It is widely recognised that children in foster care are at very heightened risk of poor outcomes, and there are major gaps in the evidence base regarding intervention and prevention strategies to reduce that risk and promote children’s well-being and development. Thus, there is an urgent need for rigorous research to support the provision of effective interventions to improve outcomes for children in foster care. A key domain that is particularly relevant and important in the foster care context is the quality of caregiving and the security of attachment, which is a logical target when attempting to promote foster children’s outcomes and reduce the risk of emotional and behavioural problems.13 Currently, there is very limited evidence regarding effective programmes for specifically addressing the attachment needs of young children in foster care. At the same time, many researchers working in this field have noted the serious challenges to outcome research inherent in the systems around children in care, which no doubt explains, in part, the relatively scant and methodologically weak evidence base.53 This feasibility study focused on adapting a parenting intervention designed to improve attachment security to meet the needs and contexts of children in foster care with symptoms of RAD, and conducting initial feasibility work to assess the extent to which a full-scale RCT might be possible. In this chapter, we summarise the main findings and consider the prospect of a definitive trial.
Summary of findings
Phase 1 of the study aimed to refine the Video-feedback Intervention to promote Positive Parenting and Sensitive Discipline, Foster Care (VIPP-FC) programme to the context of children in foster care in the UK. The Manual Development Working Group implemented a series of quite minor changes to the intervention protocol that focused on improving the suitability of the video-feedback intervention for the UK foster care context, and endorsed the content and approach of the intervention. In general, the expert group considered the VIPP-FC to be highly appropriate for use with foster carers and foster children. The enhancements suggested by the expert reference group included changes to the wording of the manual to reflect the fact that some children were cared for by special guardians and kinship carers, adjustments to the use of language to reflect the British context, reduction in the total number of visits from seven to six and a shorter gap between the sessions, adding a games booklet to help carers engage in play, inclusion of a goal-based outcome measure in the intervention to mirror processes routinely used in UK CAMHS and changes in the manual to acknowledge the sometimes time-limited nature of foster care.
Phase 2a of the study consisted of a small case series to road-test the recruitment and screening procedures, as well as the modified VIPP-FC, and to examine its clinical suitability for children with RAD symptoms in real-world clinical practice. The implementation of the case series was challenging, but highlighted many crucial practical issues that needed resolving prior to proceeding to the feasibility trial. A key issue arising from this initial work was that the return rate of screening questionnaires was comparatively low, which was compounded by the fact that a significant proportion of cases could not be approached, as they were deemed inappropriate by the local authority. This is an important consideration for future work because initial estimates of potential samples based on cases held by the local authority may significantly overestimate the effective recruitment pool. Nevertheless, at this stage in the study, the proportion of foster carers returning screening questionnaires was not below the original target, even though the amount of follow-up and support required to achieve this proportion was greater than we had expected. The process observations during this phase highlighted the critical importance of social worker involvement and engagement in the project, which is not easy to achieve because of their heavy case loads. Having a ‘champion’ of the study within the local authority, a high level of visibility of the research team, and additional administrative support to help local authorities contact foster carers and identify cases were noted as important potential solutions to problems with identification and engagement of eligible families.
To further enhance our understanding of the barriers to and constraints of recruitment and intervention, the learning from the case series was complemented by qualitative interviews with stakeholders. These qualitative interviews highlighted the complexity of conducting a RCT in the context of foster care. On the other hand, stakeholders also emphasised the potential benefits of the proposed research, particularly in bringing an effective intervention to this vulnerable population and the potential of the approach to increase foster carer skill and confidence. Stakeholders were generally positive about the VIPP-FC as an intervention: it was felt that this adapted version of the Video-feedback Intervention to promote Positive Parenting and Sensitive Discipline (VIPP-SD) had great potential to be beneficial to the families with whom stakeholders worked. However, greater flexibility in delivery, and especially timing, was considered worthy of further exploration. Stakeholders generated several ideas for tackling some of the barriers to implementing a RCT. A key message was the importance of taking a relational approach, giving time and resources to building relationships with all the key players in the system, especially those at the ‘frontline’.
