Childhood asthma prediction models: a systematic review

Area of Science:

• Pediatric Pulmonology
• Epidemiology
• Clinical Prediction Modeling

Background:

• Early identification of children at risk for school-age asthma is critical for timely intervention.
• The clinical utility of existing childhood asthma prediction models remains uncertain.

Purpose of the Study:

• To systematically review existing prediction models for identifying preschool children at risk of developing asthma by school age.
• To assess the performance and characteristics of these prediction models in clinical practice.

Main Methods:

• Systematic review of prediction models for asthma development in children aged 4 years or younger with asthma-like symptoms.
• Included studies assessed asthma between 6 and 12 years of age.
• Identified 12 models across four cohort types: healthcare visits, parent-reported symptoms, high-risk, and general population.

Main Results:

• Four basic models used non-invasive predictors (family history, allergies, symptom severity).
• Eight extended models incorporated additional clinical tests, primarily specific IgE.
• No single model demonstrated superior predictive performance for both identifying and ruling out asthma development simultaneously.

Conclusions:

• Existing asthma prediction models for preschool children show variable performance.
• Accurate prediction of asthma development in preschool children with wheeze remains difficult, highlighting limitations in current tools.
• Further research is needed to develop more robust and clinically applicable prediction models.