Abstract
Looked after children often experience greater mental health challenges than children in the general population. Consequently, there has been a call for greater focus on early preventive interventions and priority access to specialist mental health support for them. Brief mental health screening tools often provide the gateway to services and yet there is a lack of suitable assessment tools available for looked after children. The current study is the first to explore the perspectives of foster carers and clinicians working in Child and Adolescent Mental Health Services (CAMHS) in relation to the use of two brief screening tools: the Strengths and Difficulties Questionnaire (SDQ) and the Brief Assessment Checklists (BACs). Qualitative data were collected via semi-structured telephone interviews and face-to-face focus groups with 13 female foster carers and 19 clinicians working in a CAMHS setting. Thematic analyses generated six themes which were organised into three superordinate ones centred around: (1) relevance to looked after children; (2) using measures to inform care planning; and (3) current problems when using psychometric measures within the looked after children population. Foster carers and clinicians emphasised the need for a comprehensive understanding of these children given the complexity of their difficulties. There was acknowledgement that psychometric measures could facilitate and contribute to this by highlighting difficulties, providing accurate feedback to carers and clinicians, and enabling access to support. However, both groups seldom felt that current measures were nuanced enough to adequately capture the children’s needs. Yet screening tools play a crucial role in identifying the mental health needs of looked after children and facilitating access to services. Further research is needed to establish the ability of such measures to increase the sensitivity of their complex needs assessment. A number of clinical recommendations are also discussed in relation to the assessment of mental health in this population.
Introduction
The term looked after children is used in the UK to describe young people aged 0 to 18 years who are unable to live in their family home due to significant concerns about their safety and well-being. The reasons for this vary but in many cases, the child has been exposed to abuse, neglect or family violence. When a child is looked after there is a legal requirement to support him or her while living away from home (either temporarily or permanently) and the type of care provided can include foster care, kinship care, residential care or independent living.
In recent years, there has been greater focus on looked after children due to the increasing evidence that they are more likely to experience poor physical and psychosocial outcomes, such as homelessness, unemployment, teenage pregnancy and criminality (Courtney, et al., 2011; Department for Education, 2017) later in life. They are also reported to be up to four times more likely than the general child population to exhibit clinically significant mental health difficulties; a recent meta-analysis scrutinising studies from five western countries (Bronsard, et al., 2016) concluded that roughly half of such children met the criteria for a formal psychiatric diagnosis. Looked after children and care leavers are also at higher risk of self-harm as well as both attempted and completed suicide (Hjern, Vinnerljung and Lindblad, 2004; Vinnerljung, Hjern and Lindblad, 2006).
Alongside this, several studies have highlighted the long-term impact that early life trauma (e.g. maltreatment) has on brain development, physical well-being and mental health (D’Andrea, et al., 2012). Research has demonstrated that children who have been exposed to maltreatment are at greater risk for complex mental health presentations, often spanning several diagnostic criteria (D’Andrea, et al., 2012). In the neuroscience field, emerging evidence has uncovered physical changes in the brains of children who have experienced maltreatment that may indicate a hidden vulnerability to later mental health difficulties (McCrory, Gerin and Viding, 2017). In particular, the Adverse Childhood Experiences (ACEs) studies in the US and UK have greatly increased public awareness around the prevalence and impact of childhood trauma on an individual’s health (Anda, et al., 2005; Bellis, et al., 2016; Felitti, et al., 1998). Consequently, many western governments have called for services that consider the prevention and/or early treatment of ACEs.
Looked after children will have experienced several childhood traumas, including removal from their birth family, separation from their primary caregiver and in over 60% of cases, exposure to some form of abuse and neglect (Department for Education, 2017). In one US study, over 70% of those previously looked after reported five or more adverse childhood experiences (Bruskas and Tessin, 2013). Therefore, the growing evidence base regarding maltreatment must be considered within this population. To add to this complexity, there is also a high incidence of neurodevelopmental conditions in the looked after population, such as foetal alcohol spectrum disorders (Lange, et al., 2013), autistic spectrum conditions, attention deficit hyperactivity disorder (ADHD), tic disorders and/or learning difficulties (Dinkler, et al., 2017). Consequently, the importance of identifying mental health needs in these children and improving access to high quality services for them has been recognised by researchers, clinicians and policymakers worldwide.
