Residential Childcare Workers’ Knowledge of Reactive Attachment Disorder



Background:  Reactive Attachment Disorder (RAD) is associated with a history of ‘pathogenic care’ therefore residential childcare workers are likely to come into contact with children with this disorder. An ‘appropriately supportive environment’ may be a mitigating factor in terms of the disorder’s severity and duration (Corbin, 2007); therefore it is important that behaviours suggestive of RAD are recognised early and that these children are given the attention they require in the care setting. We aimed to gain an understanding of residential childcare workers’ knowledge of RAD to determine if their understanding of RAD was sufficient to recognise RAD-like behaviour.

Method:  A mixed methods study (qualitative focus groups and whole population survey) was undertaken with residential childcare workers.

Results:  Liaison with Child and Adolescent Mental Health Services was seen by residential workers as being useful and workers thought their knowledge of mental health problems had improved in recent years. However, less than half of respondents (49.1%) identified RAD from a case history given in the questionnaire.

Conclusions:  Child and Adolescent Mental Health Services staff and residential workers may need more specialised training on RAD to be able to recognise behaviours suggestive of the disorder and refer appropriately, ensuring the child’s needs are met as early as possible.

Key Practitioner Message:

  •  Residential care workers are likely to come into contact with Reactive Attachment Disorder in their work
  •  Residential care workers view a consultative relationship with Child and Adolescent Mental Health service workers as highly supportive
  •  Although these residential workers had a good understanding of attachment theory and were also familiar with some of the symptoms of RAD, less than half of the sample recognised these symptoms as belonging to that diagnostic category
  •  It is important that children with RAD are recognised and referred to CAMHS, despite there being a poor evidence-base for treatment, because a full understanding of their complex problems is likely to lead to better management


Reactive Attachment Disorder (RAD) is characterised by persistent abnormalities in social relationships (World Health Organisation, 2007) and is associated with ‘pathogenic care’ i.e. severe parental neglect, abuse, or mishandling (American Psychiatric Association, 1994). The DSM-IV classification describes two subtypes: inhibited RAD characterised by excessively inhibited, hypervigilant behaviours; and disinhibited RAD, characterised by indiscriminate friendliness (American Psychiatric Association., 1994).

Attachment theory

Attachment theory (Bowlby, 1982), is based on the concept that infants are born with the tendency to display certain attachment behaviours that serve to increase the child’s proximity to a specified attachment figure. These attachment behaviours are activated in situations of distress or insecurity, and act as a mechanism of survival. When all goes well, an attachment figure, usually the child’s primary caregiver, provides a sense of comfort and a secure base from which the child can explore an unfamiliar and strange environment. The emotional bond formed between an infant and their carer guides feelings, core beliefs and behaviours towards themselves and others, including beliefs about their worthiness of love and care. The ability to form attachments is thought to be important in the development of a child’s self-esteem, self-efficacy and their ability to cope in stressful situations (Weinfield, Sroufe, & Egeland, 2000). In adulthood, this extends to their ability to form and the quality of interpersonal relationships, regulation of affect and their own parenting style (Sroufe, 2005). The strength and security of the parent-infant attachment, referring to the child’s confidence in the attachment figure being available and responsive when needed, is to a certain extent dependent on the caregivers’ parenting style being one of consistency, sensitivity and emotional availability (Juffer, Bakermans-Kranenburg, & van Ijzendoorn, 2005).

Attachment patterns and RAD

There has been little research on RAD but, in contrast, there has been much research about attachment patterns. It is thought that approximately 60% of infant-parent pairs are described as ‘securely attached’ and approximately 40% as ‘insecurely attached’ (Ainsworth et al., 1978). Insecure attachment is not a disorder, but can be conceptualised as a (frequently adaptive) way of dealing with stress (Crittenden, 1999). Four attachment patterns have been identified: secure, anxious-ambivalent, anxious-avoidant, and disorganised. Some authors have suggested that disorganised attachment, in which the infant displays a mixture of approach and avoidance behaviours with the caregiver in stressful situations, might be approaching psychopathology in its own right (van Ijzendoorn & Bakermans-Kranenburg, 2003). Disorganised attachment has been shown to be robustly associated with conduct problems in middle childhood (van Ijzendoorn, Schuengel, & Bakermans-Kranenburg, 1999), yet the strength of the association is modest and many children who have disorganised attachments do not develop difficulties (Minnis et al., 2006b; Green & Goldwyn, 2002). The issues surrounding the aetiology and course of RAD, and its relationship with the research-based category of Disorganised Attachment, are very complex and do not form the focus of our paper; however they are well explored in other papers on the topic (Green & Goldwyn 2002).

