Correlates of joint child protection and police child sexual abuse investigations: results from the Canadian Incidence Study of Reported Child Abuse and Neglect 2008
Volume 35 · Number 8/9 · October/November 2015
- Send feedback to editors
- Correlates of joint child protection and police child sexual abuse investigations: results from the Canadian Incidence Study of Reported Child Abuse and Neglect-2008 Download this article as a PDF (141 KB - 8 pages)
- Current Issue
- Past Issues
- Submit a manuscript
- About the Journal
- Contact the Editors
Correlates of joint child protection and police child sexual abuse investigations: results from the Canadian Incidence Study of Reported Child Abuse and Neglect-2008
L. Tonmyr, PhD (1); A. Gonzalez, PhD (2)
This article has been peer reviewed.
- Surveillance and Epidemiology Division, Public Health Agency of Canada, Ottawa, Ontario, Canada
- Department of Psychiatry and Behavioural Neurosciences, Offord Centre for Child Studies, McMaster University, Hamilton, Ontario, Canada
Correspondence: Lil Tonmyr, Surveillance and Epidemiology Division, Public Health Agency of Canada, 785 Carling, AL 6807B, Ottawa, ON K1A 0K9; Tel: 613-240-6334; Email: Lil.Tonmyr@phac-aspc.gc.ca
Introduction: Our study examines the frequency of joint investigations by child protection workers and the police in sexual abuse investigations compared to other maltreatment types and the association of child-, caregiver-, maltreatment- and investigation-related characteristics in joint investigations, focussing specifically on investigations involving sexual abuse.
Methods: We analyzed data from the Canadian Incidence Study of Reported Child Abuse and Neglect-2008 using logistic regression.
Results: The data suggest that sexual abuse (55%), and then physical abuse, neglect and emotional maltreatment, are most often co-investigated. Substantiation of maltreatment, severity of maltreatment, placement in out-of-home care, child welfare court involvement and referral of a family member to specialized services was more likely when the police were involved in an investigation.
Conclusion: This study adds to the limited information on correlates of joint child protection agency and police investigations. Further research is needed to determine the effectiveness of these joint investigations.
Keywords: child abuse, child sexual abuse, child maltreatment, police, child protection worker
- Sexual abuse is the maltreatment category most often jointly investigated by child protection and police, followed by physical abuse, neglect and emotional maltreatment.
- We found that substantiation of maltreatment, severity of maltreatment, placement in care, involvement of the child welfare court and referrals of a family member to specialized services were more likely in joint investigations.
Child maltreatment is a challenging problem that requires multidisciplinary approaches to solving it. The social determinants of health almost exclusively fall outside the traditional health sector, and this is particularly true for child maltreatment as a health issue.Footnote 1 Both police officers and child protection workers are mandated to protect children from harm, and since the late 1960s the police have actively collaborated with social workers to protect children from maltreatment.Footnote 2 Joint investigations involving child protection services and the police are consistent with a public health approach to addressing child maltreatment. In fact, the World Health Organization champions multisectoral collaboration as an important component in preventing child maltreatment and improving the health of children.Footnote 3 Such joint investigations have led to increased communication and cooperation between the police and child protection agencies, with the development of written policies, interagency agreements and multidisciplinary teams.
Although a body of literature investigating the attitudes and perceptions of interagency collaborations exists, data examining characteristics of joint investigations are limited. An increased understanding of what factors are associated with joint investigation may help inform future practice and policy standards.
