The Children

Abstract

The children’s early experiences provide a context for assessing subsequent outcomes. Data collected from case files and records presented to the courts show that before separation from birth parents, almost all 210 adoptees had experienced serious and often multiple forms of maltreatment; this was the primary reason for removal. Before entering their adoptive homes, 69% of the adoptees had had four or more adverse childhood experiences (ACEs), 32% had experienced failed reunifications and 48% had had three or more foster placements. Adverse childhood experiences before entry to care, harmful experiences in care and repeated exposure to grief and loss are likely to have contributed to the high prevalence of emotional and behavioural difficulties, displayed by 49% of the adoptees. According to our classification, 57% were at high risk of experiencing adverse outcomes in adulthood.

Much of this chapter was published in Children and Youth Services Review 96, January 2019, Tregeagle, S., Moggach, L., Trivedi, H. & Ward, H. ‘Previous life experiences and the vulnerability of children adopted from out-of-home care: the impact of adverse childhood experiences and child welfare decision-making’ pp. 55–63 Copyright Elsevier Ltd 2018. We are grateful to Elsevier for permission to reproduce text and tables here.

Introduction

Adoption from care is the final step in a complex decision-making process. The previous chapter has considered the intricate web of problems with which the birth parents struggled. In many families where parents are faced with similar problems, there are sufficient protective factors, such as supportive grandparents, a partner who does not misuse drugs or alcohol and/or skilled professional support, to ensure that children are not exposed to experiences that cause them significant harm, and they can safely remain at home (Hindley et al., 2006; White et al., 2015). The extent to which children are protected from the impact of parents’ problems are therefore also factors that influence the decision-making process (Ward et al., 2012). Before the study began, professionals had decided that the Barnardos adoptees could not be sufficiently protected from harm if they lived with their birth parents or other family members and that it was in their best interests to seek permanence through adoption.

This chapter focuses on the children’s characteristics and experiences, first while living at home with their birth parents, and then in the sometimes lengthy period they spent in out-of-home care between the initial separation from their parents and the final placement in their adoptive homes. The outcomes of open adoption from care cannot be adequately understood unless they are set within the context of these experiences, which will have shaped the children’s developmental trajectories. The chapter specifically seeks to identify whether there were factors within their personal characteristics and experiences prior to the adoption placement that were likely to increase the likelihood of maladaptive development and impact on their long-term wellbeing.

The chapter draws on data held on the Barnardos files and electronic case management system, and on the documents presented to the courts at the time of the adoption order concerning all 210 adoptees. Some data from the Pathways of Care Longitudinal Study (POCLS) (Australian Institute of Family Studies et al., 2015; Hopkins et al., 2019; Smart, 2015) are also presented in order to explore whether, and in what ways, the Barnardos cohort differed in characteristics and experience from other groups of children and young people in out-of-home care in Australia.

Children’s Characteristics

The cohort of 210 children placed for adoption included 108boys and 102 girls. Of these, 89 (42%) entered the programme as single children (although an unknown number had siblings from whom they had previously been separated) and 121 (58%) were admitted as sibling groups of two (40), three (12) and five (1). As we shall see (Chap. 4), although one of the objectives of the programme was to place siblings together, this was not always possible.

Ethnicity

Table 3.1 shows theethnicity of the adoptees. The majority (89%) were of Australian, New Zealand or European heritage, and there were five children whose parents were Asian or Australian/Asian. Adoption is not considered culturally appropriate for Aboriginal and Torres Strait Islander children and it was Barnardos’ policy to refer these children to indigenous-run programmes. One child was found to be Aboriginal after referral but before placement; at the request of the New South Wales State Government, Barnardos worked alongside the Government Department’s Aboriginal team to recruit an appropriate Aboriginal placement. No other children were known to be Aboriginal at the time of placement but four were later discovered to have Aboriginal fathers. The cohort also included 14 children who had at least one parent who was a Maori or Pacific Islander.

Table 3.1 Adoptees: ethnicity (N = 210)

Health Conditions and Disabilities Prior to Adoption

Although the Barnardos programme did not specialise in finding adoptive homes for children with health conditions or disabilities (see Chap. 1), such conditions were prevalent amongst the adopted children in the study. According to the case papers, at the time the application for an adoption order was made, 121 (58%) of the adoptees had at least one diagnosed long-term health condition or a developmental delay; 62 (30%) had one condition; 30 (14%) had two; 25 (12%) had three or four and 4 (2%) children had a combination of five health conditions or disabilities.

Table 3.2 (p. 68) shows the prevalence and nature of the long-term health conditions and developmental delays affecting the Barnardos adoptees. The most common conditions affecting children’s physical health were diseases of the respiratory system such as asthma and bronchitis, affecting 14 (6.7%) of the sample. There was also a high prevalence of congenital malformations, affecting 11 (5.2%) children, and sight (11: 5.2%) and hearing (8: 3.8%) problems.

Table 3.2 Health conditions and developmental delays (Barnardos sample N = 210; POCLS sample N = 1285)

Many of the physical health conditions may be related to children’s previous experience of abuse and neglect. Some of this had occurred during the pregnancy: 13 children (6.2%) showed signs of neonatal abstinence syndrome or foetal alcohol disorder; 2 were infected with hepatitis C and another child was recorded as having a ‘fragile skull, following a physical assault on her mother’. Other children had severe health conditions related to physical abuse they had experienced as infants, as Lisa’s case study shows. Some, but not all, of the problems with teeth or oral hygiene, affecting three (1.4%) children, may also have been related to neglect and/or physical abuse.

