Posts

Today, Premier Moe announced that he would be releasing a roadmap to re-opening the province little by little tomorrow. I believe that a cautious approach is warranted but that re-opening is worth consideration. However, I would personally waited one week more to see the results of both Orthodox Easter and especially of this recent warm snap. Some very low case growth numbers is also resulting in an excess of optimism. In fact, I was concerned in the speech from Premier Moe that the tone he struck was that we have won and not that we are winning. I believe the Premier knows very well that we have stopped the first wave but must be vigilant. And I understand balancing the need for positivity and optimism when many of us are getting more crazy than normal. But it’s important that it’s understood that we have won the first battle but there is still a war to fight. Singapore’s second wave should be a warning about how things can get out of control even where they have a lot more controls, analysis, and restrictions.

Yes, we have flattened the curve. There is no doubt about it. But we can’t lose sight that the objective was to ultimately buy time for our health care system to prepare. From what the SHA reports, they are. I’m not convinced though, that their plan really constitutes the best use of this bought time. For example, I’ve heard that additional ICU rooms that were planned to be added to some hospitals have been halted because case growth has slowed. This is backwards. Case growth slowed to give the SHA time to get those resources in place. We may not need them if we are lucky. But we didn’t go through the last 5 weeks to throw that effort away.

Today’s Motto: Proceed with Caution

With this and several other troubling stories in mind, I generated several graphs in early April. I have sat on them because I am very concerned that people may misinterpret them. They are not predictions. They are not expert models or simulations. They are extremely simplified graphs that show some very basic scenarios and what those scenarios would look like, graphically. Please don’t repost them without the rest of this article. Please don’t think they’re proof either that we’re doomed or out of the woods. The intent of these graphs is to understand how exponential growth, even at low levels, can be a hard thing to manage once the genie’s out of the bottle. Now if we could all apply this to our pre-COVID savings accounts! One other important note, the hospital capacity is the “with reduced services” number from SHA but doesn’t include excess COVID-19 beds in overflow facilities. Because, let’s just not get to that point. As I urge the province, I urge you, to: proceed with caution.

One other note of caution that needs to be really understood. We actually DON’T KNOW if recovering from Covid-19 means immunity to Covid-19. And if only a small percentage of recoveries have antibodies, as may be the case, well, all models about allowing more infections to build herd immunity are moot. And then we have to totally rethink how to keep the world running with an ongoing pandemic while we look for vaccines and ways to test entire populations at a time or at least a better understanding of risk, spread, and immunity. The following article reference pre-publication studies which have not been peer reviewed yet. Normally, this wouldn’t be worth looking at yet but due to the fact that we’re flying by the seat of our pants here, these pre-pub studies represent an educated guess.

• BBC – Can You Catch Covid-19 Twice?
• ScienceAlert – Scientists Still Don’t Know if Recovering from Covid-19 Confers Immunity

Scenario 1: What if Saskatchewan Did Nothing To Flatten The Curve of COVID-19?

By the time I created this graph, we had already succeeded in flattening the initial curve. So why produce it still? Partly as a basis of comparison against other measures taken, and partly as a defence for the measures that we did take. Now, when I say “did nothing” I mean it. As in, absolutely zero. We continued on blissfully unaware and modified zero behaviours. In truth, some of us would be staying home as much as possible and limiting social interactions as much as possible, government order or no. It is incredibly unlikely that we would blindly sail upwards of 10,000 cases in the province and not even be a little more cautious. So think of this graph as “what does 22% growth (our initial trajectory) look like”. How long does it take us to get to 100,000 cases? Why 100,000? We are assuming at that point, there would be enough recoveries, deaths, and asymptomatic or unreported cases – especially if numbers grew that quickly – that growth would slow considerably. So the ridiculous becomes absurd.

OK, all that said, here we are. 22% growth we would be crossing 100,000 cases already by May 1. Hospitalization lags initial onset of a case but a week later the province’s worst case provisions (overflow in rinks, schools) would be overwhelmed. Not to mention that many of the health care workers we depend on would have fallen ill also, supplies would be utterly absent, etc. So thank you, Saskatchewan, for being prudent and cautious!

