Emergent management of pediatric patients with fever is a common challenge. Children with fever account for as many as 20% of pediatric emergency department (ED) visits,[1] and the underlying disorders in these cases range from mild conditions to the most serious of bacterial and viral illnesses.[2, 3]
Infants younger than 2 months have unique risks for serious bacterial infections; as such, their management is discussed separately from that of older children.
Clinical guidelines have been studied, reported, and scrutinized in major journals for the past 2 decades, yet definitive conclusions are sometimes still elusive and the application to each individual case in the ED is sometimes even more frustrating for the clinician.[4] Inconsistent treatment approaches exist even in the most experienced pediatric EDs.[5, 6] A 2019 European study found that antibiotic prescriptions in european emergency departments is highly variable and physicians frequently prescribe second-line antibiotics.[7]
Some of the fears and anxiety exhibited by parents are shared by ED staff as well. Myths and misperceptions about children with temperature elevations are reported. Fever phobia is well described as existing with both caregivers as well as medical providers.[8, 9, 10, 11]
Because pediatric fever is both a high-impact and a high-frequency chief complaint, the clinician should be knowledgeable about febrile conditions that occur in a variety of age groups of pediatric patients. ED guidelines for treating children with febrile illness are used in order to standardize the approach to care.
This article discusses the appropriate ED management of young febrile children, particularly those younger than 2 years. Neonates (< 28 days) and young infants (28-60 days) are discussed as subsets of this group of pediatric febrile patients.
For more information, see the Medscape Reference articles Fever in the Infant and Toddler and Fever Without a Focus.
Determine the patient's immunization status; see the 2019 immunization schedule.[12] Simply asking whether the child is “up to date” with immunizations may not elicit enough information. Ask how many doses of pneumococcal conjugate (PCV-7) and Haemophilus influenzae type B (HIB) vaccines have been administered. It is also important to inquire about which immunizations have been administered recently.
Some children may not be fully immunized—because of compliance, finances, or perceived health risks,[13] —and thus may be at risk for the infections covered by these vaccines. On the other hand, recent vaccinations might have caused an elevation in the patient's temperature.
Some pediatric patients may have had a subjective determination of an elevated temperature by their caregivers before coming to the hospital but are afebrile when they present to the ED. Parents may report a temperature elevation in their child without having actually recorded the temperature with a thermometer.
Parental reporting of fever on the basis of subjective information (eg, touching the child's torso or extremities or feeling his or her forehead) is a reliable indicator of a fever having been present. Studies have shown that the parental assessment of fever in this situation is usually accurate.[14] On the other hand, one study found that the subjective history of fever in such infants may not correlate with subsequent fever, whereas those with an elevated rectal temperature measured at home had relatively high rates of serious bacterial infection.[15]
Home use of temporal artery thermometers have not been shown to be completely reliable indicators of febrile children.[16]
Parents may be overly concerned about possible outcomes of prolonged high temperature, or they may believe that every fever requires antibiotic therapy.[17, 18] These usually unwarranted fears have been shown to vary by race and ethnicity, as well as by the age of the child and parental education level.[9, 19, 20]
Temperature elevation may not be the only sign of sepsis in neonates and infants. Other potential signs and symptoms of sepsis unique to infancy should also be assessed. History taking is an important part of clinical decision making.
The clinician should ask the patient's parents or caretakers about the following items when they bring in a febrile or ill-appearing child:
Environmental factors, such as being in the heat for a long time during the summer and being overdressed during the winter, may indicate a risk for hyperthermia.
The neonate (birth to 28 days)
The history of the neonate is explored for possible symptoms of poor feeding, vomiting, poor social interaction, changes in the quality of crying, and possible apneic episodes. Any of these findings are reasons to consider serious bacterial infection and may warrant further investigation and laboratory evaluation.
The birth history is explored to ascertain risk factors for underlying pathology, such as prematurity, maternal infections (eg, group B Streptococcus, sexually transmitted diseases such as genital herpes simplex), and congenital or chronic disease states. Neonates at risk for congenital herpes are those born to mothers with a history of recent genital infection and high-risk sexual activity; delivery-related risk factors include rupture of membranes for longer than 6 hours and use of a scalp electrode.
