Just some numbers from Canada

One of my colleagues posted this link yesterday to a shiny app giving Covid19 testing and results for all provinces in Canada. Seems to match all other sources I have heard from.

About 43,000 tests and 600 positive. The cumulative graphs of cases by province indicates Alberta is currently having the fastest increases.

Hopefully the numbers will get on a github at some point.

Anything similar in the US yet, giving total and by state testing and results?

This post is by Keith O’Rourke and as with all posts and comments on this blog, is just a deliberation on dealing with uncertainties in scientific inquiry and should not to be attributed to any entity other than the author.

42 thoughts on “Just some numbers from Canada

  1. Any comments on this article by Ioannadis; he seems to think the lockdowns etc. are too draconian? So normally, I’d have dismissed this sort of opinion as wacko / uninformed but coming from Ioannadis I must take it more seriously. Is he just in denaial? Or are these contrarian yet credible arguments?

    https://www.statnews.com/2020/03/17/a-fiasco-in-the-making-as-the-coronavirus-pandemic-takes-hold-we-are-making-decisions-without-reliable-data/?utm_content=buffere08f7&utm_medium=social&utm_source=twitter&utm_campaign=twitter_organic

    • So normally, I’d have dismissed this sort of opinion as wacko / uninformed but coming from Ioannadis I must take it more seriously.

      Did you dismiss the need for lockdowns and closed borders in mid-January as wacko/uninformed too? That is what I advocated then. But now we see this virus is not a very big threat and the benefits of such actions are far less than they would have been back then.

      But now that is the wacko/uninformed opinion…

      • > But now we see this virus is not a very big threat

        Where do you see that? In one year there are 100k deaths in Lombardy. Less than 300 deaths per day. There have been 319 covid-19 deaths reported in the last 24 hours.

        • I’d need to see your data source for the Lombardy specific numbers (does that assume no seasonality in mortality?, does it assume uniform geographic distribution of mortality across Italy?, etc). But here they are for Italy in general:

          https://www.epicentro.iss.it/coronavirus/bollettino/Report-COVID-2019_17_marzo-v2.pdf

          99.2% of the deaths were in already sick people, 51.5% had 3 or more pre-existing conditions. So I would ask how many non-covid-19 deaths were reported during that same period?

          I do believe that people are dying of respiratory failure rather than something else they were about to die of though. I also note that in China they shipped 50 tons of vitamin C into Wuhan and already have it in their guidelines to megadose IV vitamin C before the patient goes critical. Meanwhile in the West they are still putting out news stories calling any use of vitamin C a myth. Its likely these already ill people are deficient to begin with due to the chronic illness, then the extra stress of the infection and panic puts them over the edge.

          https://mobile.twitter.com/DSM/status/1224262885729349633
          http://rs.yiigle.com/m/yufabiao/1183266.htm

          I also wonder if they are putting all the COVID-19 patients in the same place which increases the viral load. There was some news out of Italy that isolating patients helped, but I don’t know if I’m interpreting that correctly: https://www.repubblica.it/salute/medicina-e-ricerca/2020/03/16/news/coronavirus_studio_il_50-75_dei_casi_a_vo_sono_asintomatici_e_molto_contagiosi-251474302/

        • As I understand it vit C is used to recycle glutathione, which is a really powerful antioxidant used everywhere in the body. Immune reactions tend to produce a lot of oxidative damage, it’s entirely possible that the real issue is Glutathione, or perhaps both glutathione and vit c work together directly. But it’s clear that immune system interactions are important in the rapid and severe version of this disease.

        • As I understand it vit C is used to recycle glutathione

          Ascorbate has single electron redox potential of ~280 mV, then the resulting ascorbate radical to dehydroascorbate oxidation is -174 mV. Compare this to glutathione of -1500 mV and 500 mV for vitamin E. So glutathione recycles ascorbate and ascorbate recycles vitamin E.
          https://www.ncbi.nlm.nih.gov/pubmed/8434935

          both glutathione and vit c work together directly.

          Yes. Glutathione is the terminal intracellular antioxidant, vitamin E is the terminal lipid-soluble antioxidant, while ascorbate (vitamin C) is the terminal extracellular antioxidant.
          https://www.ncbi.nlm.nih.gov/pmc/articles/PMC297842/

        • I should also note there were some rumors that early on in China the recommendation was to not give vitamin C. This could be the case if once the situation becomes very severe a lot of free iron is around, since it reduces Fe3+ to Fe2+, which could lead to increased free radical generation:

          https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5131655/

          That is only rumors but they could make sense. So, anyone who wants to try this should measure the free iron levels.I only say that because people (both detractors and advocates) seem to think it is magic for some reason.

        • One of the major reasons to slow the spread of this disease is so that knowledge about effective treatments can build up so that people who get it later are much better cared for. This is a good example, and just the beginning.

        • http://dati.istat.it/index.aspx?queryid=26441

          Territorio : Lombardia

          Morti (2017) : 98 784

          By month, Jan-Dec 2017:

          11 979 / 8 494 / 8 317 / 7 544 / 8 043 / 7 195 / 7 236 / 7 568 / 7 216 / 8 057 / 8 035 / 9 100

          Looking at the month of March only, 2003-2017:

          8 471 / 7 453 / 8 457 / 7 371 / 7 614 / 7 833 / 7 959 / 7 851 / 7 686 / 8 252 / 8 064 / 7 981 / 8 837 / 8 268 / 8 317

          Expected deaths per day at this time of the year are in the 250-270 range. 320 from a single cause is somewhat concerning (more so because they may be 1000 per day soon).

        • 320 from a single cause is somewhat concerning

          Not really if there are very few from other causes. As I said:

          So I would ask how many non-covid-19 deaths were reported during that same period

        • I think your point is that it’s one hypothesis that most of the people dying are having their lives shortened by ~ a month or two… that is, they were really unhealthy to begin with.

          This is a theory, but it’s not really borne out by the data I think. Still, it’s maybe a little early to say.

          The thing that’s missing from this analysis though is the very real cost of 5 weeks in the hospital, and the all-cause mortality increase post recovery for some populations.

          Even if no-one died of this illness, if 10% of the population needs a 5 week stay in a hospital with 3 of those weeks on a ventilator… it would be worth tamping down the disease heavily and working hard to solve the long term problem.

        • https://www.reuters.com/article/us-health-coronavirus-italy-homes-insigh/uncounted-among-coronavirus-victims-deaths-sweep-through-italys-nursing-homes-idUSKBN2152V0

          “Gori said there had been 164 deaths in his town [Bergamo, fourth-largest city in Lombardy; Pop: 122k] in the first two weeks of March this year, of which 31 were attributed to the coronavirus. That compares with 56 deaths over the same period last year.”

          Thanks. But I am sure you understand that looking in the last two weeks in one town vs last year does not demonstrate this is an especially threatening illness.

          I do agree this data on all cause mortality over the last few years is probably available to local officials and they should share it with us.

        • Don’t worry, I do completely get your point! Everyone who dies was going to die anyway in the coming months or years (decades in some cases, but what does it change) and it’s not like anyone is going to die twice.

        • Don’t worry, I do completely get your point! Everyone who dies was going to die anyway in the coming months or years (decades in some cases, but what does it change) and it’s not like anyone is going to die twice.

          No, I dont think you did understand my point at all. The strongest form of my point is people are dying at the usual rate and also testing positive for this virus. Do you have evidence otherwise? Did you look until now?

        • People are dying at the usual rate?

          https://news.sky.com/story/coronavirus-italian-army-called-in-to-carry-away-corpses-as-citys-crematorium-is-overwhelmed-11959994

          Coronavirus: Italian army called in to carry away corpses as city’s crematorium is overwhelmed

          “The crematorium of Bergamo, working at full capacity, 24 hours a day, can cremate 25 dead”, said a spokesperson for the local authority. “It is clear that it could not stand up to the numbers of the past few days.”

          Coffins are now being taken to crematoriums in Modena, Acqui Terme, Domodossola, Parma, Piacenza and several other cities. Once the bodies have been cremated, the ashes will be brought back to Bergamo.

          Coffins have already filled up two hospital mortuaries and a cemetery morgue.

          Meanwhile, the local newspaper’s daily obituary section has increased from two or three pages to 10.

        • According to your info there is one town of 100k that experienced 50 dead in the first two weeks of March last year (why are we only comparing to one year?), had 150 dead this year so far in March, and has one crematorium that could handle 25 per day (what percent of people from that town usually choose to be cremated?).

          Is that right?

        • In your “people are dying at the usual rate” alternate reality, why are obituaries being published in the local newspaper at four times the usual rate? Do they want to ride the coronavirus hype wave or what?

        • So it has been 2-3 pages every day for the last 10 years, or what? How long has the obituary been 10 pages? This info is so low quality…

          Why dont they share actual data with us?

        • In Wuhan, where the government cracked down. And we concluded you need to look at all cause mortality for the year in people who test positive for antibodies to really compare I thought.

