Socioeconomic risk and the longitudinal child lifetime prevalence of child protection involvement.

BACKGROUND: North American studies find that geographic indicators of disadvantage, such as concentrated poverty, significantly increase the risk of child protection involvement. Despite having one of the most extensive family support systems and progressive income redistribution policies in North America, the Canadian province of Québec still faces geographic variations in socioeconomic conditions that remain a major risk factor for child protection involvement. OBJECTIVE: This study asks how child protection involvement is distributed across socioeconomically distinct geographic areas of the province. Drawing from prior literature, we hypothesize that the highest level of child protection involvement across childhood (age 0-17) is found in the lowest socioeconomic areas. PARTICIPANTS & SETTING: This is a population-based prevalence study using administrative child protection data spanning the years 2000 to 2017 across Québec. METHODS: We constructed cumulative risk life tables of first instances of child protection events (report confirmation, compromised security or development, and out-of-home placement). Prevalence rates were mapped onto 10,650 Census dissemination areas divided into three tiers according to a validated socioeconomic status (SES) index. RESULTS: The highest childhood prevalence of confirmed child protection reports, finding of compromised security or development, and out-of-home placement was found in the lowest SES areas. Rates in low SES areas can be over twice the rates in high SES areas. CONCLUSIONS: Area-level socioeconomic vulnerability remains a robust predictor of child protection involvement even in a socially progressive context. Our findings underscore that without targeted pediatric and family services, as well as poverty-alleviation programs for high-need families in high-need areas, even well-intentioned systems may fall short of reaching the families most in need.

Reunification trajectories in Quebec: Acknowledging chronic family challenges to support stability.

BACKGROUND: While family reunification following out-of-home placement is a goal of child protection policy, complex family needs may not be met at the point that child protection systems reunify families. Permanency legislation creating maximum placement timeframes prompts questions regarding what families need to be supported in stably reunifying following a child's removal from the home. OBJECTIVE: We aim to identify clinical risk factors salient for initial placements and placements following a child reunifying with their family to inform successful reunification and improve children's stability. PARTICIPANTS AND SETTING: The study includes a representative sample of children in Quebec with a child protection investigation in 2008 (N = 3051) followed for nine years. METHODS: Cross-sectional clinical data from the Quebec Incidence Study (QIS) on Evaluated Child Protection Reports (2008) were linked with longitudinal administrative data from 16 provincial child protection agencies. Canadian Census data (2006) were used to create a factorial index measure for poverty. Chi-square automatic interaction detector (CHAID) decision tree analysis was used to compare risk factors salient for initial placements (n = 1120) with post-reunification placements (n = 455). RESULTS: For the placement sample (n = 1120), significant factors were: attachment issues, caregiver drug use, child's suicidal thoughts, child's self-harming behavior, and academic difficulties. Of the children who reunified with their families (n = 847), over half (n = 455; 54%) returned to out-of-home placements. Certain factors remained significant for placement after reunification: academic difficulties, attachment issues, and caregiver drug use. The CHAID model fit estimates suggest 70.9% (SE = 0.008) accuracy predicting out-of-home placement following child protection investigation and 58.2% (SE = 0.017) accuracy predicting re-placement following family reunification. CONCLUSIONS: Complex needs among families most likely to experience reunification breakdown indicate potential service gaps. When legislated placement timeframes prompt quick resolution of family challenges, these analyses can contribute to policy discussions regarding clinical family challenges that impact stability.

Childhood Prevalence of Involvement with the Child Protection System in Quebec: A Longitudinal Study.