Neonates who present with irritability, seizures, respiratory distress, jaundice, or a characteristic vesicular rash should be considered at risk for neonatal herpes. Note that 10-50% will not develop skin lesions during the course of their illness.[21]
The history is also explored for previous diagnostic studies and their results.
The following questions might be helpful:
There is no significant association between shivering and serious bacterial infections.[7]
It can also be helpful to ask what has been done at home to help control the fever. Was an antipyretic given at home? If so, what dose was given?
In some cases, the clinician finds that child received an inadequate does of antipyretic medication. Over-the-counter medications may not clearly list the correct weight-based dose for children younger than 2 years. Some product instructions simply state "call physician" or "seek medical care." Parents should be educated that the steadily changing weight of their child will result in a need to periodically update the correct dose of medication.[22, 23]
According to the 2003 clinical policy of the American College of Emergency Physicians (ACEP), response to antipyretic medication does not change the likelihood of a child having a serious bacterial infection and should not be used for clinical decision making.[24]
A septic hip clinical decision rule published in 2019 demonstrates low probability of a musculoskeletal infection with zero or one clinical predictor present and moderate predicability with all five predictors. Fever, refusal to bear weight/pseudoparalysis and C-reactive protein > 2.0 mg/dL are found to be the most highly correlated with musculoskeletal infections.[25]
The physical examination of the febrile child is directed at locating a source of or reason for the temperature elevation, with specific attention to potential serious bacterial illnesses. Alternatively, an elevated temperature is not the only vital-sign irregularity that may indicate a potential infectious problem. Hypothermia may be a presenting vital-sign abnormality in septic neonates.
Thermometer use varies between oral, rectal, or axillary. Ear-probe thermometers may not be as accurate as rectal thermometers in the neonate; some study results suggest that operator error is the main reason. Axillary and rectal temperature measurements may also vary widely in neonates.[26] A rectal-probe thermometer is probably most likely to result in an accurate assessment of a neonate's temperature.[27, 28, 29]
Temperature elevation may not be the only sign of sepsis in neonates and infants. Other potential signs and symptoms of sepsis unique to infancy should also be assessed.
The infant's or child's interactions with the parent or caregiver are easily observed while the history is obtained. The following observations should be noted:
The capillary refill time is generally thought to be the quickest assessment of early hypoperfusion. Capillary refill time is faster to obtain than a blood-pressure measurement and is particularly helpful in a loud or busy ED. A delay in the capillary refill time (>2 seconds) indicates hypoperfusion of the skin. Shunting of blood from the capillary beds in the skin is an indication of increased systematic vascular resistance (SVR). An increase in SVR is generally thought to occur early in the course of pediatric hypovolemia. Hypovolemia can result from obvious conditions, such as blood loss and vomiting and diarrhea, or from more subtle reasons, such as tachypnea and sweating.
A 2014 study evaluated the significance of four physical exam signs when looking for pediatric septic shock prior to hypotension: cold extremities, mental status, capillary refill, peirpheral pulse quality. It suggested that the presence of at least once sign was significantly associated with organ dysfunction and the relative risk increased with the presence of at least two signs. The sensitivity of these exam findings ranged from 8-54% and specificity 84-98%. Altered mental status and peripheral pulse quality were significantly associated with organ dysfunction while abnormal capillary refill tim eand cold mottle extremities was not.[31]
The presence of dyspnea, tachypnea, grunting, flaring, and retractions in a patient also suggests possible serious illness and requires further exploration (eg, pulse oximetry, chest radiography).[32]
Temperature increases respiratory rate by a mean of 2.6 breaths/minute for each degree rise in temperature. Temperature and age-adjusted respiratory rate has been shown to be superior to age-adjusted respiratory rate for predicting pneumonia.[33]
Hydration status is documented to include specific signs of dehydration such as dry mucous membranes, sunken fontanelle, absence of tears when crying, and/or a lack of urine output (by history).
Persistent irritability despite feeding or inability of parents to console the child is concerning. True irritability and lethargy are physical signs traditionally associated with an ill child. Lethargy is defined as a decrease in the level of consciousness, some examples of which are failure of the child to recognize parents or caregivers, absent eye contact with the examiner, or failure to interact with the environment at an age-appropriate level.
