Attachment difficulties and disorders

Children and young people who are adopted from care, in care, or at risk of going into care are at higher risk of attachment difficulties and disorders. This may increase the likelihood of mental health conditions and poor emotional regulation. GPs play a role in managing this risk in the community in conjunction with a multi-disciplinary team and supporting referrals to secondary care. However, many GPs are unfamiliar with the terminology of attachment difficulties, attachment disorders, secure attachment and insecure attachment. This article aims to explain these terms and provide an update for GPs on the implications of the National Institute for Health and Clinical Excellence guidelines on child attachment, which focuses on looked after children.

What is attachment?

Child attachment
Human infants are vulnerable and dependent on their caregivers for nurturing and safety. Attachment theory suggests that children are predisposed to signal to their familiar caregiver or caregivers when alarmed, sick or in distress (Bowlby, 1969). If children receive care that gives them confidence that their signals will be noticed and responded to, they are described as having a 'secure' attachment. Whereas if children receive care that gives them a lack of confidence that their signals will be noticed and responded to, they are described as having an 'insecure' attachment. Attachment research has shown that early experiences of care can influence our later assumptions about the availability of others to help us when we are alarmed, sick or in distress (Fraley, 2002). Children in care, or on the edge of care, often have complex experiences of care, and they may not feel that they have available and supportive caregivers. The main reason for being taken into care, affecting 62% of children in care, was abuse or neglect (Meltzer et al., 2003). Less common reasons include family dysfunction, family in acute stress, absent parenting, parental illness, parental disability or the child's disability (Meltzer et al., 2003). All of these reasons can influence the development of child attachment, especially when they compound one another.
A variety of factors can influence the extent to which adult caregivers notice and respond to children's signals of alarm, illness or distress. Broadly speaking, these can be divided into three main factors: parental mental health; maternal and neonatal physical health; and family and social context (NHS Education for Scotland (NHSEFS), 2016). Some of the potential factors include but are not limited to: post-natal depression; puerperal psychosis; traumatic births resulting in post-traumatic stress disorder; prolonged stay in hospital; and traumarelated experiences in conflict zones. Complex pregnancies, pregnancies involving foetal abnormalities or physical deformities and unplanned pregnancies have been associated with poorer mental health outcomes (NHSEFS, 2016). These factors can be identified by the multi-disciplinary team.
The Department of Health advises that a child entering care should have an initial health assessment by a doctor and a care plan created within 4 weeks. Review assessments can be carried out by midwives, nurses, social workers, foster or residential carers. These are conducted twice a year for children under 5 and annually when over 5 years old (Carmichael et al., 2016;Department of Health, 2009). Another opportunity for medical review and consideration of the social history of a looked after child occurs during routine health checks conducted at school alongside peers.

Attachment difficulties
Attachment difficulties and attachment disorders are similar sounding terms and easily confused. Indeed, GP referrals struggle to make appropriate distinctions (Woolgar and Baldock, 2015). The term 'attachment difficulties' is not a diagnostic term; it is also not a formal research construct. The National Institute for Health and Clinical Excellence (NICE) use 'attachment difficulties' as an umbrella term for children who have either (or both) insecure attachment behavioural patterns or diagnosed attachment disorders (NICE, 2015). Insecure attachment and diagnosed attachment disorders are quite different and will be explained. In common, though, NICE (2015) advises not to give a pharmacological treatment for either form of attachment difficulty, and notes that both are expected to respond well to improving the quality and stability of caregiver-child relationships.

