We are in agreement with some of the points made in the recent article by Tregeagle, ‘Weighing up the evidence and local experience of residential care’ (Children Australia, 42(4), 240–247). For example, there can be no dispute about the high costs of residential placements or that achieving a stable residential environment is very challenging. Table 1 provides a three state cost comparison of residential placements (Ainsworth, Reference Ainsworth2017).
A lack of a trained residential workforce and limited staff career prospects, all of which lead to high staff turnover, which in turn contributes to an unstable residential environment, are also major issues. It is also beyond dispute that achieving a stable residential staff team is a mighty challenge, not least of all because of the Australian child welfare system's reliance on small group homes with low level staff remuneration for what are onerous staff roles (Ainsworth, Reference Ainsworth2018). However, apart from these essential points, we largely disagree with Tregeagle's assertions.
Population to be Served
First, in addressing the population to be served, there is the question as to whether even the most specialised and highly regarded foster care programmes – such as the Oregon multi-dimensional treatment foster care model (MTFC), which is now known simply as Oregon Foster Care (OFC) – can successfully treat all comers (Chamberlain, Reference Chamberlain2003). Such a notion has certainly been canvassed even though it can be argued that residential placements serve a different population, have a different purpose and inevitably incur different costs (Ainsworth & Hansen, Reference Ainsworth and Hansen2015). It also has to be remembered that OFC re-places young people who are not responding to the treatment model, of which there are inevitably a few back into residential placements.
It might also be asked why, if as Tregeagle asserts, all young people who are in care can be looked after more appropriately in family foster care, all the 26 New South Wales (NSW) residential service providers do not abandon this type of service in favour of foster care? Or, why the Department of Family and Community Service in NSW continues to fund residential programmes if they are not needed?
It may of course be that Tregeagle, given that she is reflecting on the experience of but one agency, Barnardos, is citing atypical experience. Or that Barnardos is more skilled at this type of practice than the other NSW service providers. Whichever way it is, we suggest that to generalise in this way and claim that the Barnardos’ experience validates the argument that all young people in care can be served in family foster care and that residential services are no longer needed is unsafe. Indeed, we have the Minister for Family and Community Services in NSW saying that ‘the closer you looked at the foster care system and outcomes for these children the more you knew that is was dreadful’ (Berkovic, Reference Berkovic2018).
This is compatible with the earlier position taken by Ainsworth and Hansen (Reference Ainsworth and Hansen2014) when they asked the question ‘Family foster care: Can it survive the evidence?’ Needless to say we think the Tregeagle position is an unsustainable exaggeration.
The International Research Evidence
Matters of definition are often important and it has to be noted that Sanctuary® cited in this article as a programme is in fact a ‘‘platform’’ as it ‘is a full systems approach that targets the entire organisation with the intention of improving client care’ (Ainsworth, Reference Ainsworth2017). Children and residential experiences (CARE) is similarly regarded as a platform as it also is an approach to creating conditions for organisational change. The term ‘’programme’’ is reserved for ‘’client specific interventions’’ (Ainsworth, Reference Ainsworth2017). Thus, Mclean (Reference McLean2016), as quoted by Tregeagle, is in error by citing both Sanctuary® and CARE as programmes. The error is compounded by the claim that ‘there is little evidence to support or distinguish between the relative effectiveness of the two models’ (McLean, Reference McLean2016, p.14). On the contrary, research support for the CARE platform developed and disseminated by the Residential Child Care Project at Cornell University earned it a Scientific Rating of 3 (Promising Research Evidence) and a rating of High Relevance by the California Evidence-Based Clearinghouse for Child Welfare (CEBC) (2018). The CEBC is a critical tool for identifying, selecting and implementing evidence-based child welfare practices that will improve child safety, increase permanency, increase family and community stability and promote child and family wellbeing (http://www.cebc4cw.org/program/children-and-residential-experiences-care/detailed). CARE is a principle-based programme designed to enhance the social dynamics in residential care settings through targeted staff development and ongoing reflective practice. Using an ecological approach, CARE aims to engage all staff at a residential care agency in a systematic effort to orient practices in ways that provide developmentally enriched living environments and to create a sense of normality for youth. CARE is organised around six principles related to attachment, trauma recovery and ecological theory.