The goal of this study, the first of its kind in Canada, was to estimate the child lifetime prevalence of child protection involvement in Quebec. Using administrative and population data spanning 17 years, we performed a survival analysis of initial incidents of child protection reports, confirmed reports, confirmation of a child's security or development being compromised, and placement outside the home for one day or more. We found that before reaching the age of 18 years, over 18% of children were reported to child protection at least once, one in every ten children (10.1%) in the province had a report that led to the finding of their security or development being compromised, and over 5% were placed outside the home. We found that neglect was a primary concern in close to half (47.6%) of cases. By using a full population dataset, we obtained a more accurate prevalence estimate than studies using synthetic cohort life tables. These findings only captured initial incidents of involvement with child protection, meaning this study does not show the extent of recurrent involvement for some children. The findings reflect prior results showing that neglect is common in initial child protection involvement but less pervasive than has been shown in incidence studies, suggesting that recurrent child protection involvement is more driven by neglect than initial incidents are.

Recurrent involvement with the Quebec child protection system for reasons of neglect: A longitudinal clinical population study.

Studies in several jurisdictions have found that families become recurrently involved with child protection systems most frequently for reasons of neglect. Child protection involvement for reasons of neglect is shown to correlate with various socioeconomic vulnerabilities. OBJECTIVE: This study, the largest of its kind in Canada, examines when and for whom recurring conditions of neglect were most likely to occur for all children involved with child protection in the province of Quebec over a span of fifteen years. PARTICIPANTS AND SETTING: Specifically, the study population includes all children whose ongoing child protection intervention in Quebec closed between 2002 and 2017 (N = 76,176). METHODS: This clinical population study uses a longitudinal research design drawing anonymized clinical administrative data from all of Quebec's child protection jurisdictions spanning 15 years, and Quebec data extracted from the 2011 Canadian National Household Survey to estimate socioeconomic vulnerability. RESULTS: Of the total population studied, 32.5 % (N = 24,816) experienced a recurrence of maltreatment during the study period, of which more than one third (N = 8707) experienced a recurrence for reasons of neglect. CONCLUSIONS: Because the association between socioeconomic vulnerability and recurrence of neglect indicates a gap in material and social supports-which child protection systems have neither the mandate nor the resources to fill-we propose additional avenues that we urge policymakers and practitioners to consider in supporting the demonstrated needs of these families.

Differences and disparities over time: Black and White families investigated by Ontario's child welfare system.

BACKGROUND: Black-White disparities in child welfare involvement have been well-documented in the United States, but there is a significant knowledge gap in Ontario about how and when these disparities emerge. OBJECTIVE: This paper compares incidence data on Black and White families investigated by Ontario's child welfare system over a 20-year period. METHODS: Data from the first five cycles of the Ontario Incidence Study of Reported Child Abuse and Neglect (OIS) (1993-2013) were used to examine trends in child maltreatment investigations involving Black and White families. Incidence rates were calculated. T-tests were conducted to assess statistically significant differences between and within cycles. Population and decision-based enumeration approaches were also used to examine child welfare disparities. RESULTS: The incidence of investigations involving White families almost doubled between 1998 and 2003, but for Black families the incidence increased almost fourfold during the same period. These increases and the difference between Black and White families in 2003 were statistically significant. The results further indicate that Black families experience disparate representation in Ontario's child welfare system over time for most service dispositions. CONCLUSIONS: Several possible explanations are offered for the study's outcome, including changes in risk related to social safety net, the threshold for risk of harm, and bias and racist institutional policies and practices. This study invites policy-makers and child welfare authorities to rethink service delivery in addressing the disparate representation of Black families in the child welfare system.

An examination of trends in child sexual abuse investigations in Ontario over time.

