Fever in early infancy:

Further understanding of the physiology of fever and meta-analysis of previous studies of febrile illness in infants under 3 months of age have contributed to a more rational clinical approach in these young babies. More detailed analysis of the use of the white cell count, lumbar puncture, C-reactive protein, urinalysis and chest radiography has improved the efficiency of investigation. The risk of bacterial infection in well-looking, febrile young infants is 5-8%. The use of the Rochester criteria enables this risk to be reduced to 1% if all criteria are satisfied. Decision analyses have delineated alternative management strategies but different environments, illness prevalence, observer experience and parent reliability need to be considered.