The final phase (phase 3) of the study consisted of a pilot RCT of the modified VIPP-FC programme. A new group of children was randomised to receive either the VIPP-FC plus CAU or CAU only. The VIPP-FC was delivered in-home by trained CAMHS practitioners or other appropriately qualified VIPP-trained interveners. Baseline and end-of-treatment measures included validated assessments of RAD and co-occurring conditions, security of attachment (including disorganised attachment), sensitivity of parenting, child adjustment, carer well-being and the CA-SUS measure of service use.
Despite implementating a range of changes to the protocol in the light of learning from the case series and stakeholder interviews, the pilot trial encountered significant challenges to recruitment that mirrored our experiences during the case series. The most significant barriers were encountered during the initial stages of contact with foster carers. The overall number of mailed-out questionnaires, n = 336, was smaller than the original target of 500, largely due to significant numbers of children registered with the local authorities being considered inappropriate or ineligible for the study. The overall response rate to the screening questionnaires was low (indeed, considerably lower than in the case series), at 29%, and varied substantially by site. We were unable to recruit the target sample size of 40, even after numerous initiatives were introduced to increase recruitment, including, in stages, allowing online screening, engaging additional sites and, finally, allowing all children in foster care to enter the study even if RAD diagnostic criteria were not met. By the end of the study, we were able to recruit three-quarters of the target sample size to the feasibility trial. Of the randomised families, 15 were allocated to the VIPP-FC and 15 continued to receive usual care.
Notwithstanding the difficulties with recruitment, many other features of the pilot trial design worked well. The screening tools performed well in identifying potentially eligible participants, demonstrating good convergence and sensitivity against the research diagnostic interview. We achieved good outcome data completeness and all of the participants who received the treatment completed all sessions. The implications were quite clear: the intervention itself and the research protocol for evaluating its impact were robust and implementable to a high degree of rigour, but recruitment problems could not be fully surmounted.
In addition to the questions regarding the feasibility of implementation of a RCT of the VIPP-FC, we undertook important feasibility work linked to health economic evaluation. This had two main aims. The first involved the modification and testing of an appropriate version of the CA-SUS for the collection of service use data in this population. The results showed that the modified CA-SUS, after some adjustments, worked well and was acceptable to both interviewers and interviewees. The second aim was to explore the recent literature to assess how realistic it might be to include a measure of HRQoL capable of generating QALYs, for use in a cost–utility analysis, in this very young age group. Results indicated that it is not currently possible to recommend any specific measures for inclusion in a future trial of the VIPP-FC; thus, the primary economic evaluation should be a cost-effectiveness analysis using the primary clinical outcome measure or using an assessment of meaningful change on the primary clinical outcome measure. It was further recommended that a future trial of the VIPP-FC may be a good opportunity to compare the construct validity and sensitivity to change of the proxy versions of existing instruments by exploring the relationship of these measures with clinical outcomes.
In the final stage of the study, foster carers who had received the modified VIPP-FC took part in qualitative interviews to explore their experiences of the intervention. In general, foster carers’ experiences of the VIPP-FC were very positive. Foster carers endorsed the importance of the research and its valuable contribution towards promoting best practice in the foster care context. Foster carers described a range of benefits associated with receiving the intervention programme, from enhanced skills to improved relationship with their child, and found it acceptable and helpful.