BACKGROUND: Child sexual abuse (CSA) rates have been declining since the 1990s (Dunne et al., 2003; Finkelhor & Jones, 2004, 2012; Jones et al., 2001). Discrepancies in contexts and measures complicate comparing CSA rates across jurisdictions and studies, and there is limited literature about trends in CSA in Canada. OBJECTIVE: Using data from the Ontario Incidence Study of Reported Child Abuse and Neglect (OIS), the only source of provincially aggregated data in Ontario, Canada, that describes child welfare investigations, this paper provides information on reported and investigated CSA over the past 20 years. PARTICIPANTS AND SETTING: The OIS uses a file review methodology; information is collected directly from investigating child welfare workers. METHODS: A sample of child welfare agencies is selected for the study, and data are collected over a three-month period. Weights are applied to produce annual provincial estimates. RESULTS: The rates of investigated CSA in Ontario decreased between 1993 and 2013, from 5.20 (95% CI [3.94, 6.47]) to 1.81 (95% CI [0.97, 2.66]) children per 1000. During this time, the rate of all child maltreatment-related investigations doubled, from 21.41 (95% CI [18.38, 24.42]) to 53.32 ([29.61, 77.03]) children per 1000. CONCLUSIONS: Unlike other forms of child maltreatment, the incidence of investigated CSA in Ontario declined since 1993. Substantiation rates for CSA investigations decreased more dramatically than the rate of all CSA investigations, which could indicate a true decline in rate or an inability to accurately identify cases of CSA.

Contrasting mental health correlates of physical and sexual abuse-related shame.

OBJECTIVE: This study represents an initial attempt to contrast behavioural and mental health correlates of shame as a result of physical abuse (PA) and sexual abuse (SA). Because they are distinctive forms of injury, it is possible that corollary shame from these injuries follows unique trajectories and ultimately results in different health challenges. METHOD: Self-report data from a survey on the health of youth receiving protective services for reasons of PA and SA was used. It included standardised measures, such as the Childhood Trauma Questionnaire, Trauma Symptoms Checklist for Children, the Brief Symptoms Inventory, the Rutgers Alcohol Problem Index, and the South Oaks Gambling Screen. New measures of abuse-related shame, maltreatment, and substance use were also employed. Linear regression analyses were performed to determine whether level of shame was linked to mental health and behaviour issues, after controlling for level of abuse. RESULTS: Results were similar for shame as a result of PA and SA victimisation. After accounting for shared variance with abuse severity, both measures were linked to a full spectrum of mental health issues, such as depression (SA r2 = 0.30, PA r2 = 0.28), anxiety (SA r2 = 0.27, PA r2 = 0.20), post-traumatic stress (SA r2 = 0.26, PA r2 = 0.19), interpersonal sensitivity (SA r2 = 0.17, PA r2 = 0.22), and psychoticism (SA r2 = 0.19, PA r2 = 0.20), but not to gambling or substance use problems. PA-related shame was associated with suicidality (r2 = 0.05). CONCLUSIONS: Keeping in mind that this was largely a cross-sectional study and that causality cannot be inferred, the results seem to indicate that youth suffering from abuse-related shame are particularly vulnerable to mental health problems, but not to efforts to numb their problematic thoughts and feelings through gambling and substance use. Shame could serve as an early indicator of which child protection recipients are most in need of preventive efforts.

Child Maltreatment Among Children With Intellectual Disability in the Canadian Incidence Study.

This study aims to compare, among a representative sample of substantiated child maltreatment cases, the characteristics of those with intellectual disability (ID) from those without ID. Using the 2008 Canadian Incidence Study of Reported Child Abuse and Neglect, 5,797 cases of substantiated maltreatment that involved children aged between 0 and 14 years were analyzed. One in 10 children (11.3%) was identified with ID. Results revealed functional problems to be higher among children with ID and their parents. Moreover, children with ID experienced more severe maltreatment, and were more often referred to ongoing child protection services. These findings suggest that maltreated children with ID are facing additional challenges that must be accounted for in service planning and delivery.

The Effects of Socioeconomic Vulnerability, Psychosocial Services, and Social Service Spending on Family Reunification: A Multilevel Longitudinal Analysis.

