Reports of sexual abuse of children in state care: A comparison between children with and without intellectual disability

ABSTRACT

Background Little is known about the nature and reactions to sexual abuse of children with intellectual disability (ID). The aim was to fill this gap.

Method Official reports of sexual abuse of children with ID in state care were examined (N = 128) and compared with children without ID (N = 48).

Results Clear signs of penetration or genital touching by male (adolescent) peers or (step/foster)fathers were found in most ID reports. Victims often received residential care and disclosed themselves. Type of perpetrator seemed to affect the nature and reaction to the abuse. Cases of children with and without ID seemed to differ in location and reports to police.

Conclusions Screening of (foster)homes seems crucial. Residential facilities should find a balance between independence of children and protection. Care providers should be trained in addressing sexual issues and sexual education, accounting for different types of perpetrators (peers/adults). Uniform reporting guidelines are needed.

KEYWORDS:

• sexual abuse
• children with ID
• children with intellectual disability
• youth care

Introduction

Sexual abuse of children is a very serious concern in many countries all over the world. According to the Report of the Consultation on Child Abuse Prevention (see World Health Organization, 1999, p. 15) sexual abuse refers to “the involvement of children in sexual activity that they do not fully understand, are unable to give informed consent to, for which they are not developmentally prepared, or that violates the standards of the society in which these children live” (Stoltenborgh, Bakermans-Kranenburg, Alink, & van IJzendoorn, 2015, p. 38). Globally, countries differ with regard to the legal obligations of reporting (suspected) cases of child abuse. Additionally, each health profession has its own regulatory body in which the conditions for when and how to report sexual abuse, but also how to react to these incidents, are determined.

On 13 December 2011, the Council of the European Union and the European Parliament adopted a new directive on combating the sexual abuse and sexual exploitation of children and child pornography to enforce the worldwide fight against child sexual abuse. In 2014, the European Union Agency for Fundamental Rights examined the provisions of the legal obligation to report cases of child abuse, neglect, and violence, and concluded that in many EU member states the lack of clear reporting procedures and protocols either delays the reporting or leads to the underreporting of these cases. Furthermore, the lack of comprehensive documents outlining the referral mechanism in place, as well as the responsibilities of each of the actors involved, has resulted in ineffective cooperation among professionals (European Union Agency for Fundamental Rights, 2014). These facts clarify that significant improvement should be made to better protect children from sexual abuse.

A discrepancy is found between studies examining the prevalence of sexual abuse of children. A meta-analytic study by Stoltenborgh, van IJzendoorn, Euser, and Bakermans-Kranenburg (2011) found a large discrepancy in prevalence estimates of sexual abuse of children between self-report studies (12.7%) and informant studies (0.4%). Besides differences depending on the data sources (self-report vs. informant studies), research shows that differences in child sexual abuse prevalence rates are associated with the country or region studied and the laws and obligations for reporting that are valid there (lowest rates in Asia; Stoltenborgh et al., 2011), the child victim’s gender (lower rates for boys; Pereda, Guilera, Forns, & Gómez-Benito, 2009), and different types of children’s disabilities (Jones et al., 2012; Wissink, van Vugt, Moonen, Stams, & Hendriks, 2015).

With regard to the latter point, it was found, for instance, that children with intellectual disability (ID) are at higher risk for sexual abuse than children without ID (Kvam, 2000; Morano, 2001; Sullivan & Knutson, 2000; Wissink et al., 2015). Research shows that prevalence rates of children with ID who have been sexually abused range from 14 to 65%, of which the lower percentages often derive from studies using data based on case files. For example, Balogh et al. (2001) found that 14% of all children in a psychiatric department for people with ID had (presumably) been victim of sexual abuse, based on file data and information known by the team of caregivers. Conversely, the higher prevalence rates often derive from self-report studies. Spanjaard, Haspels, and Roos (2000), for instance, found that 26% of the boys and 65% of the girls in a residential treatment centre for youth with mild ID and behavioural problems reported that they had been victims of sexual violence. Crucially, Sullivan and Knutson (2000) described that children with ID had a 4 times greater risk for sexual abuse than children without ID. Additionally, research shows that, of all groups of children with disability, children with ID are not only at the highest risk of sexual abuse (Wissink et al., 2015), but are also subject to its more serious and repeated forms.

Even though children with ID are considered to be at increased risk of sexual abuse, not many researchers have explained why this is the case. In a comprehensive literature review, Wissink et al. (2015) discussed what makes children with ID especially vulnerable to sexual abuse and they referred in the first place to the reported lack of knowledge of children with ID about acceptable norms regarding sexual behaviour (Akbaş et al., 2009; Briggs, 2006). Briggs (2006), for instance, found that children with ID (using the UK term learning disabilities) had wrong ideas about whether it was “OK” for adults to use children for sex and about whether it was “OK” for boys to force girls to have sex (even if boys indicated that it was not appropriate for boys to force girls to have sex, they often added “unless it’s your girlfriend”). Wissink et al. (2015) concluded that children with ID are thought to develop sexually as do children without ID, but their technical knowledge and emotional and social skills are often not fully developed to deal adequately with their sexuality. Another important factor is that these children frequently depend on care and caregivers, which makes these children even more inclined to do whatever they are asked to (Wissink et al., 2015). Moreover, as children with ID tend to express affection through physical contact, they may realise less quickly that personal boundaries have been crossed. Furthermore, the wish to be accepted by others is a factor that further contributes to the elevated risk and vulnerability for coercion and seduction (Akbaş et al., 2009). Individuals with mild ID, especially, are shown to have a strong need to belong to peers without ID and to have friends in the “normal” population (Wissink et al., 2015). These vulnerabilities add to existing challenges all children are confronted with, such as the presence of sexual stimuli and situations in daily life. Finally, when sexually abused, the abuse seems to be revealed less easily, and adults can have less confidence in reports and testimonies of children with ID about the abuse experienced (Reiter, Bryen, & Shachar, 2007; Wissink et al., 2015). Thus, it is especially important for cases of sexual abuse of children with ID, when discovered, to be reported, and to provide these children with safe environments and adequate help.