Insecure attachment behavioural patterns
Insecure forms of attachment suggest that a child has learnt that their caregiver will not be reliably available and responsive when they are alarmed or distressed. They are relationship-specific: a child can have an insecure attachment relationship with one caregiver, and a secure attachment relationship with another. Insecure attachment can only be reliably assessed after children have reached a developmental age of 9 months, as infants need time to learn how to form selective attachments to caregivers. Approximately 35% of infant-parent attachment relationships in the general population are insecure (Lewis-Morrarty et al., 2015). Insecurity can take a variety of forms, for instance highly emotionally demanding and angry ('resistant') or closed off and distant ('avoidant') responses to the caregiver when the child is alarmed. The response can vary in different settings (see Table 1.). These forms are shaped profoundly by the age of the child; but what insecure children have in common is apparent distrust in their caregiver's availability. Though it can be assessed through validated psychological assessments, insecure attachment is not a recognised diagnostic category in the Diagnostic and Statistical Manual of Mental Disorders (DSM) or International Classification of Diseases (ICD). Hence, GPs do not diagnose insecure attachment behavioural patterns. Although it is not a form of mental disorder, insecure attachment is a risk factor for later mental health problems (Groh et al., 2017). As GPs we can consider how insecure attachment influences the development of mental health problems and children's communication with their caregivers about matters, including those relating to health, that may alarm or distress them.
Insecure attachment patterns are not fixed traits of the child and can improve over time, a change that is made more likely by a supportive family environment including parental sensitivity to the child's signs of alarm, sickness or distress -and any resources or interventions that help facilitate such an environment (Granqvist et al., 2017). GPs can play a contributing role to a child's environment. However, there are many other professionals involved who can help shape a supportive environment, e.g. social workers, healthcare workers, school workers, etc. (NICE, 2015).

Attachment disorders
Attachment disorders denote two very specific and rare forms of diagnosable mental disorder identified by the ICD-10 and DSM-5. The ICD-10 terms them 'reactive' and 'disinhibited', whereas the DSM-5 terms them as being either a 'reactive' or InnovAiT a 'disinhibited' social engagement disorder. They can only be diagnosed if symptoms have started after the developmental age of 9 months and before 5 years of age. Though both fall under the label of 'attachment difficulties' in NICE (2015), a critical difference is that insecure attachment is relationshipspecific, whereas attachment disorders are not (Van Ijzendoorn and de Wolff, 1997). By contrast, attachment disorders necessarily characterise behaviour shown pervasively across care-giving relationships. GPs will seldom come across attachment disorders as these are rarely expected in the general population (Zeanah and Smyke, 2009). The reason why they require discussion is largely to distinguish them from insecure attachment, since both are included under the label of 'attachment difficulties' (NICE, 2015).

Reactive attachment disorder
A child with reactive attachment disorder (RAD) typically shows a persistent lack of care-seeking behaviour towards any caregivers, even when alarmed, sick or in distress. For example, the child may fail to seek out caregivers while obviously in pain or upset, and ward off support when offered. This may be especially notable at partings and reunions with the caregiver (American Psychiatric Association (APA), 2013; NICE, 2015).

Disinhibited attachment disorder
It is expected and understood as developmentally appropriate that children show preferences between adults from whom they seek support when they are distressed. Disinhibited attachment disorder (DAD) is characterised by strong overfamiliarity towards unacquainted adults, out of keeping with age-appropriate boundaries or cultural norms. Some of the child's behaviour may vary with age, however, it would be characteristic of disorder if the underlying non-selective attachment behaviour persists. For instance, the child could venture off with a stranger without checking with caregivers in a new place (APA, 2013; NICE, 2015; World Health Organization (WHO), 1992).
The DSM-5 differs from ICD-10's label of disinhibited disorder, naming it 'disinhibited social engagement disorder', which raises the question of whether it is a disorder of attachment or of social engagement (Zeanah et al., 2016). However, both systems agree that severely 'insufficient care-giving' is distinctive for its aetiology (APA, 2013;NICE, 2015;WHO, 1992).

