There is controversy regarding the most appropriate thermometer and the best anatomical site for temperature measurement.³ Parents often use touch to detect fever in their children. However, touch has only 50% specificity.⁴ It tends to overestimate the incidence of fever, and is more useful to exclude fever. Rectal temperature has long been considered the gold standard for routine measurement of body temperature, but it does not in fact reflect true core temperature within the pulmonary artery. Moreover, parents and patients generally prefer other temperature assessments. Nonetheless, rectal temperature remains the most widely used measure in infants under 3 months of age. Tympanic thermometers provide the most accurate assessment of core temperature, but the probe may be too large for an infant’s auditory canal. Oral temperature requires patient cooperation, and is generally unsuitable for children under the age of 5 years. Axillary temperature measurement is inaccurate and insensitive.

The definition of fever is

• 38°C (rectal or tympanic);

• 37.5°C (oral); or

• 37.2°C (axilla).

Fever: to treat or not to treat?

The drugs most commonly used for treating fever are paracetamol, ibuprofen and aspirin. The routine use of these medications in the treatment of fever has been questioned.⁵ In particular, there has been concern that the use of antipyretics may prolong viral shedding, impair antibody response to viral infection, and may increase morbidity and mortality.⁵^–⁷ Moreover, each of the commonly used antipyretics may have significant adverse effects such as hepatic dysfunction, metabolic acidosis, Reye syndrome and gastrointestinal bleeding. Treatment should therefore be focused on alleviation of discomfort or pain rather than on the height of the temperature. Either paracetamol or ibuprofen may be used. It is important to note that the use of antipyretics has not been shown to prevent febrile convulsions.

Paracetamol may be given orally, rectally or intravenously at a dose of 10–15 mg kg⁻¹ 4–6-hourly. In an unsupervised, community setting, the total daily dose should be limited to 60 mg kg⁻¹, although up to 90 mg kg⁻¹ per 24 hrs can be used under medical supervision. Single doses of 30 mg kg⁻¹ may be used for night-time dosing. Serious toxicity has been reported in children with chronic daily over-dosage, mostly occurring in children who have a febrile illness and associated anorexia, vomiting and/or dehydration.⁸ A child should be reviewed after 48 hours if regular paracetamol has been ‘required’ for this period.

Ibuprofen can be used as an alternative to paracetamol at a dose of 5–10 mg kg⁻¹ (maximum of 500 mg per dose), given 6- to 8-hourly (maximum daily dose of 40 mg kg⁻¹ or 2 g). It is recommended that it be used alone, and not in combination with paracetamol, as this practice may lead to an increase in adverse effects, including gastrointestinal bleeding, renal dysfunction and anaphylaxis.⁷ A theoretical risk of aggravating concurrent asthma has also been described, although these adverse effects are refuted in large prospective studies.⁸ There is also a concern that ibuprofen may be associated with an increased risk of necrotising group A streptococcal infections.⁹ There is no evidence that alternating paracetamol and ibuprofen is any better at reducing fever or spares the potential hepatotoxicity related to paracetamol administration.⁹

Practical approach to the febrile child

The majority of febrile children will have self-limiting viral infections. The challenge is to identify those children at risk of serious illness while avoiding unnecessary investigation and treatment of children with benign viral illnesses.

Fever in children may be classiﬁed into three groups:

• fever with localising signs;

• fever without focus;

• pyrexia of unknown origin.

Fever with localising signs

A careful history and examination will identify the source of infection in most patients. These children should be managed according to the individual condition and its severity.

Fever without focus

In a small number of children presenting with fever, no focus is found. While most will have a viral infection, a more serious illness such as a urinary tract infection (4–5%), occult bacteraemia (<1%) or meningitis (<0.2%) may be present.

Occult bacteraemia is the presence of bacteria in the bloodstream of a febrile child who has no apparent focus of infection and looks well. Diagnosis is by blood culture and exclusion of focal infection. The incidence of occult bacteraemia in febrile children has reduced dramatically to <1% since the introduction of conjugate pneumococcal vaccine.^10,¹¹

Most children who present with fever and no identiﬁable focus appear otherwise well. History should include details about immunisation status, infectious contacts, travel, diet and contact with animals or insects. A thorough physical examination should be performed, paying particular attention to general appearance (colour and level of activity) and vital signs (respiratory rate, pulse, peripheral perfusion and blood pressure).

It is difficult to assess whether a child is ‘septic’ or ‘toxic’. A simple and effective approach that is useful in the ED is a combination of ABC, fluids-in and fluids-out.¹² An infant with one or more of these symptoms or signs has a higher risk of serious illness:

• A – poor arousal, reduced alertness and reduced activity.

• B – breathing difficulty.

• C – poor perfusion.

• Fluids in – the frequency of feeding over the 24 hours prior to presentation, <50% of normal over 24 hours prior to presentation, suggesting dehydration.

• Fluids out – significantly abnormal urine output of <50% of usual output.

Other features on examination that strongly suggest a seriously ill infant include pallor, purpuric rash, high-pitched scream and bulging fontanelle.

