Residential treatment for people with moderate and severe addictions is an expensive yet popular intervention. A broad range of interventions has developed, including pharmacological, psycho-social and spiritual, either alone or in combination. However, despite decades of use and many trials, there remains little evidence to help guide decisions on what treatment is appropriate for which people and there is conflicting evidence about the relationship of time spent in treatment to successful outcomes. Moreover, patients with psychiatric co-morbidities, who are recognised as forming a large proportion of the population with severe substance use disorder (SUD), and the population most at risk and often hardest to reach, have been excluded from many large-scale trials, raising questions about the relevance of findings to clinicians. Furthermore, there is a scarcity of evidence relating to the appropriateness of these treatments for indigenous people.

Realist evaluations are being used more frequently in the health sciences because they offer opportunities to unravel the complexity of interventions through a focus on uncovering the causal mechanisms and answering the questions – what works, for whom, in what circumstances and over what period of time?

This presentation will describe how a realist evaluation is being used to answer these questions for a residential treatment programme for adults with severe substance use disorder in New Zealand. It will explain how a realist lens influences data collection and analysis; set out the findings from the first study, which developed an initial hypothesis for a theory for change, and the supporting causal mechanisms; and also highlight emerging findings from a realist synthesis of the literature that is seeking to identify potential theories that explain the outcomes from residential treatment for adults with severe substance use disorder.
 
Click here for presentation slides.