Poster Topical Area: Nutritional Epidemiology
Location: Hall D
Poster Board Number: 818
Objective: This research explored respondents' cognitive processes during a tablet-based, interviewer-administered 24-hour dietary recall (24HR) in Burkina Faso and Viet Nam, focusing on how respondents recall their previous day's diet. The results elucidate which aspects of the 24HR are challenging to report accurately and why. This information could help improve the accuracy of dietary surveys in low- and middle-income countries (LMICs).
Methods: Cognitive debriefings were conducted with 120 respondents age 18-65 years in Burkina Faso and Viet Nam after administration of a tablet-based 24HR. Respondents were equally divided between men and women and rural and urban areas in both countries. In Burkina Faso, where eating from a common pot is prevalent, a member of the study team visited the household a day in advance to drop off a standard bowl and plate. Debriefings were recorded, transcribed and translated. Data analysis used an inductive, thematic analysis methodology, identifying and describing ideas within data to generate themes.
Results: Respondents had no difficulty with the tablet-based administration of the 24HR; the tablet was not distracting. Most believed they were able to remember what they ate, but said it was easier to remember what they ate at home than outside. Respondents found portion size estimation aids (PSEAs) helpful in estimating quantities consumed. Among the PSEAs, most preferred photos. In Burkina Faso, estimating quantities consumed from a common pot was a challenge. The pre-visit in Burkina Faso helped respondents recall the foods and quantities consumed. Most respondents said the pre-visit did not affect their food consumption. However, the researcher's request to use a standard plate and bowl before the recall confused some respondents, who thought they were being asked to consume a portion equal to that of the size of the dining utensils.
Conclusion: The cognitive debriefing of respondents demonstrated the acceptability of the use of a tablet to conduct an interviewer-administered 24HR. While respondents generally felt confident in their ability to recall foods consumed the previous day with the aids that the researchers had introduced, further work to validate the accuracy of recalls in rural and urban LMIC settings is warranted.