Background
Discrepancies between parent reports and electronic health records (EHRs) challenge the accurate estimation of childhood allergic disease prevalence. This study aimed to: (1) compare parent reports of asthma, eczema, and allergic rhinitis with GP-recorded diagnoses; (2) identify factors associated with reporting differences; and (3) assess the predictive validity of parent reports for future diagnoses.
Methods
Data were analyzed from 2594 children (aged 4–5 years) in the UK Born in Bradford (BiB) cohort. Parent-reported symptoms and diagnoses from questionnaires were compared against diagnoses in primary care EHRs. Agreement was assessed using prevalence estimates and agreement metrics. Logistic and Poisson regression models were used to identify factors influencing reporting and to evaluate predictive validity.
Results
Agreement varied by condition. For parent-reported “ever-diagnosed” asthma, agreement with GP records was good (Kappa = 0.68), while for recent eczema symptoms, it was poor (Kappa = 0.07), though this improved after adjusting for prevalence (PABAK = 0.66). Parent reports were highly reliable for ruling out diagnoses. Factors including ethnicity and GP visit frequency were associated with reporting discrepancies. Parent reports at age 4–5 strongly predicted a future GP diagnosis, increasing the risk fivefold for asthma and threefold for allergic rhinitis.
Conclusion
Neither parent reports nor EHRs alone capture the full picture of childhood allergic disease. Parent reports offer crucial insights into symptom burden and future risk, while EHRs provide objective diagnostic data. An integrating approach, combining both sources, is essential for comprehensive epidemiological research and a more complete understanding of disease burden.
