Assessing cough in epidemiological studies: does it matter which question is used?

Abstract

<b>Introduction:</b> Epidemiological studies use different questions to assess cough in children. We compared parents’ answers to different questions on cough in their children and how this affected prevalence estimates. <b>Methods:</b> We analysed data from <i>Luftibus in the School</i>, a study on respiratory health in schoolchildren, conducted in Switzerland, in 2013-16. Parents completed a questionnaire with three different questions on their child’s cough used in large epidemiological studies: does your child “cough apart from colds”, “has dry night cough” and “cough more than other children”. We assessed the proportion answering ‘’yes’’ to each question, how the answers overlapped and how this differed by age, sex and presence of wheeze. <b>Results:</b> We included 3427 children (age 6-17 years, 50% girls). The prevalence of cough differed substantially depending on the question asked. 872 (25%) parents reported cough apart from colds, 390 (11%) dry night cough, and 158 (5%) that their child coughs more than other children. We found only partial overlap between the cough questions (Figure 1). Only 3% of parents answered yes to all three questions. The overlap was broader for children who also reported wheeze, but did not vary by age or sex. <b>Conclusion:</b> The questions used to assess cough in children strongly affect prevalence of cough. We therefore need to be careful when interpreting and comparing studies. <b>Funding:</b> SNSF:320030_182628; Lunge Zürich