Wilson David B, Feder Lynette, Olaghere Ajima
George Mason University Fairfax Virginia USA.
University of Central Florida Orlando Florida USA.
Campbell Syst Rev. 2021 Mar 14;17(1):e1151. doi: 10.1002/cl2.1151. eCollection 2021 Mar.
Survey research and analysis of police records, hospital emergency rooms, and women's shelters have clearly established the severity of the intimate partner violence problem and the need to find programs to address this issue. Roughly 1 in 4 women in an intimate relationship is a victim of intimate partner violence. Court-mandated batterer intervention programs (BIPs) have been implemented throughout the United States as a leading method to address this problem. These programs are also now implemented in Canada and Europe. These programs emerged from the women's shelter movement leading to programs with a strong feminist orientation, such as the Duluth Model. The programs that were developed were group-based and relied on psychoeducational methods. Their aim was to get men to take responsibility for their sexist beliefs and stop abusing their partners by teaching them alternative responses for handling their anger. More recent programs draw from cognitive-behavioral therapeutic principles or a mix of the latter with feminist components as well.
This is an update of our prior review. The aim was to assess the effects of postarrest court-mandated interventions for intimate partner violence offenders that target, in part or exclusively, male batterers. Our focus was on studies aimed at reducing intimate partner violence, above and beyond what would have been expected by routine legal procedures (e.g., probation monitoring, etc.).
We searched numerous databases and websites, bibliographies of published reviews of related literature, and a scrutiny of annotated bibliographies of related literature. Our goal was to identify all published and unpublished literature that met our selection criteria. The original review identified nine eligible studies. The updated search identified two new studies. The total sample size across these 11 studies was 4824.
We included experimental (random assignment) and quasi-experimental evaluations of court-mandated BIPs that measured official or victim reports of future intimate partner violence. Rigorous quasi-experimental designs were defined as those that either used matching or statistical controls to improve the comparability of the treated (program) and untreated (comparison) groups. The original review also included quasi-experimental designs that used treatment drop-outs as the comparison group. Given the serious selection bias of such studies, these have not been included in this update.
We coded characteristics of the treatment, sample, outcomes, and research methods. Findings were extracted in the form of an effect size and effect sizes were analyzed using the inverse-variance weight method of meta-analysis. Official report and victim report outcomes were analyzed separately as were the different design types (i.e., random assignment and quasi-experimental designs with a no treatment comparison).
The mean effect for official reports of intimate partner violence from experimental studies showed a modest (but statistically nonsignificant) benefit for the program group (odds ratio, 0.79; 95% confidence interval [CI] [0.49-1.28], = 7) whereas the mean effect for victim reported outcomes showed equal outcomes for both groups (e.g., no benefit or harm; odds ratio, 0.99; 95% CI, [0.74-1.32], = 7). The quasi-experimental studies showed a small but not statistically significant benefit for the program group on official reports (odds ratio, 0.54; 95% CI [0.24-1.22], = 7). One quasi-experiment reported a nonsignificant effect for a victim report outcome (odds ratio, 1.76; 95% CI [0.50-6.14], = 1). None of the analyses were statistically significant. Thus, there is insufficient evidence to conclude that these programs are effective. Both the official measure and the victim reported measures have potential sources of bias, increasing the uncertainty regarding any benefits or harms related to these programs.
AUTHORS' CONCLUSIONS: The findings, we believe, raise doubts about the effectiveness of court-mandated BIPs in reducing re-assault among men convicted of misdemeanor intimate partner violence. New programs and/or entirely new approaches to this important social problem should be explored.
对警方记录、医院急诊室和妇女庇护所的调查研究及分析已明确证实亲密伴侣暴力问题的严重性，以及寻找应对该问题方案的必要性。处于亲密关系中的女性约有四分之一是亲密伴侣暴力的受害者。法院强制要求实施的施暴者干预方案（BIPs）已在美国各地推行，作为解决这一问题的主要方法。加拿大和欧洲现在也实施了这些方案。这些方案源自妇女庇护所运动，产生了具有强烈女权主义倾向的方案，如德卢斯模式。所制定的方案以小组为基础，依靠心理教育方法。其目的是让男性为其性别歧视观念负责，并通过教导他们处理愤怒的替代反应来停止虐待伴侣。最近的方案还借鉴了认知行为治疗原则，或与女权主义成分相结合。
这是我们之前综述的更新。目的是评估逮捕后针对亲密伴侣暴力罪犯的法院强制干预措施的效果，这些措施部分或专门针对男性施暴者。我们关注的是旨在减少亲密伴侣暴力的研究，超出常规法律程序（如缓刑监督等）预期的范围。
我们检索了众多数据库和网站、相关文献已发表综述的参考文献，以及对相关文献注释性参考文献的审查。我们的目标是识别所有符合我们选择标准的已发表和未发表文献。最初的综述确定了9项符合条件的研究。更新后的检索确定了2项新研究。这11项研究的总样本量为4824。
我们纳入了对法院强制要求的BIPs的实验性（随机分配）和准实验性评估，这些评估测量了官方或受害者关于未来亲密伴侣暴力的报告。严格的准实验设计被定义为那些使用匹配或统计控制来提高治疗组（方案组）和未治疗组（对照组）可比性的设计。最初的综述还包括将治疗退出者作为对照组的准实验设计。鉴于此类研究存在严重的选择偏差，本次更新未纳入这些研究。
我们对治疗、样本、结果和研究方法的特征进行了编码。研究结果以效应量的形式提取，并使用荟萃分析的逆方差加权法对效应量进行分析。官方报告和受害者报告的结果分别进行分析，不同的设计类型（即随机分配和无治疗对照的准实验设计）也分别进行分析。
实验研究中亲密伴侣暴力官方报告的平均效应显示，方案组有适度（但在统计学上无显著意义）的益处（优势比，0.79；95%置信区间[CI][0.49 - 1.28]， = 7），而受害者报告结果的平均效应显示两组结果相同（例如，无益处或危害；优势比，0.99；95% CI，[0.74 - 1.32]， = 7）。准实验研究显示，方案组在官方报告方面有微小但无统计学显著意义的益处（优势比，0.54；95% CI [0.24 - 1.22]， = 7）。一项准实验报告了受害者报告结果的非显著效应（优势比，1.76；95% CI [0.50 - 6.14]， = 1）。所有分析均无统计学显著意义。因此，没有足够的证据得出这些方案有效的结论。官方测量和受害者报告测量都有潜在的偏差来源，增加了与这些方案相关的任何益处或危害的不确定性。
我们认为，这些发现使人对法院强制要求的BIPs在减少被判轻罪亲密伴侣暴力的男性再次攻击行为方面的有效性产生怀疑。应该探索针对这一重要社会问题的新方案和/或全新方法。