In the Netherlands, by law, all healthcare providers for children with ID are required to report all incidents of sexual abuse between clients, between clients and staff, and between clients and third parties to the Inspectorate of Health Care. Sexual abuse was defined here as “unacceptable sexual behaviour in combination with physical and emotional intimidation and/or relational dependency” (Inspectie voor de Gezondheidszorg, 2007, p. 2). Similarly, all youth care institutions are required to register all “calamities” and report them to the Inspectorate of Youth Care. Calamities were defined here as “events that have occurred under the supervision of the care institution that unexpectedly or unintendedly could have led to or have led to a deadly or harmful result for the child, or for others as a consequence of the child’s actions” (Inspectie Jeugdzorg, 2011, p. 7). According to the inspectorate, unacceptable sexual behaviour and sexual abuse between clients, between clients and care providers, and between clients and (foster)parents/guardians were considered calamities. Not much is known about the characteristics of the reported cases of sexual abuse of children with ID who receive health or youth care and are under custody of the state. Additionally, it is unknown whether there are differences between the cases of sexual abuse involving children with ID and cases involving children without ID. Therefore, over a 3-year period, all official reports involving children with ID as a victim sent to the two Dutch inspectorates were collected, coded, and studied, and compared with reports involving cases of sexual abuse involving children without ID. The leading questions of the current study were:

1. What are the characteristics of the reported cases of sexual abuse involving children with ID? Additionally, within the ID group, are there differences in general descriptive characteristics of the cases (frequency, type of institution, location, report to the police) depending on specific type of sexual abuse and the type of perpetrator (i.e., main characteristics)?

2. Are there differences between reported cases of sexual abuse involving children with ID and cases of sexual abuse involving children without ID in both main characteristics (specific type of sexual abuse, type of perpetrator) and in general descriptive characteristics (frequency, type of institution, location, report to the police)?

Method

Design

A retrospective file analysis design was used for the current study. All case files of sexual abuse involving a child victim (younger than 18 years) with ID who received state care that had been reported to the Dutch Inspectorate of Health Care (IHC) or the Inspectorate of Youth Care (IYC) in the years 2008, 2009, and 2010 were collected and analysed. Files were included in the study if they contained cases of children in state care, which concerns children who receive care under a family supervision order (“Ondertoezichtstelling”; child protection measure), a custodial placement authorisation (“Uithuisplaatsing”), or a relief from or discharge from parental authority or care following the Compulsory Admissions Act (“BOPZ”).

Data

In the years 2008, 2009, and 2010, the IHC and IYC received altogether 176 reports of sexual abuse involving a child victim, 128 of which involved a child with ID. The IHC registered 91 cases of sexual abuse among children with ID and the IYC 36 cases (one case of both IHC and IYC; on average, 31 a year at IHC and 12 a year at IYC). Thus, a total of 176 reports of sexual abuse involving a child victim were included (IHC = 91 cases, IYC = 84 cases, both IHC and IYC = one case). To answer the first research question, all cases involving a child victim with ID of both inspectorates were combined and included (n = 128). To answer the second research question, only the reports of the IYC were used (n = 85), because cases involving children without ID were only available at this inspectorate. In other words, all case files of sexual abuse involving child victims without ID reported to the IYC (in the same period; n = 48) were examined and compared with all case files concerning child victims with ID of the IYC (n = 37).

Procedure

This study was part of a larger study conducted by the Dutch government, under supervision of the Committee Samson (2012; https://www.rijksoverheid.nl/documenten/rapporten/2014/02/06/rapport-commissie-samson-omringd-door-zorg-toch-niet-veilig), on sexual abuse of children who had been placed in state care. After an official request, the IYC and IHC gave access to the official reports of sexual abuse they received in the years 2008, 2009, and 2010.

To report an incident of sexual abuse, healthcare institutions used a report form that can be found at the IHC website. Within 6 weeks following the report, institutions had to provide more (detailed) information about all the individuals involved in the sexual abuse incident, the nature of the sexual abuse, and the reactions of the institution to the incident. Youth care institutions were obliged to report to the IYC. Each report had to contain information about the name of the institution, the location/division, the name and function of the person who reported, the circumstances, relevant information about the involved persons and the reactions of the institute to the circumstances.

Ethical issues

In light of the confidentiality of the data, both inspectorates and the directly involved researchers and research assistants signed a covenant to guarantee the anonymity of the individuals involved. Furthermore, it was confirmed that the researchers and research assistants were not allowed to use the data for other purposes or to make the data available to others without the permission of the inspectorates. The researchers were also not allowed to contact any persons or institutes, or to take any action in this regard, and they were asked to sign a pledge of confidentiality. Moreover, the covenant obliged the researchers to take the necessary organisational and technical actions to ensure the safe processing of the data. Finally, involved care institutions’ key figures and legal experts of both the inspectorates and the Committee Samson reviewed and approved the research project and the covenant.

Coding scheme

Reports were scored on the basis of a coding scheme for the following characteristics: situation of the reported sexual abuse, type of sexual abuse, characteristics of the victim, characteristics of the (presumable) perpetrator, and reported reactions to the incident.