Lisa

Before Entering Her Adoptive Home

Lisa was admitted to hospital at the age of eight weeks with convulsions and head injuries. She was intubated and put on a ventilator, with medical examination showing a complex fracture of the skull with bruising and swelling to both sides of the brain. She also had a retinal haemorrhage, old and recent fractures to many ribs, fractures to her legs and coccyx, bruising to her face and chest and was anaemic and malnourished.

Lisa’s parents gave an explanation for the injuries; however, the medical team felt that this was inconsistent with Lisa’s presentation. Lisa was discharged after three weeks into foster care.

Just the way that [birth mother] didn’t look after me very well, all that stuff. But, she did mean well, but when she got sick, she didn’t do very much. She’s okay now, but back in the old days, she trusted on [birth father] to look after me and that was one of the worst things that ever happened.

Impact of Injuries

When Lisa was seven months old, she was placed with her prospective adoptive parents, who had been advised by medical specialists that she might never see, talk or walk. She made considerable progress after placement and became a happy, contented, alert and responsive child with an engaging smile. However, she continued to have significant medical problems:

• Eyesight: Lisa required her right eye to be patched and had ongoing assessment for a squint. She required glasses and continued to wear these in adulthood.

• Therapy: Lisa had significant gross motor delay and did not walk until she was three years old; she also had a significant delay in expressive language abilities. Lisa had daily exercises at home and saw a speech therapist, occupational therapist and physiotherapist on a fortnightly basis for much of her pre-school years.

…a long time ago when I used to be brain damaged, but I can’t remember that far, but when I used to be thrown into a brick wall and things like that, and all that kind of stuff. It was one of the worst lives that I ever had before, before I started being adopted.

A high proportion of the children also displayed emotional and behavioural problems, as has been found in other studies of children and young people in out-of-home care (see Meltzer et al., 2003; Tarren-Sweeney & Hazell, 2006). Again, many of these problems were likely to be related to the children’s previous experiences. Six children were showing evidence of affective disorders or anxiety-related problems, including one child who was described as ‘detached, emotionally deprived and suffering from chronic emotional distress’, and another who was thought to suffer from ‘attachment issues, self-loathing, nightmares, suicidal tendencies, self-blaming and anger issues’.

We assumed that because we took a child on that was not a baby, she had her experience, bad experience, the baggage she came with…. There was a deathin her birth family and she knows it all, she knew about it, she was heavily affected… she was crying almost every night for a long time. The difficulty was that she was very unsettled… she had this feeling of guilt because she survived a violent family life when her little brother didn’t. (Adoptive parent of young woman, aged 3 when permanently placed)

When she came to me, lived in a pillowcase, literally had a pillowcase over her head, and she used to walk round in a pillowcase. And when I asked her why, she said she was being an egg. And the psychologist said – used to think it’s because if she can’t see anyone, they can’t see her. (Adoptive parent of young woman, aged 4 when permanently placed)

However, the most common mental health problems (as classified according to the ICD10) were developmental disorders and/or delays. Forty-five children (21% of the sample) were showing evidence of cognitive or language delay; 16 (8%) had delayed motor skills or physical development and 30 (14%) were displaying behavioural disorders and/or emotional, social or developmental delays, including 21 (10%) children who had been diagnosed as suffering from ADHD or ADD. Altogether one in three (77: 37%) of the adoptees were showing delayed development, often across several dimensions.

When I was found, I had the mental and physical abilities of a two-month-old when I was almost two years old, and that’s something that would have continued down a very dark trajectory. (Young man, aged 1 when permanently placed, aged 25 when interviewed)

While these data reflect formal health assessments, the case files also recorded less formal evidence, indicating that 49% of the Barnardos children were displaying behavioural problems by the time they entered their adoptive homes. At least 86 (41%) of them had accessed mental health services before the adoption order was made.

Thirteen children aged between 3 and 13 had also been recorded as having engaged in criminal activity before entering their adoptive homes—we do not know in what ways the younger children had been involved. There was also evidence of educational disadvantage: 26 (30%) of those children who were aged five or more at entry to their adoptive homes were recorded as having reading difficulties. Chapter 7 explores these issues in some depth for the adoptees who took part in the follow-up, using information they and their adoptive parents provided through the survey and interviews.

Additional Support Needs

The extent to which the children in the full cohort were likely to have required additional support due to their behaviour, health and disability needs is evidenced in the levels of funding they received from the New South Wales Government before the adoption order was made. While the systems and classifications varied during the study period, in general, there were four different levels: children who were supported at the basic care level presented with very few challenges beyond what is considered ‘normal’ for their relevant age group; Care+1 indicates that a child required considered and regular supervision; Care+2 that they required comprehensive and constant support; and Care+2+ that they presented with extraordinarily difficult behaviour and required intensive wrap-around support.

Table 3.3 shows the care levels at which the children had been assessed at the time of adoption. Of the 195 adoptees for whom these data were available, nearly two-thirds (121: 62%) were supported at the normal care level; however, more than one in four (50: 26%) required constant supervision and one in seven (24: 13%) were supported at either the second highest (11: 6%) or the highest possible care level (13: 7%) when they entered their adoptive homes.

Table 3.3 Care level presented at the time of adoption (N = 210)

Comparisons with Other Populations

The data on health conditions and disabilities in the Barnardos adoptees can be compared with information on a representative sample of all Australian children collected between 2014 and 2015 through the National Health Survey (Australian Bureau of Statistics, 2015) and also with similar data collected on a current population of children in out-of-home care in New South Wales (POCLS, First Wave) (Australian Institute of Family Studies et al., 2015).¹ While these comparisons are intended to set the Barnardos’ data within a context, it should be noted that they are not exact, as the data were collected by different means: Barnardos, by extracting data from case files using a pre-determined checklist; the National Health survey through personal interviews with adults in selected homes; and the POCLS study through questionnaires completed by foster carers. Moreover, the categorisation of health conditions, though based on the ICD10, differs slightly between each study.

