.3'C randomized to acetaminophen 12.6mg/kd Q6h, ibuprofen 5mg/kg Q8h, or alternating doses of the two Q4h (all after an initial loading dose of acetaminophen 25mg/kg or ibuprofen 10mg/kg), all tx continued x 3d.  The alternating-therapy recipients had sig. lower fever and levels of perceived stress than either of the monotherapy groups (Arch. Pediat. Adol. Med. 160:197, 2006--JW)