Situation

First, a short description of each reported situation was made. Further, several variables were coded regarding the reported situation, such as time of the incident reported, type of institution that provided care to the child victim (ambulant or nonresidential vs. residential), location of the incident, frequency of the sexual abuse (once/more than once), by whom and how the incident was discovered, the possible abuse of alcohol and drugs during the incident (both by the victim and/or the perpetrator), whether verbal or physical violence had been used by the perpetrator, and, finally, the evidence for the incident (indistinct signs/clear signs/proven). Cases were considered as having indistinct signs when it concerned cases in which there were worries because a child, for instance, showed deviant sexual behaviours but with no other strong indications of sexual abuse. Other cases were considered as “indistinct” when there were strong beliefs that the child (for instance, a teenage girl with borderline characteristics) had “made up” the story or when evidence was completely lacking or contained contradictory information. Cases were coded as having clear signs, for instance, when a venereal disease was diagnosed, or other physical marks were visible, but also when a child had uttered a clear, concrete accusation. Finally, cases were considered proven when the persons involved were “caught in the act,” were convicted, or when forensic evidence was available.

Type of sexual abuse

The sexual abuse coding was based on the categorising system used by the Committee Samson (2012): (1) Intercourse (penetration), (2) Touching of genitals (including oral sex), (3) Touching, but not the genitals, (4) No physical contact, (5) Other cases of sexual abuse, (6) Not specified. If penetration (either with the genitals or with fingers, tongue, other objects) occurred, this was coded as sexual abuse belonging to the first category. The second category concerned genital touching behaviours, such as masturbating, genital–oral contact (without penetration), or with fingers. The third category included other types of physical contact (not with the genitals), like the touching of breasts and buttocks, but also, for instance, (suction) kissing. Cases of no physical contact included cases of children who, for instance, were forced to watch pornography with an adult, and who received inappropriate requests (for instance, if they were willing to perform sexual actions for somebody else).

Victim characteristics

The characteristics of the victim coded (if mentioned in the file) were gender, age, total IQ score, presence of a childcare protection measure, and ethnic background.

Characteristics (presumable) perpetrator

The following characteristics of the (presumable) perpetrator of the sexual abuse were coded: gender, age, and type of perpetrator (i.e., relation to the victim). In addition, whether or not the perpetrator had been a perpetrator or victim of abuse prior to the incident was coded if the report provided this information.

Reactions

To gain an insight into the reactions to the incident, it was coded whether a risk assessment was carried out (either prior to or after the incident) and whether extra care/treatment regarding sexuality issues was offered (and, if so, which specific type), or other reactions (like translocation, conversations, additional training for employees, education of children, adjustments in policy) or performed examinations (psychological and/or physiological) were evident from the files. Additionally, it was coded whether the relevant inspectorate provided advice (and, if so, what sort of advice). It was also registered, if the file showed that the case resulted in a prosecution, whether the police were contacted concerning the incident or whether an official report was made to the police. Finally, if the files showed that parents were informed about the incident, this was registered as well.

Analytic strategy

To test whether the characteristics of sexual abuse involving children with ID were dependent on the specific type of sexual abuse or the type of perpetrator, two-sided Fisher’s exact tests were performed in R. In addition, to test whether there were differences between reported cases of sexual abuse involving children with ID versus children without ID in characteristics of sexual abuse, two-tailed Fisher’s exact tests were performed for all cases of the IYC.

Results

Characteristics of the reported cases of sexual abuse involving children with ID

Type of sexual abuse

Of the 128 reports, 101 reported cases clearly belonged to one of the four categories of sexual abuse, whereas 27 reported cases could not be classified to one of these categories because the type of sexual abuse was not clear. Of the 101 cases, most concerned cases of intercourse (penetration; n = 47, 47%), followed by cases of touching of the genitals (including oral sex; n = 30, 30%), cases of sexual abuse without physical contact (n = 13, 13%), and touching of other bodily parts (not genitals; n = 11, 11%). In all further analyses regarding the type of sexual abuse, reports in which the specific type of sexual abuse could not be established were excluded.

Situational characteristics

In the majority of the reported cases, there were clear signs of the sexual abuse (n = 78, 61%). There were indistinct signs in only 17% of cases (n = 22). Furthermore, in 22% of cases the abuse was considered proven (n = 28). Regarding the frequency of the sexual abuse, the reported cases took place once (56%), as well as repeatedly (44%).

In 66% of the cases of sexual abuse (n = 85), children disclosed the abusive behaviour by directly telling someone about the incident, whereas in 24% of the cases the abusive behaviour was signalled by individuals around the child (care providers, family members; n = 31), and in 9% of the cases perpetrators were caught in the act (n = 11; one missing). When the children disclosed the abuse themselves, they often told a care provider or group worker (n = 56, 66%), a (foster)family member (n = 16, 19%), or a group member (peer; n = 9, 11%; three other and one missing).

If information on when the incident happened was provided (n = 56), most reported cases concerned incidents that occurred during the evening (n = 27, 48%), followed by incidents during the day (n = 20, 36%). In only 16% of the cases (n = 9) the sexual abuse occurred during the night. Finally, in 19 cases (15%) the perpetrator used verbal violence (threatening/force), and in 16 cases (13%), physical violence. In a small number of cases the files mentioned the use of alcohol (four cases; 3%) or drugs (three cases; 2%) during the incident either by the perpetrator, the victim, or both.

In 85% of the reports of both inspectorates (n = 109) the incidences of sexual abuse concerned a child with ID who was placed in residential care (vs. 15% receiving ambulant care; n = 19). The reported cases of sexual abuse occurred in different locations: within the care institution (n = 49; 48% of the sexual abuse cases), at the (foster)home (n = 17, 17%), in the home of a third person (n = 11, 11%), or outside (park or playground; n = 13, 13%; eight other [school/club/transport]; three missing).

Victim characteristics

Gender and age

In most cases the victim of the reported sexual abuse with ID was a girl (n = 91, 71%). Although the reports included children of all ages, 66% of the victims were adolescents between 12 and 17 years (n = 85), 40 cases of which (47%) were between age 12 and 14, and 45 cases (53%) between age 15 and 17. A smaller percentage of the victims encompassed children aged 6 to 11 years old (n = 38, 30%) and children aged 1 to 5 years old (n = 5, 4%). The percentage of adolescent victims was high for each specific type of sexual abuse (74% of the cases of intercourse, n = 35; 57% of touching the genitals, n = 17; 91% of touching, n = 10; 69% of the cases without physical contact, n = 9).