Variations in health conditions between the samples may simply reflect differences in the children’s ages at the time of assessment, as some conditions do not develop or become apparent in early childhood. The POCLS children were aged between 9 months and 18 years at data collection, as were 99% of the Barnardos sample; however, the National Health Survey reports on children aged 0–14. Moreover, as Table 3.4 shows, there were considerable differences in the age structure of the two samples of children in out-of-home care.

Table 3.4 Age at assessment: Barnardos (N = 210) and POCLS children (Wave One interview) (N = 1285)

The children in the POCLS sample were significantly² younger than those in the Barnardos cohort, with the largest age group (44%) being under three years old at the time of the assessment; in contrast, only 7% of the Barnardos children were under three when assessed, the largest number (96: 47%) were 6–11 years old. The Barnardos sample also had a higher proportion of teenagers (27% vs 10%).

Children in the POCLS sample were found to be reasonably similar to the rest of the Australian child population in terms of types and prevalence of the most common long-term conditions; however, rates of psycho-social disabilities were considerably higher in the POCLS sample than in the general population (Australian Institute of Family Studies et al., 2015; Australian Bureau of Statistics, 2015).

Table 3.2 (p. 68) displays the data on health conditions and disabilities from the Barnardos sample alongside similar data from the POCLS study. Data on some physical conditions that may be related to abuse or neglect in utero, such as neonatal abstinence syndrome and congenital malformations, are not available for the POCLS sample. However, where comparable data are available, they show considerable differences. The most common physical long-term health conditions for both samples (and indeed for the general population) were respiratory problems, and particularly asthma; however, the prevalence appears to have been lower in the Barnardos sample (6.7% vs 15.9% (POCLS) and 20.9% (National Survey)). This could be due to under-reporting.

Both out-of-home care samples had higher rates of mental health and behavioural problems than the general population, where the prevalence was 8.9%. The Barnardos adoptees showed a slightly (but not significantly) higher prevalence of cognitive developmental delay than the POCLS children (21.4% vs 11.8%). This difference may reflect the higher proportion of children aged under three in the POCLS group, many of whom may have been too young for incipient developmental problems to become evident.

The POCLS group had a slightly higher rate of affective disorders and anxiety-related problems than the Barnardos children; however, the Barnardos adoptees showed a slightly (but not significantly) higher prevalence of behavioural and emotional disorders (14.3% vs 9.7%). Smart (2015) found that the POCLS group were beginning to show evidence of emotional and behavioural disorders from the age of three (see also Ward et al., 2012): in the Barnardos sample, 18% of 3–5-year-olds, 19% of the 6–11-year-olds and 14% of the 12–17-year-olds had been diagnosed as having an emotional or behavioural disorder by the time they were adopted; in the two younger age groups, the prevalence was higher than in the POCLS sample, where the comparable percentages were 12% (3–5-year-olds) and 15% (6–11-year-olds). There was a slightly higher rate of emotional and behavioural disorders amongst teenagers in the POCLS group (17% vs 14%); this could reflect late entry to care and therefore longer exposure to abuse for this age group. However the difference may also reflect variations in the methodology—the Barnardos data are based on a formal diagnosis, whereas the POCLS data come from caregiver reports. The Barnardos case files indicated that nearly half of the adoptees (49%) had a behavioural problem as reported by caregivers and others, including at least 80% of those aged six and over. If the informal case file data are closer to the POCLS assessments, then the Barnardos children had far higher levels of emotional and behavioural disturbance.

The Barnardos children were assessed at the time the application for an adoption order was made, often many months after they had been separated from their birth families. They did not have longer exposure to abuse than the POCLS children—in fact, more of them were separated before they were three (68% vs 55%). However early separation may have reflected high levels of abuse, as some of the case studies indicate. The higher prevalence of emotional and behavioural disorders amongst younger adoptees may reflect these experiences, which may also have been compounded by adverse experiences in out-of-home care before achieving permanence (see below).

Further evidence that the Barnardos children were more vulnerable in terms of physical and mental health than other children in out-of-home care is shown in Fig. 3.1, which displays the number of health conditions found in the two samples. A higher proportion of the Barnardos children had at least one health condition or disability (57.6% compared with 45.5% of the POCLS sample), and a higher proportion of them had two or more conditions (28% vs 20%). Such findings reflect the extreme vulnerability of children who are placed for adoption—these are, after all, the group of children in out-of-home care who are assessed as requiring permanent protection. It should also be remembered that, for many years, the Barnardos children were selected on the explicit criteria of being ‘hard to place’, often because of emotional or behavioural problems (see Chap. 1).

Fig. 3.1

Number of health conditions and developmental delays: Barnardos (N = 210) and POCLS (N = 1285) children compared

Children’s Experiences Before Separation from Birth Families

Reasons for Separation

The primary reason recorded for the decision to remove over 90% of the children from their birth families was abuse or neglect. There were 20 children (10% of the cohort) for whom no evidence of maltreatment had been recorded. Over half (11/20: 55%) of these 20 children had been removed because an older child in the family had been abused and they were considered to be at high risk of harm. The other nine children for whom there was no evidence of maltreatment had been placed in out-of-home care because of concerns about their mother’s mental health (three children) or because their parents had voluntarily relinquished them for adoption (four children), and two children were orphans, with no relatives to care for them.

