NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Mental Health (UK). Children's Attachment: Attachment in Children and Young People Who Are Adopted from Care, in Care or at High Risk of Going into Care. London: National Institute for Health and Care Excellence (UK); 2015 Nov. (NICE Guideline, No. 26.)

Cover of Children's Attachment

Children's Attachment: Attachment in Children and Young People Who Are Adopted from Care, in Care or at High Risk of Going into Care.

Show details

2Introduction to children's attachment

2.1. What is attachment?

Infants are born equipped with a range of innate behaviours to maximise their survival. Attachment behaviour allows the infant to draw others towards them at moments of need or distress (Fonagy et al., 1995). Infants who experience a secure attachment relationship develop a reasonably firm expectation of feeling protected and safe, which in turn allows them to explore their world more confidently.

Our instinct for attachment, which is shared by most mammals, is a basic adaptation for survival in infancy. When infants (or indeed adults) are frightened, stressed, feel unwell or are under threat, their attachment system is alerted. Infants in this state will initiate proximity-seeking behaviours (such as crying, clinging, or following with their gaze in babies; more verbal or sophisticated behaviours in older children) towards their primary attachment figure (normally a parent or the main caregiver). Once proximity and reassurance have been achieved, the attachment system can be deactivated. Seeking help and the different approaches an individual uses to obtain help constitute the building blocks of the attachment process.

More broadly, attachment theory also describes the ways in which individuals handle their most intimate relationships with their attachment figures (their parents, children and life partners). But as we have developed an increasingly sophisticated understanding of the relationship between early brain development, early psychosocial experiences and developmental psychopathology, it has also become clear that the role of attachment in humans goes significantly beyond its primary evolutionary purpose, the immediate survival of an infant (Crittenden, 1999; Perry, 2009; Siegel, 2001; Van der Kolk et al., 1991). Although some researchers express scepticism about whether attachment is an innate mechanism, the majority of the field (considered broadly) accept that children have a basic, biologically rooted, need to form a lasting bond with their carers. Even if this relationship is strained for reasons such as poverty or domestic abuse, the child can form ‘attachment-like’ relationships with other adults, for example their teachers (Bergin & Bergin, 2009).

The attachment strategies that a child develops are shaped by their environment, and this has major implications for the ways in which children learn to behave in close interpersonal relationships. From birth, the interactions of an infant with their primary carers will establish a base for personality development and will mould subsequent close relationships, expectations of social acceptance, and attitudes to rejection. Through interacting with others, infants learn about their role within the relationship and in time they begin to make sense of their own psychological states and those of others (Fonagy et al., 2002).

A secure base is formed when the attachment figure provides stability and safety in moments of stress, which allows the infant to explore their surroundings. Ainsworth and others also highlight the importance of parental sensitivity for a child to form a secure base (Ainsworth, 1993). Sensitivity is measured as the parent's ability to respond to the particular needs and cues of an individual child. The parent's capacity to do this takes place, or is influenced by, the systemic context (that is, contextual stressors, personal history, couple relationship and so on).

In response to parenting behaviour, the child creates a set of mental models of itself and of others in social interactions (‘internal working models’), based on repeated interactions with significant others (Bowlby, 1973). These early attachment relations are thought to be crucial for later social relationships, the acquisition of capacities for emotional and stress regulation, self-control, mentalisation and emotional maturity. Therefore, a child who develops insecure or disorganised attachments, possibly due to neglect or being placed in numerous foster care homes, is more likely to struggle in these areas and to experience emotional and behavioural difficulties.

It is worth noting that attachment may not be responsible for all interpersonal interactions with primary caregivers. For instance, Trevarthen and colleagues have demonstrated the importance of the intersubjective relationship experience between the infant and their carer, and that this complements, but is different from, their attachment relationship experience (Trevarthen & Aitken, 2001).

Attachment is a developmental process, for which behavioural and affectional aspects have their counterpart in brain development. However, far less is known about the latter than the former (Coan, 2008). One aspect which has been studied is the association between secure attachment and lower stress reactivity.

2.2. Terminology used in this guideline

This guideline covers children (defined as aged 0–12 years) and young people (defined as aged 13–17 years) who are adopted from care (and those adopted in England who are from overseas), in special guardianship, looked after by local authorities in foster homes (including kinship foster care), residential units and other accommodation, or on the edge of care.

The term ‘attachment difficulties’ refers to an insecure or disorganised attachment or diagnosed attachment disorders. The latter may be an inhibited/reactive attachment disorder or a disinhibited attachment disorder, now termed ‘disinhibited social engagement disorder’ in the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) (American Psychiatric Association, 2013).

Disorganised attachment and attachment disorder largely do not overlap (Boris et al., 2004). Disorganised attachment (as assessed by Ainsworth's Strange Situation Procedure [SSP] by a trained, reliable coder) and an attachment disorder (as diagnosed by a psychiatric assessment) are very different phenomena. Nonetheless, in this guideline, the term ‘attachment difficulties’ is used to refer to children who have either a diagnosis of an attachment disorder or who have been placed by a reliable coder in the disorganised attachment classification.

2.3. Types of attachment difficulties

Four attachment behavioural patterns have been defined in young children:

  • secure
  • insecure avoidant
  • insecure resistant (also called ambivalent)
  • disorganised.

These patterns are relatively stable over time in the absence of changes to caregiving. This stability is underpinned by continuities in a child's ‘internal working models of attachment’ that develop as a result of early interactions between the parent and child. The insecure avoidant and resistant patterns, while less optimal, are organised attachment patterns for retaining some proximity to the attachment figure and adapted to the anticipated response of the attachment figure. In contrast, children who are classified as disorganised, appear to lack an organised strategy for achieving closeness with their attachment figure when distressed.

Although particular types of attachment pattern, especially disorganised attachment, may indicate a risk for later problems (see Section 2.9), they do not represent a disorder. Furthermore, these categories of attachment are referred to as the ABCD model, however there are other approaches that are important, and not necessarily in opposition with this framework, such as the dynamic maturational model (Farnfield, 2009).

Disorders related to attachment have been described in the literature and are defined in the DSM and The International Classification of Diseases and Related Health Problems, 10th edition (ICD-10) for Mental and Behavioural Disorders: reactive attachment disorder and disinhibited attachment disorder or disinhibited social engagement disorder.

2.3.1.1. Secure attachment

Children who have a ‘secure’ attachment are generally able to be comforted by their caregivers when distressed and to use their caregiver as what is known as a ‘secure base’ from which to explore their environment when they are not distressed.