Attachment theory does not fully account for the features of RAD e.g. it is possible for a child to display a secure attachment pattern with the parent, yet display symptoms of reactive attachment disorder (Minnis et al., 2008; Zeanah et al., 2004). Conversely, a child can be insecurely attached but not have RAD. The use of the word ‘attachment’ in the title of RAD has caused confusion and in fact plays a big part in the problems with the diagnosis of RAD, since there is the presumption of insecure attachment as the core aetiological factor. Conversely, RAD might be better conceptualised as an impairment of social functioning more broadly (Minnis et al., 2006b; Green, 2003) and, hence, as including symptoms (such as indiscriminate friendliness and hypervigilance) that are not described within the literature on insecure attachment.

Childcare workers and RAD

Childcare workers often express concerns about the difficulties of forming and maintaining relationships with children with RAD-like behaviours (Howe & Fearnley, 2003). The social and emotional problems associated with RAD may persist as the child grows older (Rutter et al., 2007) with possible life-course progression to social exclusion and personality disorder (Minnis, 2003) and there is evidence to suggest that the failure of frontline care workers to accurately identify psycho-social problems will lead to vulnerabilities in later life (Lacey, 1999). Conversely, there have been concerns about possibly adverse effects of labelling children with RAD, but various authors have argued that good assessment and understanding of RAD-behaviours and their associated difficulties is essential if appropriate management strategies are to be put in place (Byrne, 2003; O’Connor & Zeanah, 2003). An ‘appropriately supportive environment’ may be a mitigating factor in terms of the disorder’s severity and duration (Corbin, 2007) and so it is important that behaviours suggestive of RAD are recognised early and that these children are given the attention they require in the care setting.

Often those working in close contact with children (e.g. teachers, youth workers, residential childcare workers) feel burdened by the scope of work they are expected to undertake in supporting the complex needs of the children in their care, particularly in relation to mental health difficulties (Clarke, Coombs, & Walton, 2003). They may lack confidence in this area of work as they have little or no baseline mental health training (Rawlinson & Williams, 2000) and they consequently feel overwhelmed and unsupported. Multiple levels of support are essential for their development in this area, ranging from their own baseline training needs to the development of referral pathways and support networks with Child and Adolescent Mental Health Services (CAMHS) (Clarke et al., 2003). In providing improved training and supervision for frontline workers who have direct contact with children, the likelihood of identifying children and adolescents in crisis is enhanced, as is their access to specialist services when necessary (Clarke et al., 2003). There have been many recent initiatives developed across the UK to improve frontline non-specialist staff’s knowledge and skills in recognising and managing child mental health problems (Health Advisory Service, 1995). In Scotland the Scottish Government’s aim is that by the year 2008 ‘all staff directly or indirectly involved in crisis prevention, response or resolution should have appropriate (and ongoing) training and skills development’ (Scottish Executive, 2006).

To summarise, given that psychopathology is significantly more prevalent among children in the care system (McCann et al., 1996; Meltzer et al., 2002), that virtually all have a history of maltreatment (Minnis et al., 2001), and that maltreatment is thought to be the key aetiological factor in RAD, residential care workers are likely to come into contact with children with this disorder. As no previous studies have examined the knowledge of residential care staff regarding RAD we aimed to answer the following questions:

  •  What are children’s residential workers levels of knowledge of Reactive Attachment Disorder?
  •  What education or training have they had in this area?
  •  What has facilitated developing knowledge?
  •  What barriers are there to developing knowledge?
  •  To what extent does age, length of experience or educational background impact on levels of knowledge?


This was a mixed methods study using qualitative and quantitative research methods (focus groups, interviews, and a whole population survey) and examined the views of a sample of residential childcare workers in a large Scottish city. Utilising methods from differing research paradigms is a recognised way of exploring research questions fully and triangulating findings from each enhances the credibility and authority of the findings (Johnson & Onwuegbuzle, 2004).

