Introduction

This is one of the AMBIT Editorial Group - current manualizing tasks that is work in progress.

The aim is to link this page directly to the Academic references that are also provided, so as to chart the areas where AMBIT is, and is not, evidence based.

What does "evidence-based" mean?

• Evidence can mean different things to different people.
• A Court, or member of the Police will understand "Evidence" to mean either a person's testimony ("she gave evidence that...") or, in the case of "forensic evidence", something more scientifically validated and (implicitly) more likely to trusted than a mere opinion or assertion.
• A scientist's understanding of evidence would be that it is a model of reality that has not yet been disproved, and which has a high statistical probability of being true.
• However, evidence that something is true in one setting, with one population, is no guarantee that the same will be true in a different setting, with a different population.
• In the world of mental health research, there has been much written about the nature, and the strength of the "evidence" for various assertions (see here) and we are not repeating this - however, it is important that the phrase "evidence-based" does not become a means of shutting down debate. There is a helpful lecture by Peter Fonagy that you can listen to about this here.

Is AMBIT evidenced based?

This is a very key question as the need for evidenced based practice is a key principle of the AMBIT approach (see Respect for Evidence). This page of the manual will outline a number of key questions for the AMBIT approach and provide some brief starting points as to how each question needs to be addressed. Each question will then have a separate page in the manual building a range of theory and evidence with links to other parts of the manual. We would welcome contributions to this part of the manual from local teams.

Not all of AMBIT (because there isn't a single "thing" called AMBIT, amongst other reasons) is "Evidence-based" but we encourage AMBIT-influenced teams always to be "Evidence-oriented".

Is the basic mechanism of the AMBIT approach, namely mentalization, supported by developmental science?

Yes. Mentalization has a substantial theoretical and empirical base with extensive studies of the association of mentalization to attachment theory and to parent-child interaction studies. This may seem unremarkable but it is not always the case that key therapeutic concepts are grounded in developmental science. For example, the idea of a 'core belief' in CBT arises from adult theory of CBT in which the developmental basis of such types of cognition remains unknown.

Can a mentalization based intervention be appropriately applied to working with hard to reach youth?

Yes. Mentalization based interventions have been shown to be effective for adults with borderline personality disorder and there are similarities between the types of presenting problems (impulsivity, self harm, poor affect regulation) of this group with adolescents seen by teams using the AMBIT approach. More significantly, a recent RCT trial of Mentalization-Based Treatment for Adolescents (MBT-A) showed that this approach was more effective than treatment as usual.

Can AMBIT training be effectively delivered to front line teams?

Yes. The basic AMBIT training has been successfully delivered to over one hundred in the U.K., Ireland, Europe, North America and Australia. Feedback from the training experience is generally extremely positive with high ratings for relevance to work experience and usefulness of ideas to practice. This feedback has been summarised in Feedback on AMBIT Training.

Are the basic processes of how AMBIT teams work together and how they work with other agencies supported by research on organisational factors associated with effective mental health interventions?

Yes and No. Many of the principles of good practice advocated by AMBIT are shared with other models. For example, the need to have clear treatment objectives, well measured outcomes and a shared joined up approach with other agencies are not controversial. The difference for AMBIT is that it proposes that these factors have a major impact of overall clinical outcomes. The degree to which such assertions are supported by evidence is equivocal and will be elaborated in further pages in this manual. For example, the evidence that more integrated intervention leads to better outcomes was not supported by the Fort Bragg study in the USA. In this important study, integrated interventions proved to be more expensive but not superior to treatment as usual.

Is there evidence that training in AMBIT results in changes in practice for a team?

Yes. Some of the teams who have been trained in AMBIT clearly adopt a number of the components of the model in their work as a team. This is demonstrated by their on-going feedback to the AMBIT Project and to the types of local evaluations that they carry out.
FOLLOW UP SURVEY NOVEMBER 2014
In November 2014 AMBIT Leads were asked to respond to a three minute AMBIT Follow up survey. The survey aimed to find out which parts of the AMBIT wheel teams are finding helpful, and which AMBIT words and phrases have been adopted by each team (a simple 'proxy indicator' for culture shift across a team, we felt). According to the survey results 87% felt that an outisde observer would have observed a small or significant positive change in their team since the training. If your organisation allows you access to Google Drive, then you can access results here.
It is clear that the degree of adoption of the approach varies between teams. In our view, this is part of a need for local adaptation of the model to local client groups and local service ecologies.

Is there evidence for improved outcomes following the introduction of the AMBIT approach?

As far as we are aware, none of the AMBIT teams that have been trained have formally compared outcomes before the training with outcomes following the training, although there is an emphasis on developing an active engagement in outcomes monitoring in AMBIT-influenced teams (such as CASUS in Cambridgeshire Engaging a team in Outcomes Measurement - Liz Cracknell and Carol Evans) and - for instance - Lothian CAMHS have shown a range of positive outcomes (AMBIT Across a Complex Array of CAMHS services - Gavin Cullen and Fiona Duffy) and Bexley have reduced their inpatient bed use (Using AMBIT to Reduce In Patient Admissions - Sarah Harmon).

Is there evidence that AMBIT trained teams were effective, for example, in meeting commissioning targets etc?

A number of local evaluations have taken place, nearly always in recently established teams which were set up with clearly specified outcomes targets set by commissioners. Perhaps the first of these was the team in Islington known as AMASS (Adolescent Multi-Agency Support Service) which was set up to work with adolescents in birth families at risk of family breakdown. The explicit objective of t setting up this service was to reduce the number of young people coming into the care system in adolescence. On-going evaluation indicated that 85% of referrals remained with the birth family after the AMASS intervention. A number of other examples of local evaluations by CASUS, Edinburgh, Bexley have shown similar positive results against commissioning intentions, particularly around reducing length and number of hospital admissions.

Is there any RCT trial data to support the AMBIT model?

No. One of the problems for AMBIT is that (as we repeat often in AMBIT trainings) "There is no such thing as an AMBIT team" - AMBIT is heterogeneous-by-design as it fosters local adaptation and is built on a model of constant improvement. Because of this it is difficult to establish usefully a definition of "what" is being tested, although there is work on defining measures of AMBIT fidelity (see the APrAT, for instance).
A great deal of work went into preparing a trial in Cambridgeshire Social Services but this was cancelled during the initial assessment phase due to changes in the social care structure and priorities. Although this would be a very useful addition to our understanding of the benefit (or otherwise) of the AMBIT approach, we would also want to emphasize that AMBIT embraces the use of evidence based Specific interventions within its overall framework. For example, we would advocate that AMBIT teams would adopt CBT approaches to social anxiety, motivational approaches to substance misuse and family approaches to eating disorders. AMBIT could be seen as a platform from which evidence based interventions can be delivered once some of the difficulties around the young people's anxieties about help seeking have been recognised and worked with.