Looked after children (LAC) and mental health

Irrespective of circumstances, LAC appear to be disproportionately burdened by an array of adversity, disrupted development and problematic relationships. These challenges commonly manifest as attachment-related disorders, emotional difficulties and maladaptive behaviours. Given the particular vulnerability of this population and the urgency required when dealing with insecure attachment, any research involving an untreated control group would be inherently unethical. Quantitative research in this field has therefore remained sparse, with the vast majority of the literature instead opting to assess case studies and qualitative data. The retrospective nature of the present study means it is able to contribute treatment-related outcome data from a relatively large sample of LAC. Although this naturalistic design is limited in terms of fully controlling confounding variables, it avoids the risk of bias from insufficient blinding and ensures that no child misses out on potentially beneficial treatment.

A comprehensive meta-analysis conducted by Vasileva and Petermann^{Reference Vasileva and Petermann1} accumulated data from 22 samples of abused and neglected LAC under 7 years old. Across 11 studies, 38.5% of the 3211 eligible child participants were found to have either motor or cognitive developmental delays; these were consistently more prevalent in children fostered by strangers than in those looked after by kin. Similarly, 39% displayed clinically significant internalising or externalising problems, and only 35% of LAC met criteria for secure attachment. The findings of Tarren-Sweeney^{Reference Tarren-Sweeney2} indicate that when mental illness goes untreated during these early years, a child's social and emotional development may be jeopardised, and the possibility of a secure attachment appears to reduce with each placement breakdown. Children that entered care at a later age and were exposed to maltreatment were found to score significantly higher on both the Child Behaviour Checklist and the Assessment Checklist for Children, whereas having a permanent placement was highlighted as an important protective factor. Given the cross-sectional design, however, it was difficult to infer causation, and it is equally possible that children with better mental health are more manageable and thus less at risk of having placements break down.

Attachment security appears to be a crucial factor mediating the mental health outcomes of children placed in foster care. Both internalising problems and frequency of mental illness diagnoses diminished significantly for girls removed from institutional care into foster care, but not for boys, in a sample of Romanian children aged 3 and below.^{Reference Mclaughlin, Zeanah, Fox and Nelson3} Subsequent analysis using the strange situation test^{Reference Ainsworth and Bell4} revealed that the demonstrated improvements in mental health for fostered girls was fully mediated by the formation of secure attachment. Guyon-Harris et al^{Reference Guyon-Harris, Humphreys, Degnan, Fox, Nelson and Zeanah5} examined the longitudinal trajectories of reactive attachment disorder (RAD) based on semi-structured interviews with caregivers. In line with and extending previous conclusions, the sharpest declines in symptoms of RAD were found for children placed in foster care, which was sustained long-term at 8- and 12-year follow-up (73% retention). The vast majority of children with rapidly decreasing symptoms of RAD had been relatively younger when placed in foster care, whereas children entering at an older age displayed persistent and elevated symptoms. Almost all of the original group left in residential care displayed persistent symptoms.

Direct interventions for mental health difficulties in LAC

The MTFC-P, a direct intervention facilitating healthy, securely attached foster parent–child relationships, included 12 hours of training, daily telephone contact with a consultant and a weekly support group for foster carers, while the child attended weekly therapeutic play group sessions.^{Reference Fisher and Kim6} Secure and avoidant attachment behaviours were assessed at five time points with 3-month intervals. Pre-schoolers in the MTFC-P condition exhibited significantly more secure and less avoidant behaviours than those in regular foster care. Furthermore, the frequency of secure attachment behaviours increased over time for those receiving the MTFC-P intervention but decreased for the comparison group.

Study on Theraplay have shown statistically significant improvements in communication, interpersonal relationships and behaviour as rated by adoptive mothers, but problem-solving, emotional awareness and general functioning were not found to change.^{Reference Weir, Lee, Canosa, Rodrigues, Mcwilliams and Parker7} In a related study, on the other hand, weekly Theraplay sessions delivered to 20 LAC over a period of up to 8 months resulted in no significant change on any of the subscales of the Strengths and Difficulties Questionnaire (SDQ).^{Reference Francis, Bennion and Humrich8} Follow-up subgroup analysis indicated that girls showed significantly more teacher-rated improvements than boys in terms of hyperactivity, conduct and prosocial behaviour. Taken together with the findings of McLaughlin and colleagues,^{Reference Mclaughlin, Zeanah, Fox and Nelson3} this suggests that girls may perhaps be more able to securely attach to a foster family and so may benefit more substantially from direct intervention.

Dyadic developmental psychotherapy (DDP) is described as a direct, attachment-based intervention involving the delivery of traditional psychotherapy to both the child and their carer.^{Reference Mercer9} Wingfield and Gurney-Smith reported that 12 adoptive parents receiving DDP gained increased curiosity, understanding and empathy for their children, while the children reported fewer behavioural problems, improved control over emotions, better relationships with peers and improved sleep.^{Reference Wingfield and Gurney-Smith10} On the other hand, in spite of progress, a number of parents described the sessions as emotionally exhausting, uncomfortable and upsetting. Feedback from therapists delivering DDP has, however, been very positive. Turner-Halliday and colleagues^{Reference Turner-Halliday, Watson, Boyer, Boyd and Minnis11} reached out to child and adolescent mental health services (CAMHS) in the UK delivering the treatment. DDP was seen as an effective tool for directly tackling complex profiles of mental illness alongside the root causes of both internalising and externalising problems. The inclusion of parents and carers on such an emotionally stimulating journey was seen to indirectly facilitate secure attachment through an improved mutual understanding.