All Canadian jurisdictions have protocols on how and when joint investigations should take place; however, the type and level of collaboration varies, as it does in the United States (US).Footnote 4 For instance, in Alberta, under the Criminal Code of Canada, child protection workers report to the police if they think an offence has been committed.Footnote 5 In British Columbia, when a police officer has reasonable grounds to believe that a child's safety or well-being is in immediate danger, the officer has the authority to take charge of the child and notify a child protection worker as quickly as possible.Footnote 6
One example of collaboration in Canada are the Child Advocacy Centres, which have been funded by the Department of Justice since 2010 and which are currently being evaluated.Footnote 7 Similar centres in the US have been shown to be effective.Footnote 8 These child-focussed centres include teams representing law enforcement, child protection services, prosecution, mental health services, victim advocacy services and child advocacy.Footnote 8
Investigations tend to exist on a continuum: formal joint investigations, where these two groups of professionals must collaborate about certain types of maltreatment; informal joint investigations, where they may work together; and separate investigations, which state one or the other is solely responsible for the investigation.Footnote 9 Formal joint investigation protocols are most common for physical and sexual abuse.Footnote 9 The goals of these joint investigations are to (1) reduce the child's discomfort and trauma by decreasing the number of interviews;Footnote 10 and (2) to better protect the child by improving evidence-gathering through increased communication among professionals.Footnote 10 Footnote 11
Both child protection workers and police officers have raised several concerns to do with their joint involvement in child maltreatment investigations. Stanley et al.Footnote 12 found that the police have a limited understanding of what child protection workers do with the information the police give them. Similarly, police officers expressed concern about delays by child protection workers in informing them about cases,Footnote 12 as well as their attitudes, job performances and capacity to make decisions.Footnote 2 According to Holdaway,Footnote 13 child protection workers are not always available when needed, leaving the police with unresolved problems (e.g. a parent is arrested in the early morning hours and their child is in need of shelter). Conversely, child protection workers were concerned about the police moving too quickly and believed that they were not responsive to constructive criticism.Footnote 14 In addition, police culture sometimes clashed with child protection workers' anti-discriminatory and anti-oppressive attitudes.Footnote 2 Interviews conducted with children were cited as the main source of conflict between the police and child protection workers.Footnote 14 Organizational constraints, such as different working hours, dissimilar geographical areas of jurisdiction and a lack of resources added to the difficulties in collaboration. Each organization felt their professional identity could erode over time.Footnote 14
Despite these challenges, joint investigations have their advantages. Working collaboratively improves communication between police officers and child protection workers. Both professional groups share information more equally, plan more comprehensively, support each other and bring distinctive skills, knowledge and experience to an investigation.Footnote 14
The benefits of working together can also occur at an individual level. Child protection workers felt safer in potentially dangerous situations when police officers were present.Footnote 2 In addition, police officers are legally permitted to use force in particular circumstances, which justifies requests for their presence when addressing high-risk family conflicts.Footnote 13 Furthermore, working closely together promotes understanding of the other agency's function, importance and competencies. Child protection workers reported learning more about the criminal justice system from the police,Footnote 14 whereas from child protection workers the police learned more about communicating with children, including using appropriate language and incorporating toys and play to increase children's comfort in engaging with them.Footnote 15 Joint investigations can also spare the child repeat interviews.Footnote 4,Footnote 16
Knowledge about the characteristics of joint investigations is sparse. One U.S. study found that the majority of sexual abuse investigations were a joint effort (45%) compared to other types of investigations including physical abuse (28%) and neglect (18%).Footnote 17 The credibility of an allegation was strongly predicted by police involvement in the investigation. In addition, caregiver alcohol and/or drug abuse, the severity of abuse, and active domestic violence often increased police involvement. Moreover, it was more likely that services would be provided to the child/family if there was a police investigation.Footnote 17
No such study has been conducted in Canada. A case vignette study that assessed attitudes towards incidents of incest found that the police focussed on collecting evidence whereas child protection workers concentrated on the safety of the child and long-term implications for the family.Footnote 18 However, although hypothetical vignettes are important, they fail to address the complexities involved in decision making and may not reflect what actually occurs.
Given this shortage of information regarding children investigated by child welfare, the Public Health Agency of Canada (PHAC) and partners initiated the Canadian Incidence Study of Reported Child Abuse and Neglect (CIS), which provides a unique opportunity to investigate police involvement with child protection agencies using multivariate techniques.
