RESUMEN
This study's purpose was to examine changes in healthy and unhealthy items purchased following the implementation of traffic light and cartoon labelling in a small retail food venue (Café Bay) in a children's hospital in eastern North Carolina. Between October 2015 and March 2016, daily food and beverage sales from Café Bay were obtained during baseline, traffic light labelling, a washout period, cartoon labelling (on healthy foods only) and a final washout period. Pearson chi-squared tests and multiple linear regressions were used to examine effects of labelling strategies, controlling for the holidays. In unadjusted analyses, traffic light labelling was associated with significant decreases in purchases of unhealthy items purchased, while cartoon labelling was associated with increases in unhealthy items purchased. In adjusted linear regression analysis, traffic light labelling was associated with fewer unhealthy purchases and thus may have potential to decrease the amount of unhealthy items purchased in a children's hospital food retail venue.