Predicting Severe Asthma Exacerbations in Children: Blueprint for Today and Tomorrow.

Severe asthma exacerbations are the primary cause of morbidity and mortality in children with asthma. Accurate prediction of children at risk for severe exacerbations, defined as those requiring systemic corticosteroids, emergency department visit and/or hospitalization, would considerably reduce healthcare utilization and improve symptoms and quality of life. Substantial progress has been made in identifying high-risk exacerbation-prone children. Known risk factors for exacerbations include demographics (i.e., low income, minority race/ethnicity), poor asthma control, environmental exposures (i.e., aeroallergen exposure/sensitization, concomitant viral infection), inflammatory biomarkers, genetic polymorphisms and markers from other "omic" technologies. The strongest risk factor for a future severe exacerbation remains having had one in the prior year. Combining risk factors into composite scores and use of advanced predictive analytic techniques such as machine learning are recent methods used to achieve stronger prediction of severe exacerbations. However, these methods are limited in prediction efficiency and are currently unable to predict children at risk for impending (within days) severe exacerbations. Thus, we provide a commentary on strategies that have potential to allow for accurate and reliable prediction of children at risk for impending exacerbations. These approaches include implementation of passive, real-time monitoring of impending exacerbation predictors, use of population health strategies, prediction of severe exacerbation responders versus non-responders to conventional exacerbation management, and considerations for pre-school age children who can be especially high risk. Rigorous prediction and prevention of severe asthma exacerbations is needed to advance asthma management and improve the associated morbidity and mortality.