Access to mental health services
Despite the recognition of need, evidence suggests that looked after children have difficulties in obtaining specialist child and adolescent mental health services (CAMHS) (Minnis and Del Priore, 2001). One Scottish study found that despite high rates of reported mental health difficulties, only 18% of this population had contact with CAMHS (Minnis, et al., 2006). This was higher (31.2%) in a recent Norwegian study, but even then over half of those with high scores on the Strengths and Difficulties Questionnaire (SDQ) did not receive support from CAMHS (Larsen, et al., 2018). Service use increased to 40‒60% for looked after children in an Australian cohort although the definition of what constituted mental health support was much broader (Tarren-Sweeney, 2010).
So why aren’t looked after children accessing services for mental health difficulties? Given that CAMHS have historically focused on diagnostic criteria to gain entry, in the UK it has been proposed that looked after children are unable to access mental health support as they do not meet diagnostic thresholds based on the current classification system (House of Commons Education Committee, 2016). There appears to be a tension between findings suggesting that they are reportedly more likely to meet the criteria for a recognised mental health disorder (Bronsard, et al., 2016) and other evidence indicating that the emotional and behavioural difficulties exhibited by such children do not always fall within current diagnostic categories and/or adequately consider their early experiences (DeJong, 2010; Tarren-Sweeney, 2008). Several studies have demonstrated a consistent constellation of symptoms and behaviours evident within those children who have been maltreated spanning seven key areas including interpersonal difficulties, dissociative responses to trauma, inattention and impulsivity, and sexualised behaviour (D’Andrea, et al., 2012; van der Kolk, 2005).
This has led to the emergence of the theorised Developmental Trauma Disorder (DTD), which conceptualises the difficulties of children with chronic trauma histories using an attachment and interpersonal framework (van der Kolk, 2005). Those in support of DTD argue that without an appropriate understanding and diagnostic label for these particular difficulties, individuals do not receive appropriate recognition and treatment (van der Kolk, 2014). Furthermore, without a sound assessment and formulation, clinicians run the risk of applying ineffective treatment approaches. For example, Post-Traumatic Stress Disorder (PTSD) is not the most common diagnosis among children who have experienced developmental trauma (Copeland, et al., 2007) and yet they are often in receipt of interventions based on the PTSD construct (Pynoos, et al., 2008). However, DTD remains controversial and has not been accepted as a formal diagnosis within DSM-5 1 due to its overlap with existing well-established syndromes, e.g. reactive attachment disorder, conduct disorder and borderline personality disorder (for a review see Schmid, Petermann and Fegert, 2013). Some have also contended that its existence would focus too heavily on the trauma-related aspects of the condition and prevent children from accessing current evidence-based care pathways and psycho-pharmaceutical treatment options (Woolgar and Baldock, 2015).
The alternative to the DSM-5, the World Health Organization’s International Classification of Diseases (ICD), released its 11th version this year incorporating a new diagnosis of Complex PTSD. This includes three additional symptom clusters, among them emotion regulation, negative sense of self and disturbances in relationships. This diagnosis is based on adult psychopathology and its applicability to children is yet unknown. Furthermore, it requires individuals to meet the criteria for PTSD first, which may not apply to victims of developmental trauma (DeJong, 2010). This leaves professionals and carers challenged to find the means to secure meaningful help and support for children who either have several disorders or do not meet the criteria for any common ones.