Toxicity is defined as a clinical syndrome with the following: lethargy with (1) poor perfusion as evidenced by delayed capillary refill or (2) cyanosis or other signs of respiratory distress. Cold hands and feet, limb pain, and pale or mottled skin are considered "red flag" manifestations of septicemia.[34]
The presence or absence of meningeal signs is documented in older children. Clinicians should bear in mind that in some infants and younger children (perhaps younger than 12-15 months) who develop meningitis, specific meningeal signs, such as the Kernig or Brudzinski sign, may not be present.
A hemorrhagic rash is classically described as resulting from overwhelming systemic bacterial infection due to meningococcemia but may be due to other (usually serious) infections. The presence of petechiae or purpura in febrile children indicates the need for prompt evaluation and therapy.
Clinical observation scales have been developed to aid in the determination of the degree of illness.[35] Clinical observation by house staff and seasoned clinicians has produced inconsistent results over the reliability and consistency of clinical observation scales. Regardless of the clinical scale used, one predictor of overall wellness of a child is the presence of a smile.[30]
A physical finding of an isolated bacterial illness, such as otitis media or pneumonia, should not preclude the clinician from possibly pursuing a more extensive workup to exclude sepsis in the neonate.
Discharging a child from the emergency department (ED) with an unrecognized serious bacterial infection is a potential medicolegal risk for the ED physician. The risk is minimized by performing and documenting an appropriate history, physical examination, and, when indicated, a workup, as well as ensuring effective follow-up care.[36]
Several factors indicate an increased risk of bacteremia and/or sepsis, including age less than 2 months, an immunocompromised state (eg, neutropenic or underlying malignancy), being unvaccinated or undervaccinated, hypothermia (core temperature < 36.8°C, or 98°F), and hyperthermia (core temperature >40.5°C, or 105°F).[37]
Patients younger than 2-3 months have an immature immune system and are more susceptible to infections. Among febrile infants less than or equal to 60 days, there is an 8-13% incidence of serious bacterial infections, including urinary tract infections, bacteremia, and bacterial meningitis.[38]
Some pediatric patients may be potentially at a higher risk for bacterial illness than their age-matched counterparts. Special consideration for these populations should include appropriate evaluation in the ED, documentation of consultation with specialists, and chart documentation related to special needs.
Patients with inserted medical devices (eg, foreign bodes) may be at increased risk for infection related to the device. A documented discussion with an appropriate consultant may be indicated. Examples of such devices include the following:
Patients with sickle cell disease (SCD) are considered functionally asplenic. Asplenic patients are at exceptional high risk of sepsis due to encapsulated organisms, such as Streptococcus pneumonia. Because of this, patients with SCD are often hospitalized for observation and empiric treatment.[39] If these patients present with fever, they are considered at higher risk for occult bacteremia and treated accordingly. Blood cultures, CBCs, and differentials are assessed. Parenteral antibiotics against encapsulated organisms are given early.
Patients who are absolutely or relatively immunocompromised (eg, patients with underlying malignancy, patients undergoing chemotherapeutic regimens) must be evaluated for possible occult bacterial infection. Fever and neutropenia should be considered, and a consultation with the oncologist may be indicated.[40]
The approach to patients with HIV infection should be aggressive enough to determine presence of serious bacterial illness. History includes the parent’s or patient's knowledge of their level of CD4 cells and viral load. A sepsis workup is usually indicated in HIV-positive patients presenting with fever. Early consultation with the primary care provider or infectious disease specialist is indicated in the care of these patients.
Infants who were born prematurely may be at risk for infections, such as bronchiolitis or other complications (eg, necrotizing enterocolitis). Consultation with the primary care provider or high-risk neonatologist may be indicated.
Patients with cystic fibrosis are especially susceptible to pneumonia and should be liberally evaluated for such. Consultation with a pulmonologist may be indicated.
Ideally, the decision to admit or discharge these patients is established by a written protocol or guidelines previously established in conjunction with the department of pediatrics. A documented discussion with the specialist may be indicated.
The evaluation of patients presenting with a simple febrile seizure focuses on determining the cause of the fever. In children aged 2-24 months with simple febrile seizures, the risk for occult bacteremia is about the same as that of patients with fever alone.
In children who remain ill appearing and/or who have signs of toxicity, the possibility of meningitis might be considered.[41] Young children with febrile seizures, especially those younger than 12 months, may not always reliably demonstrate physical signs of meningitis.
