Mental health articles

Decision to remove a child in sexual abuse Across studies, 49% to 59% of children are removed immediately after the official disclosure, and up to two-thirds of children are eventually removed (Faller, 1991b; Hunter et al., 1990; Massat & Lundy, 1998). Often these removals are lengthy or permanent. For example, Hunter et al. (1990) found that less than half of abused children were still residing with their original families 24 months after disclosure. Only 24% of the victims had not experienced a change in living situation following disclosure, and only 44% of the children who were not immediately removed experienced no placement change at the two-year follow-up evaluation. Although most children experienced only one move, 13% of the children experienced more than five moves. Children placed in foster care had the least overall stability, with an average of five different placements. Faller (199 1 b) also found that 62% of sexually abused children in her sample were placed out of the home between the initial report of abuse and the follow-up evaluation approximately one year later. Of those placed, 2% were placed with another parent, 17% were placed with another relative, 68% were placed in foster care, and 13% were institutionalized or placed in independent living. Another study of 277 cases of sexual abuse found that only 41% of children remained in their homes after disclosure (Ryan, Warren, & Weincek, 1991). Although 7% of all children were removed for less than one week, a further 18% were placed with relatives, and 34% were placed in foster care. Several factors affect the decision of child protective service workers to place the child. One of the most consistently reported factors predicting removal is the support of the nonoffending mother (Hunter et al., 1990; Leifer, Shapiro, & Kassim,1993). For example, Leifer et al. (1 993) found that 74% of children were removed when the mother was deemed unsupportive, as compared to 15% of children who were removed when the mother was supportive.6 Everson et al. (1989) also found that 80% of children were placed in foster or institutional care when the mother was deemed unsupportive, 100% were placed with relatives when mothers were scored as ambivalent in support, and 5% were placed when mothers were deemed supportive. Further, Pellegrin and Wagner (1990) found that children of more compliant mothers experienced fewer removals. While these findings seem to support the relationship of guardian support to removals, there are indications that having a supportive guardian does not necessarily protect a child from removal. As mentioned previously, 40% to 59% of children are removed immediately after the official disclosure, and up to two-thirds of children are eventually removed. In the author’s review of guardian support (2000b), however, 65% to 95% of mothers fully or partially believed the disclosure, and across studies, 81% of mothers responded to the abuse disclosure with some or full support. In other words, vastly more removals across studies occur than there are unsupportive guardians. Indeed, Ryan et al. (1991) found that 30% of children were removed from mothers assessed as very protective after disclosure, and 37% of children were removed from mostly protective mothers. The study by Ryan et al. (1991) lends insight into why more removals occur than there are unsupportive guardians. These researchers categorized the mother’s ability to protect the child before the abuse was disclosed as no protection (32%), little protection (16%), mostly protective (16%), and very protective (19%). Another 18% of mothers were considered to have no knowledge of the ongoing abuse. Yet across studies, 75% to 95% of mothers do not know about the ongoing abuse (Faller, 1990; Margolin, 1992; Myer, 1985). That child protective services workers considered that 82% of mothers knew about the abuse before disclosure appears to be a clear indication of systemic bias towards nonoffending mothers. Further, the mother’s protective actions before disclosure was one of two factors that best explained the pattern of removals, the other being the mother’s response after the abuse was disclosed. Removals were as follows for these categories: 89% for mothers offering no protection before the report; 47% for mothers offering little protection; 39% for mostly protective mothers; 28% for very protective mothers; and 44% for mothers who knew nothing about the ongoing abuse. These findings are striking examples that (a) an extraordinary systemic bias towards nonoffending guardians exists, and (b) some children are removed even when nonoffending mothers are deemed protective. Another seemingly obvious factor relating to removal of the child is whether the offender resides in the same home with the victim. In one study, while fewer than 10% of children were removed initially when the offender did not reside in the home, more than half were removed when both the mother and father (who was the offender) were in the home, and 84% of children were removed when the offender was the only biological parent living in the home (Hunter et al., 1990). At the 24- month follow-up evaluation, approximately 50% of children had been removed when the offender was nonresident, about 75% had been removed when the offender was the mother’s boyfriend, 80% had been removed when the offender was a resident father, and 100% had been removed when the offender was the resident father and no mother was present. Yet, the relationship of the offender to the victim was only significant in multivariate analysis for the initial placement. Another study also found no relationship between the decision to remove the child and the perpetrator’s access to the child, the relationship of the perpetrator to the victim, or the identity of the perpetrator (Jaudes & Morris, 1990). While this seemingly logical factor is generally unrelated to removals, other more illogical factors are related. Two different studies found that the most important variable relating to the decision to remove the child was whether the child made the initial outcry (Finkelhor, 1983; Jaudes & Morris, 1990). Those children who purposefully disclosed the abuse were removed from the home significantly more often. Other factors related to increased risk of removal are poverty, multiple children within the family, multiple types of abuse or other problems within the family, having the case referred for prosecution and then declined, and maternal employment (Cross, Martell, McDonald, & Ahl, 1999; Finkelhor, 1983; Pellegrin & Wagner, 1990). Mixed findings have been found for race and age of the child and severity of abuse (Cross et al., 1999; Finkelhor, 1983; Hunter et al., 1990). Factors not significantly related to removal are developmental delay of the child, presence of a sexually transmitted disease, presence of physical abuse, and the child’s psychopathology scores (Hunter et al., 1990; Jaudes & Morris, 1990). In summary, removals are closely related to perceived support of the nonoffending guardian (almost always the mother )—a logical relationship. The problem comes, however, with the perceived support. The study by Ryan et al. (1991) suggests that the perceptions about nonoffending guardians held by child protective services workers are extremely biased, most likely reflecting an institutionalized bias. If workers are trained to believe that most nonoffending guardians are unsupportive, then the close relationship between guardian support and removals would ensure that some children are needlessly removed. Indeed, this appears to be the case as far greater numbers of children are removed from their homes than there are unsupportive guardians. A similar concern with the literature is that compliance in nonoffending mothers is associated with fewer removals (Pellegrin & Wagner, 1990). Perhaps it is just the word compliance, but nonoffending guardians should instead be empowered by professionals to feel as if they can and do have the ability to protect their children. Compliance, on the other hand, suggests an acquiescent guardian. Perhaps we need to consider whether our interventions disempower guardians. Finally, it is always of concern when demographic and disclosure characteristics are related to decisions to remove.

Tags: sexual abuse