Editor:

Intimate partner violence (IPV) is a major public health concern, with at least 1.3 million women abused annually in the United States. Mandatory arrest laws have led to a proliferation of IPV offenders entering the criminal justice system, and an expansion of jurisdiction-mandated programs has outpaced research efforts to assess the efficacy of interventions. While more research is needed on these programs, we can learn from a large and growing body of literature on violence intervention programs in general (which includes IPV, but also other forms of violence). The report of Sherman and colleagues¹ to Congress of interventions in use throughout the country, the review by Nation and colleagues² of 35 violence and delinquency prevention programs, and the meta-analytic review of Dowden and Andrews³ of correctional rehabilitation program studies are just a few examples.

A growing consensus among the most rigorous attempts to identify the characteristics of successful (general) violence interventions includes the following: The first is that effective programs are intensive, with participants engaged in them for as much time per day and per week as possible. The second is that they are universal, so that the program does not select among peers in a given setting, facilitating change in the culture at the same time as in the individual. The third, and most important, is that they are comprehensive and multimodal, so that participants are exposed to a range of different activities wide enough to reach them at multiple levels of functioning: affective, cognitive, and behavioral. Without considering these features, to ask which components of a program facilitate change misses the point, for the characteristic that makes any one component successful is the fact that it is interacting with, reinforcing, and reinforced by all the others.

Studies of IPV intervention studies often address length of treatment and rigor of study design,⁴ but not program intensity, universality, or comprehensiveness, as once was the critical error of evaluators of violence intervention programs in general. While IPV is not the same as general violence, the World Health Organization (WHO) has advocated a typological, unified view of violence so that the common underlying causes and manifestations (as well as how they differ) can be considered. Tailoring treatments to subtypes of violent individuals while looking at interventions from a piecemeal perspective undermines the complexity of human behavior, and ignoring available and applicable evidence is likely to lead to the wide implementation of programs that have little proven efficacy. Such an approach risks repeating the mistake that led to the conclusion, after Robert Martinson's⁵ report 35 years ago, that “nothing works.”

• American Academy of Psychiatry and the Law