Family Violence Research: Title and subTitle BreakLessons Learned and Where From Here?

Child maltreatment and intimate partner violence (IPV) are difficult to study. Both occur within the family context, often involving experiences that are known only to the survivor and the perpetrator. The hidden nature of family violence creates challenges for measurement of exposure and determinants. Yet the severity of these problems and their impact on vulnerable members of society have resulted in a rush to implement prevention or screening programs that may not have first undergone rigorous evaluation.

For example, in the case of child maltreatment, Hawaii’s Healthy Start Program—a program of home visiting by paraprofessionals introduced in Oahu in 1975—was promoted as preventing child abuse and neglect, based on results of an uncontrolled pilot study.¹ In the 2 decades that followed, this intervention was the impetus for national and international adaptations, including the Healthy Families America initiative. After widespread dissemination of the model over many years, results of a randomized controlled trial (RCT) published in 2004 concluded that Hawaii’s Healthy Start Program did not prevent child maltreatment.²

Likewise, in the case of IPV, such groups as the American Medical Association³ and the American College of Obstetricians and Gynecologists⁴ recommend that health care professionals “routinely inquire” or “screen” for exposure to IPV among their women patients, without evidence that screening does more good than harm in preventing further exposure to violence, or improving quality of life.^{5 - 6}

Despite these shortcomings, there is reason for optimism: important gains made in both child maltreatment and IPV research have improved understanding of violence prevention. An important question is: how can each benefit from the “lessons learned” by the other?

The field of child maltreatment has produced significantly more published research than the field of IPV. Broad searches of the main relevant health (MEDLINE, EMBASE, CINAHL) and social sciences (PsycINFO, Sociological Abstracts) databases from the respective database start dates to June 2005 indicated that studies focused on child maltreatment outnumber those focused on IPV by a factor of almost 4 to 1 (about 47 000 to about 12 000), with about two thirds of child maltreatment studies located in health-specific databases, while the relative contributions from health and social sciences were more equivalent for IPV-focused studies.

Of course, quantity is not a substitute for quality. Systematic reviews indicate that each field has significant evidence gaps.^{7 - 10} While there has been similar progress in examining the distribution and determinants of both forms of violence, a comparison of important breakthroughs in primary prevention as well as prevention of recurrence of violence may reveal opportunities for child maltreatment and IPV researchers to learn from each other’s work.

One prevention program that stands out in child maltreatment research for the rigor of its evaluations and the importance of its results is the Nurse Family Partnership (NFP),¹¹ a nurse home visitation program that targets first-time, socially disadvantaged mothers. It begins prenatally and extends until the child is 2 years old. The nurses focus on improving child health and development through promoting competent care by parents; enhancing the life course development of parents through such approaches as encouraging educational involvement; and assisting women to improve their pregnancy outcomes. The program is based on theories of human ecology, attachment, and self-efficacy.

When Olds and colleagues¹² published the results of the first of 3 RCTs evaluating the NFP and showed a reduction in child maltreatment and associated outcomes such as injury rates, many (such as the US Advisory Board on Child Abuse and Neglect¹³ ) were quick to recommend a national program of universal home visitation for all new parents. There was little recognition that (1) cautious replication of an efficacious home visitation program should precede widespread dissemination; and (2) such replication should adhere to the original model, including the targeted nature of the intervention, and be evaluated.

In the meantime, these researchers continued to assess the effectiveness of their intervention by conducting additional trials among diverse ethnic groups and by continuing to observe the study participants to determine long-term outcomes. This represents more than a quarter century of research—what to some might have seemed too long to wait. Yet by resisting the urge to disseminate the NFP prematurely, Olds and colleagues avoided the common pitfall of widespread implementation of a promising intervention of unknown effectiveness and generalizability. As a result of the investigators’ ongoing commitment to rigorous designs with careful and comprehensive follow-up and their perseverance in exploring the limits of their intervention, the field of child maltreatment—and child health in general—now has critical information about an effective approach to preventing child abuse and neglect. The NFP is being disseminated currently in US state and local community settings, with the aim of ensuring that replications of the program are faithful to the original model and performance is monitored. This is an important example of moving from an efficacious intervention to replications evaluating effectiveness prior to broad dissemination.

Although much has been learned about the primary prevention of child maltreatment, less knowledge has been gained about the primary prevention of IPV. A few studies have assessed the effect of IPV educational campaigns directed at adolescents. The outcomes in these studies have included change in knowledge or attitudes but not measures of exposure to violence.¹⁰ However, the major focus of most IPV research has been to prevent the recurrence of violence.

