Fever in children is a common problem seen daily in almost any busy office practice. In fact, some studies estimate that up to a third of office visits for children in the United States are the result of an elevated temperature.1

Most clinicians will pursue aggressive evaluation for the child who is younger than 3 months or who appears toxic. Such an approach is not inappropriate, given the risk of invasive bacterial disease in this vulnerable age group. However, controversy persists over the management of older febrile children—aged 3 months to 3 years—especially when they appear to be nontoxic. The vast majority of these children will have viral disease, yet some could have significant bacterial infections.2,3

How can clinicians best identify these children? And how can they determine which children can safely be sent home?

Continue Reading

Fever definitions differ

Often defined as a core (rectal or tympanic membrane) temperature >38°C (100.4°F), fever nevertheless means different things to different people. Many in clinical practice have cared for children whose parents claim that a temperature of 99°F is a “fever”—because their child’s temperature “never” goes above 97°F. Other parents may balk at the idea of bringing a febrile child into the clinic because they are convinced that it is better for the child’s immune system to fight off the infection on its own.

How temperature is measured is of no small importance. A hand held over the forehead is useless (despite what parents may claim). While old-fashioned glass thermometers are highly accurate, they are no longer in widespread use. The poor correlation between hand and thermometer is mainly due to variations in skin perfusion. A child with warm, flushed skin may in fact be afebrile (if not sweating), whereas a child with cool, mottled skin (as seen with shock or dehydration) may indeed have a high fever.

The bottom line is that fever is a sign of illness, not a diagnosis, and its presence should prompt a search for the underlying cause. How extensive this investigation ought to be depends not on the absolute elevation of temperature, but on how ill the child appears and the presence or absence of other signs and symptoms. Unfortunately, there is no evidence to support the idea that high fevers are more likely to represent bacterial infection than low-grade ones.4 Nor is there proof that a fever that does not respond to an oral antipyretic is more serious than one that does.5

Beginning the evaluation

The first step in the clinical evaluation of any febrile child is to get an overall impression of the child’s status (Table 1). This should be done as soon as you enter the exam room, or perhaps even while the child is out in the waiting room. Proceeding directly into the history and physical examination can distract you from looking at the whole patient and result in missing valuable clues that might reassure you (or push you toward hospital admission).

Is the patient alert and interactive, or is he drowsy and irritable? Does he reach eagerly for a lollipop or toy or bury his face in his mother’s lap? Is he engaged in conversation, or staring listlessly at the wall? Years ago, lack of eye contact was shown to correlate strongly with serious disease in young children.6

A thorough history comes next and may often lead to a correct diagnosis—even before examination. How long has the child been ill? What were the presenting complaints? Are others in the household or day-care group also sick? Are there underlying chronic diseases (e.g., sickle cell, diabetes, or HIV)? Ask about localizing signs and symptoms, such as runny nose, cough, otalgia, and sore throat. Has there been vomiting or diarrhea? Rash or arthritis? A limp? Urinary symptoms may be hard to elicit in a small child, but their absence does not rule out UTI.

Vital signs should be noted next.

Tachycardia, for example, might be the result of anxiety but more often signifies increased cardiac output due to fever, dehydration, or shock. A rapid heart rate could also be caused by medication, such as a dose of albuterol given hours earlier in an asthmatic child. Tachypnea should be noted as well, since it often signifies serious disease. It may be the only sign of pneumonia but can also result from fever. As you undress the child, look for retractions, nasal flaring, grunting, and head bobbing, all of which indicate respiratory distress. Look also at the skin, and assess capillary refill, temperature, and turgor. Finally, note hydration of the mucous membranes in the mouth, eyes, and nose.

Keep your differential as broad as possible as you examine the child. Even if you suspect an innocuous viral infection, your goal should be to rule out serious conditions, such as pneumonia, meningitis, septic arthritis, and UTI. Most children with fevers will have a readily apparent and easily treatable cause. The management of these cases is straightforward: a child with a draining ear will get an antibiotic, a child with adenopathy and exudates gets a culture, a hot red joint goes straight to the emergency department (ED) for aspiration.