Socio-environmental factors such as poverty, psychosocial services, and social services spending all could influence the challenges faced by vulnerable families. This paper examines the extent to which socioeconomic vulnerability, psychosocial service consultations, and preventative social services spending impacts the reunification for children placed in out-of-home care. This study uses a multilevel longitudinal research design that draws data from three sources: (1) longitudinal administrative data from Quebec's child protection agencies; (2) 2006 and 2011 Canadian Census data; and, (3) intra-province health and social services data. The final data set included all children (N = 39,882) placed in out-of-home care for the first time between 1 April 2002 and 31 March 2013, and followed from their initial out-of-home placement. Multilevel hazard results indicate that socioeconomic vulnerability, controlling for psychosocial services and social services spending, contributes to the decreased likelihood of reunification. Specifically, socioeconomic vulnerability, psychosocial services, and social services spending account for 24.0% of the variation in jurisdictional reunification for younger children less than 5 years of age, 12.5% for children age 5 to 11 years and 21.4% for older children age 12 to 17 years. These findings have implications for decision makers, funding agencies, and child protection agencies to improve jurisdictional resources to reduce the socioeconomic vulnerabilities of reunifying families.

Exploring alternate specifications to explain agency-level effects in placement decisions regarding Aboriginal children: Further analysis of the Canadian Incidence Study of Reported Child Abuse and Neglect Part C.

A series of papers using data from the Canadian Incidence Study of Reported Child Abuse and Neglect (CIS) explored the influence of clinical and organizational characteristics on the decision to place Aboriginal children in out-of-home placements at the conclusion of child maltreatment investigations. The purpose of this paper is to further explore a consistent finding of the previous analyses: the proportion of investigations involving Aboriginal children at a child welfare agency is associated with placement for all children in that agency. CIS-2008 data were used in the analysis, which allowed for inclusion of previously unavailable organizational and contextual variables. Multi-level statistical models were developed to analyze the influence of clinical and organizational variables on the placement decision. Final models revealed that the proportion of investigations conducted by the child welfare agency involving Aboriginal children was again a key agency-level predictor of the placement decision for any child served by the agency. Specifically, the higher the proportion of investigations of Aboriginal children, the more likely placement was to occur for any child. Further, this analysis demonstrated that structure of governance, an organizational-level variable not available in previous cycles of the CIS, is an important agency-level predictor of out-of-home placement. Further analysis is needed to fully understand individual and organizational level variables that may influence decisions regarding placement of Aboriginal children.

Exploring alternate specifications to explain agency-level effects in placement decisions regarding aboriginal children: further analysis of the Canadian Incidence Study of Reported Child Abuse and Neglect Part B.

OBJECTIVE: This paper builds upon the analyses presented in two companion papers (Fluke et al., 2010; Fallon et al., 2013) using data from the 1998 and 2003 cycles of the Canadian Incidence Study of Reported Child Abuse and Neglect (CIS-1998 and CIS-2003) to examine the influence of clinical and organizational characteristics on the decision to place a child in out-of-home care at the conclusion of a child maltreatment investigation. This paper explores various model specifications to explain the effect of an agency-level factor, proportion of Aboriginal reports, which emerged as a stable and significant factor through the two data collection cycles. It addresses the issue of data comparability between the two cycles and explores various re-specifications and descriptive analyses of reported models to evaluate their solidity with regards to the sampling schemes and the precise contribution of a multi-level specification. METHODS: The decision to place a child in out-of-home care was examined using data from the CIS-2003. This child welfare dataset collected information about the results of nearly 12,000 child maltreatment investigations as well as a description of the characteristics of the workers and organization responsible for conducting those investigations. Multi-level statistical models were developed using MPlus software, which can accommodate dichotomous outcome variables and are more reflective of decision-making in child welfare. The models are thus multi-level binary logistic regressions. RESULTS: Final models revealed that two agency-level variables, 'Education degree of majority of workers' and 'Degree of centralization in the agency' clarify the nature of the effect of 'Proportion of Aboriginal reports', a stable, key second level predictor of the placement decision. The comparability of the effect of this agency-level variable across the 1998 and 2003 cycles becomes further evident through this analysis. By using a unified database including both cycles and various specifications of models, the comparability was found to be robust, in addition to clarifying the precise contribution of a multi-level specification. CONCLUSIONS: This third paper in a series establishes the 'Proportion of Aboriginal reports' received by the child welfare agency as an important agency level predictor associated with a child's likelihood of being placed in the Canadian child protection system. While the more complex models give support to the notion that unequal resources subtend those results, more analyses are needed to confirm this hypothesis. Unequal resources for agencies with larger Aboriginal caseloads may explain the persistence of the results. These findings suggest that specific resource constraints related to worker education may be explanatory.

