Abstract
BACKGROUND: Childhood asthma, particularly in early life, is often underdiagnosed and poorly characterized in real-world outpatient settings due to diagnostic challenges and resource constraints. A pragmatic, scientifically rigorous, prospective cohort model is urgently needed. OBJECTIVE: We aimed to establish and present a cross-sectional baseline analysis of the Clinical Registry of Childhood Asthma, a prospective, longitudinal, and digitally enhanced cohort in outpatient settings, focusing on the diagnostic spectrum of early-life asthma. METHODS: We established the Clinical Registry of Childhood Asthma cohort and performed a cross-sectional analysis of its baseline data. We launched the cohort in March 2024 as an ongoing study, enrolling children (<18 years) with persistent cough and wheezing from a tertiary pediatric referral center in Southwest China. The study used a real-world design, integrating symptom-driven recruitment with standardized electronic medical records, structured electronic patient-reported outcomes, and systematic biobanking of residual biospecimens. Participants were classified as having confirmed, suspected, or excluded asthma based on cross-sectional baseline data. RESULTS: From March 2024 to August 2025, we enrolled 396 children (median age 4.7 years) from 2296 outpatient visits (enrollment rate 17.2%). Follow-up rates were 26.7% and 43.3% at first and second timepoints, respectively. A comprehensive biorepository was established with serum, plasma, PBMCs, and other blood cell samples (average coverage 74.0%). Most children (267/396, 67.4%) were under 6 years. Patients were stratified into confirmed (131/396, 33.1%), suspected (179/396, 45.2%), and excluded asthma (86/396, 21.7%). Suspected and excluded cases were significantly younger than confirmed cases (median 4.1/3.9 vs 6.6 years, P<.001). Comorbidity profiles differed significantly: allergic rhinitis prevailed in confirmed asthma (77/131, 58.8%), while chronic cough (64/86, 74.4%) and bronchitis (39/86, 45.3%) dominated the excluded group. Type 2 inflammation biomarkers also differed across groups, including aeroallergen sensitization, blood eosinophil count, and fractional exhaled nitric oxide (P<.001). Physician-parent diagnostic discordance was most pronounced in suspected asthma (76/134, 56.7%, P<.001). Multivariable analyses showed suspected asthma (vs excluded) was associated with respiratory infection as wheezing trigger (odds ratio [OR] 4.41, 95% CI 2.16-9.42, P<.001), family history of allergic rhinitis (OR 2.27, 95% CI 1.08-4.99, P=.03), and higher blood eosinophil count (OR 1.32 per 100 cells/μL, 95% CI 1.05-1.73, P=.02). Confirmed asthma (vs suspected) was associated with older age (OR 1.29 per year, 95% CI 1.14-1.47, P<.001), allergic rhinitis (OR 4.06, 95% CI 1.99-8.31, P<.001), and aeroallergen sensitization (OR 3.83, 95% CI 1.91-7.66, P<.001). Bronchitis was negatively associated with an asthma diagnosis across models (OR 0.30 for suspected vs excluded; OR 0.21 for confirmed vs suspected; OR 0.13 for confirmed vs excluded). CONCLUSIONS: The Clinical Registry of Childhood Asthma establishes a feasible cohort in outpatient settings that captures the diagnostic uncertainty of early-life asthma. It identifies a distinct suspected asthma subgroup and reveals significant patient-clinician diagnostic discordance, providing a valuable resource for improving disease management.