Keep an open mind

In the absence of localizing signs and symptoms, however, it is important to keep thinking in broad terms. Is this infection, connective tissue disease, or malignancy? Is this a drug reaction? Hyperthermia? Or could the fever be factitious? Remember the adage, “You won’t make the diagnosis you don’t think of,” and keep going back for clues that support or refute your hypothesis. Don’t forget to keep region-specific diagnoses in mind. In the Northeast, for example, you need to think of Lyme disease, ehrlichiosis, and babesiosis, whereas patients with a history of recent foreign travel should be evaluated for malaria, dengue, hepatitis, and typhoid. Teething is often blamed for fever, but at the very least, this should be a diagnosis of exclusion. Most experts don’t believe teething can cause fever >38.5°C (101.3°F).1

Ordering lab studies

Laboratory testing should follow. This has changed significantly over the years. Twenty years ago, children who were younger than 2 months or who had no localizing signs usually got the full “septic workup” (complete blood count [CBC], sedimentation rate, chest radiograph, urine and blood cultures, and lumbar puncture). This approach was defended as rational and necessary to avoid missing a significant number of occult bacteremia cases.7-9

Back then, the incidence of occult bacteremia in febrile children was reported in some series to be as high as 10%.9 Today, thanks to the near-universal use of the conjugate Hemophilus influenzae type b (Hib) vaccine (introduced in 1991) and the pneumococcal conjugate vaccine (PCV7) (introduced in 2001) in the United States, that number has dropped to below 1%.10,11 This has led experts to adopt a less aggressive approach, with no apparent increase in adverse outcomes.

Which laboratory studies should then be ordered, and how valuable are they? In general, you should obtain laboratory studies for children who appear seriously ill and have no clear focus of infection as well as those with other complicating diagnoses, such as HIV or sickle cell disease.

Avoiding the ED

Before embarking on an extensive workup, though, consider a rapid antigen test for influenza. In the appropriate clinical scenario, making a diagnosis of flu can prevent a trip to the ED and obviate the need for further testing.12,13

A CBC is the obvious place to start, though an elevated WBC is a less reliable indicator of bacteremia than many may imagine. In as many as 20% of children with occult bacteremia, the WBC count is <15,000/µL and has poor sensitivity and specificity.1 Nevertheless, it is still commonly ordered and will lead to further investigation (i.e., urine and blood cultures) if elevated.

Blood cultures should always be considered in children who appear ill but have no localizing indicators, especially if the vital signs are unstable or the WBC is >15,000/µL. Such children should be started on empiric antibiotics (usually ceftriaxone, clindamycin, or a macrolide) while waiting for results, and most will be hospitalized for observation.1

Interestingly, as invasive pneumococcal disease has declined in importance, cases of Escherichia coli bacteremia have increased in recent years. In fact, the incidence of E. coli and pneumococcus in positive blood cultures is now equal. Of note, one study showed E. coli could also be grown out of the urine in every single case of E. coli bacteremia, underscoring the necessity and utility of evaluating the urinary tract when other sites of infection are not apparent.10

Negative dipstick isn’t definitive

This is usually done with a dipstick urinalysis and a culture. Remember, a “negative” dipstick does not rule out UTI, especially in a sick child who is voiding frequently. The presence of WBCs in the urine and a positive nitrite test are suggestive of infection. Hematuria, of course, is nonspecific and less helpful. And while such “soft” signs as foul-smelling urine ought to be acknowledged, diagnosis should never be based on them alone.

How should urine for culture be collected?