Child protection measure

The majority of the children of the 128 cases received care under a (provisional) family supervision order (n = 99, 77%); 16% of them also had a custodial placement authorisation (n = 20). In six cases a relief from or discharge from parental authority occurred (5%), and one victim received care following the Compulsory Admissions Act (two missing).

Previous victimisation

In 66 cases (52%) it became evident from the data provided by the inspectorates that the current victim had been a victim of abuse before (different types of abuse, including sexual abuse). A total of 59 children (46% of the 128 children in reports) had been victim of sexual abuse before the current report, and seven children (5%) had been victims of physical abuse, neglect, or psychological or emotional neglect.

IQ and ethnic background

In the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; American Psychiatric Association, 1994), children with ID are classified as borderline (IQ 71–84) or having mild intellectual disability (IQ 50–70). The total IQ score of the children was not always specifically mentioned in the file reports (data of children’s IQ were available for 57 cases, 45%). Of the children with an established IQ score (n = 57), there were 35 children with a total IQ score of 50–70 (61%) and 22 children with a total IQ score of 71–84 (39%). Furthermore, 10 cases (8%) concerned a victim with an ethnic minority background, but in most cases information on ethnic background was missing.

Perpetrator characteristics

Gender

In a large majority of the reports (n = 117, 91%) the (presumable) perpetrator of the sexual abuse was a male (vs. n = 7, 5% female perpetrators; four missing). Of the 117 cases where the (presumable) perpetrator was a male, 43 reports concerned cases of intercourse (37%), 27 reports concerned cases of touching of genitals (23%), 11 reports concerned cases of touching (but not the genitals; 9%), and 10 reports concerned cases of no physical contact (9%; 26 undefined sexual abuse cases). Of the seven cases where the (presumable) perpetrator was a female, these numbers were respectively n = 1 (14%), n = 3 (43%), n = 0 (0%), and n = 2 (29%) for reports concerning cases of intercourse, touching of genitals, touching (but not the genitals), and no physical contact (one undefined sexual abuse case).

Type of perpetrator (relation to victim)

In around half of the reported cases (n = 63, 49%), the (presumable) perpetrator of the sexual abuse was a group mate or peer who was in 70% of the cases an (early or late) adolescent (seven cases in the 5- to 11-year age group; 21 cases in the 12- to 14-year age group; 17 cases in the 15- to 17-year age group; six cases of 18 years or older; 12 missing). Of the 63 group mate/peer perpetrator cases, 23 concerned a group mate/peer perpetrator who had committed sexual abuse before (37%). Next to group mate/peer perpetrators, relatively many reports concerned cases where the (presumable) perpetrator was a (step/foster)parent (n = 25, 20% of all reports; mostly male). Other perpetrators were acquaintances of the parents, employees, acquaintances of the child, (foster)siblings (mostly brothers), ex-boyfriends, adult males known via the internet, “loverboys” (pimps or human traffickers), grandfathers/grandmothers, and unknown people (“of the street”).

Perpetrator’s abuse history

For some cases (n = 29, 23%) information was provided about the perpetrator’s (possible) own abuse history. For these 29 cases, for which this information was available, 20 cases (69%) concerned a case where the (presumable) perpetrator had been victim of sexual abuse him- or herself before the incident had taken place.

Reactions

Concerning the reactions within the institutions, in 67 cases (52%) a risk assessment of sexual abuse had been carried out for the victim (of which 27% took place before the incident, 37% after the incident, and 36% both before and after the incident). In 17 cases (13%) such an assessment had not been carried out and in 44 cases (34%) it could not be clearly determined whether a risk assessment had been carried out. Furthermore, in 31 cases (24%) a physical examination was performed, in two cases (2%) both a physical and psychological examination was performed, and in 95 cases (74%) no information was available on (possible) examination of the victim. In 58 cases (45%) it was reported that the victim received extra care targeting sexuality.

Other reactions to the sexual abuse incident within the care institutions varied from having a conversation with the individuals involved, transfer of the victim or offender to another department, and extra care for the victim and/or perpetrator, to educating staff and taking practical measures (such as better supervision). In 37% of the reported cases (n = 47) the inspectorate posed additional questions or further examined the case, and in 17% of the reported cases the inspectorate advised the care institution on how to deal with the incident (n = 22, 43%, or 55 cases without the advice of the inspectorate; four missing values).

At the moment the abuse was reported, it was mainly not known whether the incident resulted in prosecution (120 cases, 94%). Often, the police were contacted (25 cases, 20%) or the case was officially reported to the police (or there was an intention to; 67 cases, 52%). Finally, in 80% of the files (n = 102) it was mentioned that parents or guardians were informed about the sexual abuse.

Relationships between type of sexual abuse and general descriptive statistics (frequency, type of institution, location, report to the police) within the ID group

Two-tailed Fisher’s exact tests were performed to test whether for children with ID characteristics of the sexual abuse (frequency of sexual abuse, type of institution, type of location, and type of police report) were dependent on the type of sexual abuse. The results showed that there were no significant associations between type of sexual abuse and frequency of the abuse (p = .477), type of institution (p = .684), or the type of location (p = .194). There was, however, a significant association between the type of sexual abuse and the type of report to the police (p = .004).