The study was approved by the University of Glasgow Ethics Committee and by the local social work department approval system. Data collection took place between December 2007 and March 2008.

Phase one: focus groups

A meeting of all Unit Managers in the City (= 16) was held and discussion took place as to the best way of hearing the views of the widest possible range of residential childcare workers within the timeframe of the study. All agreed that we meet with the range of workers described in Table 1 and four Unit Managers (from the total of 16), representing the various types of units (see Table 1) volunteered to help set up the relevant groups. Four focus groups were facilitated involving a purposive sample of 30 residential childcare workers out of the total population of 265 permanent staff. The focus group participants were from various types of residential units across the Scottish city (see Table 1). The selection of participants for the focus groups ensured that we heard the views of a wide range of residential childcare workers.

Table 1.   Breakdown of focus group participants
Focus groupDateMaleFemaleStaff typeUnit type
120/12/0724MixedYoung people’s centre
217/01/0817Main grade staffHybrid group from across the different units
317/01/0845MixedSecure unit previously involved in pilot project on attachment
428/01/0825Managerial staffHybrid group from across the different units

Potential group participants were approached by one of four unit managers who had been identified at the unit managers’ workshop regarding the study and asked if they would like to be involved. Consent forms and information sheets were sent to all 30 participants in advance of the groups, giving an outline of the study and answering questions about taking part. Written consent was obtained in person from all at the time of the focus group. Focus groups took place within residential units at staff meetings over a one and a half hour period and were facilitated by two researchers. A topic guide was generated in keeping with the research questions of the study and was used alongside a presentation of a case vignette depicting a child with symptoms of RAD to prompt free discussion within the group. Groups were audio recorded using a digital mini-disc recorder and microphone and transcribed fully by a research secretary.

Phase two: whole population postal survey

A questionnaire was developed using the information gained from the qualitative exploration of the subject in phase one. In an area such as this, about which virtually nothing is known, this is a robust method to determine which quantitative questions to ask (Creswell, 2009). Developing a questionnaire from qualitative findings is a recognised type of mixed methodology that we have used in a previous published study (Blower et al., 2006). The questionnaire focused on three of the main research questions: levels of knowledge of RAD, education and training, and facilitation of knowledge. The questionnaire consisted of six sections covering 32 items. A section on information about the respondents was included to gather details about age, sex, length of employment in residential childcare, and qualifications. The questionnaire included a brief case history depicting a child with symptoms of RAD and asked the respondents to select from a list of clinical issues that may explain the behaviour of the child. They were then asked to rank features from the case history in order of importance (1 being highest) in helping them reach a conclusion in previous section. ‘Abuse and neglect’ and ‘multiple placements’ were included because evidence of ‘severe parental neglect, abuse or mishandling’ is a diagnostic requirement for RAD. To measure the importance placed on various factors affecting the participant’s knowledge, they were asked to mark a cross on a 7cm visual analogue scale from ‘not important’ to ‘very important’. The results were later transformed into a percentile score. The questionnaire was piloted with 10 medical students, and was found to be user friendly and easy to complete. Unit managers were phoned to establish the sample size. The questionnaire was then sent to all permanent staff in the 16 residential units (= 265) with a stamped addressed envelope. A reminder telephone-call was made after 2 weeks.

Data analysis: focus groups

Typed transcripts were analysed independently by three members of the research team (LF, MF and MM) using a process of thematic content analysis (Burnard, 1991), which aims to produce a detailed and systematic recording of the themes addressed in the focus group discussion and to link these themes under an exhaustive category system. Each evaluator identified emerging themes from the narrative data and reduced these to core themes reflecting the participants’ experiences.

Following independent data analysis, the evaluation team met as a group to compare and contrast interpretations of the narrative accounts and develop a common view on the central themes identified from the narrative data and the main sub-themes within these.

Data analysis: survey

Statistical analysis was undertaken using SPSS for Windows, version 15.0. Descriptive statistics and frequency analysis were used to explore levels of knowledge and factors that improved knowledge of RAD. Logistic regression analysis was undertaken to determine which variables might affect levels of knowledge of RAD. Each variable was added to the model sequentially, to determine if any were acting as confounders.