We used CIS data in this study in order to examine the frequency of joint child protection worker and police investigations into sexual abuse cases compared to other maltreatment types, and to examine the association of child-, caregiver-, maltreatment- and investigation-related characteristics in joint investigations, focussing on investigations involving sexual abuse. Although joint investigations can occur for all types of child maltreatment, we focus on sexual abuse since all jurisdictions across Canada have developed joint protocols for sexual abuse.Footnote 5,Footnote 6 Footnote 19-Footnote 28
The data used came from the Canadian Incidence Study of Reported Child Abuse and Neglect-2008 (CIS-2008).Footnote 29 The CIS was developed by PHAC with its partners to obtain estimates of child maltreatment reported to child protection agencies across Canada. The CIS was approved by McGill University's Ethics Committee. Details of the methodology of the CIS are outlined elsewhere.Footnote 29 In brief, the study used a multistage, stratified, cluster design to acquire a sample of investigations from all 13 jurisdictions in Canada. For 2008, data were collected from October 1 to December 31, 2008, from 112 of 412 agencies. Child protection workers provided information for each child investigated using a standardized CIS three-page form four to six weeks after the initial report to the agency. Researchers provided a half-day training course on how to complete the form. Researchers were also available to answer any questions throughout the data collection period.
The selection of variables was guided by the work of Cross et al.Footnote 17 as well as by hypotheses about factors that would increase the likelihood of a joint investigation (see Table 1). 29 Types of maltreatment included exposure to intimate partner violence (IPV), emotional abuse, neglect, physical abuse and sexual abuse. In the CIS, up to three were collected for each child.
Abbreviation: IPV, intimate partner violence.
Note: The selection of variables was guided by the work of Cross et al.Footnote 17 as well as by hypotheses about factors that would increase the likelihood of a joint investigation.
a We have described the relationship in those investigations where sexual abuse was the sole category of maltreatment.
|Joint investigation||Level of police involvement in the current child welfare investigation: investigation only, charges considered and charges laid|
|Sexual abuse||The child had been sexually molested or sexually exploited. This could include penetration, attempted penetration, oral sex, fondling, sex talk or images, voyeurism, exhibitionism, exploitation and other sexual abuse|
|Physical abuse||The child was physically harmed or could have suffered physical harm because of the behaviour of the person looking after the child. Codes included shake, push, grab, throw, hit with hand, punch, kick, bite, hit with object, choke, poison, stab and other physical abuse|
|Neglect||The child has been harmed or the child's safety or development has been endangered as a result of a failure to provide for or protect the child. Codes included failure to supervise, leading to physical harm; failure to supervise, leading to sexual abuse; permit criminal behaviour; physical neglect; medical/dental neglect; failure to provide psychological treatment; abandonment; and educational neglect|
|Emotional maltreatment||The child had suffered, or was at substantial risk of suffering, emotional harm caused by the person looking after the child. Codes included terrorizing, threatening violence, verbal abuse, belittling, isolation, confinement, inadequate nurturing, exploitation, corrupting behaviour and exposure to non-partner violence|
|Exposure to IPV||The child directly witnessed violence between intimate partners; was indirectly exposed to violence (overhears but does not see; or sees some of the immediate consequences, e.g. injuries; or the child is told or overhears conversations about the assault); and was exposed to emotional violence between intimate partners|
|Multiple type||Any type of sexual abuse co-occurring with another category of maltreatment|
|Primary caregiver substance (alcohol and/or drug) abuse present||If the child protection worker or another worker diagnosed, disclosed, observed, had documented in the file or at the conclusion of the investigation thought it was likely that the primary caregiver abused alcohol and/or drugs|
|IPV confirmed/suspected||If the child protection worker suspected or confirmed that the primary or secondary caregiver was a victim or perpetrator of domestic violence|
|Alleged perpetrator age||40 years or younger vs 41 years or older|
|Alleged perpetrator sex||Male/female|
|Alleged perpetrator||Primary caregiver, secondary caregiver or "other person"a|
|Substantiated maltreatment||Child maltreatment was confirmed after an investigation by a child protection worker|
|Harm||Physical and/or emotional harm|
|Severity||The child required therapeutic treatment for mental/emotional harm and/or medical treatment and/or their health/safety was seriously endangered|
|Referral||At least one referral for services for any family member|
We included only child maltreatment investigations in our analysis. Investigations focussed on "risk of future maltreatment" were excluded, as were investigations involving youth over 15 years as 15 is the oldest age for which child protection services are provided in some jurisdictions. Given that little is known about the characteristics of sexual abuse investigations that involve police, this analysis was largely exploratory. Proportions, or means and standard deviations (SD), were described for all variables of interest. Colinearity between independent variables was tested using chi-square or correlations. Bivariate associations between each variable and the outcome were tested with logistic regression. Due to the large sample size, the criterion for significance was as follows: the increase in predictive power between the model with only a constant and the model with the single variable had to equal or exceed Cox and Snell R² of.01. Independent variables that had a significant bivariate relationship with police involvement were considered for entry into the model. Various models were tested to determine the most parsimonious model that would explain the greatest proportion of variance in police involvement. In all models, variables were entered in one step. Analyses were run in SUDAAN (SUDAAN for Windows, version 7.5.3, Research Triangle Institute, NC, USA), which makes variance adjustments for the correlated data resulting from the survey's design. The nesting variables were agency and family.
