Template talk:COVID-19 pandemic data/Australia medical cases chart

WikiProject COVID-19

I've created WikiProject COVID-19 as a temporary or permanent WikiProject and invite editors to use this space for discussing ways to improve coverage of the ongoing 2019–20 coronavirus pandemic. Please bring your ideas to the project/talk page. Stay safe, --Another Believer _(Talk) 18:19, 15 March 2020 (UTC)

Recoveries are definitely wrong

On the 16ths of November there were 191,614 total cases, 1,898 total deaths and 17214 active cases. https://www.health.gov.au/sites/default/files/documents/2021/11/coronavirus-covid-19-at-a-glance-16-november-2021.pdf

Unfortunately the "Recoveries" is listed as 174400, which is wrong because it is counting deaths as recoveries. I do not have the time to go through and fix all the data points. As such, I will add the latest data point correctly and hope someone sees this data error, and has the time to go through each release and double check from here: https://www.health.gov.au/resources/collections/coronavirus-covid-19-at-a-glance-infographic-collection

Jamesjansson (talk) 08:02, 23 December 2021 (UTC) James

Recoveries Look Wrong

People usually recover within 14 days. So as the number of cases grows one would expect the number of recoveries to grow, where that is not being shown.

Where exactly is the data being retrieved from?

And what does "Recovered" actually mean? If it means recovered from Hospital, then the main should be Cases Admitted to Hospital. And what happens to people that leave the country?

The current chart suggests that very few people recover, which is clearly not the case. Unless decent data can be found I think it needs to go. There is very little useful data on Covid-19 for Australia.

I will move it to the bottom until this is clarified. Tuntable (talk) 22:43, 2 April 2020 (UTC)

Fixing recovery data

The recovery data is apparently from just a few states, and not NSW, the biggest. So the chart suggests that most people never recover, which is plain wrong.

There is a good discussion of recovery on https://www.smh.com.au/national/you-need-both-stories-who-adviser-calls-for-national-coronavirus-recovery-tally-20200402-p54ght.html

Some complain that the government is not tracking recoveries properliy. But Professor Allen Cheng, a physician at The Alfred hospital, said 10 days from the point of displaying symptoms was a conservative buffer for most patients to have recovered and no longer be infectious.

"He argued a recovery tally was not the best use of resources as the pandemic moves towards its peak in May or June. "I'm not sure it's that useful. You can infer it - if a patient hasn't died or isn't in hospital after 10 days, you can assume they've probably recovered," Professor Cheng said."

So a simple estimate of recoveries is to take the number of cases 10 days ago, minus those that have died. Not perfect, but much better than excluding the most populous state. Tuntable (talk) 00:58, 6 April 2020 (UTC)

You can't just fabricate data just so a chart looks better. The data is partly lacking but inferring a figure and placing it in that chart is misleading, even if a footnote is attached stating it is an estimate. The data on Wikipedia should match what is publicly available, it is not a forum for original research. See WP:NOR Ljgua124 (talk) 01:30, 6 April 2020 (UTC)

Well, the data has been created based on a cited methodology, and is probably fairly close the the real number. The previous data was plain wrong.

That said, I would be happy for you to completely remove the data, although others object because this type of chart seems to be common in covid-19 articles. The only thing I am not happy with is to have data that is blatantly wrong.

Perhaps another approach is to create a chart of total and recoveries that only included the states for which recoveries is known. But that would be a bit esoteric and not belong in the lead.

At the end of the day, when the epidemic is over, everyone should either be recovered or dead!

Tuntable (talk) 04:11, 6 April 2020 (UTC)

The previous data was not wrong, it was incomplete. We now have more data - this source, stating that the recoveries are 2,432. The fabricated data should be removed. I also think the legend should be altered to state 'Tested' as is done on 2019–20 coronavirus pandemic in mainland China. This would be far more accurate than stating 'Active cases', as well as including data that has been fabricated. Ljgua124 (talk) 10:02, 6 April 2020 (UTC)

If you want to update based on that link then that would be excellent. Unfortunately they provide no historical data, so it would require using Archive.org. But if you want to do the work and get actual figures from Archive.org then be my guest, but go back so that there is no sudden jump. Personally I think they will tell a similar story. I also do not really trust the government figures, and suspect that they under report recoveries given the article above criticizing the lack of recovery data. Alternatively, if you can find better data about average recovery times that would be good, Singapore has some data.

If you do that, also track number of cases in hospital and number of cases in ICU. Also excellent data that should be charted, and probably far more reliable than any other data.

But for now I will just use the 10 days approximation unless somebody wants to take a different but sound approach. Tuntable (talk) 23:39, 6 April 2020 (UTC)

Actually I had a look and there is no historic data for those figures on Archive.org. If the numbers build up a bit then we could blur the data somehow. Tuntable (talk) 23:46, 6 April 2020 (UTC)

I have also extended the average recovery time to 12 days. So it should now roughly track the other figures. Tuntable (talk) 00:06, 7 April 2020 (UTC)

An estimate cannot be included alongside data that has actually been reported - it is misleading. I have replaced the estimates with data that has actually been reported to maintain consistency and accuracy and to avoid confusion and misleading the reader. Ljgua124 (talk) 07:54, 8 April 2020 (UTC)

Victoria Recovery Data

https://www.brisbanetimes.com.au/national/australia-to-publish-coronavirus-recovery-tally-to-lift-optimism-20200406-p54hm6.html

Indicates a consistent 10 day lag.

Still nothing historical for Australia. Tuntable (talk) 03:17, 7 April 2020 (UTC)

Ifographic

There are some numbers on https://www.health.gov.au/sites/default/files/documents/2020/04/coronavirus-covid-19-at-a-glance-coronavirus-covid-19-at-a-glance-infographic_3.pdf But they only show current numbers, not history. So I will use them if I catch them, otherwise interpolate. Tuntable (talk) 23:42, 7 April 2020 (UTC)

Removed bad data; And good data looks wrong

No point in arguing about the best recovery data prior 5 April. Just removed it. Clearly cannot put a few states on same graph to represent the total, the jump at 5 April looks ridiculous.

The Info graphic data also looks very dubious. Victorian data which looks good over a long time shows a steady 10 lag, whereas the Australian data suggests a 15 day lag. We shall see.

And yes, Wikipedia articles are supposed to be meaningful. Not just random facts mangled together. Tuntable (talk) 23:21, 8 April 2020 (UTC)

Reference dates etc

I've just updated the dates for <ref name="health-current-status"> ∆10 Jan. → 19 Feb. (only ≈ 5 weeks wrong!) & DMY dates. It seem this template page is the source for that reference on the Pandemic & Timetable pages etc. I know when updated once before on another page it caused a cite error(?), I think there were effectively, two references named "health-current-status", one here & also on the actual page? Hopefully it won't be an issue, if it is I'll revert. 220 of ^ßorg 03:36, 19 February 2021 (UTC)

Seems to have worked. 220 of ^ßorg 03:38, 19 February 2021 (UTC)

COVID-19 pandemic template data updated

I've added a total number of daily cases from January 2022 on-wards as on the webpage list of https://www.worldometers.info/coronavirus/ but in the web achieve as it shows more correct data at best. Also I've updated the recoveries data from 1st January. I will update the recoveries data anything before January 1st 2022 whenever I have spare moment. The total number of deaths, recoveries and active cases report equals the total number of cases. For example: (deaths + recoveries + active cases = total number of cases).

From 09th of September 2022, start of weekly reports instead of daily reports. Thuanma (talk) 13:29, 19 September 2022 (UTC)