The child with a febrile seizure should be monitored for some time in the ED, and findings on serial examinations should be documented to differentiate children who are ill with occult disease from children who may be safely discharged home with close outpatient follow up.[42]
Lumbar puncture (LP) may be deferred in a patient who, after appropriate use of an antipyretic, exhibits normal activity. LP is reserved for the younger patient, the patient already receiving antibiotics, or the patient whose physical findings cause concern.
Guidelines have been applied to neonatal emergency medicine. Traditionally, a febrile neonate (temperature >100.4°F [>38°C]) undergoes a full sepsis workup, which includes a CBC count, urinalysis, blood culture, urine culture, chest radiography, and diagnostic LP. The age group that is defined for this workup may vary. At minimum, the guideline is applied to neonates younger than 28 days, and in some institutions or regions of the country, it may be applied to infants as old as 60 days.
A full evaluation for sepsis in febrile neonates is warranted because their risk for serious bacterial infection (SBI) is relatively high. Even among those who otherwise meet low-risk criteria, the rates of SBI are about 5%.[37, 36] The most common serious bacterial infection in infants younger than 24 months of age is urinary tract infection, at a prevalence of 5-7%. The prevalence is as high as 20% in uncircumcised male infants.[43]
About 10% of febrile neonates who have documented respiratory syncytial virus (RSV) infection or clinical bronchiolitis have SBI[44] ; most of these are urinary tract infections (UTIs). This infection rate is about the same as for those without documented RSV infection or clinical bronchiolitis; therefore, a full evaluation for sepsis should be performed for these patients.
For infants younger than 28 days, the median CSF values were as follows:
One study determined that adding polymerase chain reaction (PCR) testing for enterovirus to routine CSF testing reduced the duration of hospitalization and antibiotic use for young infants.[45]
Another study determined normal CSF profiles among a large cohort of febrile infants with atraumatic LPs, negative CSF bacterial cultures, and negative enteroviral PCR testing.[46]
Evaluation of febrile neonates for herpes simplex infection should be considered when the following risk factors are present:
Laboratory studies to screen for herpes infection may include the following:
Most febrile infants aged 5-8 weeks who present to the ED warrant a full evaluation for SBI because they are difficult to evaluate clinically and because ED physicians cannot ensure adequate follow-up.[47] Kupperman et al. recently derived and validated an accurate prediction rule to identify febrile infants 60 days and younger at low risk for serious bacterial infections (SBIs), using three variables: a normal urinalysis, absolute neutrophil count < = 4090/microliter and serum procalcitonin < = 1.71 ng/ml. Their rule had a sensitivity of 97/7%, specificity of 60% and negative predictive value of 99.6%.[38]
The threshold for performing an LP in these infants should be low; many of these infants will warrant CSF examination. Patients who were born at < 37 weeks gestation, had a prolonged NICU stay, have other comorbidities or have been on recent antibiotics, are at higher risk for bacterial meningitis.
LP may be deferred in well-appearing, immunocompetent, full-term febrile infants of this age group who seem to have a viral infection.[43]
In select cases, at the clinician's discretion, an LP may be omitted in well-appearing infants who have blood and urine studies obtained and when the following apply:
Patients who appear to have a viral infection A recent study defined median CSF profiles among uninfected infants in this age range.[46]
Urine studies are important in this age group, as UTIs are a common source of SBI.[48, 49, 50] Additional studies such as stool cultures are performed selectively. Chest radiographs do not need to be obtained routinely in those without signs of respiratory disease.[51, 52]
In children who have been immunized against pneumococcal disease, the routine use of CBC and blood culture are no longer recommended.[53, 54, 55] If a CBC is obtained, interpretation of the total WBC remains controversial. Distinguishing between a high total WBC count versus a very high WBC count may leave the physicians with even more questions about the relevance of this laboratory result.[53, 56, 57]
The rates of bacteremia and invasive pneumococcal disease have dramatically declined since the licensing of the 7-valent pneumococcal vaccine (PCV7) vaccine in 2000.[13, 58, 59, 60, 61, 62, 63] This vaccine protects against the 7 pneumococcal serotypes that cause 85% of invasive pneumococcal disease in children, including nearly all of the serotypes that are highly penicillin resistant.