Attachment among children in care
Among looked after children in the UK, 2.5% show pervasive symptoms of attachment disorders, with a further 18% showing some symptoms (Meltzer et al., 2002). Most children in care have experienced inconsistent or unstable care. This will likely have hindered their opportunities to develop secure attachments: many may therefore show attachment difficulties, even if only a small minority show symptoms of attachment disorder (see Table 1). Factors that make insecure attachment especially likely in this population include exposure to abuse or neglect prior to entering care, changes of

DAD
Over-familiarity toward unacquainted adults; may wander off with strangers.
InnovAiT placements, or waiting longer for adoption. Factors that increase the likelihood of secure attachment to a caregiver include early adoption, supportive and responsive care, and stability of care-giving environment (NICE, 2015). Adults considering fostering will be encouraged to receive training in sensitive care and effective non-coercive discipline prior to having a child placed in their care so as to encourage stability and longevity of placements.
Looked after children have rates of presentation to their GP that are similar to those of the general population; however, they have additional mental health needs (Meltzer et al., 2003). Almost half of children in care meet the criteria for a psychiatric disorder, whereas only one-in-ten children who are not in care meet the criteria (Carmichael et al., 2016). In 2002, 45% of 5-17-year-olds looked after by the local authority were found to have mental health disorders: 37% showed conduct disorders, 12% emotional disorders and 7% hyperactive disorders (Meltzer et al., 2003). The House of Commons has advised that these mental health statistics be updated and published in 2018 (Carmichael et al., 2016).

Assessing attachment difficulties
Box 1 is adapted from NICE (2015) and lists areas to focus on during a comprehensive assessment for attachment difficulties. This assessment can be carried out by social workers, key workers, clinicians and others who work with children and young people on the edge of care.
GPs may meet different members of the same family and may then assess the complex social background, while maintaining confidentiality between patients. The first consultation with a child may be at the 6-8 week post-natal check and, though achieving continuity of care is generally very challenging, when possible it can improve understanding of complex cases. Careful history taking should be documented with descriptive language and the examination could include assessing the behaviour while interacting with the caregiver and others. Pooling information from multiple sources (e.g. the school, the social worker) will aid this process.
Simple advice can be offered by GPs that can help caregivers to notice and respond to their children's non-verbal and verbal signals of alarm, sickness or distress. With young children this may include sharing emotions, turn-taking, mirroring the infant's emotions, helping the infant with strong emotions, and recognising the infant as an individual whose signals matter (NHSEFS, 2016). By contrast, the clinician may be concerned if the infant never seems to look at or communicate with their parent; this behaviour can have a variety of causes, but one may be that the child has learnt that communications are unwelcome. The clinician may also be concerned if the infant looks alarmed or distressed by their own parent. Seeking support, supervision and debrief is important for GPs where such issues arise during assessment.
If the comprehensive assessment flags up concerns about issues that are impacting the child's wellbeing and development, a referral is appropriate. There are no clearly defined red flags in the guidance from NICE (2015) so clinical acumen is important in analysing individual cases. Depending on the local Clinical Commissioning Group the referral may be sent to Child and Adolescent Mental Health Services (CAMHS), the community paediatricians or social services. In some areas, CAMHS will only accept referrals if there is a mental health concern, and community paediatrics may need to review first to rule out developmental disorders (NICE, 2015).
The majority of children in care or on the edge of care do not require a referral to secondary care: GPs will see many families where children experience attachment difficulties of varying degrees of intensity, mostly insecure forms of attachment. However, insecure attachment is not a diagnosis, or in itself a form of pathology. It also requires a validated psychological assessment for confirmation.
The House of Commons Education Committee noted challenges faced with CAMHS referrals. Some children who were referred to CAMHS who did not meet their criteria were turned away. The report noted 'the young person receives a message saying you are not a priority', which could affect their selfesteem (Carmichael et al., 2016). In some areas CAMHS do not offer treatment if a looked after child does not have the stability of a permanent placement. In one case a young person had waited 30 months for CAMHS access due to lack of stability and work is being done to improve this situation. The review also stressed the importance of a multi-agency approach, to relieve some of the burden from CAMHS, which in some areas has had budget cuts (Carmichael et al., 2016).
Children who are struggling, but are not eligible for referral, could benefit from a multi-disciplinary approach, depending on their age, e.g. involving the school and health visitor, or for young infants the community midwife, or a non-governmental organisation (NGO). Health visitors and children's social care settings may triage the case and could potentially assign a health visitor or social worker to visit in the community, as Box 1. NICE advice.
NICE advises that a comprehensive assessment for attachment difficulties in children and young people, in all health and social care settings includes: . Personal factors, e.g. child's relationships InnovAiT well as offering parenting and behavioural interventions. The school can provide a safe space for the child if they are distressed and also assign a key worker as first point of contact for the child. NICE also advises that the school has a role in reducing exclusion and monitoring for absences (NICE, 2015). Some NGOs, such as the Coram charity can offer parental interventions and creative therapies for children. Children and their families who are found to be unsuitable for referral can be monitored in primary care, with escalation as appropriate if any new concerns about physical or mental health arise. If the caregiver of the child is a patient of the GP, he/she could review the caregiver's mental and physical health. If not they could advise the caregiver to see their own GP or seek support from one of the adoption services for looked after children.
With many different professions assisting a child in care or on the edge of care, the NICE guideline stresses the importance of a case management system to coordinate care and treatment (NICE, 2015). Ideally this would be led by the same key worker, although in practice continuity is not always achieved.

