What scientific studies have been done to study child placement/adoption?

October 2003

http://www.scie.org.uk/publications/knowledgereviews/kr02-summary.pdf

This review, by Dr Alan Rushton from the Institute of Psychiatry, King’s College London, set out to map published research literature on the adoption of children from public care, to identify gaps, and to suggest future directions for research.

The study, and this summary, will be of interest to people working in the field of adoption, and to those undertaking, commissioning or planning research.

The mapping shows that:

• Small-scale research has been conducted into agencies’ contribution to effective practice in recruitment, assessment, matching and preparation, but it does not provide a sound basis for choosing between alternative kinds of practice.
• Outcomes of placements with adoptive parents, early or late in the child’s experience of being looked after, have been studied but information on teenage placements is missing.
• More needs to be known about users’ view of services – especially the views of black and ethnic minority adopters, and birth parents.
• There are no comparisons of the cost-benefits of different packages of support services.

The study recommends:

• A large-scale, multi-disciplinary investigation of placement outcomes and factors influencing them, across all placement options.
• An urgent study of the levels and types of post-placement /post-adoption contact with the birth family.
• Evaluation of adoption support services, involving both short examinations of preparation and training, and randomised controlled trials of more specialist interventions.
• Relatively small-scale and fast studies concerning policy implementation and consumer views, which could make use of questionnaires and focus on a cross-section of the relevant user group.

Introduction

The review was deliberately restricted to a particular group of adopted children – children who were previously in public care – and to a particular kind of literature. Only published research literature is covered – practice and policy papers, inspection reports and websites, which would be covered by a full review, are not included. The review identifies the key research issues, but was not intended to examine individual studies critically or comprehensively. Existing reviews are highlighted where available.

The policy context for research in the adoption field is complex and raises strong feelings. Recent seminars funded by the Nuffield Foundation, for example, showed up major differences of opinion and policy towards adoption between countries. Some countries take a radical line against breaking ties with birth parents, or ending parental rights with birth families, when a child is in need or at risk. Sweden, for example, does not allow the adoption of children from care without parental consent. These countries favour a policy of family preservation, or placement with relatives but without adoption. But these strongly-held positions are rarely backed up by independent research data on outcomes.

In the UK, recent political initiatives have encouraged the greater use of adoption as a solution to the care of children who cannot live with their birth families. This drive for ‘permanence’ has been welcomed by many, but has also given rise to debate. Where children have lingered uncertainly and for far too long in the care system, ‘permanence’ is clearly the top priority. But not all children need the same solution. They may be ‘children who wait’, but they may not be waiting specifically for adoption. It is also argued that this concentration on adoption may discourage practitioners from taking an integrated view of all the available placement choices, to ensure that the best plan is made for each individual child.

Adoption research has grown over the last 40 years in quantity and quality. But it will always remain a challenging field of enquiry because of the complex concepts being examined, and because adoption policy and practice is a constantly ‘moving target’ for researchers. This review was designed to indicate the scope of contemporary adoption research, and to suggest what further studies are needed to provide a better evidence base for policy and practice.

Findings

The findings of the review are arranged in terms of five stages in the adoption process: recruitment and preparation, adoption outcomes, the child’s problems, contact arrangement with birth families, and adoption support.

1. Recruitment, assessment, matching and preparation

The recruitment of suitable, new adoptive parents is obviously fundamental to a successful adoption policy. More research is needed (among many topics) on:

• the nature of the initial contact with the recruitment agency and its relationship with follow-through; and
• comparative outcomes of placements with traditional and non-traditional groups.
• the cost-effectiveness of major recruitment drives, and other forms of recruitment;

The government is currently reviewing the adopter assessment process with the aim of improving the fairness, transparency and consistency of assessments across agencies. No studies, to date, have collected data at the point of assessment and related them to placement outcome. The implication of the work which has been done suggests that pre-placement assessment can screen out obviously unsuitable applicants, but that outcomes cannot be predicted at this stage. A research group at Coram Family/Great Ormond Street is trying to establish whether the attachment style of the new carers contributes or not to the child’s growth of attachment.

