COMPARISON OF RESPONSES TO AN ASTHMA SYMPTOM QUESTIONNAIRE (ISAAC CORE QUESTIONS) COMPLETED BY ADOLESCENTS AND THEIR PARENTS

The primary objective of the study was to determine the impact of the identity of the respondent (parents versus adolescents) on prevalence estimates of asthma symptoms in Swiss adolescents. In addition, factor s influencing agreement between parents' and adolescents' responses to the same questions were analysed. One thousand three hundred and seve nty-four (78.4%) adolescents, aged 14 years, self-completed a question naire at school based on the International Study of Asthma and Allergy in Childhood (ISAAC) core questions on wheezing and asthma. The same questions were incorporated into a questionnaire to be completed by th e parents at home. The adolescents' self-reported prevalence rates of current asthma symptoms and ''asthma ever'' were significantly higher than those obtained from the parental questionnaires. 856 (62.6%) pare ntal questionnaires were filled in by parents without the help of the adolescents, 460 (37.4%) were completed by parents and adolescents and 51 (3.7%) were completed by the adolescents without the parents. Prev alence rates were higher when parents and adolescents completed the qu estionnaire jointly than when questionnaires were completed by parents alone. The level of agreement between parental and self-completed que stionnaires was moderate to low (kappa coefficients 0.22-0.68). Agreem ent between parental and adolescents reports of asthma symptoms was be st when questionnaires were completed jointly by parents and adolescen ts, when the adolescent was a girl, when a family history of asthma wa s recorded, when the adolescent was a non-smoker, and when the parenta l education was high. We conclude that the higher reporting of prevale nce rates of current asthma symptoms by adolescents compared to report ing by their parents demonstrates the need to take the respondent to a questionnaire into account when comparisons are made between prevalen ce studies. The results also suggest that factors related to the famil y milieu influence symptom reporting. (C) 1998 Wiley-Liss, Inc.