2.3.1.2. Insecure avoidant attachment

Children who develop an ‘avoidant’ attachment pattern are thought to maintain proximity to their caregiver by ‘down-regulating’ their attachment behaviour: they appear to manage their own distress and do not strongly signal a need for comfort. Most importantly, when reunited with a caregiver after a brief separation, these children may be quite distant, and tend to avoid contact with the caregiver. It is worth noting that these behaviours are observed during the SSP when the child is exposed to a stressful situation (separation-reunion procedure) and avoidant children are not necessarily avoidant all the time. Nevertheless, avoidant behaviour can be observed in the home using the Attachment Q-Set tool.

2.3.1.3. Insecure resistant (ambivalent) attachment

Children who have a resistant (ambivalent) attachment pattern are thought to maintain proximity to their caregiver by ‘up-regulating’ their attachment behaviour: when they are separated from a caregiver, they may become very distressed and may be angry, and resist contact when the caregiver returns, and not quickly calmed when comfort is offered. These children are less confident in terms of exploring their environment and may be wary of strangers.

2.3.1.4. Disorganised attachment

In addition to children being classified as secure or insecure, infants under 20 months can also be rated in terms of the extent to which observable behaviour suggests a disruption at the level of the attachment system, using the Main and Solomon (1990) indices of disorganisation and disorientation (Main, 1990). Examples of this behaviour include the infant approaching but with the head averted or with fearful expressions, oblique approaches or disoriented behaviours such as dazed or trance-like expressions or freezing of all movement (Lyons-Ruth & Jacobvitz, 2008). Such a disruption is understood to mean that the infant is not able to resolve their distress within the context of their relationship either by signalling their anxiety to their caregiver, or by directing their attention away from them. Where the unresolved disruption of the attachment system is regarded as substantial and/or pervasive, the coder gives a higher rating, and considers the infant for inclusion within the disorganised attachment classification. With increasing age, these children's disorganised behavioural pattern may evolve into compulsive caregiving or coercive controlling behaviours towards their primary carers.

It is important to note that behaviours reflecting disorganised attachment are only observed during an assessment, like the SSP, and may not be displayed by the child in their home (unlike insecure attachment), and that disorganised attachment may only be short-lived and can be resolved once the child is reunited and in a stable relationship with their primary caregiver. Neither the behaviour described by the Main and Solomon indices, nor a classification of disorganised attachment in the SSP, can be used in any valid way to assess a child for maltreatment. Although correlated with it, maltreatment cannot be inferred from infant disorganised attachment. Conversely, it is possible for children who are abused not to show disorganised attachment (for example, if the abuse is less severe and less frequent).

Other children, such as those on the autistic spectrum, can also exhibit disorganised attachment in the absence of maltreatment. Finally, some children will show disorganised attachment when they are frightened for their carer, for example when a parent is terminally ill or subjected to violence from another individual (typically, domestic abuse).

2.3.1.5. Attachment disorder

The term ‘attachment disorder’ refers to a highly atypical constellation of behaviours indicative of children who find it extremely difficult to form close attachments.

Reactive attachment disorder (RAD) refers to a consistent and pervasive pattern of behaviour in which a child shows extremely withdrawn behaviour, particularly a marked tendency to not show attachment behaviour toward caregivers (not seeking proximity when distressed, and not responding when comforted), accompanied by a general lack of responsiveness to others, limited positive affect and/or episodes of marked sadness, fearfulness or irritability. The diagnosis requires that there is clear evidence of pathogenic care, such as severe neglect or repeated changes in caregivers (for example, through multiple foster care placements or institutional care), and the difficulties should be evident before the age of 5.

Disinhibited attachment disorder (currently in ICD-10) (World Health Organization, 2010), which has more recently been relabelled as ‘disinhibited social engagement disorder’ (DSED) in DSM-5 (American Psychiatric Association, 2013), refers to a marked and pervasive tendency to not show appropriate cautiousness with respect to unfamiliar adults and a failure to be sensitive to social boundaries. Examples include going off willingly with a stranger with no hesitation, overly familiar verbal and physical interaction with a stranger and limited or absent checking back to a caregiver when in a new place. As in the case of RAD, DSED is only considered when there is clear evidence of a history of pathogenic care.

2.4. Assessment and measures of attachment in childhood and adolescence

The assessment of patterns of attachment is complex. Attachment is assessed for its quality or pattern, not quantitatively for its intensity and there are different ways of assessing attachment that are appropriate to different ages on the basis of observed behaviour, representation of attachment relationships and coherence of the child's account regarding their attachment relationships. Based on longitudinal studies and concurrent assessments using different methods, there is an assumption that the different methods are measuring the same concept.

The SSP, which is used between the ages of 12 and 24 months approximately, assesses the infant's attachment behaviour towards their attachment figure following a significant stressor of separation. There is also a similar assessment of a more prolonged separation for preschool-age children. Another observational method uses Q-sort to assess attachment behaviour during a more prolonged observation period. Representations of attachment patterns of children between the ages of 4 and 8 years can be assessed using the child's verbal and toy-enacted completion of beginnings (‘stems’) of a number of stories that depict stressful scenarios involving a child and their parents (MacArthur Story Stem Battery [MSSB]; Manchester Child Attachment Story Task [MCAST]). For older school-age children, attachment is assessed by verbal and non-verbal responses using 2 different procedures. In the dynamic maturational model of attachment the child is presented with a number of pictures of stressful scenarios and is invited to describe the response of a hypothetical child as well as their own responses to the scenarios. In the Child Attachment Interview (CAI), which extends into adolescence, the child or young person is invited to describe their relationships with their caregivers in various stressful situations. In this procedure, the focus of assessment is the coherence of the child or young person's account, as well as their demeanour during the interview.

There is widespread recognition regarding the importance of addressing attachment difficulties in older children and adolescents who have experienced highly disrupted care, but currently the literature is lacking clear consensus about how these should be defined and measured (Kay & Green, 2013) and very few prospective studies have addressed the factors that cause them.

