Tailoring service and system design for families known to child protection: A rapid exploratory analysis of the characteristics of families

The Royal Commission also highlighted the specific needs of Aboriginal children and children from culturally and linguistically diverse backgrounds who require culturally appropriate and tailored family support responses. The term Aboriginal as used in this paper is inclusive of Aboriginal and/or Torres Strait Islander people. Early intervention and prevention services had largely been designed based on research with families of Anglo descent, and while research has repeatedly identified the need for culturally designed and tailored intervention (Family Matters, 2020), the proportion of families who may need such responses has remained unclear (Kaur, 2012; Lewig et al., 2009). This rapid exploratory analysis enabled estimates of the proportion of families identified as CALD or with Aboriginal child. Ethics approval was obtained from the Aboriginal Health Research Ethics Committee (Protocol #04–17-718) and the University of South Australia Human Research Ethics Committee (Application ID: 0000036590). The research was also governed by an Aboriginal Leadership Group who guided research design, coding, and interpretation of findings. The final sample included 280 families who had at least one child in the family reported to child protection in a metropolitan region of Australia within a six-month period. The sample was identified using a three-stage process (see Figure 1). In stage one, unit record child protection administrative data for all 1619 children reported to child protection between 1 July and 31 December 2016 who resided in a metropolitan region (17 postcodes) were extracted. This region has a usual population of 151,758 people, with approximately 20 per cent (32,413) being children and young people (aged under 19) (Australian Bureau of Statistics, 2018). The 17 postcodes ranged on the Index of Relative Socio-economic Advantage and Disadvantage from the 4th to 80th percentile relative to other postcodes suggesting a balance of socioeconomic areas (Australian Bureau of Statistics, 2018). Stage two involved the selection of a 20 per cent random sample of the 1619 children reported to child protection. Twenty-three per cent of children in the random sample were identified as being of Aboriginal descent, which is a representative proportion of Aboriginal children known to child protection in that region. Stage three involved the identification of family groups for the random sample. Family group identification included all biological and legal sibling relationships (including full siblings, half siblings, step sibling, adoptive siblings, and the siblings of siblings) contained in the child protection casefile records. This resulted in the formation of 280 family groups. Unit record administrative data. These data were extracted by the child protection department for all reports to child protection for children in the families, including details of the date of the report and the children who were the subject of the report. Information about child protection orders for all children in the family were also extracted. Casefile data. Child protection relationship information and intake reports were extracted from the electronic case management system. Relationship information for index children, and all their siblings were extracted, and used to identify and create family groups. To identify CALD (as defined in this study as non-Aboriginal and/or non-English speaking families), case notes had to explicitly mention a cultural or linguistic background of families or the need for an interpreter. Rapid exploratory analysis utilising frequency and descriptive statistics were used to examine: the proportion of CALD families; the proportion of families with an Aboriginal child; the extent of child protection involvement for individual index children (aged under 18 years) reported over the six-month study period and their lifetime (i.e. since birth); and the extent of child protection involvement for families over their lifetime (i.e. the number of unique reports over the lifespan of the family since the birth of the eldest child). Level of child protection reporting was identified and grouped into five groups of unique reports (based on a frequency distribution): one report; two to four reports; five to nine repots; 10 to 20 reports; and more than 20 reports, consistent with categories used in previous research (Delfabbro et al., 2010). The study jurisdiction incorporates reports relating to the same incident for a child into one report narrative to avoid counting the same issues twice. For other children in the same family, only unique reports were included if they occurred on separate dates. The 280 family groups comprised between one and 16 children, with a median of three children per family (see Table 1). Twenty-seven per cent of families included at least one Aboriginal child. The families identified as CALD represented 17 per cent of families. Within the six-month period, 39.5 per cent of index children were reported to child protection two or more times (see Table 2). Children had a median of one report (mean = 1.7, SD = 1.0), and had between one and seven reports within the six-month period (Figure 2). When the extent of involvement was examined over the life of the index children, 82.7 per cent were reported two or more times. The median number of reports per child rose to five (mean = 7.9, SD = 8.7), with between one and 51 reports. Almost a third of index children had between two and four reports, and almost a quarter between five and nine reports (Figure 3). The 280 families, identified from the sample of 324 children, were the subject of a total of 8498 reports to child protection about the same or different matters, of which 4435 were reports about unique matters. A total of 86 per cent had some level of re-reporting for unique matters over the lifespan of the children in the family. Families had a median of 7.0 reports (mean = 15.8 reports, SD = 21.4), with total number of reports ranging from one to 126 (Figure 4). Only 14 per cent of families were reported for one unique matter, while over a quarter of families had more than 20 reports about different matters to child protection (see Table 3). Of the 280 families, 15.3 per cent (n = 43) were reported to child protection for the first time during the study period (July to December 2016). The sample also includes 33 families (11.8%) reported during the study period, with a child or children who had been the subject of a child protection order in this or another Australian jurisdiction. In response to both the Royal Commission and emerging evidence, this exploratory study aimed to provide policymakers with a rapid analysis of the extent of child protection involvement for individual children and their families, and the characteristics of families reported. A point of difference in this research is the examination of the extent and patterns of involvement with child protection over the life of the family. The extent of child and family involvement with child protection is high. The present study identified that 40 per cent of children were re-reported within a six-month period and 83 per cent over their lifetime. The level of child protection reporting is higher than national outcome measures which utilise specific performance indicators over relatively short periods (Australian Government Productivity Commission, 2019; Children's Bureau, 2021). For example, in Australia, the proportion of children with a substantiated report who had a subsequent substantiation ranged from 2.7–11.2 per cent within three months, and 