Focus groups

The subject of RAD was not spontaneously raised by participants in any of the focus groups. Four main themes were identified, relating to the research questions regarding the facilitation of and barriers to developing knowledge.

Theme one: relationship with CAMHS.  Participants found working alongside CAMHS beneficial in their working practice. A number of sub-themes supported this finding.

Consultation and joint working.  Direct consultation with CAMHS resulted in a diagnosis being made in difficult cases and enabled better access to information. It helped develop staff knowledge of mental health, their understanding of behaviours and helped them develop strategies to help the young people.

…interactions with (CAMHS) usually end up with an increase of knowledge

…to have a diagnosis and have information that can help with understanding…to use that information to develop strategies…to deal with it in the most appropriate manner

One participant commented that one of the main benefits of working with CAMHS was simply to be reassured that they were doing a good job with regards to their work with the young people.

…when we were doing team consultations… a lot of the time the team just want a bit of validation…that what you are doing is right, ‘that’s good what you are doing, keep on doing it, don’t under-estimate anything’. I think…a lot of the time that was actually more helpful than anything.

Theme two: reflective practice.  The opportunity and capacity to reflect on practice appeared to be an important factor for knowledge development.

Recognising symptoms.  Reflecting on everyday experience and knowledge acquired over many years working with young people, the residential childcare workers learn to recognise patterns of behaviour and symptoms. Familiarity of cases allows them to build on their knowledge and link theory to practice when a new case arises. When presented with the case vignette of a typical child displaying symptoms of RAD, many residential childcare workers commented that the behaviours were regularly witnessed in the units they worked in.

I think we see a lot of things all the time, it is just identifying it…label I suppose, put pieces together and form a bigger picture, because individually it is one piece of behaviour but you gradually build a picture up and it is glaringly obvious after while.

Seeking help.  An important factor in furthering knowledge and enhancing learning was recognizing individual limits and having the self awareness to seek assistance.

You cannot fix everything…you have to be realistic in what you can do… you have to know when it is time to talk to somebody else

…acknowledge that young person will need specialist help somewhere down the line.

Theme three: organisational culture

The focus groups highlighted the importance of working in an environment supportive of thoughtfulness.

Planning, clear criteria and goals.  The participants felt that, previously, residential childcare had been chaotic at times, with rapid turnover of residents. Procedures have now changed, with the system having a more planned approach to care resulting in an improved working culture. Careful planning prior to the young person entering the unit allowed for background information to be reviewed and for staff to be prepared for the role they will play in the child’s care.

…prior to coming in the exit plan should already be made and we should know where the kids are going…we tend to plan well ahead.

…because there is a referral system in place I think there is much more…bringing together information and planning.

However, despite a progression towards planned admissions, sometimes there is still a lack of continuity of care and young people arrive at the unit with no information.

Young people arrive with no information…so…you start building up information and that can take several days.

Team based support.  The residential childcare workers highlighted the benefits of having a working environment which encouraged team support, supervision and discussion. Discussing cases with colleagues and working through any difficulties together was something that the individuals involved found important in furthering their knowledge.

We discuss their behaviour obviously everyday at change over…talking about what things have happened…looking for why that might be happening

…supervision as well…you don’t realise how much you know until you get into a conversation with someone else who is struggling…you can help them with it in supervision or you can point them in the right direction

Staff were encouraged to take opportunities while at work to enhance their learning if a topic came up that they wanted to find out more about. The working environment allowed for this.

There is time every shift that you can take protected time as long as it is booked in advance and you are not stepping on anyone else’s toes.

… shifts where people can actually then have time to reflect and balance and look over things it is very important.

Theme four: formal and informal training.  Training was important to the participants in supporting the development of their knowledge for everyday practice.

Basic training, knowledge and experience.  The basic training that is now required to be a residential childcare worker was an important factor in the acquisition of baseline knowledge of mental health and behavioural disorders.

…there shouldn’t be anybody working here who doesn’t have at least an HNC level and SVQ level of knowledge so it is only a question of building on that, tweaking it or developing areas of interest

…the majority of people now have qualifications. Part of the courses that you would do attachment, psychology…It is all in there, there is a lot of theory, plus people have done their SVQ so there is a lot of theory, like knowledge base

The subject of on-going training was a recurring topic of discussion, with residential childcare workers highlighting its importance in an ever-changing working environment. Courses helped to gain more specific knowledge of particular aspects of mental health, and to keep their knowledge up-to-date.

