Statistical analysis

Descriptive analyses of the sample included comparisons (frequencies) of socio-demographic characteristics between migrant groups and the settled majority, and the testing of any differences between groups (using χ²). The logistic regressions (with CIs at 95%) were run for all psychotropic medications and all migrant groups compared to the settled majority, using four models. The first model was unadjusted, the second was adjusted for demographic factors (age, gender, and marital status), the third was adjusted for physical health, and the final model was adjusted fully for socio-demographic factors (employment, car availability, housing tenure, and urbanicity). These models provided an overall analytical picture of differences in prescriptions and the impact of socio-demographic and economic factors. Next, the adjusted models were applied to different psychotropic medication types broken down by migrant group in order to observe any group differences, and prescription type differences, in the use of psychotropic medications.

A sensitivity analysis was conducted that included only people who were born in Northern Ireland in the settled majority group. No significant difference was found in the number of psychotropic prescriptions between different definitions of the settled majority (p = 0.10).