A recent observational measure of attachment disorganisation has been introduced that is coded from a 15 minute interaction between parent and adolescent (Goal-Corrected Partnership in Adolescence Coding System [GPACS]; Obsuth et al., 2014). The GPACS has shown promise as a measure of attachment among at-risk adolescents, in that it is related to disorganisation in infancy, as well as to current unresolved Adult Attachment Interview (AAI) states of mind. The GPACS has also been robustly related to current maladaptation in adolescence, including increased depressive symptoms, dissociative symptoms, borderline personality disorder features, suicidality, and overall psychopathology on a standard psychiatric diagnostic interview(Obsuth et al., 2014) (Lyons-Ruth et al., 2014; Vulliez-Coady et al., 2013) It also relates significantly to abusive behaviour in romantic relationships. However, further work is needed in other samples to be sure these results will replicate broadly.

For each of these methods or procedures, there are coding manuals with variations for each of the methods.

Attachment may also be assessed indirectly by examining the primary caregiver's sensitivity to the child, particularly in response to the child's distress or fear, because a significant association has been found between maternal sensitivity and child security of attachment.

Attachment disorders are typically assessed using structured interviews with carers, and may be supplemented by questionnaires and direct observation of the child or young person's behaviour.

2.5. How common are attachment difficulties?

It is estimated that around two-thirds of children in population samples have a secure pattern of attachment across cultures (Van Ijzendoorn & Kroonenberg, 1988; van IJzendoorn et al., 1999), although this falls rapidly to around one-third in disadvantaged populations (Carlson, 1998; Weinfield et al., 2004) and less in maltreated populations.

Estimates suggest that around 8–10% of children are insecure-ambivalent (van IJzendoorn et al., 1999) and around 9% of children are insecure-avoidant (van IJzendoorn et al., 1999).

Around 15–19% of population samples (De Wolff & van Ijzendoorn, 1997) to 40% of disadvantaged populations (Carlson, 1998; Weinfield et al., 2004) and as many as 80% of maltreated populations (Carlson et al., 1989; Cyr et al., 2010) are thought to have a disorganised attachment.

The prevalence of attachment disorders in the general population is not well established, but is likely to be low (Minnis et al., 2013; Skovgaard et al., 2007). RAD and DSED are seen at substantially higher rates among young children raised in institutional care or exposed to severe abuse or neglect.

2.6. The causes of attachment difficulties

A significant body of research has investigated the causes of variations in the attachment patterns shown by infants and young children. The evidence is quite clear that the causal factors giving rise to security versus insecurity are distinct from those influencing the development of attachment disorders (Rutter et al., 2009). Each of these is considered separately below.

There is widespread recognition regarding the importance of addressing attachment difficulties in older children and adolescents who have experienced highly disrupted care, but currently the literature is lacking clear consensus about how these should be defined and measured (Kay & Green, 2013) and very few prospective studies have addressed the factors that cause them.

It seems clear from the research literature, however, that attachment difficulties are almost always caused by inappropriate parenting; behavioural genetic studies show very little genetic influence on attachment patterns, so that it is rare to observe significant attachment difficulties in the context of normatively sensitive and responsive parenting. Thus, children with attention deficit hyperactivity disorder (ADHD) and autism usually have secure attachments to their parents.

There is a growing literature on the neurobiological effects of child abuse and neglect. Child maltreatment is stressful, often repetitive or persistent and may be traumatogenic. The abusers or those neglecting the child are most often also primary attachment persons. Given the strong causal association between child maltreatment and attachment difficulties, it is likely that the neurobiological changes associated with maltreatment will be found in children with attachment difficulties. However, these changes are not explanations of the attachment difficulties.

2.6.1.1. Attachment security/insecurity

A key issue concerns the extent to which attachment security versus insecurity reflects the influence of the environment (including the behaviour of the caregiver), rather than the child's genetically-based behavioural and emotional dispositions. Several twin studies have consistently indicated that attachment security in infancy and toddlerhood is almost exclusively influenced by the environment, and minimally by genetics (Bokhorst et al., 2003; Roisman & Fraley, 2008), consistent with the emphasis within the field of attachment research on the preeminent role of parenting.

Early intensive observational work conducted by Mary Ainsworth (1969) identified variation in parental sensitivity in particular as a critical variable in determining the child's attachment security as assessed in procedures like the SSP. Broadly speaking, parental sensitivity refers to the tendency of a parent to be aware of a child's more or less subtle cues and communications, particularly (though not exclusively) those relating to distress, to interpret those cues accurately and to respond contingently to them with an appropriate response. Sensitive parenting is typically characterised by harmonious, smooth and responsive interactions in which the parent is able to read – and therefore be attuned to – the child's behaviours and cues, to accurately imagine what the child's feelings, thoughts and focus of attention might be and to respond appropriately. It is generally not considered to be equivalent to warmth, and in research studies these 2 constructs may or may not be correlated, depending on the way they are measured and the populations concerned (Mesman & Emmen, 2013). Parental sensitivity can only be measured properly by direct observation of interactions, ideally over a significant period of time, and in more than 1 context or occasion. Insecure attachment is generally considered to be associated with parenting that is insensitive, either because the parent's behaviour is intrusive (not following the child's cues, rigid or forcing the direction of interactions), rejecting (negative response to, or discouraging of the child's bids for contact or comfort), hostile, withdrawn or the parent is inconsistently available to the child.

Cross-sectional and longitudinal associations between these parenting features and attachment insecurity have been observed in numerous studies in a wide range of social, clinical and cultural contexts; for a narrative review see (Belsky & Fearon, 2008). Meta-analysis of these studies suggest that the average association is highly statistically significant, but small in size (De Wolff & van Ijzendoorn, 1997), which suggests that typical assessments of sensitivity do not capture all of the causal factors, either due to measurement error, or because other factors are involved. Crucially, intervention studies focused on improving sensitivity have been successful in improving rates of secure attachment, which suggests that sensitivity is a causal factor in attachment security, not just a correlate of it (see Bakermans-Kranenburg et al., 2003).

2.6.1.2. Disorganised attachment

A substantial number of studies have found that standard assessments of sensitivity do not reliably predict disorganised attachment (van IJzendoorn et al., 1999). Instead, existing studies indicate that disorganised attachment is associated with a cluster of parenting behaviours that include ‘frightening/frightened’, extreme intrusiveness, unmarked frightening facial expressions, unusual vocal tone and dissociative behaviour. Several studies have also indicated that a broader range of ‘atypical’ parenting behaviour may be involved, including affective communication errors, role/boundary confusion, and withdrawal (see Jacobvitz et al., 2006; Out et al., 2009). Disorganised attachment has also been observed at high rates in samples of infants and young children who have been exposed to maltreatment (Cyr et al., 2010; van IJzendoorn et al., 1999a) and, to a lesser extent, among children who have been adopted or are in foster care. Less research has been done to rigorously test in intervention studies the causal nature of these observed associations than that concerning the role of parental sensitivity.

