133 Molesworth Street
PO Box 5013
Wellington 6140
20 September 2022
New Zealand
T+64 4 496 2000
P McKenna
By email: [FYI request #19938 email]
Ref:
H2022010827
Tēnā koe,
Response to your request for official information
Thank you for your request under the Official Information Act 1982 (the Act) to Manatū Hauora
(Ministry of Health) on 23 August 2022 for clarification on information you received under
reference: H2022007735. Each part of your request is responded to below.
You state: “The vaccination status of cases was classified incorrectly for approximately a
week following changes introduced with our reporting upgrades on 6 July 2022. This was
due to a coding error that resulted in vaccination status of cases being reported as it was
a week after they tested positive for COVID-19, rather than a week before they tested
positive for COVID-19. “
1a. From this I gather that the correct method of recording is if someone tests positive,
their vaccination status is not taken as it is on the day of positive test, but instead, it is
taken as what it was 1 week earlier. So if, for example, an unvaccinated person gets their
first dose on the 1st of July and tests positive on the 7th of July, the case will be counted
as an unvaccinated case. Is this assessment correct?
Yes, this assessment is correct.
1b. I can see how this would result in a jump of 43k cases for the boosted since that is
what would be expected given the case rate of boosted people since February (43k is
roughly 1 weeks worth of cases). I can also see how 1 dose and 2 dose would drop, since
some 2 dose people would be counted as boosted, and some 1 dose as 2dose. But the
increase in unvaccinated does not seem to make sense since no one can be vaccinated
before a test and then unvaccinated after it. The movement should be UP the vaccination
'food chain' not down it. The quoted paragraph, as it relates to unvaccinated, seems to
say that 11,421 cases were erroneously labelled ‘unvaccinated’ because that is what they
were one week after positive test, when they should have been labelled ‘vaccinated’
because that was what they were one week before positive test. Obviously one can not
become physically unvaccinated so are these people subject to reclassification for some
reason, or is there another explanation?
The vaccination status of cases was classified incorrectly for approximately one week, following
changes introduced with our reporting upgrades on 6 July 2022. This was due to a coding error
that resulted in vaccination status of cases being reported as it was a week after they tested
positive for COVID-19, rather than a week before they tested positive for COVID-19.
The change in reporting was due to a switch in data sources to new data which provides insight
into how many hospitalisations were for COVID-19, rather than just with COVID-19.
Vaccination status figures changed in one direction due to the coding error when the definition
was changed on 6 July 2022. Roughly a week later, there was a second change in vaccination
figures in the opposite direction as the error was corrected. The second change involved
reclassifying people who received vaccinations in the week before testing positive from a
classification with more doses to a classification with fewer doses. For example, someone who
received their first dose of a vaccine three days before testing positive would be classified as
‘partially vaccinated’ prior to the updated reporting. This is because their vaccination status
seven days after testing positive would have been recorded as partially vaccinated. When the
definition was updated, they would be correctly classified as unvaccinated in the reporting, due
to having received zero doses seven days before they tested positive.
You did not address Question 2 except by referencing the spreadsheet. This sheet
appears to provide the hospitalisations and ICU admissions for the days leading up to,
and immediately after the change in definition which is nice, but my question refers to the
disproportionate reductions seen across the different cohorts. The deleted cases
represent those who had a positive test result but went to hospital for a non-covid reason-
correct? If you were to reduce these deleted cases as a ratio to the number of covid cases
in the respective cohorts, it demonstrates a significant excess hospitalisation among the
vaccinated numbering in the thousands- a very high percentage. This gets worse if you
consider that the case rates for unvaccinated appear to be very much under reported
(they are about half the case rate of 2 dose and boosted).
This is a very small group of hospitalisations - about 1% of all hospitalisations, so it is only
a small signal. But it is a very alarming one. If one were to extrapolate, even only a small
fraction of this increased rate, to the other 99% of hospitalisations the results would be a
large number of excess hospitalisations. And we are in fact seeing hospitals stretched to
capacity so the hypothesis I am hinting at is at least plausible.
2a) Has the Ministry identified this? 2b)Is the ministry concerned by this? If not, why not? If so, what is being done, or could
be done to further investigate?
4) My reason for Q4 was that if the MoH did record vaccination status it would be easy to
check if this effect was taking place across all hospitalisations. It may be that MoH is not
collating or reporting on this, but do hospitals collect this data at time of admission?
The hospitalisations that were removed from our COVID-19 hospitalisation figures are cases
where individuals were hospitalised
with COVID-19 but were not hospitalised
for COVID-19.
As the COVID-19 vaccine is effective in preventing hospitalisation, a greater proportion of
vaccinated patients in hospital were not hospitalised due to their COVID-19 infection, but for
other reasons.
Therefore, the vaccinated numbers dropped more than the unvaccinated numbers. This is due
to that unvaccinated people who went to hospital with COVID-19 were more likely to be there
from COVID-19. This is expected.
I trust this information fulfils your request. Under section 28(3) of the Act, you have the right to
ask the Ombudsman to review any decisions made under this request. The Ombudsman may
be contacted by email at:
[email address] or by calling 0800 802 602.
Nāku noa, nā
Dave Henderson
Group Leader, Intelligence, Surveillance and Knowledge
Public Health Agency, Ministry of Health
Page 2 of 2