As can be seen in Table 1, a high percentage of cases for which the type of sexual abuse was intercourse (penetration) was reported to the police (79.5%), and in a low percentage of these cases no contact had taken place with the police (4.5%). For instance, the proportion of cases of sexual abuse for which the police was informed or an official report was drawn was 1.3 times as high for cases for which the type of sexual abuse was intercourse versus no physical contact cases (relative risk = 1.3; see Table 1). To give an indication of the pattern of the association, we considered the standardised residuals. None of the standardised residuals were very large (the standardised residuals ranged from 0.2 to 2.1 in absolute value), providing not a strong evidence against independence.

Table 1. Type of sexual abuse and type of report to the police (child victims with ID).

	Officially report to police n (%)	Informing the police n (%)	No contact with the police n (%)	N
Type of sexual abuse
Intercourse	35 (79.5%)	7 (15.9%)	2 (4.5%)	44
Touching of genitals	14 (48.3%)	7 (24.1%)	8 (27.6%)	29
Touching (not genitals)	4 (36.4%)	3 (27.3%)	4 (36.4%)	11
No physical contact	4 (36.4%)	4 (36.4%)	3 (27.3%)	11
Total N (%)	57 (60.0%)	21 (22.1%)	17 (17.9%)	95

Relationships between type of perpetrator and general descriptive statistics (frequency, type of institution, location, report to the police) within the ID group

Two-tailed Fisher’s exact tests were performed to test whether within the ID group the general characteristics of sexual abuse (i.e., frequency of sexual abuse, type of institution, type of location, and type of police report) were dependent on the type of perpetrator (categorised as peers/family/employee/acquaintance/unknown). The results showed that there were significant associations between the type of perpetrator, on the one hand, and the frequency of sexual abuse (p = .015), the type of institution (p = .001), the type of location (p = .0001), and type of report to the police (p = .043), on the other hand.

Regarding the frequency, as displayed in Table 2, in cases for which the (presumable) perpetrator was a group mate/peer, the sexual abuse tended to occur only once (67%). When the perpetrator was a (foster)family member or an employee, the sexual abuse often had a repetitive nature (64% and 73%, respectively; see Table 2). That is, the proportion of cases of repeated abuse was twice as high for cases where the perpetrator was a (foster)family member or an employee versus cases where the perpetrator was a peer (relative risk = 2.0; see Table 2). However, none of the standardised residuals were large (the standardised residuals ranged from 0.1 to 1.4 in absolute value), providing not a strong evidence against independence.

Table 2. Type of (presumable) perpetrator and frequency of sexual abuse, type of location of abuse, and report to the police (child victims with ID; N and % within type of perpetrator).

	Frequency of sexual abuse			Type of location of abuse					Report to the police
	Once n (%)	Repeated n (%)	N	Inside institution n (%)	School n (%)	(Foster) Home n (%)	Other n (%)	N	Official report to police n (%)	Informing the police n (%)	No contact with the police n (%)	N
Type of perpetrator
Peer	41 (67.2%)	20 (32.8%)	61	47 (77.0%)	1 (1.6%)	1 (1.6%)	12 (19.7%)	61	26 (44.1%)	18 (30.5%)	15 (25.4%)	59
(Foster)Family	8 (36.4%)	14 (63.6%)	22	1 (4.3%)	0 (0.0%)	20 (87.0%)	2 (8.7%)	23	15 (68.2%)	4 (18.2%)	3 (30.0%)	22
Employee	3 (27.3%)	8 (72.7%)	11	6 (50.0%)	0 (0.0%)	4 (33.3%)	2 (16.7%)	12	6 (60.0%)	1 (10.0%)	3 (30.0%)	10
Acquaintance	15 (53.6%)	13 (46.4%)	28	1 (3.7%)	3 (11.1%)	6 (22.2%)	17 (63.0%)	27	23 (82.1%)	2 (7.1%)	3 (10.7%)	28
Unknown	2 (100%)	0 (0.0%)	2	0 (0.0%)	0 (0.0%)	0 (0.0%)	2 (100.0%)	2	1 (100.0%)	0 (0.0%)	0 (0.0%)	1
Total N (%)	69 (55.6%)	55 (44.4%)	124	55 (44.0%)	4 (3.2%)	31 (24.8%)	35 (28.0%)	125	71 (59.2%)	25 (20.8%)	24 (20.0%)	120

Regarding the location of the abuse, as displayed in Table 2, when the perpetrator was a peer, the location of the abuse was more often inside the institution (77%, standardised residual = 3.9) and less often at home (1.6%, standardised residual = –3.6). When the perpetrator was a (foster)family member, the abuse generally took place at home (87%, standardised residual = 6.0). Further, when the perpetrator was an acquaintance, the location of the abuse was less often inside the institution (3.7%, standardised residual = –3.2), and more often at a different location; for example, outside in a park or playground or in another person’s house (63%, standardised residual = 3.4). The large standardised residuals of these cells suggest these patterns are stronger than independence would predict (the standardised residuals of the other cells ranged from 0.3 to 2.9 in absolute value). Table 2 further shows the percentages across the types of contact with the police and the perpetrator type. As can be seen in Table 2, a relatively high percentage of cases for which the perpetrator was an acquaintance were officially reported to the police (82.1%). However, none of the standardised residuals were large (the standardised residuals ranged from 0.0 to 1.6 in absolute value), providing not a strong evidence against independence.

Table 3 presents the percentages of incidents per type of perpetrator and type of care institution. As can be seen in Table 3, in cases of children receiving ambulant care, family members (parents, step/foster parents, brother/sisters, grandfathers/mothers; mostly male) more often (47%) committed the sexual abuse (i.e., standardised residual = 3.1). The standardised residuals of the other cells ranged from 0.2 to 2.0 in absolute value.

Table 3. Type of (presumable) perpetrator and general type of institution (child victims with ID; N and % within type of institution).

