Pediatric Annals

CME Article 

Fever Without a Localizing Source

Jennifer L. Trainor, MD; Julie Kim Stamos, MD


Fever is one of the most common reasons for a visit to the primary care provider. Often, fever is associated with other symptoms, which make the diagnosis obvious. However, fever without a localizing source is a problem that is perplexing to health care providers and parents.


Fever is one of the most common reasons for a visit to the primary care provider. Often, fever is associated with other symptoms, which make the diagnosis obvious. However, fever without a localizing source is a problem that is perplexing to health care providers and parents.

Jennifer L. Trainor, MD, is with the Division of Pediatric Emergency Medicine, Northwestern University Feinberg School of Medicine, Chicago. Julie Kim Stamos, MD, is with the Division of Pediatric Infectious Diseases, Northwestern University Feinberg School of Medicine, Chicago.

Dr. Trainor and Dr. Stamos have disclosed no relevant financial relationships.

Address correspondence to: Jennifer Trainor, MD, Children’s Memorial Hospital, Northwestern University Feinberg School of Medicine, 2300 Children’s Plaza, fax: 773-880-8267; or e-mail:

Fever is one of the most common reasons for a visit to the primary care provider. Often, fever is associated with other symptoms, which make the diagnosis obvious. However, fever without a localizing source is a problem that is perplexing to health care providers and parents.

In this article, we focus on fever in older infants and toddlers (3 to 36 months) because the management of fever in the neonate has been reviewed extensively elsewhere.

As pediatricians know, fever is not a disease itself, but rather a symptom of an underlying illness. In the first several years of life, the average child will experience multiple infections per year, frequently accompanied by fever. With increasing numbers of children in childcare facilities and parents who work outside the home, the desire to identify and eradicate the source of fever as quickly as possible has placed enormous pressure on parents and health care providers. An efficient evaluation or a quick fix is often pursued.

“Fever phobia” is rampant among parents of all socioeconomic levels, and the medical provider’s approach to its evaluation and management often serves to fuel parental concerns rather than to reassure them. Evidence that fever causes brain damage or permanent injury of any form is lacking, yet even the most educated parents still believe that fever is harmful to their child. Therefore, parents bring their children to emergency departments or urgent care facilities, worried that fever alone indicates a serious problem warranting urgent laboratory evaluation or treatment.

The cornerstone of evaluation of fever in infants and young children remains a thorough history and complete physical examination. Clues to potential etiologies may be present in the vital signs. Tachypnea out of proportion to other vital signs may be a clue to pneumonia. Tachycardia out of proportion to fever and hydration state may be an early sign of sepsis. Certain clinical findings, such as rash, nonpurulent conjunctival injection and mucous membrane changes, may indicate incomplete Kawasaki disease. Decisions regarding evaluation should be based more on the overall impression of the child’s clinical state, the reliability of the family, and access to follow-up rather than the presence or height of fever.

Fever Without Source and Occult Bacteremia

In the 1980s and early 1990s, well-appearing febrile infants and toddlers from 3 to 36 months without an obvious source for their fevers were frequently evaluated in emergency departments. In an attempt to prevent focal complications of bacteremia, which, at that time, was primarily caused by Haemophilus influenzae type b (Hib), physicians attempted to identify infants who were not yet ill-appearing but had occult bacteremia. During that period, prospective studies revealed that approximately 3% of well-appearing infants and toddlers with fever more than 102.2°F had unsuspected bacteremia. Bacteremia had a relatively high rate of localizing to an area of the body (eg, buccal cellulitis, periorbital cellulitis, pneumonia, meningitis) and subsequently causing serious illness. In cases of bacteremia caused by Hib, 10% to 20% of initially well-appearing children with occult bacteremia had progression to serious focal bacterial infections.1

In 1993, a practice guideline was disseminated recommending routine testing and treatment for occult bacteremia in previously healthy infants and toddlers.2 This led to the reflexive testing of thousands of well-appearing infants and toddlers to find and empirically treat those who had unsuspected serious bacterial illness. Since the advent of the Hib conjugate vaccine in 1988 and its subsequent widespread use in the United States, bacteremia from Hib has been virtually eliminated.3

