Understanding the effects of neighbourhood characteristics and ethnicity on long-term outcomes of psychosis

Background Healthcare inequalities among people from black and ethnic minority backgrounds persist UK-wide and internationally, particularly in psychosis - a distressing, disabling and costly mental health condition characterised by hallucinations, delusions, and cognitive changes. Minority ethnic groups experience higher rates of psychosis, treatment delays, and involuntary hospitalisations.

Existing research, mainly from urban settings, highlights population-level social environmental factors in psychosis aetiology.

However, gaps remain about social environment features (deprivation, inequality, ethnic density, population density) that affect longer-term psychosis recovery among ethnic minority populations across rural and urban settings.

Predicting influences of socio-environmental factors and identifying people at risk of poorer outcomes can enable more effective prioritisation of limited NHS resources and reduce overall costs.

Aims, objectives, and hypotheses This fellowship aims to elucidate inequalities in psychosis treatment and recovery by examining the relationship between treatment experiences, service use and clinical outcomes; and ethnicity and neighbourhood characteristics (e.g., deprivation, ethnic density, population density, rural/urban status).

Objectives are to conduct: A qualitative systematic review (SR) on psychosis treatment experiences of patients/carers from minority ethnic backgrounds. A meta-analysis of clinical/social longitudinal psychosis outcome studies, examining ethnic group differences.

Longitudinal analyses of de-identified clinical records to (a) assess ethnic differences in readmission rates, hospital admission length, diagnostic course, mortality and treatment, (b) compare these outcomes by rural and urban status, (c) examine associations between neighbourhood characteristics (deprivation, population density, ethnic density) and outcomes from (a).

Longitudinal study hypotheses: Readmission rates, inpatient admission length, mortality, diagnostic course and treatment will vary by ethnic group, with poorer outcomes in black African, black Caribbean and Mixed ethnicity patients compared to white British patients. Rural status will predict reduced access to treatment compared with urban populations.

Lower ethnic density and higher deprivation will predict higher readmission rates and longer admissions, across rural and urban areas.

Methods Objective 1: qualitative SR examining perspectives of patients and carers from minority ethnic backgrounds about psychosis treatment. Papers will be analysed using thematic synthesis.

Objective 2: quantitative SR and meta-analysis of longitudinal clinical/social psychosis outcome studies by ethnic group.

Objective 3: de-identified NHS electronic health records of psychosis patients (South London, n=558; Cambridgeshire, n=c.300) will be examined at first presentation and 5, 10, 15-year follow-up.

Primary outcomes are: hospital admission (including compulsory admission), treatment (medication, psychological therapy), diagnostic course, and mortality.

Social environment data (ethnic density, population density, deprivation) will be included via the Office for National Statistics. PPI Patient and carer perspectives are embedded throughout. A Lived Experience Advisory Panel will meet regularly, contributing to study design, interpretation, and dissemination.

Timelines Work package Months 1: Training and Development 1-30 2: PPI 1-72 3: Qualitative SR Completed prior to fellowship 4: Meta-Analysis 1-6 5: Empirical Studies 7-53 6: Interpretation and Dissemination 54-72 Impact and dissemination Results will highlight differences in treatment, service use, and outcome across ethnicities and urban/rural areas.

The sample is larger and follow-up longer than many previous studies. Findings will inform development of tailored and targeted preventative, psychosocial interventions. Outcomes will be disseminated widely across academic and community settings.