2.6.1.3. Attachment disorders

Attachment disorders are observed almost exclusively in conditions that represent extreme departures from normative care, including extreme neglect and institutional care. In particular, a diagnosis of RAD (according to the DSM-5) is only given when children have experienced pathogenic care, meaning a persistent disregard of the child's emotional or physical needs, or repeated changes in primary caregivers (for example, in foster care or within institutions). It is notable that no cases of RAD have been identified in the literature in which neglect was not clearly present (Zeanah & Gleason, 2014). DSED, although not currently defined as a disorder of attachment in the DSM-5, has been associated with a similar set of highly disturbed early caregiving experiences, and requires the same pathogenic care criteria to be met as RAD. Both of these disorders are observed at relatively high rates in children within institutions, children adopted out of institutions and in some children in foster care, although they do not represent the majority (Zeanah & Gleason, 2014). Relatively little is known about the precise environmental processes that are responsible for the emergence of RAD or DSED. There is some suggestion that the effects of harsh or negative parenting on the development of RAD may be mediated by gene expression (Minnis et al., 2007). Although DSED is no longer defined as an attachment disorder in DSM-5, there is some disagreement in the literature about this (and for the purposes of this guideline, it is included in the definition of attachment disorders).

2.7. When do attachment difficulties start and how long do they last?

Studies of normative development suggest that clear selective attachment bonds become evident sometime between the ages of 6 and 9 months, as indicated by preferential seeking of comfort from selected individuals, distress triggered by being separated from them and stranger wariness (Schaffer, 1966). Prior to that, early interactive processes most likely important for the subsequent development of attachment are clearly observed (for example, mutual eye contact, social smiling, contingent interactions, provision of contact and comfort), although remarkably little research has investigated in detail the role that these play in the formation of attachments. Standard assessments (like the SSP) are generally used from the end of the first year and can reliably categorise attachment patterns and behaviours. It is therefore generally accepted that insecure or disorganised attachments can be clearly observed at 1 year of age, although it is not straightforward to conclude that they are not present earlier, and the parent–infant interaction patterns that are believed to give rise to them are certainly present, and measurable, earlier than that.

Similarly, structured interview techniques and related observational procedures are used at this age to assess the presence of attachment-related disorders and their associated behaviours (Zeanah & Gleason, 2014). Furthermore, diagnosis of RAD requires that signs of RAD must have been present before the age of 5 years. For both RAD and DSED highly insufficient care must have been present and would typically have occurred in infancy or early childhood.

The question of how stable attachment patterns and disorders are is a complex one, partly because there is a general lack of measurement tools that can reliably assess attachment in the same way for all age groups. Nevertheless, there is some consensus on the following key points. First, attachment patterns in infancy and early childhood show some stability over time, but are also open to change. Second, short- and medium-term change in attachment patterns (for example, from insecure to secure) tends to be linked to changes in caregiving (for example, from relatively insensitive to relatively sensitive), or other family circumstances (for example, marital difficulties or separation). Third, long-term stability in attachment security (that is, from infancy into late adolescence or adulthood) is limited, but later attachment outcomes are related to a broader assessment of the quality of familial experiences occurring right across childhood, for example, quality of care, divorce and parental wellbeing; see Groh and colleagues (2014).

RAD shows relatively high stability in the short-term (a requirement for diagnosis) in the context of a stable environmental context, for example within an institutional care setting or treatment-as-usual foster care (Gleason et al., 2011), but resolves quite quickly when appropriate stable attachment figures are provided, for example in foster care with suitably trained foster carers (Rutter et al., 2009). By contrast, DSED shows quite high levels of persistence over time in studies that have been conducted to date both in early childhood and into adolescence (Zeanah & Gleason, 2014), even when appropriate foster care has been in place for some time. However, it is important to note that the great majority of these studies have focused on children previously raised in institutions and less work has examined the stability of DSED or RAD in the context of children who entered foster care, or were adopted, from non-institutional circumstances. An example of the work that has been conducted on children in foster care, the children showed high levels of indiscriminate friendliness, a symptom or RAD, but they had experienced serious maltreatment and numerous placements (Pears et al., 2010). Thus, it is still unclear how likely symptoms of RAD or DSED are found in children living in a stable, loving foster care placement or who were adopted.

2.8. What mental health problems and behaviours are associated with attachment difficulties?

In people with RAD there is an increased prevalence of anxiety, fears and phobias. DSED is not necessarily associated with a diagnosable mental health problems, although as might be expected, there is an increased risk of both internalising (anxiety, depression) and externalising problems (conduct and aggressive problems) and a risk that social disinhibition may lead the individual to become abused by unscrupulous older children and adults, and to go down the path of antisocial behaviour, drug misuse and promiscuity. For both RAD and DSED, there can be disturbances of emotions and behaviour that will be associated with the neglect or abuse that led to them, including emotional dysregulation and poor temper control, leading to oppositional defiant disorder, and dysregulated mood disorder. In the English Romanian adoptee study, there were 4 specific patterns associated with severe neglect/privation: a quasi-autistic syndrome, ADHD, social disinhibition similar to DSED, and impaired cognitive ability (O'Connor & Rutter, 2000). Thus, a child with the disinhibited attachment picture could have any or all of the other 3 mental health problems.

The association of insecure attachment patterns with mental health problems is more complex. Here an insecure attachment pattern will be taken to include a disorganised attachment as well as avoidant and ambivalent patterns. A meta-analysis of a large number of studies found no increase or a modest increase in prevalence of all kinds of mental health problems associated with avoidant and ambivalent patterns, but significant and greater increases in mental health problems (particularly externalising problems) among children displaying disorganised attachment (Solomon & George, 2011). Children with earlier disorganised attachment frequently develop coercive controlling or compulsive caregiving behaviour.

One particular mental health problem that seems to have a higher prevalence than the others is oppositional defiant disorder/conduct disorder. Meta-analyses suggest that around 55% of children with oppositional defiant disorder/conduct disorder have any pattern of insecure attachment (compared with around 30 to 40% in controls), of whom about 30% have disorganised attachment (compared with 15% in controls); thus in children with oppositional defiant disorder/conduct disorder, the odds ratio (OR) of having disorganised attachment is nearly 4-fold.