The role of psychometric measures in looked after children
Psychometric measures play an important role in screening for mental health difficulty and evidencing the effectiveness of services (Happell, 2008). As such, they have become routine in CAMHS in the UK, Australia, New Zealand, Denmark and Norway (Kwan and Rickwood, 2015). These tools also regularly aid decisions about access to services. The most commonly used, the SDQ (Goodman, 1997), is a short questionnaire designed to measure emotional and behavioural difficulties in children aged between two and 17 years. It is considered to have high sensitivity and specificity in several clinical and general child populations (Goodman, et al., 2004) and has been the measure of choice for looked after children in many western countries. However, there are limited studies investigating its utility and validity for this population (Jee, et al., 2011; Marquis and Flynn, 2009) and a UK looked after children Expert Working Group recently concluded that the SDQ alone is not an effective way of measuring the emotional and behavioural needs of these children (Social Care Institute for Excellence, 2017).
Despite the popularity of the DTD construct in some areas, only a handful of measures have been designed to capture the psychological sequelae associated with developmental trauma as highlighted by a literature review (Denton, et al., 2016). Current measures were criticised for using samples that were ill defined (e.g. failing to differentiate between victims of single- and multiple-trauma histories) or focusing on one particular trauma type (e.g. sexual abuse) (Denton, et al., 2016). Furthermore, several potentially useful tools do not have published research data to evidence their reliability and validity. For example, the BERRI is a tool used clinically within parts of the UK; although the measure is claimed by its developers to have good clinical utility, currently there are no published data to demonstrate its usefulness, validity or reliability as a tool for use with looked after children.
In fact, following a systematic search only two measures which consider their specific needs were identified: the Assessment Checklist (Tarren-Sweeney, 2007; 2013b) and the Brief Assessment Checklists (BACs) (Tarren-Sweeney, 2013a). Both measures were developed in Australia within a large looked after children population to assess the frequency and severity of emotional and behavioural difficulties theoretically underpinned by a developmental trauma framework. The measures have since been validated in a Dutch population (Goemans, et al., 2018) and have been used in several studies of looked after and adopted children (DeJong, Hodges and Malik, 2016; Goemans, van Geel and Vedder, 2018).
Despite good internal consistency, the BACs have not been evaluated within a UK population or from the perspective of foster carers or clinicians working in the field. Given the lack of research investigating current looked after children targeted measures, this study aimed to explore the following research question: How do foster carers and clinicians understand the use of psychometric measures, such as the SDQ and the BACs in the assessment of looked after children?
Method
The present study was part of a larger research project that set out to evaluate the validity of the SDQ and the BACs in a looked after children population. The choice to include the SDQ was partly due to its popularity as a clinical tool and also owing to its established validity and reliability in non-looked children populations (Goodman, 2001; Goodman, et al., 2000). The BACs were selected because they were designed specifically for looked after children. Qualitative data were collected via semi-structured telephone interviews or face-to-face focus groups with foster carers and clinicians. All discussions were audiotaped and transcribed verbatim by the researchers.
Participants
Foster carers
The study was advertised to foster carers through 19 local authorities, 15 independent fostering agencies/charities in the UK and nationally via social media platforms. For inclusion in the study, a foster carer was defined as an individual approved to provide temporary care to a child by a local authority or independent fostering agency (Department for Education, 2017). Those with a child aged four to 17 years, currently in placement for at least four months, were asked to take part in an online survey that related to the larger research project. At the end of the survey, participants were invited to express an interest in the present study by providing an email address. Researchers contacted interested respondents to arrange either a face-to-face or telephone interview. Owing to distance only two were able to attend a face-to-face meeting held at the university. Participants were reimbursed for their travel and parking costs.
Clinicians
CAMHS clinicians were recruited from one NHS Trust in the South of England. They were required to have at least one-year post-qualification experience and to have worked in the service for a minimum of 12 months to ensure that they had sufficient experience to comment on the utility of questionnaires for this population. Participants provided informed consent and ethical approval was gained from the University Faculty Ethics Committee and the NHS Trust Research and Development department. The clinicians were invited to take part in the study via a circulated email that included details of the study and an information sheet and consent form. Participants opted into the study by emailing the researcher. Focus groups were arranged at two separate NHS CAMHS sites where clinicians were invited to look at the measures and discuss them.