Table 3.5 shows the children’s experience of maltreatment before separation from their birth parents. Thirty-six (17%) had experienced one form of abuse, with the most frequently reported being neglect (23), followed by emotional abuse/psychological harm (11) and physical abuse (2). All other children had experienced composite patterns of abuse (polyvictimisation): 87 (42%) had experienced two forms of abuse, 44 (21%) children had experienced three forms and 23 (11%) had experienced all four. The combined data show that neglect and emotional abuse were the most common forms of maltreatment, experienced respectively by 164 (78%) and 151 (72%) of the sample; 72 (34%) children had experienced physical abuse and 47 (22%) had been sexually abused. This latter group may have been particularly vulnerable to poor outcomes, for children who have experienced sexual abuse are at particular risk of experiencing multiple placements in care, poor attachments to parent figures and disrupted adoptions (Smith & Howard, 1991; Nalavany et al., 2008).

Table 3.5 Experiences of abuse before separation (N = 210)

In Chap. 1 we noted that the public discourse on adoption sometimes overlooks the issue of child maltreatment. It is therefore important to be aware of the extensive abuse experienced by the Barnardos adoptees before removal from home. The high proportion of adoptees who had experienced polyvictimisation indicates that these children were at very high risk of significant harm. We do not have detailed information concerning the exact nature of abuse experienced by all the Barnardos children, but we do know that the cohort included immobile infants who had received multiple fractures, toddlers whose physical and emotional needs were grossly neglected because of parental drug addiction, and children who had been loaned out to paedophiles in return for cash. The following reports from adoptive parents are not atypical:

His original reason for going into care was a choice of his mother’s to let him go with two men – be taken away when he was a tiny, little two-year-old, and that he was very, very seriously abused and was hospitalised when he came back – they brought him back. They had him for 10 or 12 days or something, and he was hospitalised…

…So he’s got a history of lots of physical abuse, emotional abuse and probably sexual abuse as well. (Adoptive parent of a young man, aged 6 when permanently placed)

As I understand it, by the time she was six weeks, she was already in hospital with multiple fractures. And then it was a very scattered sort of situation between various members of her birth family: her uncles and her grandfather and her mother and father… And there were about 12, 11, 13 moves before she went to [foster mother] when she was three and a half. (Adoptive parent of young woman, aged 4 when permanently placed)

Although there are no directly comparable data, it is likely that the experiences of the Barnardos adoptees were more severe than those of the POCLS cohort because the latter includes the whole range of children placed in out-of-home care. The vast majority of the Barnardos children were adopted because they could not safely return to birth parents; within five years of placement away from home, 26% of the POCLS children who had received final care and protection orders, and 81% of those who had not received a final order had been reunited, although almost one in four (22%) of the latter group had returned to out-of-home care (Hopkins et al., 2019).

Adverse Childhood Experiences

Chapter 1 has discussed some of the research on adverse childhood experiences (ACEs) that can have a long-term negative impact throughout the life trajectory. These include all forms of abuse and neglect, as well as growing up in a dysfunctional household; the two are interrelated (Felitti et al., 1998). Factors that are indicative of household dysfunction include parental substance misuse, mental health problems, domestic abuse and/or criminal behaviour patterns. As Chap. 2 has shown, these factors were prevalent in many of the birth families: at least 112 (53%) children had lived with a mother or stepmother who was the victim of domestic abuse; 135 (64%) had lived in households in which there was substance misuse; 85 (41%) with a household member who had mental health problems and 50 (24%) had experienced the imprisonment of a birth parent. These are all known risk factors for the recurrent maltreatment of children (Hindley et al., 2006; Ward et al., 2014; White et al., 2015) and are likely to have contributed to the initial decision to place the child away from home.

The records of the Barnardos cohort include data on nine out of the ten ACEs identified as being related to adverse adult outcomes.³ Table 3.6 shows the cohort compared with a normative Australian sample (7485 Australian adults interviewed at the outset of a longitudinal community study of psychological health in the Canberra region (Rosenman & Rodgers, 2004)).

Table 3.6 Prevalence of adverse childhood experiences: Barnardos adoptees (N = 210) and normative Australian sample (N = 7485)

First, the data provide strong evidence of the vulnerability of the Barnardos adoptees: before entering the Find-a-Family programme almost all (208: 99%) of them had had one or more adverse childhood experiences; just over two-thirds (145: 69%) had had four or more and 27 children (13% of the sample) had had seven or more. Moreover, these data are likely to be an underestimate, as information collected from case files depends on the quality of reporting and a lack of evidence concerning adverse factors does not necessarily mean they were not present.

Second, Felitti et al. (1998) found that exposure to four or more adverse childhood experiences significantly increased adult propensity to health risk factors and premature death (see Chap. 1 for further details). The data indicate that the Barnardos adoptees were markedly more likely to be subject to these increased risks than a normative population. Nearly 60% of the Australian population studied by Rosenman and Rodgers (2004) had not encountered any of the key adverse childhood experiences and less than one in five (17%) had encountered four or more. More than four times (69%) as many of the Barnardos adoptees had reached this critical threshold. These data, which demonstrate the extreme vulnerability of a relatively high proportion of the adoptees prior to placement with their adoptive parents, need to be taken into account when assessing the adoption outcomes.

Age at Notification

For many of the children, maltreatment was identified and notified at a very early age. As Table 3.7 shows, half of the cohort had been identified before they were six months old (102: 50%), 60% (123) before their first birthday and nearly three-quarters (74%: 153) before they were two. In fact, 20 (10%) children had been identified before they were born. However, 19 (9%) children were not identified until after their fifth birthdays, including one child who was 11 and 3 children who were 12. Although some of these children may have lived in stable, nurturing homes before the abuse began, others may have been subject to chronic maltreatment for many years before action was taken (Brown et al., 2016).

