In advance of our June excess mortality work coming out tomorrow, here is a thread on the latest COVID-19 data in Australia. #COVID19Aus
Here's a table showing the various bits of data we know about COVID-19 deaths.
Note with the change to weekly reporting, I have needed to estimate surveillance deaths for the last day of Sept (last 3 days for Qld).
This thread might be a good read for some of my new followers, explaining how Covid deaths are recorded in Australia, concentrating on the difference between daily surveillance reports and the ABS statistics.
Surveillance numbers dont change, other than adding a new month.
ABS figures increase as deaths are registered, impacting the most recent months more than older months.
Compared with last month, a lot more coroner referred deaths were registered this month (49 cf 4 last month).
Here is the chart of surveillance deaths per month.
September is much lower than July and August.
This graph compares the ABS and daily surveillance reports to the end of August. Before Jan-22, the two sources match closely but there are considerable differences in 2022.
The main reason for the big August difference (and to a lesser extent July), is because there are a lot of deaths that occurred in those months that wont have been registered yet, hence wont appear in the ABS data.
Here’s the same data presented on a cumulative basis since the start of the pandemic. I’ve also shown the ABS data including an allowance for late registrations (grey dashed line, using the same pattern of registrations as seen in the last six months).
Daily surveillance deaths to the end of August were just under 14,000.
The ABS data shows 11,400 deaths, and I estimate this will increase to 13,200 by the time all the late registrations come in.
Its unclear why surveillance deaths are higher than estimated ABS deaths:
- "historical" reporting in surveillance data confuses timing
- there could be lots of coroner-referred covid deaths awaiting decision as to cause (so arent in ABS data)
- my estimates could be wrong🙃.
Deaths *from* vs *with* covid. Overall, the ABS data shows that 9,428 (82%) of deaths are *from* covid and 2,013 (18%) *with* covid. The proportion *with* covid increased significantly with the Omicron wave and was 25% in May-July, increasing to 30% in August.
For those who died *from* Covid, overall 79% had a pre-existing condition. This proportion has increased for deaths registered in the last four months and is currently sitting at 85%.
Chronic cardiac conditions and dementia remain the most common pre-existing conditions.
For those dying *with* Covid, what were the underlying causes of death? Cancer, circulatory system diseases (which includes heart disease) and dementia are the top 3. It is likely that Covid accelerated the death of these people in many/most cases.
The end. Any questions?
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Our latest analysis of excess mortality is now out, covering all cause mortality to the end of June 2022, plus COVID-19 only mortality for July-Sept. #COVID19Aus#excessdeaths
TLDR: +11,200 excess deaths in first half of 2022 (13%) actuaries.digital/2022/10/06/cov…
Our excess deaths are measured relative to pre-pandemic expectations of mortality.
Includes allowance for continuation of mortality improvement/decline for each cause of death.
Allows for changes in the size and age composition of the population.
Excess mortality in the month of June 2022 was again very high, almost as bad as January 2022.
Needed to change my y-axis for this one.
The Ministry of Health (MoH) in Singapore released this very interesting report on excess deaths over the weekend. moh.gov.sg/docs/libraries…
The report covers excess deaths from the start of the pandemic until 30 June 2022.
Covid deaths in Singapore and Australia per head of population have been broadly similar.
In measuring excess deaths, the MoH have used the 2019 standardised death rate as their baseline. Their measurement allows for changes in the size and age distribution of the population. Any excess represents a worse mortality RATE compared with 2019. This is good.
The NSW epi report for the week ending 10 Sept is now out. Note that this is the week ending just before they switched from daily to weekly reporting, so nothing new on that front. #Covid19Aus#Covid19NSW
You can find the full report here... health.nsw.gov.au/Infectious/cov…
Cases are down another 29% this week. Excellent.
While PCR testing is down again (-15%), so is positivity (at 7% this week). Good.
Only a few highlights this week as trajectories haven't changed, and I'm a bit fried after the excess deaths stuff this week:
- cases are down 21% and all age groups moving similarly; good
- PCR testing is down, but positivity rate is also down; good
- hospital admissions down 15%; excellent. The biggest % fall we've seen in a long time
- people in hospital down 6%; good. Hopefully this % will be even bigger next week as the lower admissions flow through
- 126 deaths; never good. But the lowest number in the seven weeks
Our latest estimates of excess deaths are out, covering all cause mortality to the end of May 2022, plus COVID-19 only mortality for Jun-Aug. #COVID19Aus#excessdeaths
TLDR: +8,500 excess deaths in 5 months to May (12%) actuaries.digital/2022/08/31/cov…
Note our excess deaths are measured relative to pre-pandemic expectations of mortality.
Includes allowance for continuation of mortality improvement/decline for each cause of death.
Allows for changes in the size and age composition of the population.
Another month of huge excess deaths in May. Only one week in 2022 to date has been within our 95% confidence interval