	Ambulant n (%)	Residential n (%)	Total N (%)
Type of perpetrator
Peer	3 (15.8%)	58 (54.2%)	61 (48.4%)
Family	9 (47.4%)	13 (12.1%)	22 (17.5%)
Employee	1 (5.3%)	11 (10.3%)	12 (9.5%)
Acquaintance	6 (31.6%)	23 (21.5%)	29 (23.0%)
Unknown	0 (0.0%)	2 (1.9%)	2 (1.6%)
N	19	107	126

Between-group differences in incidents’ characteristics

A comparison was made between sexual abuse reports involving children with ID and reports involving children without ID regarding main characteristics (specific type of sexual abuse, type of perpetrator) and general descriptive characteristics (frequency, type of institution, location, report to the police) using the data of the IYC (from 2008 through 2010; age victim below 18 years).

Two-tailed Fisher’s exact tests were performed to test whether there was an association between the involvement of children with or without ID and the characteristics of the sexual abuse. The results showed that there were no significant differences between the children with or without ID in terms of the type of sexual abuse (p = .139), the (presumable) perpetrator of the abuse (p = .110), and the frequency of the abuse (p = 1.000). This implies that the specific type of sexual abuse, the perpetrator of the abuse, and the frequency of the abuse were not different for either children with or without ID. However, the results indicated a significant difference between children with or without ID and the type of institution (p = .002), type of location (p = .018), and type of report to the police (p = .0004).

First of all, there were significant differences between the reports of children with ID and those without ID across the two types of care institutions (residential vs. ambulant; see Table 4). As can be seen in Table 4, compared with the cases involving children without ID, a relatively high percentage of the cases involving children with ID occurred in institutions providing ambulant help (30% for children with ID vs. 4% for children without ID). That is, the proportion of cases that occurred in institutions providing ambulant help was 7 times as high for children with ID versus children without ID (relative risk = 7.0; see Table 4), and the proportion of cases that occurred within residential care was 1.4 times higher for children without ID versus children with ID (relative risk = 1.4; see Table 4). However, none of the standardised residuals were large (the standardised residuals ranged from 0.8 to 2.8 in absolute value), providing not a strong evidence against independence. Moreover, for both groups of children, the percentage of cases within residential care was much higher than the percentage of cases in ambulant care (70% of the cases in the ID group with ID and 96% of the cases in the group without ID).

Table 4. Type of victim and type of institution, type of location of abuse, and report to the police (N and % within type of victim).

	Type of institution			Type of location of abuse					Report to the police
	Ambulant n (%)	Residential n (%)	N	Inside institution n (%)	School n (%)	(Foster) Home n (%)	Other n (%)	N	Official report to police n (%)	Informing the police n (%)	No contact with the police n (%)	N
Type of victim
Children without ID	2 (4.3%)	45 (95.7%)	47	28 (60.9%)	2 (4.3%)	10 (21.7%)	6 (13.0%)	46	26 (57.8%)	0 (0.0%)	19 (42.2%)	45
Children with ID	11 (29.7%)	26 (70.3%)	37	11 (29.7%)	1 (2.7%)	18 (48.6%)	7 (18.9%)	37	28 (82.4%)	3 (8.8%)	3 (8.8%)	34
Total N (%)	13 (15.5%)	71 (84.5%)	84	39 (47.0%)	3 (3.6%)	28 (33.7%)	13 (15.7%)	83	54 (68.4%)	3 (3.8%)	22 (27.8%)	79

Regarding the location of the abuse, as displayed in Table 4, for a high percentage of cases involving children without ID the abuse took place within the youth care institution (61%). For cases involving children with ID, a high percentage of the abuse cases occurred in their (foster)homes (49%). That is, the proportion of cases of which the abuse took place within the youth care institution versus somewhere else was 2 times as high for children without ID versus children with ID (relative risk = 2.0), and the reverse pattern was visible for cases for which the abuse took place within (foster)homes versus somewhere else (relative risk = 2.2; see Table 4). However, none of the standardised residuals were large (the standardised residuals ranged from 0.3 to 1.6 in absolute value), providing not a strong evidence against independence.

Lastly, we examined the percentages of official police reports for both the group with ID and the group without ID. As can be observed in Table 4, a high percentage of cases of sexual abuse of children with ID (as victims) were officially reported to the police (82%). Further, in cases of children without ID, for a high percentage of cases there was no contact with the police at all (42%). That is, the proportion of cases of abuse that were officially reported to the police versus cases where the police were only informed or there was no contact with the police was 1.4 times as high for children with ID versus children without ID (relative risk = 1.4; see Table 4). Moreover, the proportion of cases of which there was no contact with the police versus a form of contact with the police was 4.8 times higher for children without ID versus children with ID (relative risk = 4.8; see Table 4). However, none of the standardised residuals were large (the standardised residuals ranged from 0.9 to 2.1 in absolute value), providing not a strong evidence against independence.

Discussion

The aim of the current study was to examine cases of sexual abuse that had been reported to the Dutch IHC and IYC, and to describe the nature of the cases as well as how institutions responded to the reports of the sexual abuse incidents. The cases of sexual abuse that were reported to the inspectorates varied from hands-off to hands-on sexual abuse situations (including intercourse). The majority of the sexual abuse reports involved intercourse (penetration; 47% of all reported incidents of sexual abuse) or touching of the genitals (including oral sex; 30%). Cases of touching (apart from touching the genitals) and cases without physical contact were reported less often. It is unclear to what extent these types of cases occur less often, are less frequently reported, or were among the unspecified cases. Wissink, Moonen, van Vugt, Stams, and Vergeer (2012) found that for employees working in youth care institutions it is often unclear which situations of sexual behaviour they are supposed to report (and which not) and, if so, to which authority. As a result of that study and related studies (Committee Samson, 2012), the Dutch government invested in the development of clearer legislation and guidelines on how to cope with sexual abuse in care institutions (see, for instance, “Kwaliteitskader voorkomen seksueel misbruik in de jeugdzorg”; a Dutch code for professionals concerning alertness and reporting of domestic violence and child abuse that was implemented in July 2013 in the Netherlands; Jeugdzorg Nederland, 2013). Additionally, several other measures were undertaken to better protect children, such as the development of an action plan called “Children Safe,” which provides tools to prevent and stop child abuse, measures to signal child abuse at an earlier stage and to limit the negative consequences, and a special task force on child abuse (see also http://www.vooreenveiligthuis.nl/veilig-thuis) was ordered to monitor the development of these measures. Additionally, a five-step plan was developed with information for teachers, doctors, and youth care workers about what to do when they encounter a case of suspected abuse (see https://www.rijksoverheid.nl/onderwerpen/huiselijk-geweld/inhoud/meldcode). Prolonged attention to sexual abuse among children in state care is important considering the serious consequences of sexual abuse (including re-victimisation) for these already vulnerable children.