Table 3.7 Age at notification of abuse (N = 210)

Months Between Notification and Separation

Rousseau and colleagues (2015) monitoredthe development of 129 children who were placed in out-of-home care in France for 20 years and found that the length of time between notification and first placement had a more significant impact on children’s long-term outcomes than the age at which they were separated. This finding is consonant with those of other studies which have found that the extensive mental health problems often displayed by older children in care are largely an artefact caused by late-placed children entering care with high levels of pre-existing disturbance (Tarren-Sweeney & Hazell, 2006; Ward & Holmes, 2008).

The French study found that children separated from birth parents within 10 months of notification of maltreatment had significantly better outcomes than those placed in out-of-home care more than 15 months after the authorities had been alerted. Table 3.8 shows the time in months between notification of abuse and first separation from birth parents of the 205 adoptees in the Barnardos cohort for whom data were available.

Table 3.8 Months between notification of abuse and separation from birth parents (N = 210)

The mean length of time between notification and separation was 11.5 months (sd = 18), and just over half the adoptees were removed within 4 months of notification. Within this group were 51 adoptees (25%) who experienced less than a month’s delay between the notification of abuse and separation from birth family, including 30 (15%) who were separated on the same day that the notification was received. More than two-thirds (141: 69%) of the children were separated within ten months of notification; the French study would indicate that these children were likely to have better outcomes than those for whom the gap between notification and separation was longer. Nearly one in four (50: 24%) of the children were separated more than 15 months after notification and were therefore likely to have less satisfactory outcomes. This group included 30 children, 15% of the cohort, who continued to remain with their birth parents for more than two years. One of the most difficult decisions a social worker has to make is whether to place a child in out-of-home care. With the benefit of hindsight, once a child has been placed for adoption, it may seem evident that the decision to separate took too long. However, the long time periods between notification and separation are likely to represent repeated attempts to support the parents in the hope that the situation will change and that the family can remain intact (Ward et al., 2012).

Age at Separation from Birth Families

Studies of children placed with adoptive parents following gross deprivation in Romanian orphanages have found that those who overcame the consequences of severe early deprivation were significantly more likely to do so if they were placed in a nurturing environment before they were six months old (Rutter et al., 2007; Sonuga-Barke et al., 2017). The extreme deprivation experienced by the Romanian adoptees was exceptional and care has to be taken in using these findings as comparators for other populations (Rutter et al., 2007). Nevertheless, there are several indications that the first 24 months of life are a sensitive period for childhood development, with the first 6 months perhaps being particularly crucial (Zeanah et al., 2011). Numerous studies have also found that children who are permanently separated from abusive families at an earlier age tend to have better outcomes than those who are separated later (see Rousseau et al., 2015; Tarren-Sweeney & Hazell, 2006; Ward et al., 2012).

Table 3.9 shows the ages at which the adoptees were first separated from their birth parents. A high proportion had indeed been separated within this sensitive period: 53 (25%) had been separated before they were six months old, including 13 (6%) who were separated within a month of their birth. More than one in three (77: 37%) had been separated before they were one, and over half (113: 54%) before they were 24 months old. On the other hand, more than one in ten of the children (22: 11%) were aged seven or older when removed from an abusive home.

Table 3.9 Age at first separation from birth parents (N = 210)

Months Between First Separation and Permanent Placement with Adoptive Parents

Selwyn, Wijedasa and Meakings (2014) found that children in England and Wales who spent more than 2 years in out-of-home care before entering their permanent placement with adoptive parents were significantly more likely to experience a disruption (i.e. leave their adoptive home before they were 18 years old). The mean length of time between first admission to care and admission to a permanent placement with an adoptive family was 27 months (sd = 28) for the Barnardos cohort. However, as Table 3.10 shows, there was a wide range of timeframes, with just over a third of the cohort reaching their permanent placement within a year (74: 35%), including nine children (4%) who were there within a month. Just over a third of the cohort (77: 36%) waited for more than two years. Twenty-eight children, 13% of the sample, waited for more than five years, including eight who did not reach their adoptive family until more than eight years after they had first been placed away from home. During this often extended period, as we shall see, there were a number of factors that may have increased the children’s vulnerability.

Table 3.10 Months between separation and permanence (N = 210)

Age at Reaching Final Placement

There is considerable evidence concerning relationships between children’s age at permanent placement and outcomes of adoption. Numerous studies have shown that the older children are at final placement, the greater the risk of adverse outcomes (Coakley & Berrick, 2008; Evan B. Donaldson Adoption Institute, 2004; Festinger, 2014). Van den Dries et al.’s (2009) meta-analysis of data from studies concerning attachment in adopted children found that those who were permanently placed with adoptive carers before their first birthdays were significantly more likely to form a secure attachment than those who were placed later. Zeanah et al.’s (2011) overview of evidence concerning sensitive periods of child development argued that:

current studies show age-at-adoption cutoffs (after which deficiencies are reported) to be 6, 12, or 18 months (parent-rated behavior problems, security of attachment), 15 months (expressive and receptive language), 18 months (parent-reported executive functioning), and 24 months (IQ, security of attachment; EEG coherence). (Zeanah et al., 2011, p. 11)

Further research has identified another cut-off point, indicating that adoptive placements made after the child is four years old are significantly more likely to disrupt than those made earlier⁴ (although the rate of disruption does not follow a linear progression) (Selwyn et al., 2014).

Table 3.11 shows the ages of the Barnardos adoptees when they entered their adoptive homes. Nine (4%) were permanently placed before they were six months old; 30 (14%) before their first birthdays; and 50 (24%) before they were two. However, just over three-quarters of this cohort (76%:160 children) were permanently placed outside the optimal timeframe for adoption placements, increasing the chances of less than satisfactory outcomes. Almost half the sample, 104 (49%) of the children, were aged four or older at the time they entered their adoptive homes and their placements would have been significantly more vulnerable to disruption than those placed earlier (Selwyn et al., 2014). This group includes 51 (24%) children who were between six and nine years old and 20 who were ten or older when permanently placed; 5 of them were teenagers.

