The Integrated Data Infrastructure (IDI) is a collection of de-identified administrative datasets (e.g., on health events, justice contacts, education enrolments, tax paid) that have been linked at the person-level for the whole New Zealand population. The IDI ‘spine’ dataset contains records for all those (ever) resident in New Zealand (either temporarily or permanently), and is formed by triangulating information from birth records, immigration records, and tax records. All other administrative datasets link directly to the spine and in the process create a whole population cohort with more-or-less complete information of contacts with each government sector. Both deterministic and probabilistic have been used, with false positive rates estimates to be < 2%.

The IDI is a resource for use by government and academic researchers, but cannot be used for either operational or regulatory purposes. IDI data is made available through strict access conditions, e.g., both the researcher and project must be approved, attempts at re-identification are strictly prohibited, and access is only available via a Statistics New Zealand data lab (no data is ever released outside of the datalab environment and access is not available outside of New Zealand). Privacy impact assessments of the IDI have indicated a low risk of disclosure, and qualitative research has indicated a favourable public perception.

Initial users of the IDI have tended to come from the government sector, though academic researchers are beginning to explore its use. For example, in the health sphere, the ‘Virtual Health Information Network’ has been set up to facilitate the use of whole population health data, and this groups includes >50 health researchers, each with an interest in undertaking research using the IDI. Methodological research is needed on potential biases in IDI data and how these might be overcome or accounted for.