Covid19 Reflections #21

Insight

One of the key numbers that many people use to track how well we are going on containing Covid19 is the number of confirmed cases. I obsessively watch it myself. Today (14 April) there are 6,400 confirmed cases in Australia, which is only a 0.5% increase on the day before. The average daily growth rate over the past week is only just over 1%. So if this keeps up, the Australian cases of Covid19 will definitely reduce – as all of the current infections either recovery (mostly) or die (currently around 1%).

So that’s good news right? But confirmed cases depend on cases being tested. And while it is a lot easier to get a test here in Australia than in the US, there are still criteria to decide whether you get a test:

Testing is recommended for a person with fever (≥38°C) or history of fever (eg, night sweats, chills) or acute respiratory infection (e.g. cough, shortness of breath, sore throat) who meets one or more of the following criteria:

• a close contact in the 14 days prior to illness onset with a confirmed case

• international travel in the 14 days prior to illness onset

• a cruise ship passenger or crew member who has travelled in the 14 days prior to illness onset

• a healthcare worker

• an aged care or other residential care worker

• in a geographically localised area with elevated risk of community transmission – see Locally acquired cases and clusters in NSW.

So you still have to have symptoms of Covid 19 to be tested. The graph below shows that Australia and New Zealand are testing a lot of people compared with the number of confirmed cases – more than 50 people have to be tested to get a confirmed case. But I would still feel more comfortable if there was genuine random testing happening in the community.  The best completely controlled population we have is the Diamond Princess (the one that was marooned off Japan in January/ February). This study shows that around half of all positive tests on that ship were asymptomatic (328 out of 634). And this study of a small town in Italy (Vo) suggests an even higher percentage of asymptomatic cases – around 70%.

So until we are randomly testing people, the headline number of cases you see, even in Australia and New Zealand, with high levels of testing, will be far from a complete set.I would rather be in a country with Australia’s level of testing than in fellow english speaking countries the US or the UK, where 20% (the US) or 30% (the UK) of tests are positive. There must be a lot of cases out there in the community in those two countries (and I’ve followed a few of them on twitter who are pretty sure they’ve had Covid19 but have had no way to test).

Our World in Data is my goto place for world wide statistics on Covid19. Others like John Hopkins, but I find Our world in data has a nicer interface.

Link

One of my favourite statistical/mathematical bloggers Cathy O’Neil (author of Weapons of Math Destruction) has an article today about all the problems with the data around Covid19.

Appealing as it may be to keep count, the true numbers might not be knowable until much later. Testing needs to be done systematically, even on asymptomatic people. For deaths, precise numbers might never emerge. It’s possible to estimate using the number of unexpected deaths compared to a year earlier. But even that’s not ideal, because lockdowns might suppress other kinds of deaths — traffic accidents, for example — by forcing people to stay at home.

Don’t get me wrong: Watching the official data is not a complete waste of time and attention. The numbers can give some sense of what’s happening — as long as we recognize their flaws.

If you read the article, you’ll particularly recognise the point about testing, as I fear I have posted about this too often (including today!). The availability of testing in an individual country (or even locality) makes a huge difference to the number of cases. But as O’Neil says, as long as you understand the flaws, the numbers are useful. But they aren’t infallible.

Life Glimpses

This week I’m leading a webinar with (!) 400 people. So I’ve been a bit focused on my sound and picture quality. We’re getting there (those people who were on virtual meetings with me today would say I’m not getting there very fast, but hopefully soon…). While my prepping was not bad over the last few months  (I was looking for face masks without success in January, buying up big on hand sanitiser in February, bought one of the last containers of yeast in Woolworths in March), we completely forgot to check our video conferencing equipment. So I’m relying on an 8 year old iMac computer’s sound and camera system which leaves a little bit to be desired.

My tech support (geekinsydney) has hopefully found one of the few webcams left in Sydney (at first he thought Alice Springs had the only one left in Australia) so next week we’ll be in much better shape.

Bit of Beauty

Today’s bit of beauty comes from a pot plant next to the footpath I’d walked past every day for a month before geekinsydney stopped and took this lovely picture of it. I don’t know if it is edible, but it sure looks tasty (and hot!).