Prior to routine use of the pneumococcal vaccine, occult bacteremia occurred with an incidence of 3-5% in children younger than 24 months with fever. Studies in the 1980s-1990s showed the rate of occult bacteremia was as high as 5%. In the 21st century, studies show a decline in the rates to as low as 0.5-1%.[64] This change is most likely due to the increasing rates of pneumococcal vaccinations.[65] E.coli is now the leading cause of bactermia, UTIs and bacterial meningitis in young children.[43]
Before widespread PCV7 use, approximately 60-70% of all cases of occult bacteremia were caused by Streptococcus pneumoniae. In 2010, the US Food and Drug Administration (FDA) licensed a 13-valent pneumococcal vaccine (PCV13), which includes protection against additional strains of pneumococcus, including serotype 19A. This PCV13 vaccination supplanted the PCV7 vaccination, presumptively causing a further decline in invasive pneumococcal disease.[66, 67] The routine use of pneumococcal vaccine has essentially made rates of occult bacteremia a classic historical discussion.[63]
Before routine use of the HIB vaccine, HIB accounted for 20% of occult illness, but this cause also decreased in frequency after the vaccination became routine in the 1990s.[68, 69]
An increase in the total band count or erythrocyte sedimentation rate (ESR) is not more predictive of occult pneumococcal bacteremia than an elevated WBC count alone. Before the routine use of the pneumococcal vaccine, a WBC count above 15,000/mm3 had been reported to be 70% sensitive for predicting occult bacteremia from pneumococcus.[70]
Children in this age group who present with a higher temperature (>102.9°F [>39.4°C]) may be at risk for occult bacteremia, especially if they are underimmunized or immunosuppressed. By definition, occult means that the patient exhibits no other signs or symptoms suggesting the etiology of the temperature elevation.
Other less common etiologic agents are Neisseria meningitides and (especially in patients with sickle cell disease) Salmonella species.[58] Herpes and community-acquired methicillin-resistant Streptococcus aureus (MRSA) are now emerging as more common pathogens in neonates.[71, 72, 73, 74, 75, 76, 77, 78]
Children younger than 24 months presenting with the clinical syndrome of bronchiolitis are at a lower risk of bacteremia and UTIs. Therefore, routine urine and blood culturing in previously healthy children presenting with fever and bronchiolitis is usually not indicated; extreme fever or ill appearance may be indications to obtain a blood culture. Testing of the nasal secretions for a viral etiology such as respiratory syncytial viral (RSV) or influenza A/B may be helpful.[79]
Rates of UTI may be lower in patients with bronchiolitis than those without any fever source.[80]
Viral infections are the most common cause of fever in young children aged 2-24 months. Well-appearing children with unremarkable histories and physical examinations may be discharged home without laboratory testing or presumptive use of antibiotics. An LP should be considered for those who are irritable, lethargic, inconsolable, or toxic appearing.
During summer months, children with fever and no other signs may have an enterovirus infection. Some studies report the incidence as high as 50% in febrile children in the ED. Enteroviral infection is a clinical diagnosis for the emergency physician. No specific laboratory testing is indicated.
Diarrhea most commonly has a viral etiology in this age group. The presence of diarrhea with blood or mucus or recent use of antibiotics may be indications for ordering stool cultures to assess for a bacterial etiology.
The routine acquisition of CBC and/or blood cultures to screen for occult bacteremia in immunocompetent children is strongly discouraged for the following reasons:
Similarly, the presumptive use of broad-spectrum antibiotics is strongly discouraged because of low bacteremia and complication rates, high costs, adverse drug reactions, and increasing rates of antibiotic resistance due to indiscriminate antibiotic use, as well as because the outcomes for febrile young children not treated presumptively with parenteral antibiotics are uniformly good.
For clinically ill-appearing children, a sepsis workup to include a blood culture should be obtained. This is especially true in children ill enough to warrant an LP as described above, those exhibiting signs of shock, or those with a petechial rash. Use of serum markers for sepsis, including a lactate level and procalcitonin, have been described as being helpful in stratifying the sickest children.[83, 84, 85, 86] There is still no widely accepted clinical decision rule or screening test to identify serious bacterial infections in children. Some markers which have been studied include white blood cell count, absolute neutrophil count, band count, c-reactive protein, interleukins, and procalcitonin.[43]
The febrile child older than 24 months (who has been previously fully immunized) is primarily evaluated by obtaining a history and performing a physical examination.
