Background: Post hoc analysis of clinical trials shows blood eosinophil counts (BEC) predict future preschool wheeze attacks; however, prospective usefulness in a clinical setting is unreported. We assessed the feasibility of point-of-care (POC) eosinophil measurements in preschool wheezers and related BEC to symptoms, lung function, and utility in predicting attacks.
Methods: Children aged 1-5 years with recurrent wheeze underwent finger-prick sampling during the outpatient clinic for POC eosinophils, forced oscillation technique (FOT) and/or spirometry, and symptom score (TRACK questionnaire). The utility of BEC and/or the other tests in predicting wheeze attacks in the subsequent 3 months was analysed by comparing those with and without an attack and using a predictive decision tree (DT) model.
Results: Seventy-three children (median age 4.27 years) were recruited; BEC were higher in atopic children (median 0.5 × 109/L vs. 0.3 × 109/L non-atopic, p < 0.01). BEC moderately correlated with FOT reactance bronchodilator reversibility z-score changes (r = 0.495, p = 0.005), but no other lung function measures or TRACK score. 68/73 (93%) children were followed up at 3 months. 29/68 (43%) children had > 1 wheeze attack requiring unscheduled healthcare attendance. Absolute and %eosinophils at the baseline visit were higher in those who had an attack (median 0.5 × 109/L vs. 0.3 × 109/L, p = 0.03 and median 6% vs. 4%, p < 0.01). The DT model showed children with BEC ≥ 4% and TRACK score < 75 were more likely to have a future attack (probability 0.63).
Conclusion: POC blood eosinophils were feasible in a clinical setting. Our preliminary data suggest elevated BEC with a low symptom score predicts a wheeze attack within 3 months.
Keywords: biomarker; eosinophils; machine learning; point‐of‐care test; preschool wheeze; risk factor; wheeze attacks.
© 2025 The Author(s). Allergy published by European Academy of Allergy and Clinical Immunology and John Wiley & Sons Ltd.