…training courses come up at different times for various people…and that includes some of the reasons for behaviour, attachment theory and mental health difficulties.

I think training is ongoing…because the clients change all the time

Responses were overwhelmingly positive about the benefits of training: courses helping them to link theory to practice.

…what the extra training does it helps with reinforcement of your own understanding and it helps to bring your instinctive practice more into your awareness.

…working on a particularly challenging behaviour it can be very difficult to see past that behaviour to what is actually behind it…with all the work we have done on attachment it has helped us to better understand.

Some participants felt that training on specific disorders was vitally important for the recognition of symptoms and for appropriate referrals to CAMHS to be made.

…you would probably never identify it and therefore refer it on to anyone else that could make the diagnosis because all you see is the challenging behaviour but if you don’t have the knowledge to say ‘oh gee you know that’s Asperger’s or autism’…So…familiarisation training or specialist knowledge training how to recognise the signs and symptoms of the varying mental health difficulties…that’s where it would help us to then to refer on to specialist resources.

Making further training available to all was a difficult task, providing a barrier to furthering knowledge. The reasons for this were budget restraints, availability of popular courses and covering staff shifts.

it is down to resources…there is too many people wanting to go on these courses or there just isn’t the money about for courses at this time

Staff sharing learning.  Participants discussed ways that information can become available to them at work. Staff share learning and knowledge with each other by distributing resources for learning on specific topics of their interest.

certain members of staff could be allocated a kid and through that relationship they might become a little bit specialised in that (disorder)…and bring that back to the staff team …information slowly filters informally.

Survey findings

A questionnaire was developed following the focus groups interviews and was aimed primarily at answering the three research questions regarding levels of knowledge of RAD, education or training have they had in the area and factors impacting on levels of knowledge.

Description of sample

One hundred and eight questionnaires were returned out of 265 sent, giving a response rate of 41%. The mean age of residential childcare workers in our sample was 40.4 years (range 22 to 63). Thirty-six percent were male and 64% female. The mean years working in residential childcare was 10.3 years (SD 8.3). The most common qualification attained was a Higher National Certificate (38.9%), while 22.2% had a university degree and 9.3% a postgraduate degree. Of the sample, 9.3% also had a specific social work qualification, while 5.6% reported having no qualifications, We were unable to compare data on responders with that of non-responders since, due to data protection concerns, there was no means of gathering information on the characteristics of the survey sample as a whole.

Main findings

Knowledge of RAD

Ninety-eight percent of respondents identified ‘attachment issues’ from the case history. Slightly less than half (49.1%) identified Reactive Attachment Disorder (Fig. 1).

Figure 1.

 Selection of mental health issues from case history by respondents

Twenty three percent ranked the core features of RAD (abuse and neglect, multiple placements in early life, overfriendliness and freezing) as most important in the case history. Sixty-two percent of respondents ranked abuse and neglect as the number one factor from the case history for importance, while 18.5% thought multiple placements in early life was most significant.

Logistic regression was undertaken to determine any patterns relating respondent characteristics to recognition of RAD, however, there were no significant associations between age, length of experience or educational background and identification of RAD from the case history.

The type of residential unit (secure/open) was cross-tabulated with whether or not participants selected RAD from the case history and Pearson’s Chi-square test revealed no significant difference between groups (59% secure versus 48% open, = .403).

Education and training

The respondents gave day-to-day experience the highest average score (84/100) for factors that helped their thinking about mental health. Their training and qualifications scored second highest, with an average score of 79/100. Involvement with CAMHS and discussion with colleagues both scored 78/100.

Only 18% of those who returned the questionnaire felt that they had had enough training on RAD, which scored the lowest of all the topics listed (Figure. 2).

Figure 2.

 Perceived adequacy of training on mental health issue


The overall aim of the study was to gain an understanding of residential workers’ knowledge of Reactive Attachment Disorder and to suggest which elements of their work affect their acquisition of knowledge.