Specific workup and/or treatment is based on the clinical findings and suspicion of disease.[87, 88, 89]
UTI is a common cause of SBI in infants and young children.[90, 48, 49, 50] Clinicians should have a lower threshold for ordering UA and urine cultures than for screening for bacteremia, especially for those with risk factors.
For boys, risk factors for UTI include age younger than 6 months and being uncircumcised.[91] For girls, risk factors for UTI include the following:
In particular, clinicians should have a lower threshold for UTI screening for white girls; in 2 separate studies, UTI rates were as high as 17% in this group.[91, 92]
Urethral catheterization is the method of choice to obtain urine to avoid contamination with skin flora. Urine collected in a bag placed on the perineum is often contaminated and should not be evaluated for bacterial culture.[93, 94]
A 2018 study found a combined analysis of leukocyte esterase and, or, nitrites yielded a sensitivity of 86% and a specificity of 80% for the diagnosis of UTIs in clean-catch samples of febrile infants less than 90 days old, suggesting that this is an accurate screening test for UTIs in febrile infants of this age group.[95] Limiting catheterization to children with positive screens from urine bag specimens reduced catheterization rates among febrile young children ages 6-24 months from 63% to < 30% without missing UTIs or prolonging ED stays[96] .
Suprapubic aspiration of the bladder for urine should be discouraged. Urethral catheterization is less invasive and less painful and is associated with a higher yield of urine.
An enhanced UA, with a hemocytometer cell count and Gram stain of unspun urine, is more sensitive than a standard UA.[97] In febrile patients aged 2 months to 2 years, a preliminary diagnosis of a UTI can be made if there is a positive urine leukocyte esterase, nitrites, leukocyte count or Gram's stain.[43] A urine culture should be sent when starting antibiotics in these patients.
In bacteremic children with UTI, blood cultures and urine cultures are likely to have identical organisms with identical antimicrobial sensitivities.[98]
A study by Tzimenatos et al that included an analysis of data from 4147 febrile infants ≤60 days old reported that for the 289 infants with a UTI and colony counts ≥50 000 CFUs/mL, a positive urinalysis regardless of bacteremia showed sensitivities of 0.94; 1.00 with bacteremia; and 0.94 without bacteremia. Specificity in all groups was 0.91.[99]
A retrospective study by Shaikh et al that included 482 children younger than 6 years old with a first or second UTI found that renal scarring was associated with a delay in the initiation of antimicrobial therapy. 35 children (7.2%) developed new renal scaring and the median duration of fever before initiation of antibiotic therapy in the group with renal scaring was 72 hours compared to 48 hours in children with no renal scarring.[100, 101]
Diarrhea in children is commonly caused by viral organisms and is usually not considered a major source of fever. However, diarrhea with blood or mucus is an indication for further testing for fecal leukocytes, which suggests invasive bacterial etiologies.
If fecal leukocytes (>5 per high-powered field) are present, a bacterial etiology is suggested and cultures are indicated. This finding may indicate infection with species of Salmonella, Campylobacter, Shigella, Yersinia, or toxic strains of Escherichia coli. The final diagnosis can be made only by obtaining stool cultures.
During winter months, children presenting with low-grade fever, vomiting, and diarrhea should be considered possibly infected with rotavirus. Children in day care centers are at increased risk for rotavirus infection. Rotavirus vaccine may change the incidence of this clinical etiology as well.
Chest radiographs are most useful for patients with high fever accompanied by focal pulmonary examination findings. Chest radiographs should not be routinely performed for fever alone, in the absence of lower respiratory tract findings or hypoxemia.
A study from Spain found that 13% of young children with fever above 39.0°C and a peripheral white blood cell (WBC) count of greater than 20 × 109/L had occult pneumonia.[102] However, just 20% of the radiographs were interpreted by radiologists, and the authors estimate that just one third to one half of the children had received the pneumococcal vaccine.
Given the very low positive predictive value of the WBC count, its routine acquisition in the setting of high fever should be discouraged. On the other hand, if the WBC is obtained and is markedly elevated, consideration should be given to obtaining a chest radiograph, in the absence of an alternative source of fever.
In febrile neonates and young infants, a chest radiograph may only be a routine part of a sepsis workup in the presence of respiratory signs in neonates (eg, rales, grunting, flaring, retractions, hypoxia) or lower respiratory tract findings (eg, cough, tachypnea) in infants. An increased respiratory rate is the earliest indicator of respiratory distress and should be considered in the overall decision to obtain a chest radiograph.
