Chapter Eight - Non-Zen and the Art of
Pandemic Twitter Haverism
In this chapter we'll have a look at how the Twitter debate developed from May 2020 onwards, but I first want to provide some background data showing the stage of the pandemic in Sweden at this time. Tegnell stated that he thought ICU admissions was the best parameter to track the development of the pandemic, because, unlike the number of confirmed cases, it isn't influenced by external factors like the number of tests carried out. As Donald Trump said in July 2020, "When you test, you create cases." - The Donald has a way with words, as always, but at least he was correct about the underlying fact: the more you test, the more cases you will find. It sounds obvious now, but a lot of people didn't understand this and made some really spurious arguments based on the number of confirmed cases in various countries. This was also the topic of a discussion between Tegnell and the "mysterious German journalist" at the press conferences (see chapter 4). I don't remember the context exactly and I can't find the relevant press conference now, but the nub of the matter was that because Germany tested more, they would also discover more mild cases. Anyway, here's a graph of the Swedish ICU admissions from March 1 2020 until the end of 2024 from the Swedish Intensive Care Registry (https://www.icuregswe.org/en/data--results/covid-19-in-swedish-intensive-care/), with the worst week highlighted (week 15) - the use of week numbers as a time reference is a frustrating Swedish custom; after nearly 15 years abroad I lost all grasp of the concept and when people now say things like "I'm going on holidays week 35" it makes my eyes glaze over with incomprehension, but a quick check in the calendar for 2020 shows that week 15 comprises the period April 6-12.
It should be pointed out that this shows the number of "vårdtillfällen", which in English seems to be called "episodes of care", which is not the same as the number of patients. If a patient is moved from one ward to another, this will be registered as two episodes of care. The difference becomes clear in this overview table: total episodes of care 14,581, total number of patients: 11,365.
And the difference between the table and the graph (14,581 vs. 14,476) is the episodes of care from January 1 until December 12 - 105, proving that Anders Tegnell and Björn Olsen were correct. The former said more than once that this is a disease we will have to learn to live with and the latter stated in a TV interview at an early stage that the pandemic would eventually become a "regular" virus that we won't really care that much about - let's for the moment disregard all the other nonsense he spouted during the course of the pandemic and that there's still a small sect who are convinced that covid-19 will mean the end of civilisation.
The reason why I brought up this graph is because of the following tweet by Mattias Wallén from April 10 where he comments on a podcast discussion produced by the editorial board of Svenska Dagbladet:
At the time I thought he was involved with the pod and had participated in the discussion, but I believe this is his personal illustration of the arguments brought forward by, i.a., Robin Fondberg and Joacim Rocklöv. Unfortunately both Fondberg and Rocklöv have deleted their Twitter profiles, regarding the former it's unfortunate because she was a sensible and knowledgeable voice in the debate and one of those scientists (she's a post doc at the department of clinical neuroscience at Karolinska Institutet) who didn't lose their heads during the pandemic, regarding the latter it's unfortunate because all his daft Twitter takes are no longer there for our general amusement, but never fear, some receipts have been kept and will be dealt with in good time. Anyway, let's have a look at the two graphs side by side. For the sake of simplicity I've reduced the graph from the Intensive Care Registry to the first half of 2020 (we'll get to the second wave in due course) and I've used the metric occupancy per day - it gives a truer picture of the pressure the health care sector was under. I would have used it above as well, but I couldn't find it when I wrote that section...I've also highlighted the day when ICU occupancy peaked - On April 26 there were 558 patients in ICU wards in Sweden with the diagnosis covid-19:
So the first thing we can conclude is that at the time when Wallén wrote this tweet on April 12 (Sunday of week 15, cf. above), ICU *admissions* were peaking, *occupancy* kept increasing for another two weeks or so, due to the fact that people admitted often needed to be kept in the ICU wards for two weeks or so, or in some cases even longer. The label of the y axis in Wallen's illustration is "ICU admissions per day", illustrating Rocklöv's belief that at some stage there would be 5,000-9,000 ICU patients per day. I believe this should be occupancy rather than admissions and Smedman makes the same argument in a reply to Wallen's tweet - one of the few things me and Björne can agree on. Either way, it doesn't really matter. Look at the figure in the bottom left hand corner of this admissions graph from the Intensive Care Registry for 2020 up until July 31: Total 3,471. (See chapter five for more about Naucler's and Rocklöv's "advanced model").
In his tweet Wallen says "In just 7-14 days we'll know more". A month later, on May 11, I tweeted:
Note the amount of traction my tweet generated...I see now that gronvita also commented on Wallen's tweet a good six hours before me and did get a reply, although in the end it seems the discussion was inconclusive.
In the real world outside the Twitter bubble, there are a few events I think should be added here for context. On April 16 Tegnell said in an interview on Norwegian national TV that there would be full immunity in Stockholm "some time in May", something which seemed strange and would later be used as "evidence" against him. Here's a link from the Norwegian broadcaster's website with the clip and a long article about the situation in Sweden, including Naucler's doomsday predictions: https://www.nrk.no/urix/tegnell_-_-det-kan-bli-flokkimmunitet-i-stockholm-i-mai-1.14984679. What Tegnell actually says is that their data modellers believed this, based on the data available and that these predictions were only as good as the data they're based on, but even this seemed wrong and the winter wave would later show that the seasonal effect probably to a greater extent rather than immunity was the driving force behind the improving situation in late April and the beginning of May. Here's an article from On April 21st, FHM released a report in which it was incorrectly stated that there were 1,000 cases for each confirmed case of covid-19. I read the incorrect version of the report when it was released and with my limited mathematical proficiency, it seems the underlying data was correct and that the model did not in fact predict that only 0.1% of all cases were discovered. I later received a first hand account of what had gone wrong from Disa Hansson, who was one of the data modellers at FHM at the time, but I don't remember the ins and outs of it now and in all honesty, I think it went over my head even then. On May 5 Johan Giesecke, one of Anders Tegnell's predecessors as state epidemiologist, published an article in the Lancet called The Invisible Pandemic, which caused a lot of stir in what would later become the "zero covid" community, which we'll also get to of course, in due course.
Tegnell's spurious statement about immunity on Norwegian television notwithstanding, it was clear that the curve had been flattened and was now heading in the right direction. And as David Olsson points out, Tegnell provided a better explanation of his reasoning regarding herd immunity at the press conference the next day:
On Twitter, there was no sign of a flattened curve in the discussions and arguments. Today I can't help but wonder how I would have reacted then if I'd known that I'd get sucked into the Twitter war for another two and a half years and how preposterous it would become. After all, the previous autumn I had deleted my account from 2012 because I didn't want to waste my time on futile political discussions (see chapter one) and here I was again...However, it should still be noted that, at this stage and with a few exceptions, the tone was still reasonably civil. Smedman threatened to block me, because of this tweet in which I expressed my annoyance at his nitpicking of everything the Swedish Public Health Agency did and said, while seemingly accepting everything from Ferguson et al. at face value.
(I used the phrase "strain at a gnat and swallow a camel" in an earlier tweet in that thread to describe his behaviour. Today I learned that this is a quote from Matthew 23:24, and now it seems blindingly obvious that it would be from the Bible.) Smedman apparently though my tweet was bordering on the amoral. I didn't understand what he meant and asked for a clarification; it never came, but he also didn't block me.
During May and June 2020 there were some notable firsts: On May 10 I referred to the State Epidemiologist as "Tengan" for the first time. I'm not sure where I picked it up and because "tengan" is the third person plural subjunctive form of the verb "have" in Spanish (tener), trying to find when it first appeared on Twitter or elsewhere on the internet would be as difficult as it would be for a camel to go through the eye of a needle (to continue with the biblical references rather than opting for the more obvious haystack simile).
Also worth noting here is how I use one of Tegnell's more dubious statements (see above) as an argument in his favour. This particular part of his reasoning is of course fully correct - the models are only as good as the data they're based on - but there's definitely some pro-Tegnell spin on my part here.
On May 12, I tweeted my first comment in a thread started by Eric Feigl-Ding, "NO TO SWEDISH MODEL", replete with dubious graphs and "facts". Unfortunately whoever I replied to has since either deleted their tweet or, more likely, deleted their profile. Note how matter-of-fact my reply was, and how much attention it garnered...
On May 26 "veteprotein" made his first of many appearance in my feed, in a thread started by Emanuel Karlsten whose weekly reports on the pandemic provided a good overview of the developments and raised aspects not covered elsewhere in reliable media, here regarding how deadly the pandemic had been so far.
In the same thread the second person who in their profile say they're a civil engineer (Smedman being the first) also appeared for the first time. I had made the argument that a comparison between the flu and covid-19 might be quite reasonable after all, all things considered. Lindberg stated that the number of deaths was high, even with all the measures introduced and asked how high the mortality would have been without any measures, which was a reasonable argument.
Another (better?) way of looking at it could be that thanks to the measures, the mortality had been lowered to levels that didn't make 2020 stand out from other months with high mortality during the last 30 years. Very soon Lindberg turned to complaining about "the world's highest taxes" (which is wrong, by the way) and that Sweden didn't even try to "protect its weakest 'members'" and then moved on to asking about the costs of people following the advice from the Public Health Agency of staying home if they felt any kind of flu-like symptoms or in "voluntary quarantine". When I asked about the costs compared to a lock-down, it was "hard to quantify" and my tweet about the €800bn EU rescue fund was left without comment.
I concluded that he would have complained no matter what the government did. These days he's moved on from the pandemic and the focus of his tweeting now seems to be complaining about Muslims and the Social Democrats (who were booted out of government in the 2022 elections). I see no reason to change my assessment of the motive behind his tweets.