Disorganised attachment, and to a lesser extent avoidant and resistant attachment patterns, are associated with externalising problems (anger, aggression), more so in boys. Avoidant attachments are associated with internalising problems (depression, anxiety, social withdrawal, somatic complaints) in both boys and girls. Disorganised, insecure avoidant and resistant attachment patterns in both boys and girls are associated with later poor social competence with peers.

Perhaps because of the overlap with maltreatment, it is not uncommon for professionals to use the term ‘attachment difficulties’ to cover a wider pattern of behaviour that might include the sequelae of maltreatment or be otherwise experienced relatively commonly by children in the care system or adopted from care. This may mean people conflate attachment difficulties with other developmental problems, such as:

  • aggression, oppositional or defiant behaviours
  • hyperactivity, poor concentration and risk-taking
  • lying, stealing and manipulative behaviours.

Also, the apparent overlap in the behaviour of a child with attachment difficulties and a child with a different neurological condition, may lead to a child being misdiagnosed (with conditions such as ADHD or Williams syndrome), before the extent of the attachment and trauma issues have been recognised. Thus it is important that healthcare professionals take into account all manner of explanations and causes during an assessment that may lead to a single or dual diagnosis.

In conclusion, for insecure attachment patterns, any mental health problem is likely to be more common, but particularly among children with disorganised attachment. However, this is not to say that the attachment difficulty has led to the behaviour problem; rather, it is much more likely that the disturbed parenting has had effects on making the child more anxious, more frustrated and aggressive, less able to comfort themselves and more emotionally dysregulated, and physiologically more prone to become rapidly emotionally aroused and to take longer to calm down and return to a more normal physiological and mood state. More severe neglect may also affect a child's neurological configuration (and continue to affect it into adolescence) and their attention span and ability to make social relationships (Cozolino, 2014; Siegel, 2001; Van der Kolk et al., 1991).

2.9. How do attachment difficulties manifest in education, healthcare, social care and criminal justice settings?

2.9.1. Education

Attachment may be an important influence on pupils' academic success and wellbeing at school. First, security of child-parent attachment has been found to influence a number of areas of child development that are extremely important in the school setting – self-regulation (controlling one's behaviour, sustaining attention, controlling emotions), willingness to take on challenges and persist in the face of setbacks, social competence with peers and less aggressive behaviour (Bergin & Bergin, 2009). Furthermore, children can and do form relationships with teachers that have an attachment quality to them, and serve a similar function of creating a feeling in the child of safety and security. The quality or security of that relationship in turn may influence the child's emotional wellbeing and engagement with learning. For young people with attachment difficulties, the challenging business of learning and coping in the classroom can be very difficult.

Data from the Department for Education for the success of looked-after children in education – many of whom will have attachment difficulties – show a very significant gap between their outcomes and those of non-looked-after children. In 2013 only 15.3% of looked-after children achieved 5 or more A* to C grade GCSEs (General Certificates of Secondary Education) including English and maths, compared with 58% of non-looked-after children (Depatment for Education, 2013) and the attainment gap in 2014 for the percentage achieving 5 or more GCSEs or equivalent at A* to C grade including English and maths is 40 percentage points (Department for Education, 2014). Nationally, in English and maths, approximately 70% of all children make 3 levels of progress from the end of key stage 2 (age 11) to the end of key stage 4. For looked-after children these percentages in 2013 were 32.6% in English and 29.2% in maths. Looked-after children were twice as likely to be permanently excluded from school and nearly 3 times more likely to have a fixed-term exclusion than all children. Unsurprisingly, around half of all looked-after children aged 5–16 years were considered to be ‘borderline’ (12.8%) or ‘cause for concern’ (36.7%) in relation to their emotional and behavioural health based on their Strengths and Difficulties Questionnaire (SDQ) scores in 2014 (Department for Education, 2014).

Behaviours associated with attachment difficulties, such as disruptive behaviour in the classroom, difficulties forming relationships with teachers or positive peers, and difficulty in establishing a moral code (linked to their value of an authority figure), are commonly seen in schools. Some children may display clinginess to teachers; older children may have difficulties with boundaries. Other children may be quiet and not engage because they are internalising their issues, and because they appear to be coping they could be overlooked.

For teachers, it is really important to be able to ‘read’ these behaviours and respond appropriately. It is a concern that the majority of teachers will not have covered such issues in their training.

It is important to note that an awareness and understanding of children's attachment difficulties should not obviate the need to examine additional reasons for a child's difficulties in the educational setting, such as physical health problems (sight, hearing) and specific learning and reading difficulties, which tend to be masked by more overt behavioural or emotional difficulties and are under-diagnosed in looked-after children. It is also important to manage the sequelae of trauma and maltreatment. For children who have been maltreated or exposed to trauma, learning is more difficult, as the normal and necessary ‘fight, flight or freeze’ response is triggered very easily. Changes of placement often involve a change of school. The resulting sense of dislocation and disruption to relationships, the need to negotiate new settings and relationships and to enter already established friendship groups increases levels of stress and decreases the ability to learn.

2.9.2. Healthcare

Healthcare settings cover a wide range of care provision, including primary, secondary and more specialised settings, for both mental and physical heath.

Direct manifestations of attachment difficulties may be observed when a child does not show distress in situations when this might be expected, does not seek comfort or shows difficulty in accepting comfort from a carer when frightened or feeling threatened. Another aspect that may be observed is a child's indiscriminate friendliness and approach to strangers, as might occur in an inpatient healthcare setting.

Attachment difficulties are also correlated with a range of emotional and behavioural problems, which will be noted in healthcare settings or for which the child may be referred, especially to child and adolescent mental health services (CAMHS). These difficulties include both internalising (such as anxiety, depression, social withdrawal and somatic complaints) and externalising problems (such as difficult and challenging behaviour or aggression and threatening behaviour). However, possible attachment difficulties cannot be assumed to be present, but, upon assessment, may be considered as part of a formulation of a child's difficulties.

Neglectful, unresponsive, insensitive or hostile parent–child interactions may be observed in healthcare settings. While these may lead to attachment difficulties, the latter cannot be assumed to be present, although if these parent–child interactions are persistent, it is likely that they will have led to attachment difficulties. Thus, it could be said that attachment difficulties are markers of some form of maltreatment within the family and that the child's behaviour is a survival response that, if left untreated, will become a hardwired, stress reaction (that is, an attachment difficulty is an indicator not an end diagnosis).