Measures
The SDQ parent version (Goodman, 1997) is a 25-item screening measure of mental health difficulties completed by carers of children aged four to 17 years. Items pertain to five subscales: Emotional Problems, Conduct Problems, Hyperactivity/Inattention, Peer Relational Problems and Prosocial Behaviour. Questions take the form of short phrases such as ‘considerate of other people’s feelings’ and ‘rather solitary, tends to play alone’. The SDQ uses a three-point scale of ‘Not True’, ‘Somewhat True’ or ‘Certainly True’. Scores are summed to give a Total difficulties score, the maximum score being 64 with scores above 14 considered ‘borderline’ and above 17 ‘abnormal’. Within the general child population, it is reported to have good reliability and validity (Goodman, 2001; Goodman, et al., 2000). A one-point increase in Total score is correlated to an increased risk of psychiatric diagnosis with odds ratios between 1.14 and 1.28 (Goodman and Goodman, 2009). Given these psychometric properties it has become the most widely used child and adolescent screening measure for mental health difficulties.
The Brief Assessment Checklists for Children (BAC-C; 4‒11 years) and Adolescents (BAC-A; 12‒17 years) (Tarren-Sweeney, 2013a) are 20-item psychometric measures designed to measure common emotional and behavioural difficulties specific to looked after children that are not present within existing measures such as the SDQ. The measures were developed with data from state-wide mail surveys of foster parents and kinship carers and from the state child welfare database measuring a large number of developmental, pre-care and in-care study factors. Difficulties include interpersonal, attachment-related difficulties, insecure relating and trauma-related anxiety and dissociation rated by the child’s carer. Questions include ‘Fears you will reject him/her’ or ‘Lacks empathy/guilt’. The first 16 questions are scored on a three-point Likert-style scale from 0 to two and the final four questions are scored on a three-point time-rating scale. Ratings on the measure are totalled to give an overall score. Scores exceeding the cut-off score of five out of 20 indicate that ‘further clinical assessment is warranted’; and good reliability and validity were reported for both measures when they were developed (Tarren-Sweeney, 2013a).
Procedure
A semi-structured interview schedule was designed for both participant groups comprising broad, open questions related to the utility of psychometric measures for looked after children with prompts identified for specific items (Smith, 2008). The interview schedule was developed within the research team of clinicians and researchers (two qualified registered psychologists and two trainee psychologists) prior to interviews and re-evaluated following the first couple of interviews, as recommended by Braun and Clarke (2013). No constraints were placed on the focus of the discussion, allowing all participants to contribute their unique perspectives.
Data analysis
Data were analysed using thematic analysis as outlined by Braun and Clarke (2013). A critical realist ontological position (Cook and Campbell, 1979) was adopted for analysis, interpretation and evaluation. A ‘theoretical’ approach (Braun and Clarke, 2006) was selected for coding and analysis due to the researchers’ primary goal of investigating a specific qualitative research question. A two-step deductive-coding process (Braun and Clarke, 2006) was adopted due to the primary goal of investigating a specific research question. Initially data were coded separately by two researchers (CF and RD) using a ‘complete coding’ approach to identify data of interest or relevance to the research question (Braun and Clarke, 2013). Data were then reviewed for latent content, producing both semantic- and researcher-derived codes. Initial patterns were provisionally identified through consideration of all the secondary codes and initial themes were mapped visually to explore potential implicit or hierarchical relationships between them.
Given the similarity of the initial themes between foster carer and clinician data, the decision was made to combine the dataset and one researcher (CF) continued to map these themes visually. Differences between the two participant group perspectives were noted. Following this, the data and codes were revisited to explore how well the candidate themes fit the data. Themes were also checked against the data and discussed between all authors over the course of the analytic process. The researchers’ (CF and RD) subjective influence was considered throughout data gathering, coding and analysis via a reflective journal.