Placement decisions and disparities among Aboriginal children: further analysis of the Canadian incidence study of reported child abuse and neglect part A: comparisons of the 1998 and 2003 surveys.

OBJECTIVE: Fluke et al. (2010) analyzed Canadian Incidence Study on Reported Child Abuse and Neglect (CIS) data collected in 1998 to explore the influence of clinical and organizational characteristics on the decision to place Aboriginal children in an out-of-home placement at the conclusion of a child maltreatment investigation. This study explores this same question using CIS data collected in 2003 which included a larger sample of Aboriginal children and First Nations child and family service agencies. METHODS: The decision to place a child in an out-of-home placement was examined using data from the Canadian Incidence Study of Reported Child Abuse and Neglect-2003 and a reanalysis of CIS-1998 data (Fluke et al., 2010). The CIS-2003 dataset includes information on nearly 12,000 child maltreatment investigations from the time of report to case disposition. The CIS-2003 also captures information on the characteristics of investigating workers and the child welfare organizations for which they work. Multi-level statistical models were developed to analyze the influence of clinical and organizational variables using MPlus software. MPlus allows the use of dichotomous outcome variables, which are more reflective of decision-making in child welfare and facilitates the specific case of the logistic link function for binary outcome variables under maximum likelihood estimation. RESULTS: Final models revealed the proportion of investigations conducted by the child welfare agency involving Aboriginal children was a key single agency level predictor of the placement decision. Specifically, the higher the proportion of investigations of Aboriginal children, the more likely placement was to occur. Contrary to the findings in the first paper (Fluke et al., 2010), individual Aboriginal status also remained significant in the final model at the first level. CONCLUSIONS: Further analysis needs to be conducted to further understand individual and organizational level variables that may influence decisions regarding placement of Aboriginal children. There is also a need for research that is sensitive to differences among, and between, Métis, First Nations and Inuit communities. Results are not generalizable to Québec because data from this province were excluded.

Shifting definitions of emotional maltreatment: an analysis child welfare investigation laws and practices in Canada.

OBJECTIVE: Although there is growing evidence that the emotional dimensions of child maltreatment are particularly damaging, the feasibility and appropriateness of including emotional maltreatment (EM) in child welfare statutes continues to be questioned. Unlike physical and sexual abuse where investigations focus on discreet incidents of maltreatment, EM is not as easily defined and delimited. Through a review of legislation and child welfare investigation practices in Canada, this paper examines (1) whether Canadian child welfare services respond to EM with the same level of perseverance as with other forms of maltreatment and (2) the extent to which the introduction in 2008 of a more specific EM taxonomy distinguishes between EM and family problems that could lead to EM. METHOD: Following an analysis of the legislative framework for EM across Canada, investigations practices in Canada are examined using data from the 1998, 2003 and 2008 cycles of the Canadian Incidence Study of Reported Child Abuse and Neglect (CIS). Using data from the 2008 cycle, EM investigations are compared to other maltreatment investigations for all of Canada (N=15,980). Changes in EM investigations over time are then compared using data from the three cycles of the study, excluding Québec because of limited data availability in 2003 (N=5,360 in 1998, 11,562 in 2003 and 14,050 in 2008). RESULTS: EM is included as a form of reportable maltreatment in all provincial and territorial statutes in Canada. Over 11,000 cases of EM were substantiated in Canada in 2008, at a rate of 1.86 cases per 1,000 children. While EM investigations were substantiated at a lower rate as other forms of maltreatment, a higher proportion of EM cases were referred for specialized services, kept open for on-going child welfare services, lead to an out of home placement, and lead to an application to child welfare court. Using a broad definition of EM the number of investigations classified as EM in Canada, excluding Québec, nearly tripled from 1998 to 2003. In 2008, using more specific definitions focusing on caregiver definitions, the number of investigations classified as EM nearly returned to their 1998 level, with nearly twice as many cases being classified as risk of future maltreatment. CONCLUSION: EM is a well established category for child welfare intervention in Canada, however, more emphasis should be given to distinguishing between EM and family problems that place children at risk of EM.

