Underdiagnosis of childhood asthma: A comparison of survey estimates to clinical evaluation

Abstract

Diagnostic patterns play a role in asthma prevalence estimates and could have implications for disease management. We sought to determine the extent to which questionnaire-derived estimates of childhood asthma reflect the disease's true occurrence. Children aged 6-12 years from Katowice, Poland, were recruited from a cross-sectional survey (N = 1822) via primary schools. Students were categorized into three mutually exclusive groups based on survey responses: "Asthma" (previously diagnosed asthma); "Respiratory symptoms" (no previous diagnosis of asthma and one or more respiratory symptoms during last year), "No respiratory symptoms" (no previous diagnosis of asthma or respiratory symptoms). A sample of children from each group (total N = 456) completed clinical testing to determine asthma presence according to GINA recommendations. Based on the survey, 5.4% of children were classified with asthma, 27.9% with respiratory symptoms, and 66.7% with no respiratory symptoms or asthma. All previously known 41 cases of asthma were confirmed. New diagnoses of asthma were made in 21 (10.9%) and 8 (3.6%) of subjects from the "Respiratory symptoms" (N = 192) and "No respiratory symptoms" (N = 223) groups, respectively. The overall prevalence of childhood asthma, incorporating the results of clinical examination, was 10.8% (95% CI: 9.4-12.2), compared to the questionnaire-derived figure of 5.4% (95% CI: 4.4-6.5%) and affected females more than males. Asthma prevalence was underestimated in this population possibly resulting from under-presentation or under-diagnosis. This could have potential implications for proper management and well-being of children. Questionnaire estimates of prevalence should be considered carefully in the context of regional diagnostic patterns.