Lessons Learnt About Conducting a Multilingual Nutrition Survey in Switzerland: Results from menuCH Pilot Survey

This paper informs about the implementation of the first trilingual Swiss nutrition pilot survey and lessons learnt in terms of recruitment, participation, data collection feasibility, and data management. The population-based cross-sectional nutrition pilot survey took place between June and November 2013. Six trained dietitians interviewed 276 adults aged 18-75 years residing in the cantons of Bern (German), Vaud (French) or Ticino (Italian). Food consumption was assessed with two non-consecutive computer-assisted 24-Hour Dietary Recalls (24HDR), applying a trilingual version of GloboDiet (R) adapted to specific requirements of Switzerland. The first interview was face-to-face and included anthropometric measurements while the second was by phone. Quality controls consisted mainly in the descriptive analysis of data at food level, and the observation and rating of 21 interviews (4%) by coordinators. Net participation rate was 29%. Participants and non-participants were similar: mean [+/- SD] age was 49 +/- 16 and 47 +/- 16 years, and women proportion 49.6% and 49.8%, respectively. Training and data collection proved feasible and deliverable in the six months using the newly developed survey instruments. Dietitians followed the standard operating procedures. Quality controls on food consumption data showed comparable results between face-to-face and phone 24HDR, and across dietitians (median number of reported food items per 24HDR: 27). Procedures to transfer and clean food consumption data were developed. The implementation concept proved applicable in the trilingual Swiss context. Additional resources were planned for increasing participation rate and facilitating data cleaning.