A roll of the dice over COVID-19
It’s a little dizzying to read on the one hand La COVID-19 frappe fort dans le Grand Montréal, Sacre-Coeur cancer patient dies of COVID-19, Un centre pour déficients intellectuels entièrement infecté par le coronavirus, Taux de mortalité quatre fois plus élevé à Montréal qu’à Toronto and on the other Legault says he wants to gradually reopen Quebec – while still calling for more support from the armed forces.
Well, we’ve seen this debate already in the comments on the blog, and it’s playing out around the world. But it’s a roll of the dice given we don’t know enough yet about immunity to the coronavirus, whether everyone who’s exposed to the virus is immune afterwards, and if so, for how long.
Chris 10:29 on 2020-04-23 Permalink
It’s a roll of the dice either way.
JaneyB 10:56 on 2020-04-23 Permalink
We won’t know those things for probably another 18 months at least. We will continue to have scattered glimpses of factors, immunity, vulnerabilities etc so that’s all we have to rely on. Only experts will be able to make sense of those. We see chaos, they see probabilities and trends.
The rate of growth (of infections as known through tests) has been about 5% per day for about 10 days now, and that’s with a focus on testing health institution workers and patients. It was 30% per day around the beginning of the lockdown, with testing focused on travellers. Also 80% of confirmed deaths have been in seniors’ residences. Lots of sickness still but the growth decrease and case concentration is why there is talk of gradual, limited opening – with the distancing measures, of course.
I read somewhere (maybe here?) that Montreal normally has 3x the flu rate of Toronto so that sounds like there is a cultural component to our rates.
DeWolf 12:46 on 2020-04-23 Permalink
Public health authorities all say we’re at the peak, and I guess this is what the peak looks like. I doubt the headlines will be as scary in a few weeks. Besides, it sounds like the de-confinement will be extremely gradual. Legault is talking about starting from the less affected regions and working towards Montreal. I bet we’ll see baby steps in early May but no big change in restrictions until the end of the month or even the beginning of June.
David2200 16:20 on 2020-04-23 Permalink
Based on very early antibody testing out of NYC today, their health dept estimates possibly a million people in their city have been infected, putting the mortality rate at approximately 1/10 of 1%.
mare 18:43 on 2020-04-23 Permalink
@ David2200
No it isn’t 1/10 of 1%
This puts the NY State fatality rate around 1% and not 10% as current data of confirmed cases vs deaths
Also these tests are skewed for many reasons
• testing kits have large false results
• they tested supermarket shoppers (so people who where outside among other people and are more at risk to get infected)
• they excluded people under 18 for consent reasons
• mostly suburban population
• no old and at risk people (who stay home and aren’t infected)
More from
Ashish Jha, Director, of the Harvard Global Health Institute
https://twitter.com/ashishkjha/status/1253399114559062016
david701 19:30 on 2020-04-23 Permalink
That guy clearly knows his stuff, but there’s a bunch of other twitter experts explaining that their stratified sampling method is leading to gaps that mean we don’t have enough info, but should make us feel pretty optimistic. At any rate, we’ll know more as the data continue to roll in. The 0.1% rate was mentioned by a New York City health official. Just like the 0.1-0.2% range was mentioned by a Los Angeles health official. It’s not like they know less than the people on twitter or Montreal.
More interesting data show that pretty much everyone who croaks has a co-morbidity. https://time.com/5825485/coronavirus-risk-factors/
This might go part of the way to explaining why ventilators won’t save us, if we’re that far gone: https://www.cnn.com/2020/04/22/health/coronavirus-ventilator-patients-die/index.html
Kevin 19:36 on 2020-04-23 Permalink
Sweden has retracted its plan, saying it was based on terribly mistaken assumptions about infection rates and antibodies. https://www.forbes.com/sites/davidnikel/2020/04/22/sweden-health-agency-withdraws-controversial-coronavirus-report/#7d1730043494
Raymond Lutz 21:23 on 2020-04-24 Permalink
“Public health authorities all say we’re at the peak” ??? Who’s saying that? As wrote JaneyB, Quebec cases are growing steadily by 5% EACH DAY SINCE TWO WEAKS! AB, ON, BC are on a perfectly exponential course too! see canadian trends plots. FYI, doubling period in days = log(2)/log(1.05) if daily growth is 5%
JaneyB 13:03 on 2020-04-25 Permalink
Growing 5% per day is a decrease from growing 30% per day. That means we’ve passed the peak and are on the downside. In this exponential world, growth is not incompatible with decrease. If we can get to say 2% per day for the next year, that would be great – for the crisis management of the healthcare system, which was always the primary goal. We will be reaching an infection rate of 70% of the population one way or another so that will not be a good experience for individuals and families. A vaccine would be wonderful – a full year or more from now – but it might not even be possible or it might only work for one season. Frustrating.
Raymond Lutz 17:16 on 2020-04-25 Permalink
“in this exponential world, growth is not incompatible with decrease” JaneyB, did you check my graph? Semi-log plot for Québec has been a STRAIGHT LINE for the last 19 days. “We will be reaching an infection rate of 70% of the population one way or another” Wow.. this is a lot… Your source? I rather found “Coronavirus ‘could infect 60% of global population if unchecked” (Guardian)
“Growing 5% per day is a decrease from growing 30% per day” Meh… 30% lasted about 2 days. Time derivative is quite noisy so I hacked this fit: daily growth with exponential fit.