Multiagency cosultation for LAC mental health

In an attempt to better combat the wide range of complex and overlapping symptoms of behavioural, mental, social and cognitive disorders presented by LAC, a small body of research has explored the utility of multi-agency consultation. Upon recounting a number of particularly challenging cases where successful outcomes only emerged in the wake of a concerted effort from all available support, Sprince^{Reference Sprince12} concludes that child therapists have an obligation to appropriately tailor treatments and advice to the specific network of individuals and organisations responsible for the child's care. The consultation model differs from regular therapeutic interventions in that rather than directly treating the child, their network of carers, teachers, social workers and therapists collaborate to tailor solutions to the child's specific needs, difficulties, history and environment.

Swann and York^{Reference Swann and York13} refer to this multi-agency intervention as ‘THINKSPACE’, succinctly illustrating the creation of a space where clinicians, service providers and stakeholders can come together to share their diverse perspectives, knowledge and skills to conjure up a holistic picture and systemic solution. Consultation sessions dedicate around an hour and a half to each child, beginning with the construction of a narrative outlining their development, diagnoses and family history. Attendees then contribute their unique perspectives, theoretical knowledge and practical experience to generate realistic and appropriate solutions. Swann and York emphasise the importance of using everyday language, having only one key interviewer and avoiding giving the primary carer direct instructions.

Qualitative data for evaluation of the consultation model have been collected from social workers and clinicians by Dimaro, Moghaddam and Kyte.^{Reference Dimaro, Moghaddam and Kyte14} Feedback from 138 social workers indicated that a vast majority felt that their goals had been suitably addressed by the collaborative consultation sessions, particularly with regard to assessing concerns, understanding behaviours, understanding relationships and planning next steps. However, 37% of those hoping for effective parenting techniques and 41% of those hoping to work more effectively with staff, agencies and local services felt their goal had not been properly addressed. Subsequent focus groups with 12 of the clinicians illuminated a number of key themes. Primarily, they uniformly felt that the sessions allowed them to provide useful, diverse and practical support. They highlighted the utility of clearly defining roles and objectives and spoke positively about the wider systemic effects their input could achieve.

The sole quantitative trial evaluating the effects of consultation on the mental health of LAC was described by Callaghan et al.^{Reference Callaghan, Young, Pace and Vostanis15} Psychiatrists, psychologists and therapists from all three tiers of a UK CAMHS came together to collect collaborative feedback from carers of 45 LAC aged between 4 and 17 years. Outcomes were assessed based on scores on the Health of the Nation Outcome Scales for Children and Adolescents and the SDQ, completed at baseline and 5-month follow-up. Although improvements in total SDQ scores did not quite reach significance, scores on an emotional problems subscale did. This data does rationalise further, more rigorous evaluation of the model but is limited in that it did not include an adequate comparison group, and thus the authors responsible for collecting data were not blinded to experimental procedures. Furthermore, of the 39 carers that completed service satisfaction feedback forms, only 51% felt the intervention offered had been efficacious, and 28% felt their young person had not shown any improvement.

A number of key themes have emerged from the literature exploring mental health presentation and interventions among LAC. Primarily, the formation of secure attachment has consistently been highlighted as a crucial factor mediating mental well-being among this population of particularly vulnerable children. This conclusion emerges from the research contrasting foster care with residential care,^{Reference Mclaughlin, Zeanah, Fox and Nelson3,Reference Guyon-Harris, Humphreys, Degnan, Fox, Nelson and Zeanah5} the success of treatments targeting child–carer interactions,^{Reference Weir, Lee, Canosa, Rodrigues, Mcwilliams and Parker7} and the discrepancy between outcomes for boys and girls,^{Reference Mclaughlin, Zeanah, Fox and Nelson3,Reference Weir, Lee, Canosa, Rodrigues, Mcwilliams and Parker7} and it is aligned with the apparent predictive power of placement stability and age at entry into care.^{Reference Tarren-Sweeney2}

Ethical considerations present a major barrier to quantitative research in that withholding potentially efficacious treatment from a vulnerable child can never be justified. However, uncontrolled quantitative data have pointed towards the efficacy of direct work,^{Reference Fisher and Kim6,Reference Weir, Lee, Canosa, Rodrigues, Mcwilliams and Parker7} as has the more abundant qualitative research.^{Reference Wingfield and Gurney-Smith10,Reference Turner-Halliday, Watson, Boyer, Boyd and Minnis11} The viability of indirect interventions, on the other hand, is far less well evidenced, with the sole quantitative data-set revealing no significant change in SDQ scores^{Reference Callaghan, Young, Pace and Vostanis15} despite promising conclusions drawn from qualitative data.^{Reference Dimaro, Moghaddam and Kyte14}

Therefore, in this study, we aim to test the following key hypotheses.

1. (a) Consultation with direct intervention is more efficacious than consultation on its own (H1).

2. (b) Consultation on its own is associated with significant SDQ-rated improvements relative to baseline (H2).

3. (c) A lower total number of address changes and longer time in current placement (H3), younger age at point of referral (H4) and female gender (H5) predict more positive outcomes.