The purpose of screening for IPV is to identify abused intimate partners and assist them to avoid recurrence of violence within the intimate relationship.¹⁴ However, few studies have evaluated evidence-based interventions to which abused persons can be referred. Consequently, recent evidence-based systematic reviews have found insufficient evidence to recommend universal screening.^{9 - 10 ,15} Some advocates and scientists argue that universal IPV screening should be implemented because it raises public awareness and may therefore lead to a willingness to overcome social stigma and seek help. Such goals may be important but do not necessarily translate into reduction of violence.

Many researchers in the field of IPV recognize the need for more rigorous evaluations of interventions to help women¹⁶ ; the call for better studies is not new.¹⁷ In fact, a particular strength of IPV research has been the diversity of perspectives brought to bear on the issue. For example, the use of qualitative research methods, in particular from nurse researchers, has led to a better understanding of women’s experiences of violence and the process involved in addressing and resolving the abuse.^{18 - 19} Application of these findings has led to the development of innovative clinical tools for assessment and referral.²⁰ Outside of health care, traditional experimental designs have been used in criminology and psychology to evaluate approaches to reducing recurrence of IPV—for example, criminal justice approaches such as restraining orders have been found to prevent recurrent abuse,²¹ although batterer treatment programs²² have had mixed, but generally negative, results.²³

In this area, child maltreatment research has much to learn from IPV researchers: despite the greater amount of published research in child maltreatment compared with IPV, few studies using rigorous designs have been conducted to evaluate reduction in recurrence of child abuse and neglect. One of the few studies to evaluate an intervention aimed at preventing recurrence of child physical abuse and neglect showed that an intensive program of home visitation provided to families involved with the child protection system was not effective.²⁴

Two IPV studies deserve mention. First, a study by Dunford²² of men in the US Navy who assaulted their wives evaluated 3 interventions (group sessions for men, group conjoint sessions, and rigorous monitoring with individual counseling) compared with a control group. None of the experimental groups showed any benefits in reducing IPV. Despite its negative results, this study has much to offer the family violence field: the methods were exceptional, loss to follow-up was minimal, and both self-report and official records were used in determining outcome. The second study, an RCT by Sullivan and Bybee²⁵ evaluating a postshelter advocacy counseling program also had much to offer: the follow-up was 2 years and attrition was minimal. Previously, there had been concerns about the ability to maintain an adequate sample when following women who had experienced IPV, yet Sullivan and Bybee had a 95% retention rate at 2 years. The results of this study showed that the intervention was effective in reducing violence and improving quality of life among those women in the experimental group.

Gelles²⁶ has suggested that both the child maltreatment and IPV fields have faced public and political apathy and a system resistant to change due to entrenched social structures. While both areas face challenges in building research capacity, the study of abuse of older persons has even fewer researchers committed to the field. Given the overlap in methodologic challenges, it seems that family violence research could benefit both from closer collaboration among those working in its various subfields and from consultation with methodologists to assist in using the best possible methods to evaluate interventions.

Perhaps more important than individual strategies and methods is to take a broader view of the field as a whole. The complexity and interrelatedness of family violence means that researchers must move beyond working in silos to ensure that lessons learned from one area can be readily transferred and adapted to other areas and help to unite the field of family violence research.

The impact of such aspects of family violence as its intergenerational nature is just beginning to be understood. For example, exposure to violence as a child—whether by experiencing maltreatment or by witnessing IPV—is related to being in an abusive intimate relationship as an adult²⁷ and can affect later parenting.²⁸ It is less clear how abusive relationships in younger adulthood (both child maltreatment and IPV) play out later in life. Are spouses or grown children abused earlier in life more likely to become violent toward an elderly parent or partner when the balance of power, due to illness or infirmity, shifts? Much research is needed in this area.

An examination of the current state of these areas of family violence research, including a comparison of some of the advancements in each field, can allow researchers to collaborate across subfields, with the goal of moving forward to a more comprehensive understanding of the risks, experiences, and consequences of these forms of violence, and, ultimately, better interventions for individuals and families. Clinicians, researchers, and policymakers need to promote and support rigorous studies in all areas of family violence and apply their results—whether positive or negative—in the development of health, behavioral, and social service interventions to benefit those exposed to violence across the lifespan.