Using a quasi-experimental design, CARE had an impact on the prevention of aggressive or dangerous behavioural incidents involving youth living in group care environments in 11 agencies (Izzo et al., Reference Izzo, Smith, Holden, Norton-Barker, Nunno and Sellers2016). Measures included monthly administrative reports of behavioural incidents and the Organizational Social Context (OSC). Results indicated that there were significant programme effects on incidents involving youth aggression toward adult staff, property destruction, and running away. Effects on aggression toward peers and self-harm were also found, but were less consistent. In addition, the quality of interactions between the young people and adults improved, as did the young person attachment measures using the Inventory of Parent and Peer Attachment as an anchor in the surveys (Sellers, Reference Sellers2017).
Another interrupted time series study examined the impact of CARE on the interactional quality among staff and youth in therapeutic residential care (TRC) (Nunno et al., Reference Nunno, Smith, Martin and Butcher2017). Data were collected over 12 years and divided into a 6-year baseline phase prior to the start of CARE in January 2009 and a 6-year implementation phase. Measures include the OSC and behavioural report incidents. Results indicated that CARE implementation reduced the prevalence of critical incidents, and that reductions are sustained following the 3-year implementation period.
The Sanctuary model® has been listed with a Scientific Rating of 3 (Promising Research Evidence) and a rating of Medium Relevance by the CEBC since 2006. The link to the listing is: http://www.cebc4cw.org/program/sanctuary-model/detailed. A recognition that trauma is pervasive in the experience of human beings forms the basis for the Sanctuary model's focus not only on the people who seek services, but equally on the people and systems who provide those services. Sanctuary has been used in organisations that provide residential treatment for youth, juvenile justice programmes, homeless and domestic violence shelters as well as a range of community-based, school-based and mental health programmes.
In a randomly assigned intervention in residential treatment units, the Sanctuary model showed positive outcomes (Rivard, Bloom, McCorkle, & Abramowitz, Reference Rivard, Bloom, McCorkle and Abramowitz2005). Measures included the Child Behaviour Checklist (CBCL), the Trauma Symptom Checklist for Children (TSCC), the Rosenberg Self-Esteem Scale, the Nowicki-Strickland Locus of Control Scale, the peer form of the Inventory of Parent and Peer Attachment, the Youth Coping Index and the Social Problem Solving Questionnaire. No significant differences were found between groups at baseline or at 3 months. At 6 months, there were a few differences showing a positive effect for the Sanctuary model. Young people in the Sanctuary model units scored lower on a measure of coping strategies that tend to increase interpersonal conflict or minimise or exaggerate interpersonal issues. He/she also exhibited a greater sense of personal control as measured by the Locus of Control Scale. Finally, he/she reduced use of verbal aggression, while control participants scored higher on verbal aggression over time. Staff also completed the Community Oriented Programs Environment Scale (COPES) which assesses aspects of the functioning of the therapeutic community. There were no significant differences between conditions at baseline and at 3 months. At 6 months, units using the Sanctuary model scored significantly better on the total scale and on the subscales of Support, Spontaneity, Autonomy, Problem Orientation and Safety.
Another study in 2015 indicated that the girls’ secure juvenile justice facility at North Central Secure Treatment Unit Girls Program (NCSTU) in Pennsylvania was a safer place for both residents and staff in 2012 after Sanctuary implementation (Elwyn, Esake, & Smith, Reference Elwyn, Esaki and Smith2015).
Tregeagle then again cites McLean (Reference McLean2016) who claims that overseas models deal with different population groups, such as children who have mental health issues. Yet, one author of this response is linked to a NSW residential programme where, in 2017, the percentage of young people at the point of admission to the programme who had a mental health diagnosis was 67.6 per cent. While this is evidence from only one agency, it is unlikely that this finding is dissimilar to other residential programmes in NSW.
