Diagnostic variation for febrile children in European emergency departments

The study aimed to explore the use of diagnostics for febrile children presenting to European emergency departments (EDs), the determinants of inter-hospital variation, and the association between test use and hospitalization. We performed a secondary analysis of a cross-sectional observational study involving 28 paediatric EDs from 11 countries. A total of 4560 children < 16 years were included, with fever as reason for consultation. We excluded neonates and children with relevant comorbidities. Our primary outcome was the proportion of children receiving testing after primary evaluation, by country and by focus of infection. Variability between hospitals and effects of blood testing on patient disposition were explored by multilevel regression analyses, adjusting for patient characteristics (age group, triage level, appearance, fever duration, focus of infection) and hospital type (academic, teaching, other). The use of routine diagnostics varied widely, mostly in the use of blood tests, ranging from 3 to 75% overall across hospitals. Age < 3 months, high-acuity triage level, ill appearance, and suspicion of urinary tract infection displayed the strongest association with blood testing (odds ratios (OR) of 8.71 (95% CI 5.23–14.53), 19.46 (3.66–103.60), 3.13 (2.29–4.26), 10.84 (6.35–18.50), respectively). Blood testing remained highly variable across hospitals (median OR of the final model 2.36, 1.98–3.54). A positive association was observed between blood testing and hospitalization (OR 13.62, 9.00–20.61).

Conclusion: the use of diagnostics for febrile children was highly variable across European EDs, yet patient and hospital characteristics could only partly explain inter-hospital variability. Focus groups of participating sites should help define reasons for unexpected variation.