2.9.3. Social care

Within social care settings, children and young people may be placed in a variety of placement types (that is, adopted home, foster care, residential care or kinship care) with varied contact arrangements and levels of insight about why they no longer live with their family of origin. They are often cared for by people who have not had specific training about attachment difficulties, who may perceive the child's behaviour simply as ‘problem behaviour’ and struggle to connect it to their past experiences or to respond with consistency and sensitivity. Children in care settings may show 1 of 2 patterns of relationships with their carers that may be a cause of concern. One group consists of children who are likely to become very agitated in their new surroundings, as well as with their new carers, especially if they have experienced disrupted placements over a short period of time (McDonald & Millen, 2012). For them, the world of relationships will have become unpredictable, to the point where their 'best' strategy for survival is to be unpredictable themselves. This offers them a way (albeit short-term) of being noticed. Consequently, these children externalise their behaviour and tend to be aggressive, demanding and hostile. The second group consists of children who are, in some respects, more worrying because they appear to internalise their distress and trauma. They tend not to show their feelings and can become superficially compliant and undemanding. They appear to 'settle in well', but underneath the facade these children are often in turmoil and experience considerable distress.

Both of these groups of children who have experienced highly troubled attachment relationships in the past can struggle to trust adults (Barton et al., 2011). Confusingly, they can become very demanding if they are offered a genuinely secure base and safe haven in, for instance, an adoptive home. They are not used to adults being predictable, kind and nurturing, so they inadvertently reject the very people they need in order for them to grow and develop emotionally, and to help them survive traumatic childhood experiences (Rivard et al., 2005). Adoptive parents, special guardians, foster carers, kinship carers, residential staff and birth parents may all need additional support to help them understand these behaviours and to prevent them from jeopardising placements.

2.9.4. Criminal justice system

Young people in contact with the youth justice system are known to have higher levels of mental health problems (Chitsabesan et al., 2006) and other unmet needs than their peers (Chitsabesan & Bailey, 2006). Although less is known specifically about attachment difficulties in this population, many of them have either been looked-after children or have had multiple carers (Harrington et al., 2005), and they have had a high level of exposure to traumatic events (Abram et al., 2004), all of which may be associated with attachment difficulties. Additionally, they have often had multiple education placements and are likely to have come into contact with many professionals, either directly as a result of their offending behaviour, or as an indirect consequence (due to placement breakdown and so on). A recent policy change (Legal Aid, Sentencing and Punishment of Offenders Act 2012; (LASPO, 2012) explicitly acknowledged their need for additional support, and now young people remanded either to custody or to the care of the local authority are deemed to be looked-after children.

This instability of relationships with primary caregivers, and the sheer number of professionals with whom they have had contact, means that young people within the justice system often have difficulty in trusting professionals they meet. As a result professionals may find that these young people constantly ‘test out’ the relationship in a number of different ways, or that it is difficult to engage them at all. It may take multiple contacts with a young person before they feel willing to engage at any level with a new professional. Some young people may focus on short-term gains within any encounter with a professional, and hence initially engage well, but the professional may have difficulty sustaining the engagement when difficult topics are broached or the young person feels challenged. Some young people have found that escalating their behaviour is an effective way to regulate relationships with professionals, as the immediate behaviour (rather than underlying issues) becomes the focus of the interaction.

Professionals may well have difficulty establishing relationships with these young people (probably at the end of a long chain of contacts with professionals), meaning that it is harder to work with this population. They may present with low empathy or escalate behaviours when challenged, have problematic relationships with staff and peers and are likely to make multiple transitions that will exacerbate problems. Behavioural problems arise in residential/custodial settings (escalation to make problems go away or in hope of ending placement). Children and young people will also present with the same difficulties that would occur in any residential setting (see the section on social care above).

2.10. Perspectives and experiences of care-leavers and carers: daily life, family and relationships

As looked-after children's relationships with previous caregivers are often disrupted and unreliable, they are unlikely to have experienced secure and stable attachments. By the very nature of entering the care system, another attachment has been disrupted. This leads children to perceive parental figures as unreliable and incapable of providing protection. Children who experience insecure attachments begin to develop defensive behavioural techniques to protect themselves from a world of insecurity and hostility (Howe et al., 2001a). Moreover, if children experience high levels of arousal and have no strategies in which to deal with them, they will face further problems, such as sleeping and eating problems (McNamara et al., 2003).

For those who have lived in the care system, each loss of caregiver or placement is a big change, and even where the feelings are not evident to an observer or are covered with challenging behaviour, these are losses that lead to a grieving process. Having nobody that you can trust and confide in becomes a fact of life, and you learn to invest less in each subsequent relationship. To the child, it feels like the problems (including placement breakdowns) are all because of something wrong with them, leading to intense feelings of shame, sadness, anger or isolation. Learning to trust in relationships again enough to share their experiences and feelings can be a slow and often painful process that needs to be given time and support. But it is the most important experience for these young people.

Birth family relationships, although not always healthy, are often very important to children who are looked after. Contact can be de-stabilising as it may bring up old wounds, but it can also be reassuring to know that relatives are still alive and care enough to come (Sinclair, 2005), this is particularly important if the care is short-term, or there are plans for the child to return to their birth family. However older children tend to make their own decisions and arrangement about the amount and type of contact they want with their birth families (Selwyn, 2004). Young people may be very concerned for the wellbeing of parents or siblings (particularly where there have been issues with alcohol or substance use, self-harm, domestic violence). Maintaining a relationship with siblings can be an important source of identity and shared experiences, as well as the longest relationships in people's lives.

Adopted children with attachment difficulties can have further difficulties in many aspects of daily life. They need to be and feel safe, to live in a caring, nurturing and structured home. Their attachment difficulties, and their behaviour needs to be fully understood by their adoptive parents, educators and supporters. The number and quality of foster care placements and previous maltreatment will impact on the attachment pattern that they bring into their adoptive placements (Sinclair et al., 2007). It is important to them to control many areas of daily life and this can often be difficult for parents, teachers and supporters to understand. Adopted children, will present with many overlapping difficulties (Schmid et al., 2013b), but regardless of age and the length of time that they are in their adoptive families, they need their parents to be attuned to all of those needs. If parents do not receive consistent support and education to be sensitive to their child's attachment needs, adopted children can – and often do – present with challenging behaviours (Selwyn et al., 2014); and even when adoptive parents are sensitive to the child's needs, the child may still go on to develop those behaviours.