The findings of the current study provide information about the extent, nature, and circumstances under which sexual abuse incidents can happen. For instance, the reported incidents often occurred in the evening (48%) or at daytime (36%) but less often at night. Additionally, most reported incidents in this study took place within a care institution (44%) or at the home of a (foster)parent (25%). The results also indicated that the frequency of the abuse was associated with the type of (assumed) perpetrator. Descriptive statistics indicated that, in cases of abuse by an adult perpetrator (parent, step/foster parent, employee of care institutions; mostly male), the sexual abuse seems to occur more often repeatedly, whereas in cases of abuse by other children (group mates/peers) the abuse seems to occur more often once. It is possible that the cases of an adult perpetrator are those of more intended sexual abuse, whereas the cases where a group mate/peer is involved are more often incidents, with changing victims, of sexual experimentation that have got out of hand (Rushbrooke, Murray, & Townsend, 2014). Therefore, it is important to discuss with employees, educating them about what is regarded as “normal” sexual experimenting behaviour (considering the abilities and age of the children involved), and what should be considered as deviant and unacceptable behaviour. More agreement and clarity about the boundaries of sexual (experimenting) behaviour for children with ID of different ages is essential. Systems like the “flag system,” a method to distinguish between experimentation and sexually deviant or abusive behaviour, could be helpful for staff members (see, for instance, Janssens, Felten, & Frans, 2010). Examples of possibilities that Briggs (2006) mentioned are training care providers to recognise potentially dangerous situations, establishing rules about sexual behaviour of children, and improving the understanding of children’s rights in relation to sexual abuse.

In their meta-synthesis, Rushbrooke et al. (2014) also described the difficulties that (both professional and informal) caregivers of people with ID experience in relation to the issue of sexuality. They found that caregivers often experienced fear, uncertainty, anxiety, and a lack of competency when dealing with sexuality issues, possibly resulting in a reluctance to make decisions or to get involved in supporting sexual needs. Caregivers also experienced barriers to discussing sexuality, including embarrassment. Rushbrooke and colleagues (2014) found that training was considered very important in order to enable caregivers to feel more confident in dealing with issues of sexuality and relationships, and to feel less fearful of the consequences of making decisions about sexual matters. Employees should receive special training in providing such a complex task of supervision and education (Wissink et al., 2012).

In the majority of the reported cases involving a child with ID, the sexual abuse occurred between peers (56%). This was the case for all types of abuse (51% of the incidents of intercourse, 63% of touching of the genitals, 73% of other types of touching, and 42% of incidents without physical contact).¹ Most of the group mate/peer perpetrators were adolescents (aged 12 to 18; a much smaller percentage concerned younger children). This finding is also in line with the findings of Rushbrooke et al. (2014), who reported that caregivers felt especially uncomfortable with the developing sexuality of adolescents with ID. Care providers of adolescents with ID believe they have to find a balance between being a protector against the risks associated with sexuality, on the one hand, and being a facilitator of sexual expression and pursuit of relationships, on the other hand. Finding and continuously re-evaluating this delicate balance is a complicated task for caregivers, which deserves more attention in training programs and in policy.

Furthermore, the reports indicated that at least 37% of the group mate/peer perpetrators had committed sexual abuse before, and, for at least 56% of these group mate/peer perpetrators, a sexual abuse risk assessment was conducted either before or after the incident (Wissink et al., 2012). It seems that a sizable group of perpetrators had already been identified to be at risk for sexual abuse before the incident. Rushbrooke et al. (2014) suggested that caregivers are reluctant to work with individuals with ID and who need support with sexual matters because of the fear of what might happen if sex education is provided. As a result, these individuals are sometimes left without support.

The results of the current study also showed that step/foster family member perpetrators are more prevalent in cases involving children receiving ambulant care. Most of the cases with family member perpetrators occurred in the (foster)home, whereas most of incidents with group mate/peer perpetrators occurred inside the institutions. Based on all the descriptive results together, a rough general picture can be drawn of two broad types of sexual abuse involving children with ID: one group of incidents with adolescents, often previously abusive or abused peers of children within (residential) care institutions that most often happened once, and another group of incidents with adult perpetrators abusing children with ID of a repetitive nature that took place outside the institutions, sometimes within the (foster)home. As the type of perpetrator seems quite determining for the characteristics of the incident, but also for the reactions to the incident, more research is needed to determine whether perhaps specific programs should be developed to prevent these different types of sexual abuse committed by a different group of perpetrators.

The results of our comparison of the reports involving children with ID with the reports involving children in comparable settings but without ID indicated that children with ID in youth care were relatively more often abused in the (foster)homes, whereas children without ID in youth care were more often abused inside the care institution. Special attention for the possible risks in the foster home situations of the children with ID is very important to further diminish the risks for sexual abuse. Hereby, one could think of better screening of foster families and using extensive risk assessments to screen for risks in the family, including, if feasible, extended family members like stepfathers, grandfathers, uncles, frequently visiting friends of the family, also with kinship foster care placements. In their latest annual report, the Dutch Inspectorate of Youth Care also strongly recommended making better use of risk assessment instruments in order to judge the safety and risks within a vulnerable child’s living conditions (Inspectie Jeugdzorg, 2014).

