Challenges and lessons learnt from conducting a health survey in an ethnically diverse population.

Aim: To document the challenges of developing and executing an interviewer-driven survey questionnaire in an ethnically diverse and deprived social housing area undergoing reconstruction in Denmark. Methods: The survey was initially planned to be conducted in three waves (before, during and after reconstruction of the housing area), based on a questionnaire covering health, wellbeing and social relations. The first two waves took place in 2018 and 2019 (invited n =~600 per wave) and the third wave will be conducted once the reconstruction is finalised. The approach to the third wave is under revision by the research team. The questionnaire was translated from Danish into the seven most common languages in the housing area. A bilingual interview team went door to door interviewing residents. Field notes were collected systematically during each wave to document the process. Results: The response rates were 35% (n =209) and 22% (n =132) for waves 1 and 2, respectively. There was an overall decrease in response rates between waves 1 and 2 for all language groups, but particularly for Arabic and Turkish-speaking men. The most frequently stated reasons for non-participation included illness and language barriers. The key lessons learnt were that overcoming linguistic and cultural barriers to conducting research among residents in this social housing area requires time and resources. Conclusions: Several challenges are associated with conducting a survey in ethnically diverse and deprived social housing areas. Documenting the challenges and learning from experience are both important, in order to include this hard-to-reach population in health research. [ABSTRACT FROM AUTHOR]