B.2 Data and methods
The majority of data used in this appendix was provided to the Inquiry on an anonymised and aggregated basis by the Ministry of Health and Health New Zealand | Te Whatu Ora. Key measures included in these datasets – and their sources – are outlined in Table 1.
Table 1. Key measures used in epidemiological overview
| Measure | Data Source | Notes |
|---|---|---|
|
Population |
Health Service User 2022 population
|
This dataset includes all people enrolled with a primary healthcare provider or who received services from a healthcare provider in New Zealand in 2022. While it covers a very high proportion of the population living in the country at that time, it does not include individuals who had no contact with the health system in that year. It may also have included some people who were not living in New Zealand but who had received healthcare in the country at some point during the year.1
|
| COVID-19 cases | EpiSurv (national notifiable disease surveillance database), National Contact Tracing Solution |
Until February 2022, new cases of COVID-19 infection were detected via PCR tests conducted by health workers. From late February 2022, most new COVID-19 cases were detected through self-administered rapid antigen tests (RAT tests) with members of the public asked to self-report any positive test via an online portal. The proportion of COVID-19 infections being detected and reported declined under the new testing regime – meaning new cases are likely to be underestimated from March 2022 onwards, with possible differences in detection by age, ethnicity, deprivation and other characteristics. |
| COVID-19 hospitalisations | National Minimum Dataset (hospital events) | Hospitalisation for COVID-19 was determined retrospectively based on the diagnostic codes relating to the specific hospital admission. It does not include people admitted to hospital for other reasons who were found to also have COVID-19 infection, unless that infection subsequently became a contributing reason for their hospital stay. |
| COVID-19 deaths | National Health Index database, national mortality data | Deaths attributed to COVID-19 are deaths where COVID-19 was listed as either the underlying or a contributing cause of death. |
| COVID-19 vaccinations | National Immunisation Register | While eligibility for COVID-19 vaccination was initially limited to those aged 16 and older, the age-threshold had been expanded to include 12–15-year-olds by the time the vaccine rollout had reached younger agegroups (in the latter part of 2021). Calculating vaccination coverage is complicated by younger individuals becoming eligible during the period under study (so moving from outside to inside the eligible population). For this reason, vaccination coverage in this appendix is usually calculated for the population aged 15 years and older, based on the age people were at the beginning of 2022. |
Categorisation of demographic factors (age, sex, ethnicity and socio-economic deprivation) is based on information recorded in the National Health Index dataset.2 Age was calculated at 1 January 2022, based on a person’s date of birth. Sex is recorded as male or female.i Ethnicity is self-identified and has been categorised as prioritised ethnicityii (Māori, Pacific and non-Māori non-Pacific or ‘Other’) for consistency with analyses undertaken by the Ministry of Health.3 Deprivation refers to neighbourhood socio-economic deprivation, based on a person’s residential address (on 1 January 2022) linked to the 2018 New Zealand socio-economic deprivation index.4 For presentation purposes, deprivation is categorised in three groups (least deprived/New ZealandDep deciles 1–3, mid-range deprivation/deciles 4–7, most deprived/deciles 8–10) or as quintiles (from quintile 1/least deprived to quintile 5/most deprived).
Data on COVID-19 cases detected at the border versus in the community were sourced from the Ministry of Health’s public COVID-19 data website.5 Dates of policy changes regarding border restrictions and application of ‘lockdowns’ (that is, settings-based restrictions) were sourced from the official COVID-19 timeline developed by the Department of the Prime Minister and Cabinet.6
Data for cross-national comparisons (of COVID-19 deaths, excess mortality and the stringency of policy measures) was obtained from Our World in Data.7 Data on other major causes of death in New Zealand was obtained from the Global Burden of Disease Study 2021.8
Most figures present numbers (of COVID-19 cases, hospitalisations or deaths) as an incidence rate or numbers per head of population for a given time period. Where incidence rates are compared between population groups (defined by ethnicity or deprivation), the data is standardised for age. (This is in order to filter out any differences due to the different age-structures of the groups being compared.) Rates were standardised to the World Health Organization World Standard Population. The majority of data visualisations presented in this appendix were undertaken by the Inquiry secretariat using R statistical software.
Comparison of Years of Life Lost (YLL) to different diseases / under different counterfactuals (Figure 7) was undertaken by Professor Tony Blakely using Microsoft Excel. YLLs to different diseases for the population of Aotearoa New Zealand were taken from the Global Burden of Disease Study 2021.9 YLLs for COVID-19 deaths were estimated based on Datta et al (2024)10 and Milkovska et al (2024).11 For the counterfactual scenario of New Zealand having no vaccination, figures generated by Datta et al (2024) from standard lifetables were used to estimate that each person dying from COVID-19 would have had an average of 11.2 years of remaining life had they not become infected with COVID-19 (i.e. 74,500 YLL divided by 6,650 deaths).12 We scaled this estimate to account for the greater prevalence of co-morbidities in people dying from COVID-19 – meaning they would have fewer years of remaining life expectancy compared with the average person of the same age. Milkovska et al (2024)13 estimated that a person dying of COVID-19 had on average 30 percent fewer remaining expected years to live compared with someone of the same age who did not die from COVID-19. Based on this estimate, we adjusted Datta et al’s estimate of 11.2 years down to 7.8 YLLs per COVID-19 death. The burden of morbidity due to COVID-19 in New Zealand was estimated using the ratio of YLDs to YLLs from Howe et al (2023),14 who estimated the burden of disease from COVID-19 in an Australian study. Findings from Datta et al (2024) were also used to estimate YLL under the counterfactual of New Zealand having had no COVID-19 vaccination but otherwise the same strategy and timeline for moving out of elimination and removing border restrictions.
Comparison of risk for COVID-19 hospitalisation and death (Tables 2 and 3) was undertaken by the Public Health Agency, Ministry of Health. Risk ratios and 95 percent CIs were estimated using Poisson regression with robust standard errors. Analyses were undertaken using STATA MP/18.0 (StataCorp, LLC) statistical software. Data presented here is preliminary but was shared with the Inquiry in order to inform its findings.
i The National Health Index records sex with categories limited to male, female, unknown and indeterminate (this last is largely used in relation to newborn babies). At the time of writing, it does not include a category for gender.
ii ‘Prioritised ethnicity’ means that – for presentation of data on ethnicity – people are assigned to a single ethnic group in a given order of priority, even if they identified with more than one ethnic group. The priority commonly used in Aotearoa New Zealand is Māori, Pacific Peoples, New Zealand European and other ethnic groups.