Application of Asthma Prediction Tools in a Cohort of Infants with Severe Bronchiolitis

Area of Science:

• Pediatric Pulmonology
• Asthma Epidemiology
• Clinical Prediction Models

Background:

• Severe bronchiolitis in infancy is a major risk factor for developing childhood asthma.
• Early and accurate asthma prediction is crucial for timely intervention.
• Existing prediction tools require evaluation in high-risk infant populations.

Purpose of the Study:

• To assess the predictive accuracy of the Asthma Predictive Index (API), modified API (mAPI), and Pediatric Asthma Risk Score (PARS) for asthma at age 6 years.
• To evaluate these tools using infancy data in a cohort of infants hospitalized with severe bronchiolitis.
• To identify the most effective tool for predicting asthma in high-risk infants.

Main Methods:

• A multi-center cohort study of infants under 1 year old hospitalized with severe bronchiolitis between 2011 and 2014.
• Infancy data was used to calculate scores for API (stringent and loose criteria), mAPI, and PARS.
• Asthma diagnosis at age 6 years was determined via parent-reported data.

Main Results:

• The prevalence of parent-reported asthma at age 6 years was 37%.
• Positive prediction rates varied: API (stringent) 21%, API (loose) 51%, mAPI 11%, and PARS 34%.
• Area under the receiver operating characteristic curves (AUC) ranged from 0.57 for mAPI to 0.66 for PARS, indicating moderate predictive ability.

Conclusions:

• The Pediatric Asthma Risk Score (PARS) demonstrated the highest predictive performance among the evaluated indices for asthma at age 6 in this high-risk cohort.
• Current asthma prediction tools show limitations in accurately identifying all high-risk infants.
• Development of improved asthma prediction tools is needed to guide preventative strategies for infants with severe bronchiolitis.