Late Breaking Abstract - A symptom-based algorithm to improve asthma diagnosis in early childhood:the CHILD Cohort Study

Abstract

Background: Our objective was to develop a short, symptom-based algorithm to screen symptomatic 3-year old children to detect those at risk of persistent wheeze at 5 years. Methods: Parents of children participating in the CHILD Cohort study completed health questionnaires at multiple time points. Children underwent structured clinical assessments at ages 1, 3 and 5 years. Diagnoses of asthma by a study pediatrician were documented at 3 and 5 years. Results: At age 3 years, four operational definitions of asthma were compared: (a) clinical assessment by an expert asthma clinician, (b) parental report of physician-diagnosed asthma, (c) modified Asthma Predictive Index (Castro-Rodriguez et al ARJCCM 2000; Guilbert et al JACI 2004) and (d) a symptom-based algorithm using questionnaire data. Among 2354 children, a classification of “definite asthma by algorithm,” compared to “no asthma by algorithm,” had the strongest associations with study physician-diagnosed asthma at age 5 years (OR 50.4, 95% CI 29.7-89.3) and persistent wheeze to age 5 years (OR 239, 95% CI 114-583). The algorithm identified more cases of definite asthma and more unique cases not identified by the other definitions. Notably, of 79 children with persistent wheeze to age 5 years, the algorithm identified 72 (91.1%), whereas in-person clinical assessment identified 49 (62.0%), and parental report of physician-diagnosed asthma only 33 (41.8%). Conclusions: A short, simple symptom-based algorithm identified most children at age 3 years with likely asthma and at higher risk for persistence to age 5 years. The algorithm could be incorporated in electronic medical records to earlier alert physicians to asthma diagnosis.