For the toilet-trained child, a clean-catch specimen may be adequate, but for the diapered child, catheterization is the gold standard. Since many clinicians will hesitate to catheterize a child in the office, bag specimens are still widely accepted. This may not be such a bad thing. Data show that one needs to collect 21 catheterized specimens to avoid one ambiguous bag result.14

Given that catheterization is relatively simple to perform, however, certain patients still merit this approach. These include children with a history of multiple unexplained fevers or unconvincing bagged culture results and children with vesicoureteral reflux in whom a positive culture will determine whether to proceed with surgery or long-term antibiotic prophylaxis.

When to order a chest x-ray

Deciding whether or not to order a chest x-ray can be more difficult. However, recognizing the limits of the physical examination in children and looking at the big picture can help. A child whose breathing is not labored and who has clear lungs with a pulse of 60 beats per minute is not likely to have pneumonia, even in the presence of a cough. Conversely, an ill-appearing child with an elevated respiratory rate and mottled skin might, despite clear lungs on auscultation. Recent guidelines call for a chest radiograph if there is evidence of respiratory distress, oxygen saturation <95% on oximetry, or WBC >20,000/µL.15,16

When to do a lumbar puncture

A lumbar puncture should be considered in any child who appears seriously ill and has no localizing signs and symptoms. Despite what many clinicians have been taught, very few children with acute bacterial meningitis present with a triad of fever, headache, and vomiting, and not all children with meningitis will even display nuchal rigidity.17 Clinicians must maintain a high index of suspicion for this diagnosis and should not be afraid to err on the side of caution.

Such caution notwithstanding, an astute provider will pay close attention to the history of the illness. For example, a child who is still running around the waiting room after three days of fever and headache is more likely to have a viral upper respiratory tract infection than meningitis and might best be served by an office test for influenza.

If you suspect a benign process and decide to send a child home for observation, make sure that the parent or guardian can provide good care. Is he or she an experienced parent? Worried? Disinterested? Preoccupied with other children or work? Are you confident you can count on this individual to call the clinic if the child’s condition worsens? Or are there issues, such as substance abuse or inadequate social support, that will render good follow-up impossible? How well you know the child’s family and social situation may affect your decision to let that child go home and lead you to insist on hospital observation.

Is treatment really needed?

Whether fevers should be treated at all is a question open to debate. There is evidence that elevated body temperature actually helps fight infection.18 And contrary to popular belief, there is no evidence that fever in and of itself causes harm in the way of brain damage or hearing loss.5 Fever therapy was common decades ago and even led to a Nobel Prize in 1927 when an Austrian psychiatrist used malarial fevers to treat tertiary syphilis.

In other words, fever is not the enemy, and parents should be educated to stop worrying about missing a dose of acetaminophen or waking a peacefully sleeping child at 2 am to force down another dose of ibuprofen. If fevers are treated because of pain, discomfort, or dehydration, parents should stick to adequate doses of either acetaminophen or ibuprofen and should not try to use both or to alternate them.5

Nonpharmacologic measures, such as undressing the child (or at least removing long sleeves and heavy blankets) and offering frequent drinks, should not be ignored. Tepid baths, which may induce shivering and thermogenesis, are no longer recommended but are not particularly harmful either, as long as alcohol is not used for sponging down the child (alcohol may be absorbed and lower the seizure threshold).

Febrile seizures—convulsions associated with fever and occurring without any other apparent cause—are a feared complication, but recent evidence shows that even these are not as significant as once thought and cannot be prevented even by the continuous use of antipyretics.5 Neurologic sequelae are exceedingly rare, and the majority of simple febrile seizures do not lead to chronic epilepsy. A thorough investigation (including lumbar puncture) is warranted if meningitis is suspected or the child is younger than 1 year, but routine electroencephalography is not indicated and cannot help predict recurrence.19

Clinicians should remember that while fever itself may not require treatment, the cause of any fever can still be serious.