Differential diagnoses
In general, a key difference between attachment disorders and the following differentials is the history of grossly inadequate and/or unstable care. The prevalence and incidence of autistic spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD) is higher than attachment disorders and are more likely diagnoses. However, the clinical reality is that children can present with an unclear developmental history and symptoms that overlap, as well as children with ASD and ADHD presenting with co-existent inadequate or unstable care. For example, children with ASD may show significant levels of restricted and repetitive behaviour, as can be found in attachment difficulties and RAD (Davidson et al., 2015;Sadiq et al., 2012). Children with ADHD or DAD may have difficulties understanding social boundaries and may show indiscriminate behaviours towards strangers/unfamiliar adults (APA, 2013;WHO, 1992). This ambiguity may leave GPs unsure when and where to refer the child.
In answer to these concerns, any suggested neurodevelopmental disorder such as ASD/ADHD will have to be excluded first before a diagnosis of attachment difficulties or disorders could be confirmed by a specialist. The ASD/ADHD pathway often requires a secondary referral to community paediatrics. Most of the referrals come via schools, and this helps provide the paediatrician with information about behaviour in several settings, which helps with diagnosis. The GP can also refer, although further information may need to be collected from the school. The community paediatricians can organise an outpatient clinic appointment to assess the child and their caregiver. These assessments could include the use of resilience or vulnerability matrices to help predict and improve protective factors for children (Daniel and Wassell, 2002) or the Coventry Grid Interview to look at the overlap between ASD and attachment difficulties (Flackhill et al., 2017).
Ultimately the diagnosis is beyond the scope of general practice, but having some knowledge of the differentials will improve referrals and confidence in not referring. If concerned and unsure about the referral process, which is understandable as these cases are often very complex, helpful telephone advice for GPs can be gained from the community paediatrician, CAMHS or social services. In secondary care mental health professionals will use observational procedures or clinical interviews to help make the differentiation (Giltaij et al., 2017;McKenzie, 2017). It can be important for secondary care to differentiate between the diagnoses, not only to access the appropriate intervention, but also, potentially, to alter the prognosis. For example, children with RAD may show symptomatic improvement in a new environment, whereas language impairment and stereotypic behaviour may persist in children with ASD.

Safeguarding and child protection concerns
Safeguarding is the action taken to promote the welfare of children and protect them from harm such as abuse or maltreatment. This includes neglect, which is the ongoing failure to meet a child's physical or psychological needs likely to result in the serious impairment of the child's health or development. Any concerns regarding neglect, will require a safeguarding referral. Reviewing the General Medical Council (GMC)'s guidance can assist making a decision about safeguarding (GMC, 2018). In any child presenting with behavioural concerns, safeguarding should be considered as part of the diagnostic work-up. This is essential in children presenting with possible attachment difficulties, especially when considering the impact of caregiver(s) responses to the needs of the child.
It is important to note that attachment difficulties do not imply that maltreatment has occurred or that the caregiver is 'to blame'. Caregiver's may have a mental, physical or social situation affecting their ability to respond to the child's alarm or distress and unrelated to maltreatment. During discussions with caregiver(s) and the child it may be helpful to discuss any concerns of 'blame'.