Even during the Hib era, Streptococcus pneumoniae accounted for most cases of occult bacteremia. Fortunately, most cases of pneumococcal bacteremia resolved spontaneously without treatment. Only a minority of children with occult bacteremia secondary to S. pneumoniae went on to develop focal complications. Before the introduction of the heptavalent pneumococcal vaccine, rates of occult bacteremia had already been reduced to 1.5%. Current figures indicate an overall prevalence of occult bacteremia in healthy, vaccinated infants and toddlers of 0.7%.4 Still, these numbers may over-represent occult bacteremia because they include children with a focus of infection on examination or urinary tract infection with concomitant bacteremia. In a study performed in the Northern California Kaiser Permanente Health System done after the introduction of the heptavalent pneumococcal vaccine, the rate of occult bacteremia was found to be only 0.25%.4

Given the current low rates of occult bacteremia and its sequelae, well-appearing infants (older than 3 months of age), and toddlers with fever without a localizing source should no longer be routinely subjected to laboratory investigations to detect occult bacteremia. False positive rates of blood cultures range from 1% to 5% in published series, thus the false-positive rate easily exceeds the true-positive rate.4,5,6,7 Recent studies have shown that more than 70% of organisms identified from blood cultures in febrile but otherwise healthy infants and toddlers were contaminants.4,6,7 The cost in health care dollars and parental and health care provider time for follow-up of these false-positive blood cultures are clearly significant.

On the other hand, infants and young children with ill appearance, altered mental status, very abnormal vital signs and poor perfusion, including mottling or cyanosis, represent an entirely different risk group and should be thoroughly investigated for bacterial illness. In these types of patients, a complete blood count, blood culture, urinalysis, and urine culture should be performed at a minimum.

If the child is hemodynamically stable and a lumbar puncture is indicated by clinical examination, cerebrospinal fluid should be sent for analysis and culture as well. Empiric antibiotic therapy pending culture results is standard practice in these acutely ill patients. Unfortunately, white blood cell indices have not proven to reliably include or exclude serious bacterial illness.4 Thus, the practice of deciding to send cultures and/or treat with antibiotics based on the total white cell count or absolute band count is not a prudent strategy.

Urinary Tract Infection

Urinary tract infection is the most common serious bacterial infection in infants and young children. Deciding which children with fever to test for urinary tract infection requires knowledge of the baseline rate of disease in the population of interest. It is sensible only to test infants and toddlers whose pretest probability of urinary tract infection (UTI) is high enough to warrant a bladder catheterization. Urine cultures sent from perineal bag samples are more likely to represent contaminants than true positive cultures.8 Bladder catheterization should always be performed when a clean catch of urine cannot be obtained.

A recent meta-analysis that pooled published studies of infants and toddlers presenting with fever, revealed that the overall prevalence of UTI was 7% (95% CI, 5.5–8.4) in the 0–24-month age group. Rates, however, vary based on age and sex. The pooled prevalence rates of UTI in girls aged 3 to 6 months, 6 to 12 months, and older than 12 months were 5.7%, 8.3%, and 2.1%, respectively. Interestingly, the rate of UTI was higher among white infants than among black infants.9

As part of its series on the Rational Clinical Examination, the Journal of the American Medical Association published a meta-analysis of studies on UTI in infants and toddlers in 2007. Diagnostic algorithms for febrile male and female infants and toddlers, which were stratified by age and other risk factors, were derived from these studies.9

Risk factors included a previous history of UTI, temperature of more than 102.2°F, fever without a localizing source, ill appearance, suprapubic tenderness, fever of more than a 24-hour duration, and nonblack race. If the probability of UTI was less than 2% based on age and risk factors, the authors did not recommend immediate urinalysis and culture but rather a follow-up examination in 24 hours. For rates more than this threshold, testing was recommended.9

According to the pooled analysis, any child younger than 24 months with a fever more than 102.2°F, and no localizing signs on physical examination has a rate of UTI exceeding 2%. Fortunately, a dipstick urinalysis can stratify children into risk groups to help guide which children require empiric treatment.

For example, an uncircumcised non-black boy aged 3 to 23 months with fever for more than 24 hours has a pretest probability of UTI of 10% to 25%. If the result of his urinalysis show no nitrites or leukocyte esterase, his risk of UTI is only 2% to 6%. In this situation, a practitioner might await the results of a urine culture to decide on treatment. If, however, nitrite and leukocyte esterase are positive, his risk of having a UTI is 75% to 90%. Empiric treatment to cover gram-negative organisms would then be prudent.