Results
A total of 13 UK foster carers agreed to take part: two via group interview and 11 via telephone interviews (Table 1). This was approximately 4% of the total sample of foster carers in the larger study. All foster carers were female and were given pseudonyms during analysis to protect their anonymity. Two clinician focus groups comprised nine participants (two males, seven females) and were referred to as C1, C2, etc. to protect their anonymity (Table 2). Thematic analyses generated six individual themes, which were organised into two superordinate themes, as summarised in Figure 1.
Participant information for foster carers.
Participant information for CAMHS clinicians.
A diagrammatical representation of the themes that emerged from foster carer and clinician transcripts using Thematic Analysis.
Within the narrative, foster carers and clinicians emphasised the need to gain a comprehensive understanding of the needs of looked after children. The BACs and SDQ were believed to aid this process. However, potential shortfalls and problems around how information from these measures are used to facilitate support were discussed. Points of divergence between foster carer and clinician perspectives are highlighted within the text.
Theme 1: Relevance to looked after children
Within this theme, there were two subordinate ones describing the arguments for and against the ability of the BACs and SDQ to capture the key emotional and behavioural difficulties of looked after children. Foster carers suggested that the measures were important for organising how information about the children was captured and for facilitating a process of communication between themselves and professionals. For clinicians, the measures were described as useful in providing a framework from which to base discussions with foster carers. In particular, the BACs were identified as especially important for junior members of staff who had less experience of working with attachment and trauma issues. On the other hand, the children were recognised as having complex needs that were difficult to capture and this limited the utility of both measures.
Measures can ‘flag up’ key difficulties
Foster carers frequently talked about questionnaires as a way to raise concerns or ‘flag up’ (Sarah) difficulties that a child might be experiencing. Gathering critical information at the outset was also highlighted as important by clinicians. Many felt the measures had a role in providing a ‘snapshot’ (C2) of information about a child at the referral stage that could help the professional to think ‘in a sharper, more focused way’ (C5). A number of clinicians also suggested that the information provided by short questionnaires can be useful as a ‘screen’ (C4), particularly when completed by multiple informants.
Almost all of the foster carers praised the BACs for including difficulties that they felt were not acknowledged by the SDQ. These were described as relating to problems with ‘emotion’, ‘relationships’, ‘abuse’, ‘trauma’ ‘sexualised behaviour’ and/or ‘attachment difficulties’ and for some, seeing them in a measure was ‘comforting’ (Barbara) and ‘reassuring’ (Lisa): …it appears to be more geared to my child… it’s got questions on there about the sexualised behaviour and the intensity of new relationships… and emotions that it hasn’t got on the other one. (Norah) I think the BAC one seems to be more around emotion. Whereas the other one, there’s a lot more physical behaviours… (Laura) If they were acting out, then the other one [SDQ] would probably be better. (Barbara)
Measures cannot capture the complexity of looked after children
Foster carers and clinicians alike raised concerns about the brevity of the questions while also acknowledging the challenges of attempting to summarise complex difficulties experienced by looked after children. The BACs and SDQ were considered to lack sufficient detail and this resulted in the majority of foster carers stating that they would prefer to talk to a professional. Some also wanted to provide more information in order to feel understood, such as Karen when discussing the SDQ: ‘I ticked two boxes and wrote notes…’ Several clinicians suggested that looked after children were likely to score highly on questionnaires and, as such, wondered how useful they would be when attempting to ‘gauge the seriousness of the difficulties or differentiate between children who are fighting for resources’ (C1).
