Psychological treatment outcomes have been shown to differ systematically between therapists, and more recently, between the clinical services providing treatment. However, the above effects could be confounded by the effects of other forms of social, clinical, or geographic clustering, which have been under-researched to date. In particular, differentially deprived neighbourhoods could partly account for apparent differences between services’ outcomes. This mixed-methods thesis aimed to
1) quantify neighbourhood effects on treatment outcomes, simultaneously with therapist and service effects,
2) investigate the extent to which deprivation might account for (or explain) these cluster effects, and
3) qualitatively understand the potential mechanisms by which neighbourhood effects might occur.
Treatment outcomes included clinical effectiveness (post-treatment symptom severity, recovery), and treatment dropout.
First, sophisticated multilevel models were used to analyse data from around 1,000,000 patients across 55 services, 15,000 practitioners and 20,0000 neighbourhoods. Significant neighbourhood effects were found, independent of therapist and service effects. Neighbourhood effects accounted for 1-2% of outcome variance, comparable to service effects in some contexts but smaller than therapist effects. Deprivation factors significantly explained neighbourhood effects but not service effects. Neighbourhood effects were primarily explained by a complex shared effect of deprivation and baseline severity factors. Understanding this effect required the development of a novel method to visualise explained variance. Findings were similar across outcomes (i.e. both for clinical effectiveness and dropout), and across treatment types (low intensity and high intensity treatment).
Next, data from the quantitative results were used to identify deprived neighbourhoods with either a) above expected or b) below expected clinical effectiveness outcomes (recovery rates). Interviews were conducted with a total of 12 participants from both neighbourhood types to understand 1) how neighbourhood and local deprivation might influence clinical effectiveness, and 2) the differences between these types of neighbourhood that might be targeted to improve effectiveness. Themes included lack of agency, experiences of threat versus safety, the importance of social support and community, and the quality of local environment. Neighbourhoods with below average effectiveness outcomes were particularly characterised by increased interpersonal/violent crime and antisocial behaviour, lack of community support, and abandoned, littered environments with less access to public facilities. Escaping the neighbourhood was an important theme in below average effectiveness neighbourhoods.
These findings suggest that the neighbourhood where a patient lives (and the associated deprivation they experience) matters for their treatment outcomes, and that there may be a range of factors beyond the therapy room that could be potential intervention targets to reduce treatment inequalities.
