Abstract Introduction Electronic cigarette (EC) use has increased rapidly among adolescents in Latin America, with Colombia showing a prevalence of 32%. Their popularity is largely driven by design appeal and flavor diversity, while awareness of health risks remains limited. No school-based educational programs in Colombia have been designed to modify knowledge and perceived harm related to electronic nicotine delivery systems or heated tobacco products. This study aimed to determine the impact of a structured educational intervention on knowledge, risk perception, and EC use among students aged 10-17 years in Bogotá, Colombia. Methods A cohort study was conducted between 2023 and 2024 in public and private schools in Bogotá. The intervention consisted of six educational stages (active and passive strategies) over 12 months. Anonymous pre- and post-intervention questionnaires (Microsoft Forms) assessed knowledge, risk perception, and EC consumption. To link pre- and post-responses from the same individual while preserving anonymity, a probabilistic record linkage algorithm (Fellegi-Sunter with Expectation-Maximization) was applied. Matching rules included identical school/sex, similar grade (+1 tolerance), Jaro-Winkler initial similarity, and temporal proximity (month-year). Borderline matches were reviewed by two independent evaluators, retaining only consensus pairs. Pre-post comparisons were performed using the Wilcoxon signed-rank test (α = 0.05). Analyses were conducted in R. Results A total of 1,274 pre- and 1,152 post-surveys were collected. The algorithm identified 298 unambiguous 1:1 matches and 318 borderline pairs (total=616). After manual review, 395 confirmed matches were included in the final analysis. The mean age was 15.1 years (SD 1.56), 55.4% female. Knowledge and perceived risk scores significantly improved following the intervention. Students’ perception that ECs cause addiction (p 0.01), and that using ECs helps manage stress (p = 0.03) or aids in smoking cessation (p = 0.04) changed significantly toward risk-aware attitudes. Knowledge of nicotine’s effects on the brain (p = 0.02) and awareness of being targeted by EC advertising (p 0.01) also improved. Conclusions This study demonstrates that a school-based educational program can significantly increase knowledge and risk perception related to EC use among adolescents in a middle-income country. The integration of anonymized data and AI-based matching models offers an innovative, ethical, and replicable framework for behavioral intervention evaluation. Expanding similar programs could help prevent early nicotine initiation and promote respiratory health among youth in Latin America. This abstract is funded by: None
Ascencio et al. (Fri,) studied this question.