This article reviews research through the mid-1990s on the treatment of children who have been sexually abused, outlines the challenges faced in doing this kind of research, and offers suggestions for future research.
After searching the literature and contacting researchers in the treatment field, the authors identified 29 studies upon which this review is based. The studies fall into three broad categories: pre-post designs, quasi-experimental designs, and experimental designs. They review each group of studies separately.
In pre-post design studies, the children under treatment were evaluated at two or more points in time. All but one of the 17 studies in this group found that treatment lessened children's depression and fearfulness. But symptoms like anger and sexually acting-out were more resistant to change. A drawback of this type of research is that it cannot tell us conclusively whether improvement is actually due to the treatment or simply a result of the passage of time. Some studies in this group used an enhanced design to try to deal with this problem, by also measuring the children's functioning at multiple points before the treatment began.
The second type of study-quasi-experimental designs-compares two groups, but the members of the groups are not equivalent. For example, one study compared a group of children who underwent treatment with a group of children whose parents refused permission to have their children treated. The treated group did better, but it may be that the untreated group might have had more difficulty recovering even if they had received treatment. Because the groups of people studied in quasi-experimental designs do not match, it is difficult to say whether improvement is due to treatment or some other important difference between the two groups. The authors reviewed five studies of this type, and results were mixed. Three studies found no advantage for children receiving therapy as opposed to those who were not treated. This could mean that the therapy offered was not particularly effective, or might imply that the children were receiving something as valuable-parental support, for example-at home.
Experimental designs randomly assign children with similar characteristics to different groups. They may compare children who are treated versus those who are untreated, or compare two or more different types of treatment. Of the studies reviewed in this article, those comparing treated children to those who had not been treated found significant differences. Comparisons of different types of treatment showed mixed results, however. For example, one study found that group treatment improved self-esteem more than individual treatment, but symptoms of anxiety and depression did not improve more. Strong differences were found, however, in children treated in structured abuse-focused groups than in those who received supportive therapy.
While the picture presented here is a mixed one, evidence from two of the experimental designs and the enhanced pre-post design studies, coupled with the body of knowledge about effective psychological treatments for children in other areas, suggests that treatment for sexually abused children can be effective. Some conclusions that also can be drawn are that sexualized behavior is difficult to treat; some children do not improve as much or as quickly as others; and the effectiveness of some treatments has been "oversold." Other challenges that researchers and clinicians face include the diversity of children's reactions to sexual abuse; treatment drop-outs; the issue of asymptomatic children; and the importance of the family context in reactions to and treatment of child sexual abuse.