The systematic and holistic assessment of the children is supported by Kirby and Hardesty’s guidance on conducting detailed assessments of looked after children; and Quinton and Murray have recently discussed the assessment of the emotional and behavioural development of children looked after away from home. Practitioners also need:

• a widely accepted, easy to administer, brief assessment tool for a range of uses (the proposed Integrated Children’s System promises to provide a common approach to the assessment of children’s developmental progress); and
• further investigation of the concept of ‘readiness for placement’, already used by practitioners.

A recent review of the evidence on matching concludes that research has yet to provide clear indicators of what constitutes a good match. Independent research is needed into the matching process to establish whether factors before placement can be identified which increase the likelihood of good outcomes. One of the most contentious issues in matching the child or children to new parents is whether new parents should be selected according to racial and ethnic background. The arguments are summarised and the research comprehensively reviewed in Rushton and Minnis: it is now generally agreed that agencies should try to make the closest ethnic match whenever possible.

Future research interest is now more likely to focus on samples of transracially placed children and their adult adjustment and identities.

The practice literature indicates that models of preparing the new family vary considerably. They need to be evaluated for their relevance and effectiveness, and this will be made easier once standard pre-placement packages are more commonly used. It is also important to learn more about the ways in which adopters’ own children react to the arrival of a new child, and to understand better what adoption might mean for them and for their extended family.

Although preparation of children, especially ‘Life Story Work’, has been described and promoted by practitioners, there are no studies on how this subsequently affects the child’s development and placement. Research is needed:

• on a large scale, using a relatively standard and clearly specified method of intervention;
• into the cultural competence of agencies in their recruitment, assessment, preparation and support activities; and
• to establish which models are better received for different groups.

In order to counter drift in the care system, new permanency time-frames are being introduced, but it will also be important to undertake research to establish whether compromises are made in assessment, matching and preparation and consulting birth parents’ wishes in order to meet exacting targets.

In each of these areas (recruitment, assessment, matching and preparation) we still need to know whether there are different consequences for conducting the work in one way rather than another.

2. Adoption outcomes

The rates, causes and some of the immediate consequences of adoption disruption have been researched, and have found very positive outcomes for healthy infants who are adopted. But the outcomes for very young children adopted after neglect or abuse need further study.

A number of research reviews have recently been conducted on children placed from care for adoption, all of which have given some attention to outcome. These include:

• Sellick and Thoburn’s review which covers studies on all forms of temporary and permanent family placement;
• Adoption Now, which covers recent Department of Health (DoH) commissioned research;
• Rushton’s review of outcome literature; and
• Cohen, on what determines adjustment in children.

About a dozen studies have been published examining outcomes for children placed with non-relatives, after infancy. These have shown disruption rates of about 20 per cent, and even where children are placed in middle childhood, adoption appears to work well for most. However, the picture is much less positive for adolescents, and the full story is much more complex.

Adoption researchers agree that measuring outcomes is very difficult and is probably becoming more so as approaches involving multiple indicators are being developed. Reduction in the disruption rate should certainly remain as one key service target. But new developments have been taking place in assessing outcome in different kinds of placement, including adoption. As research teams try to measure outcomes for a variety of placement choices and are devising more useful classifications than just disruption or survival, there is a risk that they may produce differing classification systems, leading to lack of comparability of findings.

Selwyn, Sturgess, Baxter and Quinton are currently engaged in a longitudinal study of children placed for adoption between three to eight years in Avon. This will provide important new evidence on how to predict outcomes partly because important data will be available on the children’s pre-care experiences. Possibilities exist for other studies, which follow up samples gathered some time ago, and where the adopted children are now adolescents or young adults. Studies are also needed on the adoption of particular groups – adolescents who want to be adopted, for example.

The Department for Education and Skills is currently considering ways of establishing adoption disruption/survival rates nationally by checking on the frequency with which children re-enter care and the reasons for readmission. Detailed, interview-based research is probably also necessary.

An important question is not just how adoptions turn out, but how they compare with other placement plans. Simple comparisons can be misleading, and new studies with large samples are needed in order to control for differences in the populations. Adoption should be considered ‘worth the risk’ on the basis of studies of placements made late in the child’s care career, but evidence is emerging that adoptions clearly do not work for a minority. All the time, more is becoming known about the consequences of placement decisions. But the messages from research have not been as clear cut or as easy to translate into practice guidance as some might have hoped.

