External validation of the Melbourne ASSET score for pediatric cellulitis

To perform the first external validation of the Melbourne ASSET Score for predicting the need for intravenous (IV) antibiotic therapy in pediatric emergency department (ED) cellulitis.

This prospective cohort study (January 2022-January 2024) at a Canadian tertiary pediatric ED enrolled children aged 6 months to 18 years with cellulitis. Exclusion criteria included orbital cellulitis, immunocompromised state, clinical toxicity, and inability to tolerate oral antibiotics. The primary outcome was IV versus oral antibiotic administration at 24 h. Secondary outcomes included interrater reliability and treatment failure.

Among 229 children (median age, 5.7 years; 38% female), 54% had an ASSET Score ≥ 4. At 24 h, 69% received oral antibiotics and 31% IV antibiotics. The ASSET Score demonstrated limited discrimination (AUC 0.68; 95% CI 0.61-0.75) and correctly classified 61% of cases. Sensitivity for predicting IV antibiotic use was 74% (95% CI 62-83), specificity 55% (95% CI 47-63), positive predictive value 43% (95% CI 34 52), and negative predictive value 82% (95% CI 73-89). Interrater reliability was substantial (κ = 0.66; 95% CI 0.38-0.95), with variability across individual components. Notably, 57% of children with scores ≥ 4 were successfully managed with oral antibiotics. Treatment failure occurred in 28 patients (12%), with severe complications occurring rarely and exclusively in patients already receiving IV therapy.

In its first external validation, the Melbourne ASSET Score showed limited discriminative performance as a binary decision rule for IV antibiotics, heavily influenced by successful local high-dose oral therapy pathways. It is best utilized as a risk-stratification tool alongside clinical judgment rather than a strict mandate for IV therapy.