In febrile children aged 3-24 months, pneumonia may be present even in the absence of definite auscultatory signs. An abnormal respiratory rate or pulse oximetry should alert the emergency physician to the need for a chest radiograph. Chest radiography is indicated if the child shows signs of respiratory distress, such as tachypnea; grunting, flaring, or retractions; or hypoxia, as determined with pulse oximetry. A history of prolonged fever (greater than 48 hours) or cough may also be predictors of occult respiratory tract infection.[103, 43]
In children older than 2 years, chest radiography is not routinely ordered unless a specific indication is present, such as prolonged cough, tachypnea, or hypoxia.[104] Chest radiographs may also be deferred in patients with wheezing or a high likelihood of bronchiolitis.[43]
Diagnostic LP is suggested for ill-appearing febrile infants and children associated with any of the following:
In young children, the classic signs of meningismus, such as the Kernig or Brudzinski signs, may be absent, even in the presence of CNS infection. See Pediatric Bacterial Meningitis.
Prehospital care provided by emergency medical service (EMS) personnel may vary from region to region based on regional differences in education, training, and transport times. Children with fever alone may require no specific intervention in the EMS setting, except for those other conditions that exist in the presence of fever.
Assessment of and attention to airway, breathing, and circulation is recommended first for all ill children. Securing intravenous (IV) access is part of the prehospital care, especially in toxic-appearing children. Children who have had prolonged or very high fever may become dehydrated, and assessment of the hydration status is indicated. Some children may be prone to febrile seizures. Seizure precautions should be followed in those with a history of febrile seizures.
Clinicians first need to address airway, breathing, and circulation if this has not already been done. Manage the airway (supplemental oxygen or intubation as needed) and secure intravenous access if it was not established during prehospital care. Monitor the patient's heart rate, blood pressure, and pulse oximetry readings. Administer intravenous fluids, such as normal sodium chloride solution with boluses (weight based), as needed for hypotension. Pressors should be given as needed for hypotension that does not respond to fluid.
Vital signs recorded in triage should routinely include a temperature in young children. This should be recorded for all neonates and infants presenting with fever or other potentially infectious complaints (eg, not eating well, not breathing well, color changes, behavior changes). A rectal temperature is the most accurate for infants.[105]
Nursing triage includes an accurate assessment of the patient's weight to ascertain the correct dose of an antipyretic or other medications. An antipyretic should be given as early as possible during the ED visit.
Reduction of fever is used to help comfort the child as well as provide for the optimal examination conditions and should be addressed in triage protocols. Acetaminophen and ibuprofen have both been used for fever control.[23] These drugs are sometimes alternated to achieve an overlapping period of fever control.[106, 107, 108] Some studies suggest that acetaminophen may reduce fever more quickly, while ibuprofen may have a longer-lasting effect on fever reduction.[22, 109]
In a study of children aged 6 months to 6 years whose fever could be managed at home (37.5-41°C) Hay et al, found that the use of acetaminophen and ibuprofen improved time without fever during the first 4 hours and was superior to acetaminophen alone, but not ibuprofen alone.[107]
The combination also decreased fever 23 minutes faster than acetaminophen alone, but no faster than ibuprofen alone. Time without fever during the first 24 hours was improved with the combination compared with either acetaminophen or ibuprofen. Dosage was calculated as acetaminophen 15 mg/kg/dose and ibuprofen 10 mg/kg/dose.[107]
The ED staff should be educated that a positive response to antipyretics (evidenced by a drop in the measured temperature) does not predict the absence of a potentially serious bacterial disease.[109]
Criteria for discharge from the ED do not necessarily include reduction of the child's fever to a certain level before discharge. No evidence indicates that fever reduction is necessary for the child to be discharged from the ED.
The infant with only a history of subjectively estimated fever whose rectal temperature is normal on repeated measurements and who has an entirely normal clinical evaluation may be assessed as not requiring laboratory studies. All such infants discharged home still warrant close follow-up and continued short-term monitoring of rectal temperature.[1]
Neonates with fever who undergo a routine workup to rule out sepsis should then receive 2 antibiotics as treatment for their potential septicemia: an aminopenicillin (eg, ampicillin) and a cephalosporin (eg, cefotaxime).[110] Local surveillance for ampicillin resistance is critical.[111, 112] Acyclovir should be considered in all neonates who are ill appeareing or who are at risk for neonatal herpes infection.