In relation to the specific foci of the questionnaires, the SDQ was criticised for overlooking key behaviours such as relational difficulties and trauma by foster carers, and attachment and dissociation by clinicians. In contrast, information related to somatising behaviour, self-harm and mood issues were missing from the BACs, as described by Sarah: ‘… none of the questions would necessarily draw out the symptoms of depression’. Foster carers also highlighted key difficulties that were missing from both questionnaires, such as controlling behaviour (P10, P11), bed-wetting (P8), violent behaviour (P3, P4, P11) and suicidality (P7):
I don’t think there’s a lot about control and my experience is that these children try to control lots of things because they’ve had so much control taken away… (P11)
Theme 2: Using measures to inform care planning
Questionnaires were believed to have the potential to enhance care plans by facilitating a way of monitoring outcomes, providing feedback to foster carers, planning subsequent action and facilitating access to CAMHS. However, the perspectives of foster carers and clinicians were different when considering the ability of the two measures to do this.
Facilitating access to CAMHS
Whether or not questionnaires could facilitate access to CAMHS was discussed at length by both participant groups. Several foster carers commented that looked after children were atypical and capturing their difficulties posed a challenge when completing questionnaires believed to be the gateway to services. At times, this prevented their looked after children from accessing support. In Lisa’s words:
Many suggested that children who communicated their distress through externalised or challenging behaviour were more likely to be identified by the SDQ; however, children who communicated their distress in other ways, for example, those who ‘shut down completely’ or ‘comply, do well at school and are quiet’ (Karen) were missed. Foster carers often felt they had to reach a ‘crisis’ point (Lisa) or to experience ‘violence’ from looked after children (Karen) before help was provided. As such, they hoped that the use of measures specific to looked after children, such as the BACs, would be a ‘starting point’ (Elizabeth) or ‘stepping stone’ (Laura) that would help them to ‘fight for my children’s corner’ (Laura) when requesting further support. A key advantage of using questionnaires was to have concerns ‘on the record’ (Karen) or ‘in writing’ (Elizabeth).
However, within the clinicians’ discussions, some reported that even if emotional and behavioural difficulties specific to looked after children were identified through additional measures, e.g. the BACs, the service would not be able to provide support to the family as it was not commissioned to work with complex attachment presentations or non-diagnosable mental health conditions. Thus, the assessment of these difficulties would ‘open the flood-gates’ (C5).
Using measures to monitor change
The ability of the questionnaires to track and monitor the behaviours of looked after children was highlighted as an important function. Foster carers often talked about using measures to gain objective ‘feedback’ (Elizabeth) on the child in their care and to observe the progress they have made. Some suggested questionnaires should be used routinely, for instance at six-monthly reviews, to inform ‘what therapy’ (Sarah) or determine ‘how we manage behaviour’ (Irene): I think they’re, they’re really useful… I use them because I can then look back on them and I can see if there are improvements or patterns or changes. (Norah) Wherever they go, whoever’s going to get them and read them, at least there’s some sort of, you know, a baseline of what these children are like. (Emma) You’d hope that somebody who’s been referred in and who is impulsive or lacks guilt… after some therapy, [the score] it’s going to go from ‘mostly true’ to ‘partly true’… (C4) You have to know whether or not to expect to see scores go down with attachment problems as to whether or not you even could actually use it in that way… as a routine outcome measure. (C2)
Theme 3: Current problems when using measures with looked after children
The third superordinate theme comprised two subthemes: one depicted both participant group concerns regarding interpretation of the questionnaires and the other, which emerged solely from the foster carers, related to the failure to integrate information from questionnaires into the child’s care plan.
Interpretation of the questions vary
There was a worry that items on both measures were open to misinterpretation or could be influenced by the foster carers’ level of experience and value system: If you’ve got a newly qualified foster carer… you know, yeah he craves attention or does he? No, he doesn’t come for kisses, but they don’t realise that the attention is actually the fact that he’s swinging off the roof or stealing from the kids at school and you know they don’t think, ‘What does that mean?’ (Charlotte) One of the young people that’s placed with me would share her makeup at the drop of a hat because she thinks that’s fun. Ask her to share her food and you’d get a fork in the back of your hand!