Table 3.11 Age at entering adoptive home (N = 210)

After they had reached their adoptive home, some children waited lengthy periods for an adoption order to be made. Although for young children, moving into the adoptive home will be the significant moment at which permanence is experienced, older children may be unable to feel that they belong to an adoptive family until their legal status is secure.

He wanted to be adopted from the word go. “I want to be here, and I want to be here permanently”.… as soon as he had his adoption papers, he was fine. It gave him strength. (Adoptive parent of young man, aged 2 when permanently placed)

There is some evidence to suggest that adoptions are significantly more likely to disrupt if the gap between entering the placement and the adoption order being made is longer than 12 months. This is possibly because long delays between placement and the adoption order may be indicative of adoptive parents’ ambivalence or concerns about the placement (Selwyn et al., 2014). However, this evidence comes from a British study and it may not be so relevant in an Australian context, where there may be different reasons for delays. For instance, almost half of the Barnardos cohort (88: 42%) entered their adoptive home with a plan for long-term foster care, and it was sometimes years before adoption was considered as a long-term option. Moreover, some of the adult adoptees told the interviewer that they had asked for their adoption order to be delayed because they had not wanted to upset their birth parents.

And they had all the paperwork ready and at the last point I said, “No. I don’t want to go through this”. It disappointed my [adoptive] mum and dad very much…. The reason why it stopped was because there was a major concern on my biological side – my biological mother’s side. She disapproved the adoption process and that’s something that I didn’t want to hurt my mum about, so that’s why I cancelled the adoption process, not to hurt her. (Young man, aged 9 when permanently placed, aged 40 when interviewed)

One of the Barnardos adoptees received their adoption order more than 25 years after the placement began, but this was an exception. It took between 6 months and 13.5 years from the time of placement for adoption orders to be issued for the other 209 adoptees in the cohort. Ten (5%) of them had an adoption order within 12 months of being placed, 46 (22%) within 2 years and 144 (69%) within 5 years. The other 65 (31%) children waited between 5 and 13.5 years.

Children’s Experiences Between First Separation from Birth Parents and Permanent Placement with an Adoptive Family

Not all the Barnardos adoptees had remained continuously separated from their birth parents after the initial admission to out-of-home care. Attempts had been made to reunite just under a third (68: 32%) of them. Fifty-eight (28%) of the cohort had experienced one failed restoration; seven (3%) had experienced two, and three children had been rehabilitated with their birth families and then returned to care on three occasions before being placed in permanent care with a view to adoption.

Failed restorations are known to be detrimental to children’s wellbeing, particularly if they are repeated (Farmer et al., 2011). In addition, 23 (11%) children had experienced a failed kinship placement, including 9 who also had an unsuccessful attempt at reunification with birth parents. The majority of children whose kinship placements failed had been voluntarily relinquished to the statutory agency—only five were removed through court orders. Such experiences are likely to have enhanced the perceptions of rejection already held by these vulnerable children.

Most children were first placed in out-of-home care by the statutory agency and then referred to Barnardos. While they were looked after by the statutory agency, a substantial group of children had a relatively stable care experience. However, a number had experienced frequent changes ofplacements. As Table 3.12 shows, about half (106: 50%) the children had one placement prior to entering the Barnardos programme or moved there straight from their birth parents. However, just over a third (78: 37%) had between two and four previous placements and one in eight (26: 13%) had five or more, including ten children who had experienced at least ten previous placements. In this latter group was one child who had had 15, one who had had 17 and one who had had 22 previous placements in out-of-home care.

Table 3.12 Number of placements before entering the Barnardos programme (N = 210)

Not all children moved immediately to permanent Find-a-Family placements on entry to Barnardos. Altogether, 115 (55%) children experienced further moves within Barnardos before reaching their adoptive homes. Seventy-five of these children were placed in the Barnardos’ Temporary Family Care programme prior to receiving a long-term care and protection order and then moving to Find-a-Family. Other moves occurred within the Find-a-Family programme. While 64 (30%) children had one additional placement within Barnardos before achieving permanence, 18 (9%) had three or more, including four children who had five additional placements and one child who had six.

A wealth of evidence shows that frequent changes of placement are detrimental to children’s wellbeing, not least because of their adverse impact on children’s ability to form secure attachments (e.g. Osborn & Delfabbro, 2006) and negative impact on mental health (Rubin et al., 2007). There is also evidence that a history of placement changes increases the likelihood of further disruptions, including disrupted adoptions. Selwyn et al. (2014) found that children who had experienced three or more placements before being placed with adoptive parents were 13 times more likely to experience a disruption. Table 3.13 shows the total number of placements experienced by the Barnardos children before entering their adoptive home. While 7 children (3%) moved straight to their adoptive parents, 103 (49%) had one or two interim placements and 100 (48%) had three or more. Thirty children had more than five interim placements and ten had more than ten.

I kind of just remember little things like different houses. I just every now and again – when I was younger, I used to think – dream and stuff – like have memories of just all these different houses and like different – I couldn’t really picture faces, but kind of blurred out faces sort of thing. And that’s about it.

I was scared whenever people came to the door just because I’d been through so many placements. (Young man, aged 3 when permanently placed, aged 19 when interviewed)

So when he came to us, he had already been in nine different homes – some fostered, some homes for children – nine before he came to us.