Patients with unexplained fever with a higher likelihood for serious infection include the following patient groups or conditions:

• neonates (<1 month of age);

• incompletely immunised children;

• immunocompromised (e.g. congenital immunodeﬁciency, human immunodeficiency virus (HIV), neutropenic and other oncology patients, cytotoxic drugs and steroids);

• asplenic children (congenital, post splenectomy or functional, e.g. sickle cell disease);

• patients who have received prior oral antibiotics; many of these patients have a viral infection, but meningitis or other serious bacterial infection must be considered;

• children with fever and prolonged convulsion;

• children with underlying medical conditions (e.g. cystic ﬁbrosis, structural cardiac defects, etc.);

• children with central venous devices, shunts or other foreign material.

When considering management strategies for febrile infants, three age groups are generally assigned: <1 month of age, 1–3 months and >3 months (Table 9.1.1). Infants less than 1 month of age, and those with any of the risk factors above require several investigations including full blood examination, culture of blood, urine and cerebrospinal fluid (CSF), and a chest X-ray if indicated. Empiric antimicrobial therapy should be based on the patient’s clinical illness, risk factors, and the local epidemiology of potential pathogens and their antibiotic susceptibility.

Age Investigation Management <1 month FBE; blood, urine and CSF cultures; CXR Admit
Empiric IV antibiotics: amoxicillin and cefotaxime 1–3 months Urine culture
± blood and CSF cultures ± CXR Consider admission and observation
Discharge with arranged review >3 months Consider urine culture Discharge with arranged review

CXR, chest X-ray; FBE, full blood examination; CSF, cerebrospinal fluid.

Clinical scores, such as the Rochester and Boston criteria, have been devised to identify children at low risk of serious bacterial infection.¹³ However, their utility has been questioned in the era of widespread Hib and conjugate pneumococcal vaccination.

Febrile infants between 1 and 3 months of age who appear well and do not have risk factors may not require blood tests or a lumbar puncture, although urine microscopy and culture is advisable. Those over 3 months of age do not routinely require laboratory testing or treatment, although urine microscopy and culture may still be appropriate.

There is no evidence that oral or parenteral antibiotics prevent the rare occurrence of focal infections from occult bacteraemia; instead, they result in delayed diagnosis, drug side effects, additional costs and the development of resistant organisms. What is required is a careful clinical assessment, parental education and review within 24 hours.

As urinary tract infection is the most common serious bacterial infection among febrile infants and children, urine microscopy and culture should be included in the investigation of most such children. In infants, a urine sample should ideally be obtained via suprapubic aspiration or catheter. A negative urinalysis does not exclude a urinary tract infection, which may occur in the absence of pyuria.^14,¹⁵

Other rarer causes of fever should also be considered:

• Kawasaki disease;

• connective tissue disease (e.g. juvenile arthritis);

• inflammatory bowel disease;

• malignancy (e.g. leukaemia).

Pyrexia of unknown origin (PUO)

PUO is defined as prolonged fever (usually greater than 10 days) where history, examination and ‘routine’ investigations have not identified a cause. Occasionally, fever appears to occur in a repetitive or periodic pattern. In either case, a chronic or non-infectious condition should be considered, such as juvenile idiopathic arthritis and other connective tissue diseases, inﬂammatory bowel disease or malignancy. Infectious causes include systemic viral syndromes (such as infectious mononucleosis), upper or lower respiratory infections (e.g. sinusitis), urinary tract infection, central nervous system infection, bone infection, tuberculosis, abscess (e.g. parameningeal, intra-abdominal), endocarditis and enteric infections (e.g. typhoid fever).

More extensive investigations are often required, including specific investigations for mycobacterial, fungal or viral infections. Imaging may be required, looking for occult abdominal or central nervous system collection, for osteomyelitis or endocarditis.

Kawasaki disease is an important consideration in an infant or child presenting with prolonged fever, and diagnosis is often delayed. There is a degree of urgency in diagnosis, because treatment within 10 days of onset of fever with intravenous immunoglobulin and aspirin reduces the incidence of coronary artery lesions from around 20% to around 5%.¹⁶

Empiric antibiotic therapy

With the possible exception of bacterial meningitis, where Gram stain results may guide therapy, the most appropriate antibiotic therapy in children must be based on epidemiological grounds. The most important factors determining the likely pathogens, which should be targeted by empiric therapy, are:

• age;

• presumed focus of infection;

• presence of underlying disease or anatomical abnormality;

• whether the infection is hospital or community acquired.

In addition, the site of infection may have implications for the expected penetration of the antibiotic chosen (e.g. aminoglycosides do not penetrate into abscess cavities and are inactive in an anaerobic environment).

For presumed bacterial infection (including meningitis) in the first 3 months of life, empiric treatment must cover Group B streptococci, Escherichia coli and Listeria monocytogenes infections. Recommended antibiotics are: amoxicillin plus cefotaxime.

Amoxicillin 50 mg kg⁻¹ per dose intravenous (IV) 12-hourly (week 1 of life), 8-hourly (week 2–4 of life), 4–6-hourly thereafter. Cefotaxime: 50 mg kg⁻¹ per dose IV 12-hourly (week 1 of life), 8-hourly (week 2–4 of life), 6-hourly thereafter.

For presumed bacterial infection (including meningitis) after 3 months of age, potential pathogens include Neisseria meningitidis, Streptococcus pneumoniae, Group A streptococci and Staphylococcus aureus. Recommended empiric therapy is: flucloxacillin plus cefotaxime.