Interventions for attachment difficulties
If the assessment in secondary care identifies attachment difficulties in children and young people in care or on the edge of care, NICE recommends specific interventions dependent on the child's age (NICE, 2015). The type of intervention also varies depending on the history of any maltreatment. Many of the evidence-based interventions focus on improving parental sensitivity and require tailoring to the individual child or young person's needs.
Pre-school children with or at risk of attachment difficulties can benefit from video feedback with caregivers, which will be facilitated by a trained health or social worker. This involves a video recording, often in the caregiver's home, consisting of around 10 sessions lasting 60 minutes over 3-4 months. The feedback encourages parental reflection on their sensitivity to their child's distress. It helps the caregiver to become more InnovAiT aware of the child's signals, understand their child's behaviour, and respond positively to the child's cues and how to regulate their emotions when interacting with the child. Video feedback has been shown to increase carer stability and longevity for placements for children in care (NICE, 2015). Some parents do not agree to participate in video interventions; NICE recommends multi-agency review before progressing to other interventions. Another intervention is a home visiting programme, which assesses the child-parent relationship without the use of video. It consists of 12 weekly or monthly sessions over an 18-month period. (NICE, 2015).
If a pre-school-aged child has been or is at risk of being maltreated, parent-child psychotherapy can be considered, although safeguarding concerns must have been addressed first. It involves weekly 45-60 minute sessions over the course of a year, delivered in the parent's home, provided by a therapist trained in the intervention. The intervention explores the relationship between the emotional reactions of the parents, the child's perspective and the parent's own childhood experiences (NICE, 2015).
For primary and secondary school-age children, parental sensitivity and behaviour training for parents can be considered. If a child in this age group has been maltreated or is at risk, the child may show signs of trauma or post-traumatic stress disorder. The NICE guidance on post-traumatic stress disorder should then be followed as well. In the case of a primary school-aged child with a foster carer, special guardian or adoptive parent, intensive training should be offered before and after the placement starts with group therapeutic play sessions. Particularly challenging moments for children in this age group may be 'physical and sexual development, adolescence transition, and re-awakening emotions about their birth parents or family origins' (NICE, 2015). All professionals working with children in care or on the edge of care should ensure discussions about contact with birth parents or original family are sensitive and empathetic (NICE, 2015). If the child or young person is in residential care then it is the professional carer who will participate in the intervention.

Interventions for attachment disorders
The available evidence for interventions for attachment disorders is limited. RAD and DAD have thus far focused on adoption (from institutions) into families, with enhanced care-giving thought to counter the deprivation that gave rise to the attachment disorder (Zeanah et al., 2016). Indeed, it has been suggested that RAD and DAD should be treated with provision of the same type of care-giving that fosters secure attachment in children without an attachment disorder. Signs of RAD tend to diminish substantially, and disappear in most cases, following placements into families. However, findings regarding DAD are more mixed, with some studies suggesting symptoms can persist for many years, despite adoption and provision of adequate care.

KEY POINTS
. Attachment difficulties express a child's expectation that they cannot be sure of their caregiver's sensitive responses to their experiences of alarm, sickness or distress . Attachment difficulties constitute a common modifiable risk factor for mental health problems but should not be pathologised or treated pharmacologically . Attachment disorder is a rare mental health disorder associated with a history of grossly inadequate and/or unstable care and is unusual even among looked after children . Only in extreme cases of attachment disorder do attachment difficulties constitute a medically diagnosable mental illness . Attachment difficulties are likely to be reduced by supportive family environments, and interventions by mental health and social welfare practitioners to facilitate such support . Clinicians can face difficulties in distinguishing attachment disorder and difficulties from other conditions such as ASD and ADHD as symptoms overlap Funding