Febrile circumcised boys aged 3 to 24 months without a source of infection have the lowest pretest probability of UTI of 2% to 4%. However, if a urine dipstick shows nitrites or leukocyte esterase, their risk of UTI is 15% to 34%; with both positive, the risk increases to 46% to 71%.9

Girls aged younger than 12 months with fever without localizing signs for more than 24 hours have a pretest probability of UTI of 10% to 25%. A urine dipstick negative for leukocyte esterase and nitrites reduces their probability of UTI to 2% to 6%, although either positive test raises it to 40% to 66%, and both positive tests raise the probability to 75% to 90%. For well-appearing highly febrile girls older than a year in age who have no obvious source for fever on examination, the pretest probability of UTI is 3% to 8%. A negative urine dip reduces the likelihood to less than 2%, although either positive test increases it to 15% to 34%, and both positive tests make the likelihood of having a UTI 46% to 71%.9

Thus, knowing the disease prevalence in the population to be tested makes the decision-making process of whether to catheterize a particular child for urinalysis and urine culture much easier. However, keeping these numbers in your head is extremely difficult. Familiarizing yourself with current algorithms on risk assessment for UTI in young children with fever will help your decision-making process. The meta-analysis published in the Journal of the American Medical Association has excellent flow diagrams to help guide testing and treatment.9


With careful anticipatory guidance, health care providers can convince most parents that low-grade fever is not harmful to their child. However, once the temperature reaches 106°F at home, even the most calm parent reaches for the telephone or the keys to the car. Older studies indicated that these children were at higher risk for serious bacterial illness than their less febrile counterparts; however, these studies were conducted before the advent of the Hib vaccine. Now that the whole landscape concerning occult bacteremia has changed, we explore what data exist on the evaluation of children with hyperpyrexia, which is defined as a fever of more than 106°F.

In order to answer the question of whether these children are at high risk of serious bacterial infection, investigators at Baylor College of Medicine prospectively studied children during a 2-year period (from September 1998 to September 2000) who presented to a pediatric emergency department with a rectal temperature more than 106°F. All patients had a history and physical examination performed, and a CBC and a blood culture. Nasopharyngeal viral cultures were sent on all patients. Nasopharyngeal samples were tested for respiratory syncytial virus (RSV); influenza A and B; parainfluenza viruses types 1 to 4; adenovirus; picornaviruses; rhinoviruses; enteroviruses; herpes simplex viruses types 1 and 2; varicella zoster virus; and cytomegalovirus. Urine was sent for urinalysis, and urine cultures were obtained for all children younger than age 2, as well as older children with dysuria, frequency, previous UTI, or renal abnormality.10

Although all children younger than 18 years were eligible for enrollment, no child younger than 3 months of age presented to the ED with hyperpyrexia during the study period. The overall rate of hyperpyrexia was 1 per 1,270 patient visits. The median age of patients was 17 months, with an interquartile range of 11 to 25 months. The median rectal temperature was 106.2°F. The rate of documented infections was more than 40%; however, the children were equally likely to have a documented viral illness (21%) as they were to have a bacterial illness (20%). Not surprisingly, children with pre-existing medical conditions were at higher risk of bacterial infection (36.8%) than their previously healthy counterparts (15.5%). The presence of rhinorrhea or any viral symptom was associated with a decreased risk of serious bacterial infection.

Therefore, it appears that children with hyperpyrexia are at higher risk of serious bacterial infection than their less febrile counterparts. Other than occult bacteremia secondary to S. pneumoniae (the study was performed before the widespread use of the pneumococcal conjugate vaccine), UTI was the most common source of infection. Moreover, a viral infection was as commonly found in these patients as a bacterial infection, and the age, maximum temperature, and total white blood cell count were not predictive of a bacterial versus a viral infection.10

The Role of Antipyretics

Little evidence exists that fever itself is harmful to children. In the absence of autonomic dysregulation, environmentally induced heat stroke, or malignant hyperthermia, high temperatures have not been not been shown to have ill effects.1 Although febrile seizures are a common childhood phenomenon, most are self-limited, and there is no evidence that treatment with antipyretics prevent their occurrence or recurrence.11,12

Barton Schmitt first coined the term “fever phobia” in 1980 to describe parental fears of the negative consequences of fever. Half of all parents (48%) he surveyed believed that fever would go to 107°F or higher without treatment. A similar number of parents (46%) thought fever could cause brain damage.13 In 1999, a group of Baltimore investigators at two urban primary care clinics replicated his survey to determine if caregiver beliefs about fever had changed in the intervening 20 years.