Well, it played with my anxieties – sometimes I didn’t fit in, sometimes I would be going, “Okay, I can befriendswith you, or I can hang around, but I know that I’m not going to stay here for long”. So therefore, it’s that, “Well I’m not going to be here for long so I can do whatever I want”, whereas it played on emotions because you get the anxiety. You get the depression. You’re not sticking with one group. You’re just being shifted, so it’s all of that, and there’s no sense of belonging. (Adoptive parent and young man, aged 9 when permanently placed, aged 40 when interviewed)

Table 3.13 Total number of placements before entering adoptive home (N = 210)

Children who had experienced four or more placements before reaching their adoptive homes were significantly⁵ more likely to be rated as displaying emotional and behavioural problems than those who had moved less frequently.

Children’s Vulnerability at Entry to Adoptive Homes

So far we have seen that a high proportion of the adoptees had suffered extensive and lengthy exposure to adverse childhood experiences before separation; about half the cohort also had repeated experiences of failed restoration and/or frequent moves in the months between separation and placement within a permanent family. For many children, it seems clear that pre-existing problems would have been exacerbated by harmful experiences within the care system. In addition to the data on placement instability, there is also evidence that 16 adoptees had made formal allegations of abuse while in care, though we do not know which placements were referred to in the complaints or how serious these allegations were. These harmful experiences, both before entering out-of-home care and in the period between separation from birth parents and permanent placement with adoptive carers, will have meant that many of the children were extremely vulnerable when they entered their adoptive homes.

Experiences of Loss

Furthermore, at this point many of the children would have been dealing with feelings of loss and confusion about their birth families. The vast majority (186: 89%) had already experienced their birth parents’ separation and/or divorce. At least 21(10%) had already experienced the death of a birth mother and at least 16 (8%) the death of a birth father; 3 children had experienced the deaths of both parents. Only six of those children whose birth mothers had died had contact with their birthfather. Forty-two (22%) of those children whose birth parents were alive had no contact with their birth mothers and 109 (56%) had no contact with their birth fathers by the time they were adopted; 52 (25%) children had no contact with either parent. Sixteen of these children had contact with a grandparent, 12 with other family members and 22 with a sibling. However, 13 children had no contact with any member of their birth family at the time the adoption order was made. This extensive experience of loss is likely to have impacted on the children’s sense of identity as well as theiremotional and behavioural development.

Changes over Time

To some extent the prevalence of vulnerability factors, such as those discussed above, changed over time and, specifically, as new legislation concerning adoption and permanence was implemented. In Chap. 1 we noted that policy and practice changed over the four time periods for this study: 1 January 1979–31 August 1985 (focus on young children in long-term foster care); 1 September 1985–30 June 1991 (older primary-school-aged children with behavioural/emotional challenges); 1 July 1991–30 June 2007 (wider age range) and 1 July 2007–30 June 2012 (inclusion of infants); and these are to some extent reflected in changes in the prevalence of some of the vulnerability factors. There were significant differences in the age at which children entered their adoptive homes.⁶ Those who entered their adoption placements between September 1985 and June 1991 were older than those who entered within the other three time periods, reflecting the policy over this period to offer support to older primary-school-aged children with emotional or behavioural challenges. Those permanently placed in this time period tended to have had more previous placements (mean = 4.8) and be older at separation from their birth family (mean = 54 months) than children placed at other times. After being first placed in out-of-home care, they had also waited significantly longer before moving to a permanent placement than those in the two groups that focused on younger children.⁷ This may reflect the vicious circle by which children’s behavioural problems are exacerbated by a sense of insecurity and instability as they become older and harder to place.

Children who entered their adoptive homes between July 1991 and June 2007, when the policy changed to widen the age range to include younger children, were significantly more likely to have a behavioural problem than those permanently placed at other times.⁸ These children were also more likely to have experienced more ACEs (mean = 4.7) than children who entered at other times. Children placed in this timeframe and in the previous one, when behavioural problems were most prevalent, were also significantly more likely to have been sexually abused while living with their birth parents than those who were permanently placed before September 1985 or after June 2007.⁹

Finally, those children who entered the Barnardos programme in the first phase of its adoption work waited on average longer for their adoption order than those who entered later (mean = 99.1 months). Almost all of them waited at least five years before the order was made. This reflects the specific nature of their placements, which had begun as long-term foster care and only later moved towards adoption. However, there was also considerable variance in the timeframes between adoptive placement and Supreme Court order within the other groups.¹⁰

Identifying Children with Different Levels of Vulnerability

The outcomes of the Barnardos Find-a-Family programme need to be understood within the context of the extensive evidence of adversity experienced by a high proportion of the cohort before they entered their adoptive homes. Before they were separated from their birth parents, 145 (69%) children had had four or more adverse childhood experiences, including 47 (22%) who had been sexually abused and 87 (41%) who had experienced polyvictimisation. Fifty (24%) children had remained for more than 15 months after notification of abuse with birth parents who could not meet their needs; 97 (46%) were more than two years old when first separated. Seventy-seven (36%) children waited for two or more years between separation and permanence, and during that period 10 (5%) children experienced two or more failed reunifications and 100 (48%) had three or more placements. By the time they were permanently placed, 102 (49%) children were reported to have behavioural problems and 24 (13%) of them had been assessed as requiring support at the two highest care levels. One hundred and four (49%) children had had their fourth birthdays before they entered their adoptive home, and 200 (96%) then waited more than 12 months before an adoption order was made. These 11 factors are all known to be significantly associated with poor outcomes in adulthood and/or disrupted adoption placements (Farmer et al., 2011; Felitti et al., 1998; Finkelhor et al., 2011; Nalavany et al., 2008; Osborn & Delfabbro, 2006; Rousseau et al., 2015; Selwyn et al., 2014; White, 2016); there is considerable overlap between them and many children experienced constellations of multiple risk factors.