Noticeably, the NSW Department of Families and Community Services has recently brought in two overseas models namely Multi-systemic Therapy for Child Abuse and Neglect (MST–CAN®) and Functional Family Therapy through Child Welfare (FFT-CW®) in an attempt to reduce the number of children being taken into state care (Berkovic, Reference Berkovic2018). So, why should overseas programme models of TRC (Ainsworth, Reference Ainsworth2015) and treatment that are evidence based, such as the Boys Town Teaching Family Model (TFM) (Thompson & Daily, Reference Thompson, Daly, Whittaker, del Valle and Holmes2015) or the Starr Commonwealth's Positive Peer Culture (PPC) (Vorrath & Brendtro, Reference Vorrath and Brendtro1985) or the EQUIP programme (Gibbs, Potter, & Goldstein, Reference Gibbs, Potter and Goldstein1995) not be brought in also?
Furthermore, in commenting on McLean's (Reference McLean2016) remarks about residential staff recruitment and retention, Tregeagle says ‘it is unlikely in Australia, for example, that programmes would be able to employ staff with PhDs as they do in the United States’ (Tregeagle, Reference Tregeagle2017, p. 241). In the US residential programmes invariably have a capacity greater than Australian residential programmes (Ainsworth, Reference Ainsworth2018). It is in these larger programmes that PhD qualified staff are employed – more often as a CEO or as a Director of Policy, or as a Director of Research (just like Tregeagle herself) rather than in lower level positions. This trend is just visible in Australia and it is likely to grow over the next decade. In the US, it is most unlikely that persons with PhDs will be employed in less senior positions for the obvious reason of cost.
There are also 2 US journals Residential Treatment for Children and Youth and the Journal of Emotional and Behavioural Disorders that regularly publish research studies. This goes some way to rebut the claim that there is an absence of research studies about the outcomes of TRC.
Clearly, McLean and Tregeagle need to develop a richer understanding of US programmes and the US residential care service delivery system to increase the accuracy of their assertions.
What Happens Elsewhere?
It is important that we access the knowledge of other countries, but this is linked to an understanding of the scale of residential care in countries, such as US and UK. For instance, in US, the Association of Children Living Centres (ACRC) has 160 agency members and holds an annual conference that is devoted to residential service matters including research. And in the UK, a recent Ofsted report from England (Ofsted, 2017) noted the existence of 2,061 children's homes. Ofsted is the Government inspection agency located within the UK's Department of Education and a subsequent report rated 82 per cent of these homes as good or outstanding (Schooling, Reference Schooling2018).
Even in Australia residential facilities are numerous. In NSW there are 26 residential service providers who served 670 children and young people in 2016 (NSW Family and Community Services, 2017). In QLD there are 109 residential facilities (Queensland Government, 2013) and in Victoria 191 residential services for children (Victorian Auditor-General's Office, 2014). And in the 1980s, but not now, there were small residential care associations in Western Australia (WA), South Australia (SA) and Queensland (QLD) that offered some workforce training plus TAFE vocational certificate courses in residential care. These all disappeared as the move away from the use of residential placements in favour of foster care gained pace in the 1980s. Now the hope of a trained residential services workforce, as Tregeagle indicates, is a distant dream. Given that there were over 2,000 children and young people in residential programmes in 2015 (Australian Institute of Health and Welfare, 2016), often because of an inability by agencies to find alternative community based foster care placements for every child, the Barnardos experience as reported by Tregeagle is unlikely to be repeated and is also likely to be a distant dream. In fact, the NSW child welfare system depends on a range of services, including residential services that support other parts of the service system including foster care.
In conclusion, the Tregeagle article seems to be one more attempt to argue that the NSW child welfare system does not need any residential programmes. This is a false claim. As long ago as 2005 this dream was shown to simply push some vulnerable young people out of the child welfare system into other systems that cater for homeless youth, or worse still, into juvenile justice institutions (Ainsworth & Hansen, Reference Ainsworth and Hansen2005). We might also now add accommodation in a motel room that often costs significantly more than a place in a TRC programme.
Let it be asserted that a mature child welfare system requires, and will always require, some residential programmes, though for the few not the many (Ainsworth, Reference Ainsworth2017). The issue is how is NSW and Australia as a whole going to build the programme expertise and a skilled workforce to meet the needs of an increasing number of children and young people with emotional and behavioural difficulties? Ignoring the issue will not make it go away.