Although fewer parents believed that temperature would continue to rise to potentially lethal levels if untreated, 7% still believed that temperature could rise to more than 110°F. Moreover, almost all (91%) caregivers believed that fever caused harm to their children. Half of all parents reported that they would check their child’s temperature at intervals of about 1 hour during a febrile illness, and 25% would give antipyretics for a temperature of less than 100°F. Surprisingly, 85% of caregivers reported that they awakened their child solely to give an antipyretic.14

Not surprisingly, parents over-treated fever. Fourteen percent gave acetaminophen too frequently, 44% gave ibuprofen too frequently, and 27% alternated acetaminophen and ibuprofen.14 Subsequent studies have shown that as many as two-thirds of parents alternated acetaminophen and ibuprofen when their children have fever. Eighty-one percent of these caregivers stated that their health care provider instructed them to do so.15 In a 1998 survey of pediatricians in practice, 50% recommend alternating acetaminophen and ibuprofen to parents whose children had fever.16

Although there is evidence that ibuprofen and acetaminophen are effective antipyretics,15,16 data from recent randomized controlled trials in the UK indicate that combination therapy may be superior to monotherapy with ibuprofen,17 which is, in turn, superior to monotherapy with paracetamol (acetaminophen).17 Combination therapy resulted in faster time to fever clearance and more prolonged time without fever in the first 4 hours and in the first 24 hours of febrile illness. It did not, however, reduce discomfort at 48 hours.18

The increased efficacy, however, may come at a price. Dosing regimens are potentially confusing, and even in the setting of a well-designed controlled trial with clear-written instructions, the recommended maximum dose was exceeded in 8% with acetaminophen and 11% with ibuprofen.18 Moreover, in actual practice, parents use a range of dosing intervals, some as often as every 2 to 3 hours. Given the multiple formulations and concentrations of antipyretics, the potential for dosing errors is substantial.15

Furthermore, no study of sufficient size has been performed to prove that the practice of using combination therapy is safe. There are some theoretical concerns. Ibuprofen inhibits hepatic production of glutathione, and if acetaminophen concentrations are high, the potential for hepatic and renal toxicity exists, especially in the febrile child with increased insensible losses and hypovolemia.1 Moreover, using two medications to “control” what pediatricians believe to be a relatively harmless phenomenon potentially sends the wrong message to parents.

As health care providers, pediatricians have an obligation to combat, rather than to fuel, fever phobia. Pediatricians need to emphasize to parents that fever is a useful body defense to infection and that it is a symptom rather than a disease. Caregivers need to understand that fever will persist until the underlying disease process (most of which is viral) resolves. Pediatricians must convey to parents that the object is not to “control” the fever but to provide relief of discomfort.

If the health care provider can convince parents that fever is a normal response that does not necessarily need to be treated, the provider can potentially allay their fears, as well as their overuse of antipyretics. Pediatricians should counsel parents to pay more attention to their child’s symptoms, hydration status, and overall well-being, rather than focusing on the number on the thermometer.


