Fig 1.
Flow diagram of the source dataset in the IDI (Top row), the specific data tables utilised (Middle Row) and the variables that were extracted for our analyses (Bottom Row). *Where data was not available in the Census dataset, this was supplemented from data in the mortality dataset. **Where address data was not available from either Census or mortality datasets, this was imputed from the SNZ derived variables dataset.
Fig 2.
Distribution study cohort by status at the end of the study period and deprivation measured by IMD2013.
IMD2013 Decile 1 represents the 10% least deprived areas in New Zealand and Decile 10 represents the most deprived.
Table 1.
Timing of last contact for suicide decedents and individuals alive at the end of the study period.
Table 2.
Univariate and adjusted odds ratios for base model and full model.
Fig 3.
Unadjusted (univariate) and adjusted odds of suicide associated with contact primary (A), secondary (B) and tertiary (C) services. Adjusted odds control for the Base Model variables of sex, age group, ethnic group, marital status and NZIMD (No Health). The adjusted odds ratios are also available in S1 Table.