Non-response in the West Sweden Asthma Study Questionnaires

Abstract

Introduction: The West Sweden Asthma Study is an ongoing project aiming to investigate the respiratory health in West Sweden. In 2016, a postal questionnaire in the field of respiratory epidemiology was sent out to 50 000 subjects with a response rate of 50 %. It is of great importance to investigate the population of non-responders since they may differ from the responders and thereby possibly cause non-response bias. Aim: The primary aim was to investigate the validity of the questionnaire results by comparing the responders against the non-responders, regarding demographic data and respiratory health. Methods: A sample of 700 subjects were randomized from the non-response population. They were interviewed by telephone with questions composed by a shortened version of the postal questionnaire. Results: A total of 311 non-responders participated in the telephone interview. Non responders were more often males, smokers, living in an urban area and younger than responders. Attacks of shortness of breath were more common among non-responders, while physician diagnosed chronic bronchitis, COPD (chronic obstructive pulmonary disease) or emphysema, longstanding cough and sputum production were more prevalent among responders. Post-stratification weighting for non-response presented that the prevalence estimates for the weighted data were similar to the unweighted data.5 Conclusions: Non-responders and responders differed in demographic data but not for most of the investigated respiratory symptoms and diseases. The results of prevalence data of symptoms and diseases from the postal questionnaire study are considered as representative for the population in West Sweden. Future studies are recommended to explore if digital methods can replace postal questionnaires and if register data can be used for weighting procedures regarding non-responders.