Clearly, it would not be helpful for residential workers to be making the diagnosis of RAD themselves, especially as RAD is a complex disorder with a high likelihood of comorbidity (Minnis, Everett, et al. 2006), but an ability to refer children displaying RAD-like behaviours to the appropriate services is likely to be beneficial. While labelling and the associated stigma is a concern for all child psychiatric diagnoses, it has been argued by various authors that good understanding of RAD-behaviours and their associated difficulties is essential if appropriate management strategies are to be put in place (Byrne, 2003; Minnis, Marwick et al., 2006). There are no well evidenced treatments specifically for RAD, but recognising behaviours suggestive of RAD may give these very vulnerable children entry into services and a rounded assessment may be a crucial aid to management of their difficulties. In addition, development of treatments cannot proceed unless the target population are recognised.

Knowledge of RAD

This study suggests that although participants clearly recognised the pattern of behaviours described in the case study, the diagnosis of RAD is not well known to residential childcare workers, with less than 50% of our survey sample identifying it from the case history and no-one spontaneously mentioning the disorder in the focus groups. Attachment issues were, however, readily identified from the survey case history. This compared well with the finding from the focus groups that the participants have been involved in recent training on attachment and had an attachment theory focus to their practice. This suggests that the residential childcare workers are learning about attachment and recognise its importance in child and adolescent development, but that they are less clear about the symptom profile of RAD.

By being familiar with ‘attachment issues’ or ‘attachment difficulties’ the staff in our study may be referring to attachment patterns and Ainsworth’s Classification of attachment behaviours in the Strange Situation (Ainsworth, 1979). However these are not the same as attachment disorder, which has clear diagnostic criteria as described in DSM-IV or ICD-10. For example, disorganised attachment, describes an insecure but selective attachment, whereas reactive attachment disorder can exist in the absence of selective attachment (Zeanah, Smyke, Koga et al., 2005). This is a very important difference and has possible implications for our understanding of the aetiology and prognosis of the disorder. Furthermore, disorganised attachment as classified by the Strange Situation may imply a relatively brief and context specific disturbance, whereas reactive attachment disorder implies a more severe and pervasive disturbance, with potentially negative long-term outcomes. (O’Connor et al., 2003).

RAD is one of only two psychiatric disorders (the other being Post Traumatic Stress Disorder) that use aetiology as a diagnostic requirement: the DSM-IV diagnosis of RAD requires a history of ‘pathogenic care’ (American Psychiatric Association., 1994). This is interesting in that a high proportion of our sample selected ‘abuse and neglect’ and ‘multiple placements’ as being important issues from the case history, demonstrating that they are considering the aetiological factors that may contribute to this mental health disorder; however they are perhaps not yet making the link between these aetiological factors and the specific symptoms of RAD.

Residential workers’ lack of consideration of RAD may, at least in part, reflect the differing nomenclature used in its classification and the controversy surrounding its exact nosological boundaries (Zilberstein, 2006; Chaffin et al., 2006). Some authors have also recently suggested that RAD, PTSD and other outcomes related to abuse and neglect may fall under the rubric of Complex Trauma or Developmental Trauma Disorder, which is also thought to interfere with attachment formation between child and caregiver and interferes with neurobiological development, including emotional and cognitive development (Cook et al., 2005). The term complex trauma describes the experience of multiple and/or chronic, developmentally adverse traumatic events, especially those of an interpersonal nature occurring within the care-giving system such as abuse and neglect, and the impact of this early-life exposure on immediate and long-term outcomes (Cook et al., 2005). Our findings, and those of other authors, suggest that there is the need for greater consensus on the definition of RAD to improve understanding of the disorder and for it to be more easily identified in practice (Boris et al., 2004).

Why is it important that RAD is recognised?

It is not expected that residential workers would or should diagnose RAD, but simply be able to consider that a child’s behaviour may potentially be in the RAD domain and proceed to obtain the involvement of CAMHS. For example, because indiscriminate friendliness may be common in residential care and residential workers may have little contact with children from the general population, such behaviours could easily be understood as normal rather than regarded as symptoms of potentially serious psychopathology. Children with RAD symptoms use more services and engender greater costs to society even in childhood (Minnis et al., 2006a) and clinicians suggest that RAD may be the first step on the developmental trajectory towards conduct disorder, which has life-long implications for the individual, family and society (Scott et al., 2001). A better understanding of RAD behaviour might lead to the drawing up of guidelines for frontline staff in responding to RAD-related difficult behaviour, thus supporting residential care workers in their role and improving the day to day responses to children with RAD,

There are potentially very different routes of management for children who have attachment difficulties and those with RAD. Although there are now many interventions aimed at insecure attachment (Cooper et al., 2005; Bakermans-Kranenburg, 2003), no treatment method has been shown to be effective for children specifically with attachment disorder and, as yet, no studies have investigated the effectiveness of some of the more established attachment-based interventions with children with RAD (O’Connor and Zeanah, 2003).