        • There’s nothing special about a year really, whereas the 2 weeks (or maybe it’s 3 or 4 whatever) during which you suffer from the disease are the relevant concern. Yes it’s also of interest to know if there are heightened risks post “recovery” but to first approximation, Symptomatic COVID produces a ~50x increase in death risk rate for the duration of symptoms.

          That is not in any way the same as business as usual with a minor shifting of the death time for only the very sick. The disease is clearly very bad, particularly for people over 65. Without COVID a 70 year old white woman has a life expectancy of about 20 years, with COVID she’d have a ~ 10% chance of dying within 2 weeks.

          sure comorbidities and soforth are important, but face it 2 weeks and 20 years are very big differences, they’re not all down to covid only hitting people with severe health issues.

        • The life tables available were for probability of dying while a certain age, thats the only reason one year was special.

        • You can think of the probability of dying in a given year as the integral of a rate function… let r(t) be the rate, then the probability of dying at a certain age between year1 and year2 is integrate(r(t), y1,y2)

          so long as the r(t) doesn’t change too quickly you could approximate this integral as just r(tm)*(y2-y1) where tm is the midpoint in time. so basically it’s a linear function of the duration, and this is particularly true for short durations… so you can directly compare 2 weeks by just multiplying the year rate by 2/52 Since that’s how long two weeks is as a fraction of a year.

          so the 2 week rate for COVID is vastly larger than the 2 week rate for regular life, this is all we need to know to realize that when the number of people is large and the risk is large… many will die. it’s not even close to business as usual. the exact quantity doesn’t much matter since it’s so much larger. if it were say a 1.5 or 2x increase we would need to do careful analysis, when it’s ~50x increase we know our answer without too much issue. it’s like will you die if an elephant steps on your head? the answer is yes pretty well regardless of how big an elephant or what kind of bike helmet you are wearing.

        • All-cause deaths in the province of Bergamo in the month of March (2003 to 2017):

          798 712 766 716 706 793 771 804 789 844 788 811 953 865 845

          Reported covid-19 deaths as of March 21 (i.e. with one third of the month to go): 951

          Do you need to know how many non-covid-19 deaths were reported to deduce that maybe people are not dying at the usual rate? That’s at least a nine-sigma anomaly!

          (And don’t forget that the official tally is undercounting deaths, as it was already discussed, because many people die at their own homes or in nursing homes and are not tested.)

        • I don’t have answers regarding other people’s motives [*]. Why don’t you explain what hypotetical numbers would support your claim that people are dying at the usual rate?

          [*] I don’t even know how “they” are. Do you know anyone anywhere providing that kind of statistics on a daily basis? Please share with us if you do, it will be interesting to look at the evolution of those numbers.

        • <

          Do you know anyone anywhere providing that kind of statistics on a daily basis?

          Only for covid-19, no other illness is worth it apparently.

        • Why could that be?

          Also, why don’t you explain what hypotetical numbers would support your claim that people are dying at the usual rate?

        • > Why could that be?

          Because there is a hysteria so it gets lots of clicks/views.

          > Also, why don’t you explain what hypotetical numbers would support your claim that people are dying at the usual rate?

          That isn’t really what I proposed. Its that people with weakened immune systems are dying of respiratory failure instead of from whatever other thing it would have been a few weeks/months down the line.

          If correct, that means this will be a barely detectable blip on all cause mortality when aggregated at say the monthly or yearly level.

        • I dont think you understand your own point at all. The strongest form of your point is people are dying at the usual rate and also testing positive for this virus. Do you have evidence otherwise?

        • We are only told the covid-19 deaths, which in Italy means anyone who died that tested positive. Also, since it seems to have been spreading in the hospitals there that means many deaths will be attributed to the virus just because people near death often get taken to the hospital.

          Have we been told the percent of covid-19 deaths due to ARDS in Italy?

  2. I’m sure the availability heuristic can influence our opinions, as Cass Sunstein has elaborated in the Laws of Fear. There is language in John’s article to suggest that he may be familiar with Sunstein’s viewpoints. John is highlighting that there is ‘selection bias’ in our guidelines.

    In viewing the White House Briefing today, Burke seemed to echo both John Ioannidis and Cass Sunstein. Cass Sunstein recently weighed in on the fears generated by the onset of the virus here in the US. He seemed to suggest that there were exaggerated fears. Of course, some experts on Twitter took him to task

    John though emphasizes the lack of reliable data and selection bias, in that the CDC guidelines are extrapolated on the basis of data gleaned from those who are sickest with COV-ID19. Thus neglecting that the younger majorities recover within 4-14 days.

Leave a Reply to Keith O’Rourke Cancel reply

Your email address will not be published. Required fields are marked *