Placement decisions and disparities among aboriginal groups: an application of the decision making ecology through multi-level analysis.

OBJECTIVE: This paper examined the relative influence of clinical and organizational characteristics on the decision to place a child in out-of-home care at the conclusion of a child maltreatment investigation. It tested the hypothesis that extraneous factors, specifically, organizational characteristics, impact the decision to place a child in out-of-home care. A secondary aim was to identify possible decision making influences related to disparities in placement decisions tied to Aboriginal children. Research suggests that the Aboriginal status of the child and structural risk factors affecting the family, such as poverty and poor housing, substantially account for this overrepresentation. METHODS: The decision to place a child in out-of-home care was examined using data from the Canadian Incidence Study of Reported Child Abuse and Neglect. This child welfare dataset collected information about the results of nearly 5,000 child maltreatment investigations as well as a description of the characteristics of the workers and organization responsible for conducting those investigations. Multi-level statistical models were developed using MPlus software, which can accommodate dichotomous outcome variables, which are more reflective of decision making in child welfare. MPlus allows the specific case of the logistic link function for binary outcome variables under maximum likelihood estimation. RESULTS: Final models revealed the importance of the number of Aboriginal reports to an organization as a key second level predictor of the placement decision. It is the only second level factor that remains in the final model. This finding was very stable when tested over several different levels of proportionate caseload representation ranging from greater than 50% to 20% of the caseload. CONCLUSIONS: Disparities among Aboriginal children in child welfare decision making were identified at the agency level. PRACTICE IMPLICATIONS: The study provides additional evidence supporting the possibility that one source of overrepresentation of Aboriginal children in the Canadian foster care system is a lack of appropriate resources at the agency or community level.

What predicts injury from physical punishment? A test of the typologies of violence hypothesis.

OBJECTIVE: This study examined the power of child, perpetrator, and socio-economic characteristics to predict injury in cases of reported child physical abuse. The study was designed to assess the validity of the assumption that physically injurious incidents of child physical abuse are qualitatively different from those that do not result in injury, that their generative factors are distinctive, and that the quality of caregiving in these two types of incidents is different. METHOD: A weighted, nationally representative sample of 8,164 substantiated punishment abuse cases in Canada was used. Various models were constructed and evaluated through logistic regression. RESULTS: Of six potential predictors - child age, perpetrator sex, child functioning, parent functioning, economic stress, and social stress - none predicted injury to the child. CONCLUSIONS: The findings suggest that injurious and non-injurious physical abuse cannot be distinguished on the basis of the personal characteristics or circumstances of the child or perpetrator. PRACTICE IMPLICATIONS: A common criterion for child welfare intervention into cases of suspected physical abuse is injury or risk of injury. This criterion assumes that injurious and non-injurious assaults are qualitatively different phenomena, predicted by different risk factors. In the present study an attempt was made to differentiate between injurious and non-injurious cases of punitive physical abuse on the basis of characteristics of the child, perpetrator, family, and social context. None of these factors explained the likelihood of injury, suggesting that the prediction of injury as an intervention criterion may be questionable.