Dear Health New Zealand,

Please expand on Health NZ’s previous response to OIA HNZ00061156 by breaking the numbers down into Age Range and number of doses of any COVID vaccine received by the person at the date/time of arrival at the Emergency Dept.

The response should be in a machine readable format - in a csv or Excel format is acceptable.

To recap, the original OIA read: ‘ The number of people under the age of 40 presenting to Emergency departments throughout
NZ hospitals with Chest pain or heart issues for the years 2017 to 2024 ? (Please provide figures year by year)’

Further breakdown requested in this OIA:
1) Age Ranges - under 40, 40 to 65 inclusive, over 65

2) Doses - 0,1,2,3,4,5,6,7+

3) Where a person had multiple visits to an Emergency Dept in a year please provide two columns - a total count and a unique count of NHI numbers.

Cell obfuscation:

a) Where a cell is <5 people please populate this with "<5" for confidentiality.

b) True zero cells should be shown - as zero and not marked as <5.

Please do consult as early as possible if there are any issues or questions. This a distinct OIA from my previous ones as it breaks down another person’s OIA and is in the public’s interest that it be responded to separately.

This is similar in intensity and effort to previous OIAs for data that Health NZ has successfully responded to. It may require linking the hospital dataset to the pay per dose/immunisation database via the NHI Number - or your data warehouses (eg Snowflake) may have already structured the required dataset - eg for safety monitoring and alarming.

Yours faithfully,

Chris Johnston

Dear hnzOIA,

Thanks for your reply.

I include comments below.

"The apparent large rise over the last six years is a result of incomplete data collection, not an
actual rise in chest pain ED presentations. What appears as a large rise in the number of ED
presentations due to chest pain is a direct result of the way the data was previously collected,
which did not include all districts, and was provided without the appropriate caveats and context."

The request is for the data "as is".

I welcome you providing the appropriate caveats and context. Can you please confirm that the following is the complete list of caveats and context? I notice that you have not mentioned which districts were not collecting this data in each year.

"We would note that 2024 is the only year that all districts at Health NZ were recording this data and
only a few districts (2-4) were recording this data prior to 2020. Additionally, not all 20 districts use
the same code for recording this data, so the data set is incomplete."

Health New Zealand state that:
"Chest pain on presentation to ED can be attributable to many conditions and the initial reason for
presentation often differs from the final diagnosis. The data from discharge codes – i.e. the final
diagnosis made when a patient is discharged – is the most valuable information when comparing
data for a specific condition (but is not what was asked for). "

Health New Zealand state what they think I should ask for, but the OIA is for people to ask what they want to ask for. So despite what Health New Zealand thinks I should ask for it is obliged to provide what is requested. The reason that the OIA was written this way is that this prevents public servants hiding information.

In addition, Health New Zealand could have Consulted with me about this to agree "final diagnosis" as the definition of "presented". Health New Zealand seems to be interpreting "presenting" in a temporal sense (as they enter the ER - where you say there is incomplete collection), rather than intended sense in the OIA of "why" they presented to ER - which Health New Zealand say could be answered better as "final diagnosis".

Now Health New Zealand has failed to Consult, and has gone straight to refusal. So it seems you may have breached the Act in this regard. I require you to Consult so that we can agree what might be useful. It may be that both sets of counts can help the country get to the bottom of this perplexing problem that we are all interested in.

My proposal (subject to the outcome of our Consultation) is that "presented" be provided as counts in two datasets -
1) which is classified according to the reported issue on arrival (Health NZ's current interpretation of the OIA) and
2) the Health NZ recommended "final diagnosis" (derived from ICD-10 or similar - but not itemised at that level of detail).

In summary, Health New Zealand:
- Cannot refuse to provide a reasonable request for information that is literally sitting in its database - whether or not there is data quality issues or it agrees that the "correct" thing has been requested, and
- Have recognised that it is in its interests to provide caveats and context while responding for PR and self protection purposes - and should include these not use them as a reason to refuse.
- Need to Consult and have failed to do so. Therefore this is Health New Zealand's next action. Please commence this ASAP.
- Now has a proposal (above) to consider and include in its Consultation.

This has been referred to the Ombudsman.

Yours sincerely,

Chris Johnston