Validating childhood symptoms with physician-diagnosed allergic rhinitis

Abstract

BackgroundMultiple population-based and high-risk cohort studies use parental questionnaire responses to define allergic rhinitis (AR) in children. Individual questionnaire items have not been validated by comparison with physician-diagnosed AR (PDAR). ObjectiveTo identify routine clinical questions that best agree with a physician diagnosis of AR and can be used for early case identification. MethodsChildren participating in a longitudinal birth cohort study were evaluated at ages 1 through 4 and at age 7 (n = 531) using questionnaires, physical examinations, and skin prick tests (SPT) with 15 aeroallergens (AG). Parents answered 3 stem questions pertaining to their child, including presence of nasal symptoms absent a cold/flu (ISAAC-validated question), presence of hayfever, and ocular itch. Substem questions were answered with details regarding seasonality, nasal triggers, and ocular seasonality. A global assessment of allergic diseases, including AR, was performed by a specialty-trained clinician. Percent agreement, sensitivity, specificity, and positive predictive values were assessed for individual stem and substem questions. ResultsPositive response to having hayfever and presence of ocular symptoms had the highest specificity (84% and 69%, respectively) and the highest percent agreement (74% and 68%) with PDAR. Identification of triggers for nasal and ocular symptoms had the highest sensitivity (89%). Positive predictive values ranged from 31 to 39%. Combining 2 responses with highest agreement increased specificity for PDAR to 91%. ConclusionResponses to hayfever and ocular symptoms had better specificity and percent agreement with PDAR than the ISAAC-validated questionnaire item. Combining 2 rhinitis questions sharply increases specificity and may improve diagnostic accuracy of clinical questions.