P49 Adding Parent Feeding Information to an Established Nutrition Education Program: Fidelity Across 2 States, 2 Languages in a Randomized Control Trial

Objective Determine if peer educators can simultaneously deliver 3 arms of a randomized control trial while maintaining curriculum fidelity. Use of Theory or Research Implementation Science principles Target Audience Expanded Food and Nutrition Education Program participants (n = 540, enrolled) in Washington and Colorado; experienced peer educators (n = 11). Program Description A randomized control trial with 3 arms, each with English and Spanish classes. The same educators taught Control, In-class, and Online arms. All participants, including the Control arm, received 9 lessons using the Eating Smart • Being Active (ESBA) curriculum. In-class participants watched videos on parent feeding behaviors (adapted from previous work) and participated in related educator led activities. Online participants received texts with links to 7 videos and activities between face-to-face lessons but no in-class content. Evaluations occurred at Pre-, Post-, 6- and 12 months post. To address fidelity, educators received extensive training and practice opportunities. Multiple observations in each class series and educator-supervisor discussions assessed and reinforced fidelity. Evaluation Methods Detailed class observation forms, tailored to the specific intervention arm, assessed adherence to ESBA and parent feeding information as well as any cross-contamination of parent feeding behavior information between arms. Open-ended phone interviews with educators post-program. Results One hundred and twenty-eight observations, 58 in English and 70 in Spanish, with 40-45 per arm. The Control group received no parent feeding behavior information; ESBA lessons were delivered as intended 89% of time. In the Online group, minor parent feeding behavior information was shared during 7% of the observations; ESBA lessons were followed 87% of the time. The In-class group delivered the parent feeding behavior information appropriately 84-90% of the time. Most issues were minor and, typically, involved hurried discussions or reviews of previous material. Conclusions With proper training and support, peer educators are capable of maintaining fidelity while delivering multiple arms in a randomized control trial. Researchers interested in studying educational programming delivered by peer educators should invest resources to establish and maintain fidelity for quality research.