Its presence always warrants some degree of investigation. Thanks to advances in routine pediatric immunization, occult bacteremia is much rarer than once taught. With careful examination and rational decision-making, fewer febrile children should need invasive testing or require referral to the ED for hospital observation.20


Mr. Zimmerman is a full-time practitioner with Esopus Medical, PC, an independent family practice in Rifton, N.Y., and a contributing editor to The Clinical Advisor.



1. Allen CH. Fever without a source in children 3 to 36 months of age. In: UpToDate. Rose BD, ed. Wellesley, Mass.: UpToDate; 2007.

2. Baraff LJ, Schriger DL, Bass JW, et al. Practice guideline for the management of infants and children 0 to 36 months of age with fever without source. Pediatrics. 1993;92:1-12.

3. Schriger DL. Management of the young febrile child. Clinical guidelines in the setting of incomplete evidence. Pediatrics. 1997;100:136.

4. Teach SJ, Fleisher GR. Efficacy of an observation scale in detecting bacteremia in febrile children three to thirty-six months of age, treated as outpatients. Occult Bacteremia Study Group. J Pediatr. 1995;126:877-881.

5. Crocetti MT, Serwint JR. Fever: separating fact from fiction. Contemp Pediatr. 2005;22:34-42.

6. McCarthy PL, Jekel JF, Stashwick CA, et al. Further definition of history and observation variables in assessing febrile children. Pediatrics. 1981;67: 687-693.

7. Caspe WB, Chamudes O, Louie B. The evaluation and treatment of the febrile infant. Pediatr Infect Dis. 1983;2:131-135.

8. McCarthy P. Management of the febrile infant. Pediatrics. 1992;89:1251-1253.

9. Grossman M. Management of the febrile patient. Pediatr Infect Dis. 1986;5:730-734.

10. Herz AM, Greenhow TL, Alcantara J, et al. Changing epidemiology of outpatient bacteremia in 3- to 36-month-old children after the introduction of the heptavalent-conjugated pneumococcal vaccine. Pediatr Infect Dis J. 2006;25:293-300.

11. Stoll ML, Rubin LG. Incidence of occult bacteremia among highly febrile young children in the era of the pneumococcal conjugate vaccine: a study from a Children’s Hospital Emergency Department and Urgent Care Center. Arch Pediatr Adolesc Med. 2004;158:671-675.

12. Vega R. Rapid viral testing in the evaluation of the febrile infant and child. Curr Opin Pediatr. 2005;17:363-367.

13. Benito-Fernandez J, Vázquez-Ronco MA, Morteruel-Aizkuren E, et al. Impact of rapid viral testing for influenza A and B viruses on management of febrile infants without signs of focal infection. Pediatr Infect Dis J. 2006;25:1153-1157.

14. Schroeder A, Newman TB, Wasserman RC, et al. Choice of urine collection methods for the diagnosis of urinary tract infection in young, febrile infants. Arch Pediatr Adolesc Med. 2005;159:915-922.

15. American College of Emergency Physicians Clinical Policies Committee; American College of Emergency Physicians Clinical Policies Subcommittee on Pediatric Fever. Clinical policy for children younger than three years presenting to the emergency department with fever. Ann Emerg Med. 2003;42: 530-545.

16. Bachur R, Perry H, Harper MB. Occult pneumonias: empiric chest radiographs in febrile children with leukocytosis. Ann Emerg Med. 1999;33:166-173.

17. Kaplan SL. Clinical features and diagnosis of acute bacterial meningitis in children. In: UpToDate. Rose BD, ed. Wellesley, Mass.: UpToDate; 2007.

18. Roberts NJ Jr. Impact of temperature elevation on immunologic defenses. Rev Infect Dis. 1991;13:462-472.

19. Fishman MA. Febrile seizures. In: UpToDate. Rose BD, ed. Wellesley, Mass.: UpToDate; 2007.

20. Lee GM, Fleisher GR, Harper MB. Management of febrile children in the age of conjugate pneumococcal vaccine: a cost-effectiveness analysis. Pediatrics. 2001;108:835-844.