Non-blanching rash, particularly with one or more of:

• An ill looking child
  
• Lesions larger than 2 mm in diameter (purpura)
  
• A capillary refill time of 3 seconds
  
• Neck stiffness
  

Meningitis

• Neck stiffness
  
• Bulging fontanelle
  
• Decreased level of consciousness
  
• Convulsive status epilepticus
  

Herpes simplex encephalitis

• Focal neurological signs
  
• Focal seizures
  
• Decreased level of consciousness
  

Pneumonia

• Cyanosis
  
• Tachypnoea: respiratory rate >60 breaths/min if age 0-5 months; >50 breaths/min if age 6-12 months; >40/min if age >12 months
  
• Nasal flaring
  
• Chest indrawing
  
• Crackles on auscultation
  
• Oxygen saturation 95%
  

Urinary tract infection

• Vomiting
  
• Poor feeding
  
• Lethargy
  
• Irritability
  
• Abdominal pain or tenderness
  
• Urinary frequency or dysuria
  
• Offensive urine or haematuria
  

Septic arthritis or osteomyelitis

• Swelling of a limb or joint
  
• Not using an extremity
  
• Non-weight bearing
  

Kawasaki disease

Fever for more than five days and at least four of:

• Bilateral conjunctival injection
  
• Change in mucous membranes
  
• Change in the extremities
  
• Polymorphous rash
  
• Cervical lymphadenopathy
  

Improved evaluation and treatment of febrile illnesses in children will hopefully reduce mortality. A recent national study of deaths from meningococcal disease (the leading cause of death from infectious diseases in children) showed that mortality from meningococcal disease is often associated with late identification, suboptimal treatment, and other deficiencies in health care.¹ There is also concern that the provision of care for children with feverish illnesses varies considerably across the UK. Differences in childhood mortality due to health inequality are also well recognised: another study of meningococcal disease showed that child mortality from meningococcal disease in the most deprived areas of the UK is three times that in the most affluent areas.^w1 One objective of a public service agreement issued by the Department of Health in 2001 is thus to reduce the gap in infant mortality between different social groups by 10% by 2010. Dealing with differences in the management of febrile illnesses in young children with this new NICE guideline may be one way to achieve this.

Methods
This guidance differs from previous NICE guidelines in that it deals with a presenting problem rather than a disease. Our aim was to develop a user friendly format that would also encourage implementation—its presentation as "traffic light" tables highlight levels of risk associated with various symptoms and signs.

This is the first NICE clinical guideline developed by the newly established children's guideline programme at the National Collaborating Centre of Women and Children's Health. The technical team, consisting of a clinic co-director, research fellow, information specialist, health economist, and project co-ordinator, supports the chair and Guideline Development Group (GDG) in the development of the guideline.

We followed standard NICE methodology with highly sensitive literature searches, literature appraisal, and evaluation and translation into recommendations by the GDG.^w2 Although recommendations are based on data from the highest quality level of evidence available for the clinical question, NICE no longer grades recommendations, as the most important recommendations for clinical practice do not always come from the highest level of evidence. Most of our recommendations on assessment and identification of serious illness are based on prospective cohort studies. For a problem-based guideline rather than disease-specific guideline, search terms had to be more inclusive and less precise. Most of the highest level evidence came from prognostic studies, which are not easily identified by standard NICE filters and require the use of special filters. Health economics studies were also integrated into the review of evidence where appropriate.

Where no relevant evidence was found, we used a formal, two round, Delphi consensus method, which is considered more valid than a process of arriving at recommendations within the GDG.^w3 The consensus method involved more than 50 clinicians, parents, and carers, and the GDG made final recommendations according to predetermined criteria.

Our presentation of recommendations with "traffic light" tables does not follow the standard NICE format. These tables are based on, firstly, prospective cohort studies of febrile children with non-specific symptoms and signs associated with serious illness. Most of the evidence is limited to data on infants less than six months old in a secondary care setting. No single symptom or sign is reliably associated with serious illness. The tables are also based on the validated Yale Observation Scale, which shows good correlation for children aged 3 months to 3 years who went on to develop serious illness. We felt, on the basis of clinical experience, that these data could be extrapolated to children up to the age of 5 years in the UK.

Future updates of the guideline will be produced as part of the NICE guideline development programme.^w4

Future research

• A study to confirm the normal range of heart rate for different body temperatures and to determine whether children with heart rates outside these ranges are at higher risk of serious illness.
  
• A UK study to determine the validity of symptoms reported on remote assessment for children with fever (eg via telephone assessment).
  
• Research on referral patterns between primary and secondary care for children with fever, so the health economic impact of this and future guidelines can be estimated.
  
• A UK study of the performance characteristics and cost effectiveness of procalcitonin versus C-reactive protein in identifying serious bacterial infection in children with fever without apparent source.
  
• Studies in primary care and secondary care to determine whether examination or re-examination after a dose of antipyretic medication is of benefit in differentiating children with serious illness from those with less serious conditions.
  

Guideline Development Group
Richard Bowker, James Cave, Jean Challiner, Sharon Conroy, John Crimmins, Annette Dearnum, Jennifer Elliott, Jane Houghton, Edward Purssell, Martin Richardson, Andrew Riordan, Peter Rudd, Ben Stanhope, Bridie Taylor.

Technical team
National Collaborating Centre for Women and Children's Health staff: Adebayo Akande, Michael Corkett, Rosie Crossley, Hannah-Rose Douglas, Monica Lakhanpaul, Chia-Wen Lee, Peny Retsa.

References

w1 Heyderman RS, Ben-Shlomo Y, Brennan CA, Somerset M. The incidence and mortality for meningococcal disease associated with area deprivation: an ecological study of hospital episode statistics. Arch Dis Child 2004;89:1064-8.

w2 National Institute for Clinical Excellence. Guideline development methods: information for national collaborating centres and guideline developers. London: NICE, 2005.

w3 Murphy MK, Black NA, Lamping DL, McKee CM, Sanderson CFB, Askham J, et al. Consensus development methods and their use in clinical guideline development. Health Technology Assessment 1998;2(3):1-88.

w4 National Institute for Health and Clinical Excellence. Updating guidelines and correcting errors. In: The guidelines manual. (Ch 15.) www.nice.org.uk/page.aspx?o=423088