Prinz Ronald J, Sanders Matthew R, Shapiro Cheri J, Whitaker Daniel J, Lutzker John R
Psychology Department, University of South Carolina, Columbia, SC, 29208, USA.
University of Queensland, Brisbane, Australia.
Prev Sci. 2016 Apr;17(3):410-6. doi: 10.1007/s11121-016-0631-x.
A previous article published several years ago (Prinz et al. Prevention Science, 10, 1-12, 2009) described the main results of a place-randomized-design study focused on the prevention of child-maltreatment-related outcomes at a population level through the implementation of a multilevel system of parenting and family support (the Triple P-Positive Parenting Program). The current report, prepared at the encouragement of the journal, provides additional details about procedures, measures, and design-related decisions, presents an additional analysis of the main outcome variables, and poses questions about the study and its implications. We also offer guidance about how the field can move forward to build on this line of research. From the outset, the three designated primary child maltreatment outcomes were county-wide rates for substantiated child maltreatment cases, out-of-home placements, and hospital-treated child maltreatment injuries, derived from independent data sources available through administrative archival records. Baseline equivalence between the two intervention conditions was reaffirmed. The additional analysis, which made use of a 5-year baseline (replacing a 1-year baseline) and ANCOVA, yielded large effect sizes for all three outcomes that converged with those from the original analyses. Overall, the study underscored the potential for community-wide parenting and family support to produce population-level preventive impact on child maltreatment. Issues addressed included (1) the need for replication of population-oriented maltreatment prevention strategies like the one tested in this randomized experiment, (2) the need to demonstrate that a parenting-based population approach to maltreatment prevention can also impact children's adjustment apart from child abuse, and (3) the role of implementation science for achieving greater population reach and maintenance over time.
几年前发表的一篇文章（普林兹等人，《预防科学》，第10卷，第1 - 12页，2009年）描述了一项整群随机设计研究的主要结果，该研究旨在通过实施一个多层次的育儿和家庭支持系统（即“积极育儿计划”），在人群层面预防与儿童虐待相关的后果。应该期刊的鼓励撰写的本报告，提供了有关程序、测量方法和与设计相关决策的更多细节，对主要结果变量进行了额外分析，并提出了关于该研究及其影响的问题。我们还就该领域如何在此研究基础上继续推进提供了指导。从一开始，三个指定的主要儿童虐待结果是全县范围内经证实的儿童虐待案件发生率、家庭外安置率以及因儿童虐待受伤而接受住院治疗的比率，这些数据来自通过行政档案记录获得的独立数据源。再次确认了两种干预条件之间的基线等效性。利用5年基线（取代1年基线）和协方差分析进行的额外分析，得出所有三个结果的效应量都很大，与原始分析结果一致。总体而言，该研究强调了社区层面的育儿和家庭支持对儿童虐待产生人群层面预防影响的潜力。讨论的问题包括：（1）需要复制像本次随机试验中所测试的那种以人群为导向的虐待预防策略；（2）需要证明基于育儿的人群层面虐待预防方法除了对儿童虐待有影响外，还能影响儿童的适应情况；（3）实施科学对于随着时间推移实现更大的人群覆盖和维持效果的作用。