Including child maltreatment investigations of those aged under 16 years resulted in a sample of 11 807. Table 2 includes the entire sample to show police involvement in the investigation of the five maltreatment types. Compared with investigations of exposure to IPV (reference), emotional maltreatment was 1.4 times (95% CI: 1.1-1.9) more likely to have police involved in the investigation, neglect about twice as likely (odds ratio [OR] = 1.9; 95% CI: 1.5-2.2), physical abuse 4.5 times as likely (95% CI 3.8-5.4) and investigations of sexual abuse about 21 times more likely to have police involvement (OR = 20.9; 95% CI: 16.8-26.9). About 55% of sexual abuse investigations were conducted jointly (see Table 2).
|Primary child maltreatment||Joint investigations|
Abbreviation: IPV, intimate partner violence.
|Exposure to IPV||2562||94.4||153||5.6|
Of the 11 807 investigations, 842 had sexual abuse as one of the three categories of maltreatment. Table 3 shows that over two-thirds of cases were sexual abuse without co-occurring maltreatment (68.1%), followed by sexual abuse and neglect (15.2%). Fondling (32.3%) and other sexual abuse (27.2%) occurred most frequently, followed by penetration (8.3%), sex talk or images (3.7%), oral sex (3.3%), exhibitionism (2.7%) and attempted penetration (2.0%). Voyeurism occurred in less than 1% of the sample. Information on what acts constituted "other sexual abuse" were not available.
|Distribution of maltreatment categories||N||%|
Abbreviation: IPV, intimate partner violence.
— Estimates of < 5.
|Sexual abuse only||573||68.1|
|Sexual abuse and neglect||128||15.2|
|Sexual abuse and physical abuse||43||5.1|
|Sexual abuse and emotional abuse||23||2.7|
|Sexual abuse and exposure to IPV||23||2.7|
|Sexual abuse and neglect and emotional abuse||13||1.5|
|Sexual abuse and neglect and exposure to IPV||13||1.5|
|Sexual abuse and physical abuse and emotional abuse||8||1.0|
|Sexual abuse and physical abuse and exposure to IPV||8||1.0|
|Sexual abuse and physical abuse and neglect||7||0.8|
|Sexual abuse and emotional abuse and exposure to IPV||—||—|
Investigations of female victims were twice as likely to have the police involved (OR = 2.1; 95% CI: 1.6-2.9) (see Table 4). Having more than one subtype of sexual abuse (OR = 2.2; 95% CI: 1.3-3.6), severe maltreatment (OR = 2.9; 95% CI: 1.8-4.5) or evidence of harm (OR = 3.5; 95% CI: 2.3-5.3) increased the likelihood of police involvement (see Table 5). Investigations where the alleged perpetrator was identified as "other person" (someone other than the primary or secondary caregiver; OR = 1.8; 95% CI: 1.2-2.6; primary caregiver reference category) or involved a male alleged perpetrator (OR = 1.9; 95% CI: 1.3-2.8), were more likely to involve police.