Additionally, and quite surprisingly, the cases involving children with ID were slightly more often reported to the police, perhaps because these more often concerned incidents in (foster)homes, and that in those cases care providers in institutions especially feel the need to report the case both to the inspectorate and the police. Finally, it is likely that institutions are more inclined to report to the police in cases of children with ID because the care providers themselves feel less equipped in dealing with these kinds of incidents involving children with ID (Wissink et al., 2012). In line with this possible explanation, Keilty and Connelly (2001) also found that police expressed frustration around the high number of reported sexual incidents involving people with ID. According to some of the police officers interviewed, care providers did not discriminate between minor (consensual activity) and serious incidents and seemed to regard reporting to the police as a “butt-covering exercise” (p. 286).

In all cases, obviously, it is important to be attentive when children start to behave differently or say inappropriate things. Quite remarkable, the results of the current study showed that most of the reported cases of sexual abuse involving children with ID were discovered because children disclosed the act themselves (67%). Another substantial part of the reported cases of sexual abuse came to light because staff members of the care institutions noticed signals. According to the files of the reports, diverse reactions followed. For example, exhaustive conversations with the victims and/or perpetrators, transferrals to other departments, additional accompaniment of the child or adolescent, additional education of personnel, extra supervision, and practical measures.

Before we conclude, some limitations of the current study need to be addressed. First, the results showed that most reports of sexual abuse concerned children who received care in residential institutions (84%). However, the current data are not conclusive concerning whether there are indeed more incidents in these residential institutions or whether these residential institutions simply report the abuse more often than ambulant care institutions. Nevertheless, there are indications that children in residential care institutions are at higher risk for sexual abuse (Halter, Brown, & Stone, 2007; Spanjaard et al., 2000; Timms & Goreczny, 2002; Vermeulen, Jansen, & Feltzer, 2007; Wissink et al., 2012; Wissink et al., 2015). Spanjaard et al. (2000), for instance, suggested that children in residential institutions possibly have more opportunities to experiment (with the danger of crossing borders). In line with this, Timms and Goreczny (2002) reasoned that residential placement of children and adolescents with ID could diminish normative behaviour and makes it easier to show deviant sexual trespassing behaviours. Moreover, Halter et al. (2007) also mentioned that unacceptable sexual behaviour by professionals seems to be more prevalent in residential settings, and Freel (2003) found that a sexual interest in children is relatively common among male public sector childcare workers. Finally, Vermeulen and colleagues (2007) reported that a larger percentage of children with ID in residential institutions were abused than children with ID in ambulant day care or who lived at home. In further research, more attention should be paid to the possibility that differences in risks are dependent on the nature of the therapeutic setting. The same problem, as mentioned earlier, is true for the less extreme types of sexual abuse, such as touching: Do these types really occur less often than the more extreme types or are they reported less often? A legal obligation for all professionals to report all incidents of sexual abuse could take away some of these concerns. Additionally, it is recommended that studies are conducted that use self-report data as well.

A final limitation of the current study is that the information found in the examined files of the inspectorates was not consistent or uniform. A recommendation was to develop a more consistent and uniform format for the report of incidences of sexual abuse (Wissink et al., 2012). In general, and internationally, a better and more systematic registration of incidents of sexual abuse in youth care could contribute to a better understanding of the problem, followed by more adequate measures.

Notwithstanding these limitations, the present study was the first to provide an insight into the official reports of sexual abuse of children receiving state care in the Netherlands. Based on the results, recommendations have been made to improve prevention and intervention (also see Wissink et al., 2012; Wissink et al., 2015). It is promising that since then more attention has been paid to the subject of child abuse in general, and to the abuse of children with ID specifically, both at an institutional and at a national level. This has already resulted in better education of both children and employees, more straightforward guidelines, and improved legislation. In 2016, the child ombudsmen in Europe again asked for more attention for the victims of child abuse as they felt that the number of victims should be much lower. They stated that professionals should be better trained to signal abuse and that good measures to effectively deal with child abuse should be implemented. It was estimated that each euro invested in the prevention of child abuse would deliver a 19-euro reduction in medical and criminal justice costs.

In conclusion, and based on the research findings, additional attention for good screening and risk assessments of the foster home situations of children with ID seems an important next step to diminish the risk of sexual abuse incidents outside institutions. Regarding incidents within (residential) institutions, it is especially important that those care providers who work with (adolescent) children with ID are helped to find the right balance between granting these children some level of independence and protecting them against risks. Correct (and personalised) information about sexual development in a child with ID is therefore crucial. It is also important to train care providers in how they can speak about issues around sexuality, how they can take it out of its secretive sphere, and in how they can educate children with ID about issues around sexuality (what is appropriate behaviour, boundaries, protection, saying “no,” children’s rights, etc.). It is therefore important to keep in mind the different sexual abuse situations (situations among peers and situations in which an adult commits the abuse). Finally, uniform international and practical guidelines regarding what behaviours should be reported, and, on a national level, to which instance, should be very clear to all people working and living with vulnerable children. These guidelines should not only be available, but also whether people put these guidelines and protocols into practice should be monitored. If not, more effort is needed to get these guidelines “out of drawers” and “inside the minds” of the people who are working with this already vulnerable group of children.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Dutch Committee Samson (led by former attorney-general Mr H. W. Samson-Geerlings) by order of the Dutch Minister of Justice and the Minister of Youth and the Family under Grant 2011-0058.

Notes

1 It should be noted that it is not the case that incidents between group mates or peers especially occurred with children with ID; there was no significant association between the perpetrator and whether children belonged to the group with ID or group without ID.