Linear mixed models within groups comparisons.

Education about dengue disease is fundamental for its prevention. However, the most effective methods for delivering this knowledge in schools remain unclear. In this study, we compared two widely used educational interventions on their effectiveness in increasing dengue knowledge. A cluster-randomized controlled trial, with pre- and post-evaluations was applied. Three hundred ninety 10-year-old children from primary schools in a low-socioeconomic status area of Argentina that is endemic for dengue, were assigned to one of three groups: workshop, an abbreviated workshop, or direct instruction. Dengue knowledge was assessed with a questionnaire before, immediately after the intervention, and one month later. Teacher effect, concepts taught, duration of intervention, and baseline knowledge were controlled for. A Generalized Linear Mixed Model indicated that the three interventions were equally effective in improving dengue knowledge, both immediately after the intervention (children learned 3 dengue concepts) and one month later (children retained 2 dengue concepts). However, each intervention is more effective in teaching some concepts in particular. Our results, among the very few comparing educational interventions in dengue knowledge, suggest that the optimal approach depends more on contextual factors (e.g., specific concepts taught) than on the inherent superiority of either method.