16.1–27.3 per cent over the 12-month period (Australian Government Productivity Commission, 2019). In the United States of America, a median of 5.5 per cent (ranging from 0.4% – 12.3%) of children who had a substantiated/indicated report across states had another substantiated/indicated report within six months (Children's Bureau, 2021). This highlights that these indicators may significantly underestimate the extent of re-reporting for individual children and provide inaccurate estimates of system performance for children. Eighty-three per cent of index children were the subject of more than one child protection report, and half had been the subject of five or more reports over their childhood. Over their lifetime, children had an average of 7.9 reports (median = 5), which is higher than previous research. The mean number of reports ranged from 2.0 for children born 1990/1991 in Western Australia (Bilson et al., 2013) to 4.8 for Aboriginal children born in 1991 in South Australia (Hirte et al., 2008). The proportion is also higher than other previous research, for example 40 per cent of a sample of low-income children in Missouri had two or more investigated reports, with 19 per cent having between four and 22 reports by age 16 (Jonson-Reid et al., 2012). These high rates of repeated reporting have been associated with increased risk of youth offending, poorer physical and mental health, alcohol and drug use in childhood, and higher levels of child abuse perpetration in adulthood (Graham et al., 2010; Ireland et al., 2002; Jonson-Reid et al., 2012; Stewart et al., 2002). This indicates an urgent need to focus on children currently reported to child protection in the context of cumulative harm and chronic exposure they and their siblings may have experienced. When examined at a family level, 86 per cent of families known to child protection in a six-month period were reported to child protection multiple times. On average families had 16 reports to child protection (median = 7.0 reports) higher than previous research in which families had on average 3.73 reports following families over eight years (Bromfield & Higgins, 2005). One-quarter of families were known to child protection more than 20 times throughout their life compared to 8 per cent of individual children. This supports the importance of understanding the child protection concerns for children in the context of their family (Meiksans et al., 2021; Putnam-Hornstein et al., 2015). Examining contact for individual children in isolation will underestimate involvement with child protection, and potentially the risks that children face within the family context. It also prompts further questions regarding the opportunities for early intervention to prevent further risk for children and their siblings. This sample included only a small proportion (15%) of families who were reported to child protection for the first time during the study period. This has important implications as screening decisions and priority response times for matters reported to child protection are often based on characteristics of the reported concern. However, this research indicates that over three-quarters of children and families reported to child protection have had multiple child protection concerns raised about them, and this should inform the assessment and response for practitioners including assessments of risk, chronicity of child maltreatment, and cumulative harm. Policymakers should consider how the screening of concerns and decision-making regarding appropriate responses, including alternatives to child protective responses, should be informed by the lifetime extent of reporting and the accumulation of risk and harm for these families. Future research should explore the characteristics and concerns at the time of the first report to child protection for a family to determine opportunities for early intervention and support to reduce the extent of repeat involvement over the life of the family. The sample also included a small, but not insignificant, proportion of families (11.8%) reported during the study period that included a child who had previously been the subject of a child protection order. This indicates that one in 10 families in the sample have required a response where the child/children have not been able to remain in the home. This highlights potentially unique support needs compared with families reported for the first time, such as managing contact arrangements and post-reunification assessment and supports. This is the first study to identify the proportion of culturally and linguistically diverse families from report narratives. The proportions of CALD and Aboriginal families in the child protection system require culturally safe systems and services that recognise and can respond to the diverse needs, circumstances and experiences. For child protection, this may mean appointing staff that reflect the cultural diversity in the community and working towards the cultural safety of child protection systems through consultation, improving cultural knowledge of staff, and implementing policies and assessment tools that reflect the needs of CALD and Aboriginal families. Enhanced self-determination and Aboriginal-led solutions, including the need for family-led decision-making and a greater delegation of child protection decision-making to Aboriginal-led services, are vital to improving Aboriginal child and families outcomes (Family Matters, 2020). Further research is required to understand the population of families reported to the child protection systems to improve early intervention responses. Together with the data custodian, the research team are exploring risk factors and outcomes associated with the extent of reporting; the characteristics and needs of Aboriginal and CALD families including optimal points for early intervention and support; and case conceptualisation of families to understand concerns in the context of historical concerns to plan system-level and individually tailored responses. Understanding the characteristics of these families will allow for the provision of supports and services targeted to their specific needs. A priority is to provide supports and services to families at the earliest possible time to prevent repeat reporting to child protection systems and improve outcomes for children. This rapid analysis identified that the extent and characteristics of children and families reported to child protection has implications for service design and delivery. It has identified higher rates of reporting to child protection than current performance data, and this is likely associated with poor outcomes for children. Further research is required to support system reform and responses for children and families reported to child protection. As this study relates to a metropolitan region, this limits the generalisability of this research, therefore further research in other regions is required, including child protection reporting in other jurisdictions at scale. This study was approved by the Aboriginal Health Council of South Australia, Aboriginal Health Research Ethics Committee (Protocol #04–17-718) and the University of South Australia Human Research Ethics Committee (ID: 0000036590). The authors would like to acknowledge Christine Gibson, Dr Kerry Lewig and Dr Mary Salveron for their contributions to this research including data extraction. The authors would also like to acknowledge Dr Samantha Finan for their contribution to the research including coding for the case file review. We have no conflicts of interests to declare. This research was funded by the state government in the jurisdiction in which the research was undertaken.