Lessons Learned and Recommendations from the Implementation of a Live Coding Method to Assess Acceptability of SQ-LNS Supplements in 7- to 24-Month Old Children

Abstract

Abstract Objectives To assess the feasibility of using a live coding method developed in a laboratory to a naturalistic field setting in Mexico and report lessons learned and recommendations for future use. Methods Two coders previously trained to reliability for video-coding of child food acceptance were trained on video live coding (coding video recordings of feeding sessions without pausing or playback) and live coding (LC) in the field. The live coding method was previously piloted and adapted to the Mexican context and common infant feeding practices. From 62 caregivers with 7- to 24-mo children taking part in a randomized field trial, 20 participated in a video-recorded home feeding session. The first 10 offers of food were coded. Each coder individually live coded 10 sessions and video live coded the other 10. Interclass correlation coefficients (ICC) were calculated between coders and with the supervisor to determine reliability. Results Four main categories of adaptations were identified: 1) personnel requirements, 2) participant considerations and instructions, 3) elements of setting, and 4) materials. In the Mexican context (and similar settings) we recommend: limiting the number of personnel entering the home to avoid disrupting the child's eating; offering precise instructions to caregivers on the feeding process; limiting distractions from family members; using a fix and a mobile recording device. Successful implementation of child-feeding LC protocols in the field must balance the following considerations: feasibility in the specific field context, local feeding practices, study resources and timelines, and maintaining quality data collection. All adaptations must prioritize LC best practices, including monitoring reliability during the training and LC period, across coders and a supervisor. Reliability levels in the Mexico setting were good (ICC > .80). Conclusions LC was feasible and successfully implemented in the Mexican field context. Its piloting in the field setting provided essential information to carry out necessary adaptations to the method prior to implementation. Continued use of this method requires context-specific adaptations aligned with best practice recommendations. Funding Sources Global Alliance for Improved Nutrition (GAIN).