3. The child’s previous experiences, and current difficulties

Most studies of longitudinal data confirm that the factors associated with disrupted placements are:

• older age at placement
• the child’s adverse experiences in previous environments
• the level of behavioural difficulties at placement
• placement of a child in an established family with a resident child of similar age.

Research has not shown the sex and race of the adopted child to be associated with greater risk of placement instability, and disability carries less risk of disruption than emotional and behavioural problems. But beyond these basic findings, the meaning of other associations with risk becomes harder to unravel. An important field of study of great relevance to adoption is the growing research both on pre-natal and early experience, and on child maltreatment and its consequences, for all aspects of development and especially attachment.

Although it is likely that adopted children will have a whole range of problems, three groups stand out both in the children, and in the new parents’ descriptions of the difficulties which are hardest to handle:

• Relationship difficulties: the application of attachment theory to relationships in adoption has helped in understanding the origin and consequences of insecure attachment, but there is little certainty yet about cause and effect.
• Educational problems: in an early study of late placements, it was reported that school problems became more important as the placement progressed. It is not known how late-adopted children do throughout their school years and to what extent they succeed in higher education. It is also important to know what contribution educational difficulties have on placement stability and quality of family life. Future adoption research should look at developments in the child’s life beyond the immediate family.

Behavioural and emotional problems: these are extremely common in samples of placed children, although Brand and Brinich have shown that the difference in behaviour problems in adopted compared with non-adopted children is accounted for by a small proportion of children with a high level of problems. Longitudinal research has established which problems are likely to persist for adopted children over time. This information is vital in shaping effective work with parents and with children. prospectively, not just those cases currently known by an agency to be in contact.

4. Contact arrangements with the birth family after adoption

There have been some studies relevant to parts of the question, but no research group has yet set up a study to examine comprehensively the short and long-term effects on the children and the other involved parties of various forms of contact with the birth family. The design for such a study would need to produce evidence on the effect of contact itself while accounting for the many possible and confusing variables. It would have to examine a representative sample

Impact studies suggest that adoptive families can deal with contact; there is as yet no well-conducted study of effects.

5. Adoption support, the views of service users and the evaluation of interventions

There are many gaps in the literature on consumers’ views of adoption support. The views of the adoptive parents and service providers have been much more frequently canvassed than the views of the children or the birth parents. The views of black and minority ethnic adopters have only been gathered in small-scale research. Representative samples of the parties to adoption need to be studied, rather than those who are researched because they seek help. Filling the gaps in service user research will increase the prospect of adoption services being much more carefully tailored than in the past.

There is an urgent need for more information on how to support adoptive families facing the challenge of parenting adopted children with severe and continuing difficulties. Over the coming years, many more trials must be set up to compare different interventions to support placements in difficulty. They will need to demonstrate what works best for whom, and how to replicate the interventions to achieve the same results. At the same time, we need to understand the factors contributing to outcomes and the processes of change.

Recommendations

Adoption research is not by any means complete. Even given the restrictions on the scope of this review, it is clear that both complex scientific studies, and many smaller-scale consumer-oriented and policy implementation studies are needed.

The planning of the longer-term prospective, longitudinal studies needs to be considered now to make best use of all the activities generated by the government’s adoption agenda. This can be followed by less expensive and less complex studies which have a shorter time-frame. Some initiatives may need to be in place for a few years before a useful evaluation can be conducted.

Good quality adoption research has potentially very significant implications for children who cannot remain with their birth parents. The more that practice is underpinned by secure research-based knowledge, the more the risk of disruption and unstable placements should be reduced and with it, the need for continuing health and social care services for adopted children and adoptive families.

Social Care Institute for Excellence

Goldings House, 2 Hay’s Lane
London SE1 2HB
tel 020 7089 6840
fax 020 7089 6841
textphone 020 7089 6893
www.scie.org.uk

Braille, tape and large print versions of this summary can be made available on request.

A full version of the knowledge review is available on the SCIE website and in print from SCIE.