Considerations regarding progression to a full trial
Within the TSC and among stakeholders, there was a clear steer that progressing this work is important and should be considered seriously, despite the challenges of recruitment. The results of the feasibility trial indicated very strongly that the research protocol itself was feasible. In the course of the study, we developed a robust protocol for managing consent, for intervention delivery and for assessing outcomes. We did not encounter any difficulties with randomisation. We were able to identify good candidate measures for screening and for outcome evaluation, and were able to recommend the DAI as a robust primary outcome measure, as it showed high levels of acceptability, adequate range and indications of sensitivity to change (significant reductions overall from pre treatment to follow-up). The rate of follow-up to post treatment was high and engagement in treatment was good. The fundamental barrier to progression to a full-scale trial is recruitment. As noted in Chapter 5, there are clear indications about the most important ingredients that make up a successful site and knowing those could allow the active identification and selection of such sites from the outset. Doing so ought to make consistent recruitment at the rate we observed in the final stages of the pilot study feasible. In that regard, one strategy to consider for a future trial is careful selection of sites based on research capability. However, we note a clear downside of such an approach, which is that it may bias both the populations of foster families included in the trial and the service contexts in which trial delivery would take place, thereby limiting generalisability. With that in mind, we also identified a range of factors that could significantly enhance recruitment and could provide routes to more effective recruitment in the future. Owing to a lack of resource, we could not easily implement all of these adjustments during the course of the study. Most important among these is the provision of administrative capacity in local authorities to manage and support recruitment, as well as additional capacity in the research team to allow researchers to be present in each recruitment site to a greater degree. Given the highly vulnerable nature of this population, the feasibility of most elements of the trial protocol, the high levels of acceptability of the VIPP-FC and the good evidence of its efficacy in other broadly similar populations, we believe that it is imperative to find ways to resolve these systemic problems with recruitment.
Research recommendations
On the basis of the above considerations, we make the following recommendations:
- A RCT is an appropriate design to evaluate the VIPP-FC in the foster care context.
- Drawing on evidence from this study and a number of other sources, we consider the DAI symptom count to be an appropriate primary outcome for such a trial.
- A range of other measures, including continuous RAD symptom questionnaires, the Strange Situation Procedure, measures of child adjustment, parenting sensitivity and carer burden, are feasible and appropriate secondary outcomes in a randomised trial.
- We are not able to recommend online data collection as a method for obtaining baseline and follow-up assessments.
- The primary economic evaluation of such a trial should be a cost-effectiveness analysis using the DAI or using an assessment of meaningful change in the DAI.
- In the light of the large proportion of children in foster care showing RAD symptoms and the challenging nature of recruitment, a future trial should not select participants on the basis of RAD diagnosis, but instead address the efficacy of the VIPP-FC for children in foster care in general, measuring RAD symptoms as the primary outcome.
- We obtained an estimate of the SD of the DAI of 4.16 points in the foster care population, with an upper 80% confidence limit of 5.2 points. We suggest that a difference of 2 points on the DAI is a minimum clinically significant difference, which equates to a standardised effect size of approximately d = 0.40. To detect such an effect size with 80% power at α = 0.05 and an equal allocation ratio would require a minimum sample size of 200 participants. A sample size of 266 would be required for 90% power.
- A randomised trial should not proceed until more robust recruitment procedures can be established. Recruitment should be supported by adequate resourcing within local authorities, high levels of engagement of social work teams and should involve direct contact between the research team and all eligible foster carers (e.g. during joint home visits with social workers).
As recruitment was by far the most significant barrier to a robust trial, researchers may wish to consider a range of alternative recruitment and sampling strategies. One alternative strategy suggested by our research team is to consider extending recruitment to children who have been adopted, who present with similar rates of attachment difficulties to those of children in foster care and whose adoptive parents are typically highly motivated to receive support.17 It is also noteworthy that, although we encouraged local authorities to consider recruiting special guardians and friends and family carers to take part in the trial, many did not, largely because SGOs tended to be managed by different teams. More active engagement of these groups of carers in a future trial may also markedly increase the total pool from which recruits can be drawn. We also take the view that it may be worth considering implementing a trial of the VIPP-FC as part of universal foster carer training, which would largely remove most barriers to recruitment. Feedback we received during the study from stakeholders and foster carers provided considerable initial support for this suggestion, as both groups noted that ‘live’ training in attachment principles may be more effective as a means of enhancing foster carers’ professional skills than more pedagogical approaches commonly used in the foster care sector, while also providing more immediate and direct benefit to children.