Abbreviation: SD, standard deviation.
|Sex of child|
|Substance abuse by primary caregiver|
|Mean (SD) age of child age, years||9.1 (4.2)||8.4 (4.0)||8.8 (4.1)|
a Not primary or secondary caregiver.
|Sexual abuse only vs sexual abuse with other maltreatment|
|Sexual abuse alone||313||71.5||260||64.4||573||68.1|
|Number of subtypes of sexual abuse|
|More than one||81||18.5||38||9.4||119||14.1|
|Intimate partner violence in the home|
|Sex of alleged perpetrator|
|Age of alleged perpetrator, years|
Family members as well as peers were the "other persons" most commonly involved in joint investigations where sexual abuse was the sole investigation of maltreatment (see Table 6).
a Not primary or secondary caregiver.
|Biological or step- or common-law mother/father or adoptive/foster parenta||41||20.7||30||21.1||71||20.9|
|Grandparent, uncle, aunt or other relative||48||24.2||14||9.9||62||18.2|
|Sibling/foster sibling, cousin, boyfriend/girlfriend, child's friend or peer||59||29.8||47||33.1||106||31.2|
|Family friend, babysitter/babysitter's family, neighbour or boarder||25||12.6||17||12.0||42||12.4|
|Recreational staff, maintenance staff, daycare provider, coach, teacher or other professional||7||3.5||0||0||7||2.1|
Police involvement in the investigation was associated with an increased likelihood that the case would be substantiated (OR = 4.5; 95% CI: 3.1-6.6), that the child would be placed or considered to be placed into care during the investigation (OR = 2.8; 95% CI: 1.4-5.6), that a referral to services would be made for the family (OR = 1.6; 95% CI: 1.1-2.9) and that child welfare court would become involved (OR = 3.2; 95% CI: 1.7-5.9) (see Table 7).
|Investigation-related characteristic||Joint investigations|
Abbreviation: IPV, intimate partner violence.
|Substantiated sexual abuse|
|Case previously opened for any family member|
|Referral for any family member|
|Child welfare court|
|Police involved in adult IPV investigation|
On examining colinearity between variables with significant associations with police involvement, we found that 36 of the 66 pairs of independent variables predictors were significantly related at α =.05. Most notable were the associations between alleged perpetrator type-primary caregiver, secondary caregiver, other person-and their sex and between harm and severity. Perpetrator type explained 59% of the variance in perpetrator sex, and severity explained 47% of the variance in harm. Due to the high colinearity between these variables, we tested models that included either alleged perpetrator type or sex and harm or severity and were chosen based on goodness of fit. The most parsimonious model is shown in Table 8. A female victim, a male alleged perpetrator, the presence of harm and case substantiation were significantly associated with police involvement.
|Contrast||Degrees of freedom||Wald||p value||OR (95% CI)|
Abbreviations: CI, confidence interval; OR, odds ratio.
Note: Cox & Snell R² =.16
|Model minus intercept||4||18.69||<.0001|
|Substantiated case (yes)||1||26.68||<.0001||2.85 (1.92-4.25)|
|Alleged perpetrator sex (male)||1||8.10||.0046||1.83 (1.21-2.78)|
|Harm (yes)||1||8.39||.0039||2.01 (1.25-3.22)|
|Sex of child (female)||1||9.37||.0023||1.65 (1.20-2.27)|
Our study adds to the sparse information on correlates of joint child protection agency and police investigations. Consistent with prescribed protocols, we found that sexual abuse (55%) is the maltreatment category most often jointly investigated by child protection and police, followed by physical abuse, neglect and emotional maltreatment. This is similar to findings from USA and the United Kingdom.Footnote 14,Footnote 16,Footnote 17,Footnote 30
Being female (for the victims) and male (for the perpetrators) are established risk factors for child sexual abuseFootnote 31 and are associated with police involvement. Most research indicates that females are exposed to sexual abuse more often than are males.Footnote 31 However, others argue that sexual abuse of male victims is underreported.Footnote 32
We found no differences in police involvement in sexual abuse investigations with different ages of the victims. Investigations with and without police involvement focussed on pre-adolescents. It is possible that it is easier to gather information from these children and that they may be seen as more credible sources of information.Footnote 33 Furthermore, sexual abuse of pre-adolescents may be seen as more serious and that of adolescents may be less likely to be reported.