The main concern with using such interventions with RAD children is that the theory underlying attachment-based treatment does not apply in the case of children with attachment disorder. These concerns arose after thinking about why, in an important minority of cases, children with attachment disorder who are placed in foster/adoptive care continue to exhibit a lack of selective attachment behaviour despite years of either optimal or ‘adequate’ care-giving (O’Connor and Zeanah, 2003).

The evidence base for treatments for RAD is weak and effective treatment strategies cannot be developed and disseminated until reliable methods of diagnosis exist (Minnis et al., 2006b).

Education and training

Front-line staff frequently encounter early manifestations of mental health need (Gale & Vostanis, 2003) therefore it is paramount that staff are trained accordingly. The factors that the sample identified from the survey as being most important in their learning triangulate with the main themes highlighted in the focus groups for acquisition of knowledge. Day-to-day experience, having thought-provoking discussion with colleagues and training are all thought to be important in helping the residential childcare workers reflect on and understand difficult cases. It is clear also that the consultation service offered by CAMHS is viewed as a highly beneficial resource for knowledge development and is effective in providing support for those on the frontline working with challenging children, supporting government drives to alter the structure of mental health services for children to ensure that direct support, supervision and guidance is given to non-specialist staff (Health Advisory Service, 1995). Despite participants placing much importance on the involvement of CAMHS and on having baseline knowledge of theory, our findings demonstrate that general consultation and general theory are not enough to help residential childcare workers recognise RAD. Improved training and supervision from CAMHS, providing residential childcare workers with a knowledge base and theoretical framework on which to structure thinking, should enable them to identify specific child mental health symptoms at an early stage without the unreasonable expectation that they attain expert diagnostic skills in the field (Gale et al., 2003).

Despite staff perception that they do have access to information and training on mental health, the participants of our focus groups recognised the need for more specialised training on specific disorders. This was reinforced by the findings of the survey. ‘Indiscriminate friendliness’ is the core feature of disinhibited RAD - the form strongly associated with institutional upbringing (O’Connor, Bredenkamp, & Rutter, 1999), but very few of our sample recognised ‘overfriendliness’ as being important from the case history.


The main limitation of the study is the low survey response rate of 41% which may introduce bias and information was not available to compare respondents with the whole population. It may be that those who responded had greater interest in the topic and hence greater knowledge, and therefore the levels of knowledge of RAD among residential workers may be lower than suggested by our results. On the other hand, it could also be that those who responded had less knowledge on the subject, and those with a greater knowledge did not respond because they saw nothing to gain; thus, the levels of knowledge of RAD among residential workers may be higher than suggested by our results.

A second issue is that of generalisability of the findings: qualitative data do not set out to provide generalisable findings and the fact that the quantitative data were collected only in one local authority may limit the degree to which findings can be related to practice in other areas. Some participants had also been involved in a pilot of training on attachment. It would be interesting, in future studies, to compare these findings with other services that are less attachment focused in their operation or to compare with residential childcare services that do not yet have a CAMHS consultation service available to them.


This sample of residential child care workers were familiar with attachment theory and were using it in their practice. Involvement from CAMHS was viewed as highly supportive, helping workers understand and refer on the children with whom they work. Residential workers recognise children with the constellation of behaviours described in DSM-IV as RAD, may conceptualise these problems as ‘attachment issues’ and may discuss such children with CAMHS colleagues. However, identification of RAD was low compared to their recognition of less specific issues and it is clear that both CAMHS workers and residential staff might require additional and more specialised training on RAD in order to develop a joint understanding of this disorder. Residential workers and the CAMHS workers who consult with them may have a key role in contributing to the ongoing debate about diagnostic criteria for RAD.


The authors would like to thank the residential childcare workers and unit managers involved in the study, and Irene O’Neill for her help with survey distribution and focus group transcription.