Information is not engaged with
A theme emerging solely from the foster carers was related to the perceived insignificance of questionnaires as they received little or no feedback about how these were utilised. The majority of foster carers believed that information was not acted upon and many referred to questionnaires as merely a ‘tick box exercise’ (Norah). One carer stated that she would be more likely to complete questionnaires if she believed it would add value to the children in her care. Others held a belief that the measures went into a void with no action subsequently taken: ‘I don’t think the paperwork I do is ever looked at’ (Emma) and ‘I find it a bit frustrating really …but they just don’t use it’ (Norah).
Discussion
The present study sought to contribute to the field of mental health assessments of looked after children by gathering data from the perspectives of foster carers and clinicians working in CAMHS. The thematic analysis uncovered three overarching themes that included: discussion about whether brief measures such as the SDQ and BACs can identify relevant emotional and behavioural difficulties in the children; how these measures might be useful to inform care planning; and the potential problems arising with the use of these measures with the looked after children population. It was noteworthy that both foster carers and clinicians were broadly in tune with one another regarding the merits and the challenges of psychometric measures.
Identifying mental health needs in looked after children
The findings suggest that both clinicians and foster carers have mixed views on the potential utility of the BACs and SDQ. Both groups saw their merits and would appear to endorse their continued use, albeit with some caution. There was a consensus that these measures had some use in that they could identify difficulties and monitor change over time. However, foster carers and clinicians had concerns about whether the measures were fit for purpose as it was unclear if either of them could capture and represent the breadth and complexity of children’s presenting difficulties. The theoretical frameworks underpinning the SDQ and BACs questionnaires are different. The SDQ is underpinned by a developmental theoretical framework (Goodman, 1997) while the BACs was developed with attachment and trauma theories in mind (Tarren-Sweeney, 2013a). From the participants’ perspective, neither questionnaire was able to provide the necessary and sufficient profile of the young person to be able to develop an effective care plan. More broadly, foster carers highlighted that key difficulties observed in their child were often absent from psychosocial measures, including problems in relationships (e.g. lying and stealing), seeking control, violent behaviour, toileting problems, emotion regulation difficulties, self-harm and suicidality. Furthermore, one carer highlighted that learning difficulties and other neurodevelopmental challenges are often missing in brief screening tools, despite the knowledge that they are more prevalent among looked after children (Dinkler, et al., 2017).
Many clinicians have warned against an over-reliance on narrow assessment tools for looked after children, arguing for a broader and more sophisticated conceptualisation of their needs, such as psychological formulation (DeJong, et al., 2016; Golding, 2010). A significant number of participants suggested that using a combination of different measures would be more helpful than either measure on its own. A similar conclusion was made by the UK Expert Working Group (Social Care Institute for Excellence, 2017) which highlighted that the SDQ alone was not an effective way of measuring the emotional and behavioural needs of looked after children and thereby recommended that the assessment process be supported by a broader set of measures. This was mirrored in the US, where several groups of clinicians have developed trauma-based mental health assessment models, such as the Chadwick Center for Children and Families Trauma Assessment Pathway (TAP) (2009), which aims to provide a comprehensive assessment of complex trauma using a number of measures tailored to the individual, family context and culture.
Today’s evidence in the UK suggests that the current theoretical frameworks supporting the two measures under examination need to be expanded to allow for the development of a more comprehensive measure. Constructs embedded within DTD could potentially be usefully operationalised and integrated alongside material from the SDQ and BACs to provide a holistic measure. Alternatively, the Assessment Checklists (Tarren-Sweeney, 2007; 2013b) could be investigated further. These measures provide longer and more comprehensive versions of the BACs and incorporate symptoms of DTD. For that reason, they could address some of the concerns raised by foster carers and clinicians in the present study. However, the measure is more time consuming and may pose a challenge when used as part of routine practice.