Investigations of sexual abuse by the "other person" most often included family members or peers as perpetrators. This is consistent with research that shows that in most instances the victim knows the perpetrator. For example, in a representative U.S. survey, Finkelhor et al.Footnote 34 found that acquaintances perpetrated sexual abuse 91% of the time compared to 7% for strangers and 2% for family. These findings are not directly comparable to our results as the above study is population based, whereas our study focusses on children known to child protection.
Several service-related variables were significantly related to joint investigations. Like Cross et al.,Footnote 17 we found that substantiation of maltreatment, severity of maltreatment, placement in care, involvement of the child welfare court and referrals of a family member to specialized services were more likely in joint investigations. We can only speculate on the reasons, given the nature of the data, but the mandate of joint investigations is to ensure safety and protection of children. This can involve providing evidence to the court for removal of the child to care or punishment of the perpetrator,Footnote 18 but it can also encompass the provision of services to improve parenting capacity whether it includes alcohol and/or drug counselling or receipt of financial assistance or child-focussed services. Cross et al.Footnote 17 suggested that the additional services offered in joint investigations may indicate that police involvement does not, in fact, cause a greater distress that hampers the relationship with the investigated families or change the focus of child protection agencies' mandate to assist children and families. Instead, this could be a reflection of the increased closeness of police and the public as a result of the movement towards community policing.Footnote 35
Contrary to findings by Cross et al.,Footnote 17 we did not find substance abuse and IPV of the primary caregiver to be associated with an increased likelihood of police involvement-although their findings were not specific to sexual abuse but were across maltreatment types. It may be that other caregiver risk factors, not included in these analyses, such as maternal age, caregiver mental health and the presence of a stepfatherFootnote 36,Footnote 37 are related to increased likelihood of police involvement.
Strengths and limitations
Although the CIS has several strengths, there are limitations that may influence the conclusions that can be drawn from the data. In fact, the CIS
- collects information on children reported to child protection agencies and thus underestimates the occurrence of maltreatment;
- portrays the clinical opinion of child protection workers, which are not independently verified;
- gathers data during three months in the fall, which may not be representative of the year; and
- uses a cross-sectional design, which precludes establishing causation.
The results of this analysis suggest that, of all categories of reported maltreatment, sexual abuse is most likely to involve joint investigations. Although, continued collaboration between specialized groups of child protection workers and police may be appropriate, the potential benefits of this synergistic approach need to be monitored and evaluated. Little is known about its effectiveness in Canada. Although more is known about joint investigations in the US, the Canadian context may be different due to the legal system and jurisdictional issues. Also, some authorsFootnote 30 speculate whether child protection agencies have gone too far in the investigative process and have moved away from the critical matters of prevention and treatment. This question merits attention, although the present analysis suggested that more services were provided after joint investigations. The roles of other professions on multidisciplinary teams such as those within the health sector also require further attention.
More research is needed to investigate the effectiveness of joint investigations, for instance, has there been a decline in the number of times a child is interviewed? Has information-sharing between police and child protection increased? Would it be beneficial to the health of the child to have joint investigations more often for the other types of maltreatment than happens currently? These questions need to be answered in order to increase our understanding of collaborative investigations and their potential to benefit victims of child maltreatment.
Responding to child maltreatment with multidisciplinary teams will assist in an integrated response through primary and secondary prevention of child maltreatment and associated negative health outcomes, surveillance, increased public awareness, development of best practices and evaluation of programs, especially considering that children and their families received more services when investigations were joint. It is promising to see that joint investigations take place in the intended circumstances as per provincial and territorial protocols.
The authors gratefully acknowledge the assistance of Jasminka Draca, Dr. David Hubka, Ellen Jamieson, Joanne Lacroix and Caroline Wallace.
Report a problem or mistake on this page
- Date modified: