Evaluating physical urban features in several mental illnesses using electronic health record data

Area of Science:

• Environmental Psychology
• Psychiatry
• Urban Health

Background:

• Urban environments are increasingly recognized for their potential influence on mental health.
• Previous research suggests a correlation between urban living and the prevalence of severe mental illnesses (SMI).

Purpose of the Study:

• To investigate the association between physical urban environmental characteristics and clinical outcomes in individuals with SMI.
• To identify specific urban features that predict SMI prevalence, psychiatric hospital admissions, and antipsychotic prescribing patterns.

Main Methods:

• Utilized data from 30,210 individuals with SMI from the South London and Maudsley NHS Foundation Trust (SLaM) via the Clinical Record Interactive Search (CRIS) tool.
• Examined 28 urban features and 6 clinical variables, employing five machine learning regression models and Self-Organising Maps (SOM) for analysis and visualization.

Main Results:

• A higher prevalence of SMI, psychiatric hospital admissions, and antipsychotic prescriptions were observed in urban areas.
• Machine learning models could not accurately predict clinical outcomes based solely on urban environmental data.

Conclusions:

• While urban environments are associated with increased SMI prevalence, they do not solely explain variations in psychotic disorder prevalence or clinical outcomes.
• Clinical outcomes in SMI are likely influenced by a combination of urban factors and individual patient-level determinants.