Trauma-informed services
The themes to emerge from the research suggest that existing questionnaires may not be sufficiently sensitive in identifying needs specific to looked after children. This could go some way to explain why they are often denied access to mental health services (Minnis and Del Priore, 2001). Despite recommendations that they should be able to access service provision in a timely manner, there is a serious challenge to identify measures and assessment processes that can facilitate the development of meaningful formulations. Currently, there is no clear care pathway for those who do not meet criteria for psychiatric diagnosis and yet require therapeutic input (Rao, Ali and Vostanis, 2010), leaving a population of young people unserved by mental health services.
With the increased awareness of the impact of adverse childhood experiences, there is a growing movement towards trauma-informed care which Conradi and Wilson (2010: 622) describe as ‘a wider system impacting children and families, with multiple components designed to meet the varying needs of traumatised individuals who are receiving services’. These include collaboration across public service agencies, with a grounding in contemporary knowledge and understanding of trauma and how it presents. Emerging evidence suggests that the development of trauma-informed systems in education and social care may be effective in building resilience in these children (Ko, et al., 2008) as well as offering a more integrated approach to care (Social Care Institute for Excellence, 2017). Although this is clearly positive, it adds to the complexity of ensuring that multiple perspectives are captured in order to comprehensively identify the child’s needs. If the envisaged changes to systems are also supported by the adoption of broader measures alongside the SDQ, such as the BACs or Assessment Checklists, an increased understanding of the children is likely. Embedding this within a person-centred assessment, with the commensurate discussions to explore the data, may enable the development of a more sophisticated formulation, leading to more effective interventions and associated long-term relationships with subsequent improvements in outcomes.
Limitations
The results were analysed within a critical realist approach and therefore the findings need to be considered as reflective of the participants recruited at a specific time. The foster carers (while all female) were drawn from across the country and are therefore not organised by one particular service framework and the associated access arrangements. It may be that those who were particularly interested in the use of these measures and/or the mental health needs of looked after children were more motivated to take part in the study. Furthermore, the clinician sample was recruited from one mental health trust and although clinicians had a range of roles and experiences of working with looked after children, future research should replicate the present findings in a larger sample.
Summary and recommendations
The mental health of looked after children has been highlighted as a clinical priority in several western countries. Despite recommendations that the views of children, carers and clinicians should form the heart of service design and development (Australian Government, 2011a; 2011b; NHS Taskforce, 2014), little empirical research has been undertaken to explore foster carers’ and clinicians’ perspectives on the mental health assessment of their children. The findings from the current study highlighted the complexity of this issue when reviewing two specific psychometric measures and participants seldom felt that the needs of looked after children were adequately captured by existing measures in isolation. This may go some way to explain why they are often unsupported by mental health services.
When working with children and young people whose life histories are complex and challenging, it is crucial that mental health assessments incorporate different perspectives (e.g. the child, parents, teachers and social workers) and use multiple assessment tools (e.g. attachment-based measures, trauma-specific measures, interviews and observations) to build the fullest picture of the child and their context. To aid this process, a number of assessment tools may be necessary and complex trauma assessment models such as TAP (Chadwick Center for Children and Families, 2010) can provide a useful guide for clinicians when selecting assessment tools. The current study also highlights the importance of offering carers the opportunity to write notes, expand on their answers and provide further qualitative feedback and explanations about their child during the assessment process. Finally, it is unclear what the future holds for DTD or how the Complex PTSD diagnosis will apply to children and young people in care. Although diagnosis can facilitate access to support, if individuals and their families are to be helped to understand their child’s needs, a psychological formulation may be the most useful way forward given that it includes multiple biopsychosocial elements, maintaining factors and strengths from which to develop a bespoke treatment package for the young person.
Footnotes
Acknowledgements
All authors share responsibility for the final version of the work submitted and published. Dr Frogley and Dr Denton were responsible for the collection and analysis of original data and so take responsibility for the integrity of the data and the accuracy of its analysis. Ms John and Dr Querstret were responsible for the supervision and assistance regarding the development of the study project, data analysis and interpretation, and assistance with the writing of this article. Dr Frogley drafted the final research article with the assistance of the remaining authors.