Finding the sick febrile kid

Finding children with serious illness among the multitudes who present with fever is the number one challenge in paediatric emergency medicine.
A two year prospective cohort study was conducted at the Children’s Hospital Westmead in Sydney to develop and test a multivariable model to distinguish serious bacterial infections from self limiting non-bacterial illnesses.
A standardised clinical evaluation that included mandatory entry of 40 clinical features was recorded by physicians on around 15000 febrile kids under age 5. Clinical, laboratory and radiological follow up was undertaken to identify one of three key types of serious bacterial infection (SBI): urinary tract infection, pneumonia, and bacteraemia.

7.2% had SBI – urinary tract infection 3.4%, pneumonia 3.4%, and bacteraemia 0.4%.
A diagnostic model was developed using multinomial logistic regression methods. Physicians’ diagnoses of bacterial infection had low sensitivity (10-50%) and high specificity (90-100%), whereas the clinical diagnostic model provided a broad range of values for sensitivity and specificity.
The authors suggest that a computer assisted diagnostic decision tool could be used to determine the likelihood of serious bacterial infection.
The strongest positive predictors of serious bacterial infection were a generally very unwell appearance, high temperature, chronic disease, and prolonged capillary refill time. For children with pneumonia, other predictors were coughing, difficulty breathing, abnormal chest sounds, and to a lesser extent tachypnoea, chest crackles, and tachycardia. For urinary tract infection, the presence of urinary symptoms was by far the strongest indicator, whereas for bacteraemia, tachycardia and crying were also strong indicators although an editorial points out that only 64 cases of bacteraemia occurred, so this last result should be treated with caution.
The accuracy of clinical symptoms and signs for the diagnosis of serious bacterial infection in young febrile children: prospective cohort study of 15 781 febrile illnesses
BMJ. 2010 Apr 20;340:c1594