2.11. Treatment and management of attachment difficulties in England and Wales

2.11.1. Psychological interventions

Psychological interventions for children with attachment difficulties can be conceptualised as those that directly address child attachment security, and those that address associated problems. With respect to those that address attachment security, for children still living in the family where the attachment difficulty has arisen, the first line of treatment is to improve the relationship between carer and child. The largest number of randomised controlled trials (RCTs) have been conducted in infancy, and in the meta-analysis by Bakermans-Kranengburg 2003 (Bakermans-Kranenburg et al., 2003), the conclusion was that in this population ‘less is more’, meaning that interventions that were relatively short and had a behavioural focus in improving sensitive responding of the parent and, where necessary, improving limit setting, led to the greatest increase in attachment security. In addition to this Leiden group, other major research groups who have conducted trials on interventions to increase attachment security include the Mount Hope Centre in Rochester, New York (Toth et al., 2006), the Delaware group (Dozier et al., 2006) and the Washington State group who have developed the Circle Of Security, although this has not yet been subjected to an RCT.

There is much less evidence for later developmental periods, including middle childhood and adolescence. The recently published naturalistic National Institute of Child Health and Human Development longitudinal study found that children who moved from insecure attachment to secure attachment as they grow up experienced an associated improvement in their parenting they received and in their living circumstances, again suggesting that improving parenting will lead to greater attachment security and better outcomes generally. However this process may be longer and require more support for older and more traumatised children, and relies on a stable placement with high levels of parental sensitivity.

For children who have been removed from abusive families and placed into foster care, meta-analyses suggest that the attachment security to their foster carers is similar to typically brought up children, suggesting that children do indeed have the capacity to form new trusting attachment relationships despite early abuse. This was directly tested in the study of Joseph and colleagues (2014) where intra-individual attachment security was measured, and was almost entirely insecure to abusive birth parents, but the majority were secure to their foster carers, showing that a more benign parenting environment led to secure attachment patterns. It seems that the children can benefit from more nurturing experiences and gain healthier attachments and ways of expressing their needs. However, this does not erase internal working models based on the trauma they have experienced, which can trigger challenging behaviour when under stress, especially when transitioning from late childhood to adolescence (Hodges et al., 2003; Hodges et al., 2005).

Additionally, there are a number of approaches with looked-after children that aim to improve parenting, and which may also improve attachment security. Standard parenting programmes such as the incredible years have been shown to improve sensitive responding, which is likely to lead to more attachment security. Other parenting programmes specifically for Foster carers that are based on evidence-based principles also appear to show an improvement in child attachment security (Briskman et al., 2014).

There is a wide range of other relationship-based therapies available, but none appear to have been subject to an RCT. Some are widely used within the UK and may promote secure attachment in children on the edge of care or in care. Others are abusive, not therapeutic and make unsubstantiated claims about improving brain function. Any form of ‘therapy’ involving physical restraint, coercion, the child lying or sitting on the therapist or any form of aversive stimulation (for example, ‘holding therapy’) not only has no evidence base, but is associated with harm to children and should be considered malpractice (see the Report of the APSAC Task Force on Attachment Therapy, Reactive Attachment Disorder, and Attachment Problems).

An attractive notion is that giving a child individual psychotherapy will help them come to terms with an abusive parent and so improve their attachment security in relation to that person, by enabling them to talk about them in a balanced and coherent way, so called ‘earned security’. However, to date, these ideas are untested and therefore individual psychotherapy is of uncertain value. Although use of creative and non-directive therapies is popular with this population, there is no evidence for the efficacy of any form of individual therapy done with primary school-age children in terms of addressing attachment difficulties. The evidence available shows parent–child psychotherapy or trauma-focused cognitive therapy for both the child and parent may improve parental sensitivity or attachment security in children and young people who have been maltreated with related trauma (Cicchetti et al., 2006; Cohen et al., 2004).

It is important for healthcare professionals to understand that a child's behaviour in care can be very complex and may be due to a past trauma, not necessarily the result of poor parenting provided by the foster carers or adoptive parents. Thus, complex trauma (in the presence of attachment difficulties) should be identified and addressed separately to an intervention that aims to improve the attachment between the child and their foster carer or adoptive parents.

With respect to the associated problems (such as complex trauma), standard evidence-based treatments should be offered to children with attachment difficulties, just as they should be with children who do not have attachment difficulties. Thus, with older children therapeutic techniques such as cognitive behavioural therapy (CBT), interpersonal therapy, eye movement desensitisation and reprocessing, dialectical behaviour therapy, cognitive analytic therapy, family therapy that have a proven evidence base should be used for problems that they have already shown to be effective for in other populations of children. It is important to note that these interventions have primarily been assessed in children without attachment difficulties. Thus, future research should focus on how to better adapt these interventions for this population.

It is important that there is also understanding of the child's psychological needs and a consistent, empathic and containing environment within school.

There is evidence that respite as part of an overall parent training and support package is effective with looked-after children who have previously been traumatised, showing good effects both on reducing the likelihood of placement disruption and potentially increasing attachment security (Hudson & Levasseur, 2002; Redding et al., 2000; Rhodes et al., 2001; Triseliotis, 1997). Brief periods of time out, in the context of a loving relationship, are unlikely to recreate traumatic experiences and provide temporary relief for foster parents to rejuvenate from the stress of fostering

Because it is a relatively small population group whose needs are highly complex, services often span (or fall between) health and social care, and the priority is normally to find and support stable placements for looked-after children, which should be within a family wherever possible (Winokur et al., 2014). It is often hard for families and carers to access therapeutic support due to the pressures in the public sector to limit CAMHS to working with diagnosed mental health problems, rather than the sequelae of maltreatment, but specialist therapeutic support is highly valued by participants. Sadly, it remains the case that straightforward, evidence-based interventions are often very hard to access for adopters, both because overall therapeutic provision is low, and secondly because even where they are available, they are not offered to children on the edge of care and fostered and adopted children.

2.11.2. Pharmacological interventions

Pharmacological interventions are not the mainstay of interventions for attachment difficulties. It is difficult to conceive of medication that would enhance a child's expression of their distress or which would increase the child's capacity to receive and accept comfort. However, there may be circumstances in which treating another disorder may help a parent to be more sensitive and responsive because the child's behaviour may be more manageable, which in turn may support a secure attachment.

There are medications that ameliorate some of the emotional and behavioural difficulties associated with attachment difficulties, such as ADHD or depression, but there is no theoretical explanation why this should affect attachment.

Regarding caregiver sensitivity, this could in theory be enhanced by the administration of oxytocin. To date, there have been no studies showing increases in attachment security in children in relation to use of oxytocin.

