A multi-level study of the determinants of mental health service utilization.

BACKGROUND: Until now, research has focused on neighbourhood variations in mental health services and their relationships with local attributes, such as healthcare supply and socio-economic deprivation, without controlling for individual characteristics (age, sex, income, or education, for instance). Hence, this study is a major attempt to clarify the role played by individual and local attributes in the utilization of mental health services. The aim of this study was to disentangle individual and neighbourhood effects on mental health service use.

METHODS: In this cross-sectional study, individual-level data on 423 participants with a frequent mental health disorder was recruited from the general population and linked to neighbourhood-level data at the census tract level from the 2006 Canadian Census. Neighbourhood variables included socio-economic deprivation, mean income, residential stability and the proportion of recent immigrants. Individual characteristics included gender, age, marital status, self-rated mental health and the number of diagnoses. Multi-level logistic regression was used to assess the effects of individual and neighbourhood characteristics simultaneously on mental health service use.

RESULTS: The intraclass correlation coefficient indicated that 12.26% of the variance of mental health service utilization is at the neighbourhood level. Final analysis showed that at the individual level, being female, married, or self-rating mental health less than excellent increased healthcare use. At the neighbourhood level, deprived socio-economic neighbourhood decreased health service use (OR=0.71, P<0.05), while residential stability increased use (OR=1.24, P<0.05).

CONCLUSIONS: Individual and neighbourhood characteristics determine mental health service utilization. Taking both into consideration allows better targeting of health service policy and planning and enables more accurate needs-based resource allocation. However, future research should continue to investigate the pathway through which neighbourhood affects health service utilization.