D2.145 - Parental Recognition and Consistency of Noisy Breathing Reports: Treatment and Symptom Associations

Parental reports of noisy breathing, particularly wheeze recognition, forms the cornerstone of epidemiology of wheezing disorders, early asthma diagnosis and management decisions. Despite their importance, the consistency of parental sound identification over repeated episodes and its relationship to reported symptoms and treatments remains understudied. Objective of this study is to characterize parental consistency in identifying distinct breathing noises during early childhood and examine associations with accompanying symptoms and prescribed treatments.

The Breathing Together birth cohort collected monthly parentally-reported questionnaires from birth to age five years (a total of 30,776 questionnaires from 883 participants). Parents selected the abnormal respiratory sound most similar to their child's from four unlabelled audio samples (stridor, stertor, wheeze, infant wheeze) during the episodes of reported noisy breathing; uncertain responses were categorized as "unsure." Associations between noise categories and contemporaneous other respiratory symptoms and subsequent treatments prescribed were analysed using Bayesian multivariable logistic regression (brms). Reporting consistency was quantified via transition matrices across 2,488 episodes.

Among 611 families who reported 3,099 episodes of noisy breathing, parents were unsure of the matching breathing sound in 40% of cases (n=1,228). Parents repeated the same sound label across 58% of 2,488 consecutive transitions (defined sound consistency: 52%), with 53% containing at least one unsure label in the transitions. Compared to the “unsure” reference, parental identification of wheezing was strongly associated with the bronchodilator prescription (OR 2.5 [1.65–3.88]); infant wheezing linked to bronchodilators (OR 1.62 [1.29–2.03]), antibiotics (OR 1.36 [1.02–1.81]), fever (OR 1.58 [1.26–1.97]), and cough (OR 1.96 [1.54–2.50]); stridor and stertor associated with oral corticosteroids (ORs 2.49 [1.82–3.42] and 2.24 [1.88–2.67] respectively); while stertor associated with cold symptoms (OR 2.10 [1.76–2.50]).

High uncertainty and temporal inconsistency in the parental recognition of abnormal respiratory sounds may have important implications for wheeze-based asthma phenotyping, management, and epidemiology.