  1. Avner JR. Acute fever. Pediatr Rev. 2009;30(1);5–13. doi:10.1542/pir.30-1-5 [CrossRef]
  2. Baraff LJ, Bass JW, Fleisher GR, et al. Practice guideline for the management of infants and children 0 to 36 months of age with fever without source. Agency for Health Care Policy and Research. Ann Emerg Med. 1993;22(9):1198–1210. doi:10.1016/S0196-0644(05)80991-6 [CrossRef]
  3. Centers for Disease Control and Prevention. Progress toward elimination of Haemophius influenzae type b disease among infants and children — United States, 1998–2000. MMWR Morb Mortal Wkly Rep. 2002;51(11):234–237.
  4. Herz AM, Greenhow TL, Alcantara J, et al. Changing epidemiology of outpatient bacteremia in 3- to 35-month-old children after the introduction of the heptavalent-conjugated pneumococcal vaccine. Pediatr Infect Dis J. 2006;25:293–300. doi:10.1097/ [CrossRef]
  5. Stoll ML, Rubin LG. Incidence of occult bacteremia among highly febrile young children in the era of the pneumococcal conjugate vaccine. Arch Pediatr Adolesc Med. 2004;158(7):671–675. doi:10.1001/archpedi.158.7.671 [CrossRef]
  6. Carstairs KL, Tanen DA, Johnson AS, Kailes SB, Riffenburgh RH. Pneumococcal bacteremia in febrile infants presenting to the emergency department before and after the introduction of the heptavalent pneumococcal vaccine. Ann Emerg Med. 2007;49(6):772–777. doi:10.1016/j.annemergmed.2006.10.026 [CrossRef]
  7. Waddle E, Jhaveri R. Outcomes of febrile children without localising signs after pneumococcal conjugate vaccine. Arch Dis Child. 2009;94(2):144–147. doi:10.1136/adc.2007.130583 [CrossRef]
  8. Nader S, Morone NE, Lopez J, et al. Does this child have a urinary tract infection?JAMA. 2007;298(24):2895–2904. doi:10.1001/jama.298.24.2895 [CrossRef]
  9. Shaikh N, Morone NE, Bost JE, et al. Prevalence of urinary tract infection in childhood: a meta-analysis. Pediatr Infect Dis J. 2008;27(4):302–308. doi:10.1097/INF.0b013e31815e4122 [CrossRef]
  10. Trautner BW, Caviness AC, Gerlacher GR, Dremmler G, Macias CG. Prospective evaluation of the risk of serious bacterial infection in children who present to the emergency department with hyperpyrexia (temperature of 106°F or higher). Pediatrics. 2006;118(1):34–40. doi:10.1542/peds.2005-2823 [CrossRef]
  11. Strengell T, Uhari M, Tarkka R, et al. Antipyretic agents for preventing recurrences of febrile seizures. Arch Pediatr Adolesc Med. 2009;163(9):799–804. doi:10.1001/archpediatrics.2009.137 [CrossRef]
  12. Steering Committee on Quality Improvement and Management, Subcommittee on Febrile Seizures. Febrile seizures: clinical practice guideline for the long-term management of the child with simple febrile seizures. Pediatrics. 2008; 121(6):1281–1286.
  13. Schmitt BD. Fever phobia: misconceptions of parents about fevers. Am J Dis Child. 1980;134(2):176–181.
  14. Crocetti M, Moghbeli N, Serwint J. Fever phobia revisited: Have parental misconceptions about fever changed in 20 years?Pediatrics. 2001;107(6):1241–1246. doi:10.1542/peds.107.6.1241 [CrossRef]
  15. Wright AD, Liebelt EL. Alternating antipyretics for fever reduction in Children: An unfounded practice passed down to parents from pediatricians. Clin Pediatr. 2007;46(2):146–150. doi:10.1177/0009922806293922 [CrossRef]
  16. Mayoral CE, Marino RV, RosenfeldGreensher J. Alternating antipyretics: is this an alternative?Pediatrics. 2000;105(5):1009–1012. doi:10.1542/peds.105.5.1009 [CrossRef]
  17. Hay AD, Costelloe C, Redmond NM, et al. Paracetamol plus ibuprofen for the treatment of fever in children: the PITCH randomized controlled trial. BMJ. 2008;337:a1302. doi:10.1136/bmj.a1302 [CrossRef]
  18. Nabulsi MM, Tamin H, Mahfoud Z, et al. Alternating ibuprofen and acetaminophen in the treatment of febrile children: a pilot study. BMC Med. 2000;4:4. doi:10.1186/1741-7015-4-4 [CrossRef]

CME Educational Objectives

  1. Review the current incidence of occult bacteremia in children.

  2. Determine which children with fever without a localizing source are at higher risk of having a urinary tract infection.

  3. Discuss the effects of fever on children and the potential detrimental effects of antipyretics.


Jennifer L. Trainor, MD, is with the Division of Pediatric Emergency Medicine, Northwestern University Feinberg School of Medicine, Chicago. Julie Kim Stamos, MD, is with the Division of Pediatric Infectious Diseases, Northwestern University Feinberg School of Medicine, Chicago.

Dr. Trainor and Dr. Stamos have disclosed no relevant financial relationships.

Address correspondence to: Jennifer Trainor, MD, Children’s Memorial Hospital, Northwestern University Feinberg School of Medicine, 2300 Children’s Plaza, fax: 773-880-8267; or e-mail:


Sign up to receive

Journal E-contents