Administer antibiotics as indicated:
Corticosteroid therapy is not routinely indicated in the treatment of young infants with bacterial meningitis.
Controversy continues over whether early bacterial meningitis or other severe systemic bacterial illness in these patients may not be apparent. Therefore, all such patients discharged from the ED require meticulous examination and thorough documentation.
Any discharge protocol in which patients may be observed on an outpatient basis should be developed collaboratively by the department of emergency medicine and the department of pediatrics.
A study sought to reduce time to antibiotic for children presenting to the emergency department with fever and neutropenia. The authors reported that implementing a standard approach to patients at risk for neutropenia decreased time to antibiotic which was associated with improved patient outcomes. Some of the challenges in providing timely antibiotics to children included venous access (time to effect of topical anesthetics, and difficulty obtaining access), physicians waiting on laboratory results, and antibiotic availability.[113]
The decision to admit older neonates and young infants for presumptive sepsis should be based on several factors, including their ability to be reevaluated by a primary care provider the next day, toxic appearance, need for monitoring, need for hydration or other supportive measures, or poor social situation.
Admission is warranted for febrile infants 28-56 days old; however, these patients may be discharged home if they meet all of the following low-risk criteria[114, 115, 116, 117, 118, 119, 120] :
Follow-up necessitates reliable parents with a telephone at home, with transportation readily available to them, and with follow-up care arranged with the primary physician the next day. If next-day follow-up with the primary physician cannot be arranged, the patient should return to the ED. Diagnosis-specific information sheets can enhance parental understanding of ED discharge instructions.[121]
Older children with a low-grade fever, no risk factors, no localized signs of infection, a good appearance, and no irritability may require only symptomatic treatment and close follow-up care. These patients do not routinely need laboratory evaluation, radiography, or empiric antibiotics.
For patients discharged from the ED, close follow-up with the primary care physicians should be arranged, and an ED protocol for notification of positive cultures is needed. Primary care physicians or ED physicians should notify caretakers if an outpatient has a positive blood culture. Such children, if they remain febrile, may require collection of CSF, a repeat blood culture, and hospitalization.
Primary care physicians or ED physicians should notify caretakers if a child has a positive urine culture. Older children with pyelonephritis may continue outpatient treatment if they appear well and can tolerate oral antibiotics. Clinicians should have a low threshold for hospitalizing those who are younger than 6 months with pyelonephritis.
The child should return either to the ED or to the primary care physician's office if his or her clinical presentation worsens after discharge.
All febrile neonates aged 0-28 days old should be hospitalized for presumed sepsis and treated with antibiotics until cultures of CSF, blood, and urine are all negative for growth after 48 hours.[122, 115, 123] Two parenteral antibiotics are administered during these inpatient admissions: an aminopenicillin, such as ampicillin, and either a cephalosporin, such as cefotaxime, or an aminoglycoside such as gentamicin. Infants less than or equal to 21 days should also be treated with acyclovir.
If admitted, febrile patients outside of the neonatal period are prophylactically treated with an antibiotic (eg, a third-generation cephalosporin, such as ceftriaxone), until cultures are shown to be negative.[124]
Iatrogenic risks are involved with routine hospitalizations of all febrile neonates and young infants.[114] Protocols developed in advance, in cooperation with the ED staff as well as the pediatric staff, streamline the approach at each institution.
If the initial facility is not designed for pediatric inpatient care, transfer to a pediatric facility as needed after his or her condition is stabilized in the ED and after initial workup and treatments are completed.
To ensure the quality of care, an ED policy should be established for positive culture results requiring a call back to the care provider or to the patient.
For a positive cerebrospinal fluid culture, a patient who was sent home should return to the hospital for intravenous antibiotics until identification of the organism is confirmed. Many EDs routinely admit all patients receiving lumbar puncture in the ED.
For a positive blood culture, the patient should be reevaluated by a physician. An afebrile patient who is doing well may be treated on an outpatient basis after infection with penicillinase-resistant Streptococcus pneumoniae is considered. A repeat blood culture is indicated in most cases.
For a positive urine culture, the patient may be treated on an outpatient basis, with follow-up care and repeated cultures to prove effective treatment.[125]