Dear Health New Zealand,

Under the Official Information Act 1982 I request three aggregated tables. This is the conceptual successor to RT-268 (HNZ00033452) and RT-611 (HNZ00052991), specified below so it can be actioned independently.

(A) DEATHS TABLE

Long-format, one row per non-empty combination of:

- Month of death — YYYY-MM
- Age at death band — 0–20, 21–40, 41–60, 61–80, 81–100, 100+
- Last COVID-19 dose number — 0, 1, 2, 3, 4, 5, 6, 7, 8 (0 = no prior dose)
- Days between last vaccination and death — 0–7, 8–14, 15–28, 29–60, 61–90, 91–180, 181–365, 366+, N/A (N/A where dose number = 0)
- Sex — Male, Female, Another gender, Unknown
- Ethnicity (prioritised, MoH standard) — Māori, Pacific peoples, Asian, European or Other, Unknown
- Priority-group eligibility under which the last dose was administered — aged-care resident or staff; healthcare worker; severely immunocompromised (MoH definition); disability sector (Sequencing Framework Group 2); general population by age; other; unknown; N/A where dose=0
- Underlying cause of death, ICD-10 chapter, with U07.1, U07.2, V01–X59 (accidents) and X60–X84 (intentional self-harm) on separate lines rather than pooled — chapters I (with U07.1, U07.2 split), II, IV, VI, IX, X, XIV, XVIII, XX (with V01–X59 and X60–X84 split), Other (any chapter not listed)
- Count — integer; cells <6 reported as "<6"

Temporal window: deaths from 2019-07 onwards through the most recent extract date.

(B) DENOMINATOR (VACCINATED PERSON-TIME) TABLE

Long-format, one row per non-empty combination of:

- Month — YYYY-MM
- Age band — as in (A)
- Sex — as in (A)
- Ethnicity — as in (A)
- Dose number of the most recent dose received as of any day in that month — 1, 2, 3, 4, 5, 6, 7, 8
- Priority-group eligibility under which the most recent dose was administered — as in (A)
- Days since most recent dose — 0–7, 8–14, 15–28, 29–60, 61–90, 91–180, 181–365, 366+
- Person-days — integer; cells <6 reported as "<6"

A person crossing a days-since-dose bin boundary mid-month splits person-time proportionally. A person receiving a new dose mid-month contributes the remaining days to the new dose number's 0–7 day bin.

Temporal window: vaccinations from 2020-12 onwards through the most recent extract date.

If full person-time accounting is infeasible, end-of-month head-counts split into the same bins is an acceptable substitute.

(C) PRE-VACCINE BASELINE MORTALITY TABLE

Long-format, one row per non-empty combination of:

- Year — 2015–2020 inclusive
- Age band — as in (A)
- Sex — as in (A)
- Ethnicity — as in (A)
- Underlying cause of death, ICD-10 chapter — same scheme as (A)
- Count — integer; cells <6 reported as "<6"

Format: CSV or XLSX, UTF-8.

Suppression: <6 cell suppression as used in RT-268 and RT-611 is acceptable; please indicate which cells have been suppressed or rounded.

BASIS AND ACCEPTABILITY

This request is reasonable under the OIA because:
- Data aggregation is explicitly allowed by the Ombudsman. You are required.
- This OIA requires aggregating from similar IT systems and datasets for previous OIAs already responded to by HealthNZ. Precedent.
- I am a tax resident and citizen permanently residing in NZ.
- These aggregation effort is not substantial for a data analyst familiar with the relevant systems and/or the data warehouse infrastructure HealthNZ has invested in.
- The NHI number Tues all systems together for easy aggregation of datasets.
- Providing this data is of high public interest and should already be tabulated, or able to be aggregated, as part of normal operations for safety monitoring as part of HealthNZ’s regulated roles and obligations.
- Privacy is assured by having a limit of 5 for reporting in an aggregated cell. This is standard practice and has been undertaken and complied with by many government agencies for OIAs including HealthNZ.
- The frequency of such follow-up OIA requests from myself is low and appropriate to the topic and analysis needs. Any overlap with other requests is primarily due to the delay in response by HealthNZ to those requests.
- It is worth noting that the timeframe in previous OIAs from the data extract date to the response date is often only a few days. This indicates what is possible once a decision has been made to respond.

WAYS OF WORKING

If HealthNZ has any questions or concerns regarding this OIA then you are required to Consult with me. It would be a sign of good faith, if the need arises, to consult as soon as possible so that HealthNZ has the best opportunity to meet its response timeline.

I look forward to HealthNZ’s prompt response within the regulated timeframe of 20 days.

Yours faithfully,

Chris Johnston