Preventing Child Abuse Requires Many Approaches, Not Just One

May 30, 2012 | Updated Jul 30, 2012

David Bornstein's May 16 New York Times commentary, "The Power of Nursing," highlights the important contribution that support for infants and their parents can play in minimizing the violence and poor developmental outcomes experienced by millions of children in the U.S. each year. Bornstein makes his case using a single exemplar, the Nurse Family Partnership (NFP), a free, voluntary program that partners first-time moms with nurse home visitors from early in the pregnancy until the child is 2 years old. He suggests that replicating this program, over all other options, holds the greatest promise of success.

That is where Bornstein's policy recommendation goes astray. Although NFP is a stellar example of how to design, test, and replicate an intervention, solving high-cost problems such as child maltreatment, unintentional injuries, and poor cognitive development requires more than replicating a single promising intervention.

Federal policy, in the form of the Federal Maternal Infant and Early Childhood Home Visiting Act (MIECHV), provides a $1.5-billion public investment over five years to assist states in building a comprehensive early childhood system to promote health and safety of pregnant women, children ages 0 to 8, and their families. Importantly, it also stipulates that 75 percent of the funds be allocated to programs with evidence of effectiveness based on rigorous evaluation research. Today, NFP and eight other home visiting models have been approved under the MIECHV program based on their research rigor and findings. It will take all of these programs, with their diverse approaches and attributes, and more, to make a significant impact on rates of child maltreatment in the U.S.

What are the attributes of these programs, and what are the research findings that demonstrate their effectiveness?

NFP serves exclusively first-time mothers, who represent only 40 percent of all births. Although these mothers and their children may face notable challenges, they are more likely to be younger and still able to access some support from their families, and to be pursuing education. Other interventions embrace all at-risk mothers, including the 60 percent who have already had children. The older woman birthing her third or fourth child may have a longer history of multiple partners, educational failures, poor employment history, prior child welfare involvement, and exposure to violence. The odds that these women will experience an initial or subsequent report of maltreatment are far higher than for first-time parents.

With regard to research findings, the evidence that NFP can indeed impact the number of child abuse and neglect reports is impressive but limited to one clinical trial of the model, conducted in Elmira, N.Y. with children born in the 1980s. A 15-year follow-up study of these families found that the women in the group who received the intervention had, on average, almost 50-percent fewer substantiated reports of child abuse and neglect than women in the control group (who did not receive the intervention). This was not the only evaluation study conducted of NFP that examined the frequency of injury to a child that might have resulted from maltreatment, but because the NFP's research team did not specifically examine subsequent reports of child abuse and neglect in later randomized trials of their model, we have no way of knowing if these effects were or could be replicated in other communities or with other service populations enrolled in NFP today.

The most misleading misperception in Bornstein's article, however, is that no other programs do as well as NFP or hold out as much promise in producing strong outcomes for children and their parents. That simply is not the case. After all, NFP holds no promise for the majority of women having children each year -- those who are not first-time mothers. We must offer other models if we really want to move the needle on rates of maltreatment and provide young children the support they need in this most critical developmental period.

Models other than NFP have produced positive findings in well-conducted clinical randomized trials and carefully crafted and rigorous quasi-experimental designs. A comprehensive and objective analysis of evaluative data conducted by Mathematica Policy Research Inc. determined that the nine home visiting models approved under the MIEHCV program all represent prudent investments for achieving the goals outlined in the MIECHV legislation. Although they have different logic models, staffing structures, and program content, each has achieved success in altering parental capacity and strengthening child well-being in important and meaningful ways, including reducing maltreatment and injury. These models include Healthy Families America, Parents As Teachers, Early Head Start, Home Improvement Program for Preschool Children (HIPPY), Family Check-Up, Healthy Steps, Child First, and the Early Intervention Progam. A summary of the research base supporting these and other models is available at As community planners seek to build and better coordinate their supports for newborns and their families, they would be wise to consider the relative merits of diverse approaches as opposed to adopting a single choice.

To NFP's credit, its management team has been exemplary in their commitment to continuously assessing their replication process, highlighting areas of poor performance, and then embarking on ways to improve their practice. This commitment to continuous quality improvement may be the most important criterion to use when determining where investments might be best placed. At the end of the day, there is no single, optimal best choice for a home visitor's profession, a program's content, or its service delivery method. Different populations and different communities will respond to different service approaches. The best bet we have for strengthening collective impacts is to have a range of high-quality, effective programs that continually strive to do better.

Building this type of multifaceted service network is hampered by promoting one model or one approach. Promoting one program model over others that also have viable claims of efficacy directs investments from the many to the one. Families want choice in how they secure their health care and, increasingly, in how they educate their children. The same principle needs to rule in how we reach out to and support new parents.