Variations in the prevalence of these risk factors make it possible to distinguish between those children who were extremely vulnerable to adverse life trajectories and those whose life chances had been less severely compromised before they entered their adoptive homes; later chapters in this book will explore the relationship between children’s vulnerability at the time of the adoption and their subsequent, adult outcomes. In order to facilitate comparisons, the children’s experiences were categorised on each of the 11 risk factors identified above: low risk indicated that the child’s experience in this area had not reached a level that other studies had shown to be significantly related to adverse outcomes (for instance, they had had less than four ACEs or they were less than two years old when first removed from their birth parents’ care); high risk indicated that the child’s experience had reached or surpassed the level which other studies have found to be significant; extreme risk indicated that the child’s experience had reached at least twice this level (for further details, see Appendix 3). All the adoptees had encountered at least one experience which met the high-risk level and all but 17 of them (193: 91.9%) met the extreme risk level on at least one of these factors. Figure 3.2 shows the percentage of Low, High and Extreme occurrences amongst the sample, for the 11 risk factors.

Fig. 3.2

Percentage* of Low, High and Extreme counts for each of the 11 significant risk factors that occur in the Barnardos sample (N = 210). *percentages have been rounded

Using these data, the adoptees were divided into two groups: the ‘medium vulnerabilitygroup’ included 90 (43%) children who had been categorised as low on six or more of the 11 vulnerability factors (but high or very high on others) and the ‘high vulnerability group’ included 120 (57%) children who had been categorised as high or very high on six or more of these factors. Within the medium vulnerability group is a sub-group of nine (4%) children who were categorised as low on nine or more of the variables and had no extreme scores: the ‘low vulnerability sub-group’. Within the high vulnerability group is a contrasting sub-group of 17 children who were categorised as at extreme risk of poor outcomes on six or more of the relevant factors: the ‘extreme vulnerability sub-group’ (see Fig. 3.3).

Fig. 3.3

Number of Barnardos adoptees in medium and high vulnerability groups, including extreme sub-groups (N = 210)

Conclusion

Identifying these risk factors and grouping the children according to levels of vulnerability gives us some indication of the likely risk of adverse outcome at the point when permanence was achieved. It is clear that, at the time they entered their adoptive homes, many of the children were at high risk of following negative life trajectories. They are likely to have been at greater risk than a normative care population, which includes children with a wider variation of experience and need. However, there is evidence from the research on Romanian orphans that supportive and sensitive substitute parenting can act as a strong protective factor and help many children who faced extensive early adversity move towards a more positive developmental trajectory (Rutter et al., 2007, 2009; Sonuga-Barke et al., 2017). Moreover:

resilience does not come from rare and special qualities, but from the everyday magic of ordinary, normative human resources in the minds, brains and bodies of children, in their families and relationships and in their communities. (Masten, 2001, p. 235)

Subsequent chapters explore how far children’s outcomes related to their previous experience; how far their adoptive parents were prepared for the challenges they presented; and whether the experience of adoption was part of the ‘everyday magic’ that could promote and strengthen their resilience.

Key Points

• There were 108 boys and 102 girls in the cohort, and the majority (89%) were of Australian, New Zealand or European heritage. Eighty-nine (42%) adoptees entered the programme as single children and 121 (58%) were admitted as sibling groups of two (40), three (12) and five (1).

• The Barnardos children showed a greater prevalence of health conditions and disabilities than other children in out-of-home care in New South Wales. At the time the application for an adoption order was made, 121 (58%) of the adoptees had at least one diagnosed long-term health condition or developmental delay; 62 (30%) had one condition; 30 (14%) had two; 29 (14%) had three or more. Some of these were related to the children’s previous experiences of abuse and neglect.

• Developmental disorders and/or delays affected 77 (37%) adoptees. Forty-five children (21% of the cohort) showed evidence of cognitive or language delay and 30 (14%) had a diagnosed emotional or behavioural disorder.

• Over 90% of the adoptees had been removed from their parents’ care because of serious, and often multiple, forms of abuse and neglect; there were maltreatment concerns for all but nine children.

• Most children had experienced composite patterns of maltreatment: 87 (42%) had experienced two forms of abuse; 44 (21%) had experienced three forms and 23 (11%) had experienced all four forms of abuse (neglect, emotional abuse, physical abuse and sexual abuse).

• Before entering the Barnardos programme, just over two-thirds (145: 69%) of the adoptees had had four or more adverse childhood experiences. The comparative figure for the general population in Australia is 17%.

• Children’s experiences after notification of abuse may have increased their vulnerability: 50 (24%) remained for more than 15 months with birth parents who could not meet their needs; 96 (46%) were aged two or more when first separated; 77 (36%) waited for 2 or more years between separation and permanence;104 (49%) had had their fourth birthdays before they entered their adoptive home and 200 (95%) waited more than 12 months after placement before an adoption order was made. All these time points have been identified by other research as cut-off points beyond which the likelihood of adverse outcomes is increased.

• Instability in out-of-home care is also likely to have exacerbated children’s vulnerability: before they entered their adoptive homes, 68 (32%) children had experienced failed reunifications and 100 (48%) had had three or more placements.

• Adverse childhood experiences before entry to care, compounded by harmful experiences in out-of-home care, as well as repeated exposure to grief and loss, are likely to have been factors underlying the high prevalence of emotional and behavioural difficulties displayed by the children. Case file reports indicated that nearly half (102: 49%) the adoptees had behavioural problems and 24 (13%) required support at the two highest care levels; 86 (41%) had accessed mental health services before the adoption order was made.

• The research team categorised the children according to the presence of 11 factors identified by other robust research studies as increasing the risk of adverse outcomes in adulthood. All the children in the sample had encountered at least one of these factors: 90 (43%) were categorised as at medium risk and 120 (57%) as at high risk of adverse outcomes. Combinations of risk factors were expected to correlate with adult outcomes.