INTRODUCTION Given the growing scientific evidence on the detrimental effects of bullying, several prevention programs have been implemented internationally to prevent this behavior among students. Brazil's Educational Program for Drug and Violence Resistance (PROERD) is an adaptation of US' DARE/Keepin' it REAL program, being the most widespread school-based prevention program in the country. However, it has been offered without any effectiveness evaluation. As such, this study evaluates the effectiveness of PROERD in reducing bullying perpetration and victimization among students. METHODS Two cluster randomized controlled trials were carried out with 4030 students (1727 5th graders and 2303 7th graders) in 30 public schools in São Paulo, Brazil. The intervention group attended 10 PROERD classes taught by trained police officers whereas the control group underwent no intervention. Data were collected by self-administered questionnaires using smartphones at two moments (baseline preintervention and 9-month follow-up). Multilevel analysis included two paradigms, complete cases (CC) and intention-to-treat (ITT), using Full Information Maximum Likelihood (FIML) and Multiple Imputation (MI). RESULTS AND CONCLUSION: Results show no statistical difference between groups, indicating lack of evidence on PROERD's effectiveness in preventing bullying behaviors. The insufficient number of classes on bullying prevention and the lack of cultural adaptation may explain these unexpected results. New in-depth evaluation studies concerning the program's components and process are needed.

中文翻译：

---

DARE/Keepin' it REAL 在巴西学生中的欺凌预防计划的有效性。

引言 鉴于关于欺凌的有害影响的科学证据越来越多，国际上已经实施了多项预防计划以防止学生发生这种行为。巴西的毒品和暴力抵抗教育计划 (PROERD) 改编自美国的 DARE/Keepin' it REAL 计划，是该国最广泛的以学校为基础的预防计划。然而，它在没有任何有效性评估的情况下被提供。因此，本研究评估了 PROERD 在减少学生欺凌行为和受害方面的有效性。方法 对巴西圣保罗 30 所公立学校的 4030 名学生（1727 名五年级学生和 2303 名七年级学生）进行了两项整群随机对照试验。干预组参加了 10 节由训练有素的警官教授的 PROERD 课程，而对照组则没有进行任何干预。数据是在两个时刻（基线干预前和 9 个月随访）使用智能手机通过自我管理的问卷收集的。多级分析包括两种范式，完整病例 (CC) 和意向治疗 (ITT)，使用全信息最大似然法 (FIML) 和多重插补 (MI)。结果与结论：结果显示组间无统计学差异，表明缺乏证据证明 PROERD 在预防欺凌行为方面的有效性。防止欺凌课程数量不足和缺乏文化适应可能解释了这些意想不到的结果。需要对计划的组成部分和过程进行新的深入评估研究。