Scenario 2: What if Saskatchewan Kept Current COVID-19 Restrictions In Place?

When I created this graph, it looked like cases in the province might be growing at around 3%. It has since flattened to 1% growth. That said, we are one super spreader away from being back at 3% or likely higher. One person visiting family in Calgary. What you see if that as growth in new cases dropped, growth in recoveries continued on at the previous trajectory (because recoveries lag onset of the virus). So the number of active cases drops down to basically zero before picking back up again at the new but lower growth rate of cases (3%). And then, assuming a constant trajectory, we reach the overwhelmed hospital stage around Halloween but still within the SHA plan using civic facilities for overflow. One assumes that as the number of cases and recoveries grow, the slower the rate of spread would be. So again, it’s possible that we’d be looking towards December for that threshold. The problem here is that it is too long for people to stay home, not working. Psychologically. It is also quite hard on the economy, which is obviously of secondary importance to human life, but up until fall or winter, we have a lot of excess capacity to have more cases in the public. At least, according to the SHA. And that will bring us nicely to scenario 3.

Scenario 3: What if Saskatchewan Slightly Lifts Restrictions Temporarily?

This was a theoretical scenario two weeks ago but it looks like we’re here. I put together a model where restrictions were very lightly lifted on April 21. Yesterday. Too soon but I know the government is under a lot of pressure to reopen ASAP, so I thought they’d jump on it one week after Easter. Turns out it was a week and two days. As above, you have the active cases and hospitalized lagging the initial growth trajectory so it drops and picks up to 2-3 cases (probably an error in my basic formulas), drops again and eventually starts picking up with the 6% growth rate that we assumed. Why 6%? Well, we expect that people will naturally be a bit more cautious, at least with strangers, and some of the current safeguards will remain in place. As well, we assumed a very limited relaxation of restrictions. 6% may be optimistic but I’m hoping that in light of the drop in case growth to 1% over the last week, it actually is pessimistic.

In either case, as long as there is a growth rate, we continue to march towards that point where we start running out of hospital rooms or people to infect. Growth hits levels where alarm bells are going off somewhere in June. At that point, we have around 260 hospitalized patients and almost 18,000 cases. Arguably the alarm bells should have gone off earlier, but here we are. No Canada Day gatherings. Instead, all the current restrictions are back in place. Growth slows to 3%. But not overnight. It takes time for the restrictions to have the desired effect and just in time, we curb the number of hospitalizations below the threshold in July and start seeing a decrease. But about a month later, because of the sheer volume of people, even at 3% we cross back over that threshold again and we start for the arenas. However, also around that point, we are around 100,000 cases. By a trick of exponential graphs, we actually have around 70,000 recovered cases and  58,000 active cases by August 20 but because exponential graphs.  Why does it look like there are so many more active cases? Because the space between smaller intervals is larger. If you look at the scale, there is as much space from 1-10 as from 10,000 – 100,000. So we also expect that growth would have slowed substantially by this point.

So… What Does This Mean?

Well, it means that unless there’s a vaccine or a global 4-week absolute lockdown, we can’t expect to fully stop the spread of COVID-19. But it also means that the actions we’ve taken were worth doing, if health authorities use the time wisely to obtain resources, ramp up testing, train, and prepare. It means that a temporary and partial lift of restrictions might be a good way to balance economic and social needs with protecting ourselves IF it is true that recovered cases are immune for a sufficient amount of time and if there is no way to vaccinate or treat the virus that can be tested and proven in the next two months. However, it also shows that exponential growth tends to ratchet up. When you have 10,000 active cases, even a drop to 3% still means 300 new cases in a single day. So as measures and restrictions are lifted, expect them to return sooner or later as the case count grows. Probably by summer. We should know that what happens with immunity and have a better idea of spread before we really do go ahead and let loose. We may just be accelerating our progress towards overload. But unless there is some other way of treating this virus, I’m not sure we have much choice at this point. We just have to proceed with caution. See? I told you the conclusion way back near the beginning!

Recently, an article from a US right-wing rag called The College Fix has been making the rounds. In it, an epidemiologist (I haven’t verified his credentials but they do sound good) named Knut Wittkowski, a former department head at Rockefeller University in New York, makes some claims that are questionable at best.

Now don’t get me wrong, I’m all for healthy debate and questioning assumed beliefs. We ARE going to have to have a discussion about when to start lifting restrictions and weighing the risks of continued isolation with economic and social disaster. On April 10 here in Saskatchewan, growth in total cases has dropped to 3% (and active cases will decline significantly if this continues). Right now we are buying time for hospitals to get the equipment, resources, and setup they need to allow us to lift at least some restrictions. But the College Fix article and the interview of Knut Wittkowski on which the article is based is not that discussion. It’s not balanced, reasoned, and at times defies logic.

I don’t want to link to the article or interview because garbage should be removed, not shared so they can rake in advertising dollars. But I will quote directly from it and have saved an archive PDF of the interview since these sorts of things seem to disappear or change to suit the day. I like to think that Wittkowski, who I will just call by his first name (Knut) really does believe that what he’s proposing is right. There is part of me that wonders though, if, given the alignment with Trump rhetoric and just how shallow the argument is, he isn’t aiming for a cushy appointment with the white house. Nonetheless, some selected talking points I wish to evaluate are below.

Why Flatten The Curve?

“Well, what people are trying to do is flatten the curve. I don’t really know why. But, what happens is if you flatten the curve, you also prolong, to widen it, and it takes more time. And I don’t see a good reason for a respiratory disease to stay in the population longer than necessary.”

OK. The guy’s an epidemiologist and he doesn’t know why? Even if he is ardently against flattening the curve, for whatever reason, he can’t “see a good reason” to slow the spread and reduce the overburdening on health care resources, he doesn’t deserve an interview.

For those of us that aren’t claiming to be epidemiologists, here’s a good simulation of how social distancing works and an article on why it’s needed. In essence, yes, the idea is to spread the disease infection time out to ensure that we don’t exceed our capability to treat not only coronavirus patients but also the rest of the diseases, illnesses, injuries, and transplants that go on daily. If all the ventilators are in use, you can’t have surgery to remove an early stage tumour.

Those familiar with our hospitals know that there were beds in the hallways BEFORE Covid-19 hit. We are trying to save as many lives as possible and avoid being in a place where these decisions need to get made about who gets saved and who dies.

Once we have sufficient health care capacity, yes, the approach of allowing more spread makes sense. The virus isn’t going to disappear and short of a vaccine which may be a year or more away, most of us will catch it.

At that point, we will need to pivot to protecting the at risk exclusively so that enough people have the virus that the spread grinds to a halt and it is safe for them to be in society again. These are future states, not present ones, at least not in Saskatchewan or the US. Just letting the disease run rampant should not have been the only action or more accurately inaction that our governments took.

Hide The Elderly Somewhere

“With all respiratory diseases, the only thing that stops the disease is herd immunity. About 80% of the people need to have had contact with the virus, and the majority of them won’t even have recognized that they were infected, or they had very, very mild symptoms, especially if they are children. So, it’s very important to keep the schools open and kids mingling to spread the virus to get herd immunity as fast as possible, and then the elderly people, who should be separated, and the nursing homes should be closed during that time, can come back and meet their children and grandchildren after about 4 weeks when the virus has been exterminated.”

Very, very mild symptoms. Does this sound like anyone you know (rhymes with dump) that’s always trying to minimize (or maximize) things? Some – we don’t know how many, but it may be as many as 10x the number of known cases- have no symptoms at all or mild symptoms. Maybe even “very, very” mild symptoms. It’s not that what he’s saying is wrong, it’s that he’s choosing words to paint a specific type of picture in your mind that this is some trivial little thing.

If we’re going to balance minimizing, it’s worth reading some first hand reports of people that have had coronavirus. Moving on to the next sentence, his argument is to close the nursing homes but life goes on as normal, let the kids get and spread it. Where do the elderly go? Not to family, then, obviously. One presumes some facility where all the old people are rounded up and live together for a while. What do they call those places? It’s a nursing home. But not a nursing home. That’s dangerous!

I’m not meaning to diminish the risk to the elderly in nursing homes. If I had a relative in one, I’d take them out. Even Knut implies that nursing homes are not a safe place to be right now. But many elderly still need care, and all still need food, supplies, etc. Where is this magical place we whisk them all away to that has no chance of infection from the workers that are caring for them? What does this system look like and how is it different than nursing homes now?

In Knut’s world of no restrictions, Covid-19 is spreading like wildfire outside the community, increasing the chance of infection to the isolated elderly from the people caring for them. I’m sure we could find a way to stock up a place and then also ask support workers to be quarantined with the elderly at their work for a month, so it’s not impossible, but it’s again minimizing the risk.

The Pandemic Is Over. Long Live The Pandemic

Later, the interviewer asks a telling question.

Interviewer: “You were speaking to my producer the other day on the phone, and you said, ‘The pandemic is over.’ What do you mean by that?”

Knut: “There are no more new cases in China and in South Korea. The number of new cases in Europe is already beginning to decline. The virus came later to the US, so here we see a bit of an incline, maybe, and leveling off within the next couple of days. And if we see that the cases are not increasing dramatically, that means that the number of new infections has already declined substantially and peaked about a week ago.”

Keep in mind that this interview is dated April 1 & 2. On those dates respectively, China had 35 and 31 new cases respectively, according to Worldometers. But certainly less than weeks ago to the point that it’s much closer to zero than 1000. As I’m writing this on April 10, new cases in the US are in the 25000-35000 cases per day range and appear to have peaked or plateaued these past few days.

What Knut doesn’t say is that most states have also severely limited social gatherings and are in various states of lockdown. Where would those numbers be if everyone was out infecting each other still? Obviously not flat but that is his argument. This table shows that most states declared a state of emergency about two weeks before April 1 (mid March) and some have gone further in issuing stay-at-home orders near the end of March. Both of those actions would help in flattening the curve and have nothing to do with the pandemic being over or near over.

Oh, and by the way, South Korea had 101 and 89 cases respectively on April 1 & 2. And slightly off topic, but South Korea reported today that around 100 Covid-19 patients thought to have recovered appear to have had the virus reactivate. But sure, it’s over, although he is about to say “it would have been over” except for government meddling.

What Knut’s Really About

“Well, I’m not paid by the government, so I’m entitled to actually do science. If the government, if there had been no intervention, the epidemic would have been over, like every other respiratory disease epidemic. […] I think people in the United States and maybe other countries as well are more docile than they should be. People should talk with their politicians, question them, ask them to explain, because if people don’t stand up to their rights, their rights will be forgotten.”

We get to the crux of his position here. You can’t trust government. But trust me. I’m just going to shoot from the hip and make wild statements with no basis in fact because… I get my money from elsewhere. I’m entitled to actually do science but I’m not going to use it in any of my statements here which, again, are not going to be scientific commentary, but policy commentary despite not being in politics. This is, after all, an interview all about policy and flattening the curve vs. just letting a bunch of people die unnecessarily.

People may be docile, and perhaps nowhere moreso than here in Canada, but going along with a good plan is different than throwing reason to the wolves in the name of resistance. Question, sure. Question everything. Even interviews with epidemiologists and blog posts criticizing them. Don’t give people a pass because they conform to your anti-government or anti-whatever bias.

More Trouble With Numbers

Asked how many would die in his estimation (editorial comments in square brackets):

“Okay. We have, right now, let’s take realistic numbers in the United States: we have about 25,000 cases every day, that is probably the upper limit—make it 30,000—who knows?”

So far, the peak has been about 35,000 new cases in the US in a single day. With countermeasures in place, that probably (hopefully) is going to remain the peak for the US. Not being critical here, estimation is hard, but now that we have progressed, facts are easy.

“But let’s talk about 25,000. 2% of them will actually have symptoms—that is 500 cases a day. Maybe a third or a fifth—let’s say half of them—will need to be hospitalized. That’s 250 patients a day. If they have been hospitalized for about 10 days, that means that we will have—our hospital system will have to deal with 2,500 patients every day for a certain period of time—that could be 3 or 4 weeks, and then the number will dramatically decrease again and the whole epidemic will be over.”

…

“2% of all symptomatic cases will die. That is 2% of the [25,000] a day. So that is 500 people a day, and that will happen over 4 weeks. So, that could be as high as 10,000 people.”

So far in the US, WITH social distancing and lockdowns in various states, there have been 18,000 deaths. By April 2, there had been 6,000 deaths. So 8 days after his maximum prediction, and again, WITH countermeasures in place, he is wrong by a huge degree. This is the guy you want to trust with informing policy to roll the dice and let the virus run through most of society unchecked??

Even from one sentence to the next, his definitions change. First 2% of the new cases will “actually have symptoms” then later that same 2% die. And then there’s a whole string of math based on this out-of-nowhere 2% number. If they are a known case, they probably have symptoms severe enough to get a test in the US where tests are still hard to come by or are celebrities.

CDC guidelines prioritize testing of symptomatic cases. I don’t have a number for how many of the known cases in the US are symptomatic and I’m not in the habit of making up numbers but I’m pretty confident that the number is much closer to 90% or even 100% than 2%.

Coronavirus Is NOT a Flu, Foo!

So, rerunning Knut’s numbers. Last 10 days, an average of 29,563 new cases per day. Knut says maybe 1/3 or 1/5 or maybe 1/2 need to be hospitalized and actually here we think the number is lower. But we’ll use Knut’s first guess of 1/3. Roughly 10,000 people per day. Assuming most hospital stays are 10 days (in Saskatchewan, we’re closer to a 12-day average), that’s 100,000 Covid patients added to the health care system. Not 2,500. Just a minor error.

My point isn’t to nitpick numbers. It’s that this guy is all over the place, has an obvious agenda, has already been proven to be inaccurate, doesn’t grasp why we should flatten the curve, and is speaking more from the world of myth and conspiracy than fact. He thinks this is the flu. Here’s one final bit from the interview to ponder.

Knut: “Social distancing definitely is good. It prevented the sky from falling down.”

Interviewer: “Are you being ironic?”

Knut: “Of course! I don’t know where these numbers [showing a reduction in potential deaths from social distancing] are coming from – they’re totally unrealistic. There are no indications that this flu is fundamentally different from every other flu.”

Quick interjection, an epidemiologist, scientist, or someone that cares about presenting fact would know that a coronavirus is not a flu the same way that a cold is not a flu. And if you’re thinking, “Oh, he just slipped up” he reinforces the “it’s a flu” myth 2 more times by the end of the quote. Scientists are not cavalier in their language and don’t throw words around like they don’t have meaning. He goes on to compare COVID-19 to a ‘regular’ flu and double down comparing to “other” flus.

The Virus That Hates Camping

Interviewer: “So, now we’re spending more time indoors. We’ve been told to go indoors. Isn’t that – doesn’t that help keep the virus going?”

Knut: “It keeps the virus healthy, yeah.”

Um, what?? The virus is just as happy living in your body alone on a deserted island as it is frolicking in your lungs near the TV. One might presume that he is saying that close contact from being indoors helps the virus spread, except that he has spent all the rest of this time saying that social distancing is for the chicken littles worried about the sky falling.

Interviewer: “So we should be told to go outdoors?”

Knut: “Yeah. Going outdoors is what stops every respiratory disease.”

And I will leave you to ponder that. One final note, though. The underlying assumption is that if you have COVID-19, you are immune from getting it again. That has not been proven. If immunity does occur, it’s also unknown how long that immunity would last. There are recent developments of cases in China and South Korea, thought recovered, either reactivating or catching the virus again.

Flu Part Two

Finally, to the point of “it’s like the flu.” All of us have some immunity to one type of flu variant or another. It’s a known quantity that our immune systems have dealt with in some variant or another. And there are immunizations every year for the flu variant that experts expect to hit hardest. As Knut has shown, they’re not always right. But when it comes to COVID-19, there is no immunity, and there (was) no previous variant that any of us had been exposed to.

What there is is an exponential spread through the population. Today’s 104,938 deaths are next week’s 200,000 deaths. If you are in contact, you’re likely to carry it whether you get sick or not. Flu deaths are not growing exponentially. 10,000 deaths today doesn’t become 20,000 in a week or so.

Finally, the season flu has a total mortality rate (young and old) in the US of around 0.1%. 1 person dies in every 1000. Actually shockingly high! 8 in 1000 symptomatic cases for 65+, 0.05  in 1000 (or 57 in 1 million) for under 18s.

People seem to get stuck because we don’t have a good handle on how many cases are asymptomatic. There are a lot of reasons that this is important to know. But for now, it is valid enough to use the symptomatic (known) cases causing death at least for estimating outcomes based on the number of known cases. Certainly asymptomatic cases are hard to measure regardless of the virus. So don’t get hung up that the actual rate may be lower than what is reported. It will. Of course. But that’s irrelevant right now to the people that know they have it in estimating their chances.

Even among known cases, Covid-19 mortality rates are all over the map. We are told that respiratory issues, smoking, vaping, diabetes, obesity and of course age are all factors that dramatically increase your risk of death regardless of age group. We also aren’t counting preventable deaths caused by overburden from COVID-19 as part of this mortality rate. Regardless, it seems that most think it’s at least 1% overall.

So if you’re comparing this to a flu, please reconsider. Reconsider what a moderate to bad case is like compared to a moderate to bad flu. Consider the exponential growth in cases. Consider that the world hasn’t shut down for a flu like this in over 100 years. Read some of the reports from young, otherwise healthy people that survived but suffered. Because you’re not guaranteed a pass if you’re young. You just have a much better CHANCE of surviving than your grandparents.

We’ve been watching the COVID-19 case and test numbers in Saskatchewan these past few days with fingers crossed. Today marks the third day of sustained lower growth rates and while it’s early to declare a new growth rate, it is likely that we have slowed the overall growth in the province.

Have a look at the graph below. We had a 22% growth rate (doubling the infected every 4.5 days) right up until March 30. On March 31, there was a noticeable drop off in the growth rate and it has so far continued. Two weeks back from March 30 is March 16, which happens to be the day that the province announced it was going to close schools.

Our overall average growth rate, starting from the first day with 30 cases, is now 18% (5.5 days). Averaging over that time period doesn’t really do justice to our efforts to flatten the curve, though. Over the last few days, growth has slowed to an average of 6-7% (15 days). Again, it’s early to call victory, but the point of these graphs is to try to predict where we are going to see if action is needed. We’ll be updating the graph for future use to re-project the curve starting tomorrow. But we’ve already updated the graph below for Saskatoon, adding a new 13-day trend line to reflect the slowed growth here. We are basing that line off of Mar 31, which marks 90 cases and the beginning of the lower growth trend.

Finally, the graph that is most intended to indicate the spread of infection in the province. How does it look? We’re also seeing a decrease here (or at least a flattening) in the percentage of tests returning positive. So all signs are looking good. We’ll need at least a few more days to verify that this trajectory is real, but it certainly is encouraging. It seems likely that the spread has slowed.

As always, you can find updated charts daily on our COVID-19 page. Be sure to support local businesses – restaurants that are offering pickup/delivery are some of the hardest hit thanks to low margins. Get bread from a bakery. Our medical workers are out there saving lives everyday. We can do our part and help save livelihoods, and without social contact at all. Finally, I’ll leave you with this thought. Coronavirus is showing us all what a world looks like without 1 vaccine. I don’t want to imagine a world with no vaccines.