2.12. The economic cost

Children who are on the edge of care, looked after, or adopted from care are at high risk of both insecure and disorganised attachment. In England in 2011 a majority of children were in care as a result of abuse and neglect (55%) (Curtis, 2014; Department for Education and Skills, 2005), and as many as 80% of children who have experienced maltreatment have a disorganised attachment (Carlson et al., 1989; Cyr et al., 2010).

In England gross expenditure on looked-after children was estimated to be £2.5 billion in 2013/14. The majority of expenditure was on foster care services (55% of expenditure, around £1.4 billion, caring for 51,340 children and young people), and children's homes (36% of expenditure, around £0.9 billion, caring for 6,360 children and young people) (Harker & Heath, 2014). Estimates of the average social care cost per looked-after child range from £33,634 a year for children with no additional support needs to £109,178 for those with complex emotional or behavioural needs. The cost of providing and maintaining the placement accounts for over 90% of the costs of a care episode. As well as reflecting different levels of activity from social care staff, the substantial variations in cost incurred by children with different needs reflect variations in the type and cost of placements they receive. The weekly cost per child is £2,995 for a local authority care home (2013/14 prices), £2,947 for a non-statutory care home (that is, voluntary and private sector care homes) and £700 for local authority foster care (Curtis, 2014).

In England the average weekly social services cost per child who experienced abuse/neglect is £163 if supported in their families or independently, and £756 if looked after (Curtis, 2014). The social services' costs include: the costs of field and centre staff time carrying out social services activities with, or on behalf of, identified children in need and their families; the costs of providing care and accommodation for looked-after children (and similar regular, ongoing expenditure that can be treated in the same way); and one-off ad hoc payments and purchases for children in need or their families. Similarly, the costs associated with adoption are high. The average cost per day across all adoption services (including the private and voluntary sector) is £230 (2013/14 prices). This estimate includes adoption allowances paid and other staff and overhead costs associated with adoption including the costs of social workers seeking new and supporting existing adoptive parents.

Foster placement instability is a significant problem with large numbers of children, particularly teenage children, experiencing as many as 3 moves in the first year (Ward & Skuse, 2001). Attachment and other forms of emotional disturbance are 1 of a number of factors influencing the stability of such placements (Sinclair, 2005). Multiple placements of this sort have significant cost implications. In a recent report, Hannon and colleagues (2010) explored the consequences associated with 2 care journeys, which represent the best and the worst current system. One journey was designed to reflect the experience of the very top range of 5–10% of children in care who are fortunate enough to have long-term, stable placements and supported transitions. The other scenario reflected the 5–10% of children who have a journey characterised by instability, disruption and abrupt exits. The authors found significant variation in costs: ‘Child A’ with a stable care journey cost £352,053 over a 14-year period, while ‘Child B’ with unstable care journey cost in total £393,579 over a 7-year period (a difference in total cost of £41,526). This translates to a substantial difference in annual costs per year (£23,470 for ‘Child A’ and £56,225 for ‘Child B’) once their length of stay in care is taken into account (15 versus 7 years), difference of £32,755 per year.

The authors went on to consider adult outcomes that might be associated with each scenario to estimate the possible costs to the state up to the age of 30. It was assumed that ‘Child A’ leaves care at the age of 18 following a stable placement, with good qualifications. ‘Child B’ was assumed to leave care at 16.5, with no qualifications, and with mental health problems. ‘Child A’ may cost the state £20,119 by age 30 if they go on to university and secure a graduate job. ‘Child B’ may cost the state £111,924 if they experience unemployment, underemployment and mental health problems. Between the age of 16 and 30 there is a difference between the costs of ‘Child A’ and ‘Child B’ to the public sector of £91,805, or £6,558 per annum. Greater stability and improved mental health can reduce immediate costs to the local authority by reducing social workers' time, use of expensive agency and residential placements, and therapeutic support.

Attachment difficulties are strongly associated with later problems. A review of 69 studies that examined the association between insecure or disorganised attachment and externalising problems found significantly increased risk for both insecure (Cohen's d [d] = 0.31, 95% confidence interval [CI]: 0.23 to 0.40, with larger effects for boys [d = 0.35], clinical samples [d = 0.49], and observation-based outcome assessments [d = 0.58]) and disorganised children (d = 0.34, 95% CI: 0.18 to 0.50), with weaker effects for avoidance (d = 0.12, 95% CI: 0.03 to 0.21) and resistance (d = 0.11, 95% CI: 0.04 to 0.26) (Fearson et al., 2010). Externalising problems of this nature are strongly associated with a range of later problems including substance misuse and criminality (Allen et al., 1996). Conduct disorder, substance dependency and crime impose significant social costs and harm to individuals and their victims, families and carers, and to society at large. The cost of proven offending to the criminal justice system, including the costs of police, courts, offender management teams and custody was estimated to be approximately £8,000 per young offender (in 2008/09 prices) (National Audit Office, 2011). Criminal behaviour can persist into adulthood imposing immense costs to society. For example, the lifetime costs of crime attributable to conduct disorder in childhood range from £75,000 to £225,000 per case (Health, 2009).

The authors of a recent Health Technology Assessment (HTA) report (Wright et al., unpublished) estimated the expected budget impact of screening strategies and treatment for disorganised attachment within the context of a clinical commissioning group (CCG). The authors assessed budget impact of screening and treating disorganised attachment by various target populations (for example, general population, middle class children, born into poverty, alternative caregiver [that is, adopted or fostered], and maltreated). Assuming all children born in a CCG were to be screened (a general population programme) at a certain age after birth the number of screens per year would be equal to the number of births. If the average CCG in the UK covers 264,039 individuals and assuming the general population screening strategy aimed to screen all children born in that CCG at a pre-defined time from birth, the expected cohort that could be screened in the general population would be 3,237 newborn children with a total cost of identification to the average CCG of £93,873, and subsequent treatment would cost, on average, £219,987. This suggests that the total cost to screen the general population and change disorganised attachment would approximate to £313,860 per year (2011/12 prices). The above estimates assume use of the SSP at a cost of £29 per case, average treatment cost of £2,265 per case, and expected prevalence of 3%.

Attachment difficulties and associated mental health problems during childhood therefore place a considerable financial burden on health and social care services, the criminal justice system and society. As such, it is important to identify cost-effective interventions that can help to reduce the burden to service users, their families and carers and society as a whole.

Copyright © 2015 The British Psychological Society & The Royal College of Psychiatrists.
Bookshelf ID: NBK356196