Sweden may be recording COVID-19 deaths differently than other countries

Department of Lies, Damned lies, and Statistics: “In Sweden, Will Voluntary Self-Isolation Work Better Than State-Enforced Lockdowns in the Long Run?” (Reason).

As noted previously in this blog, despite having access to schools, restaurants, gyms, and offices, Swedes have been dying at a lower rate than residents of shut-down Massachusetts (latest). But there are some European countries in shutdown that have yet lower official COVID-19 death rates. Almost nobody in Germany or Norway dies from COVID-19. Are they exceptionally lucky, doing something better, or using a different standard to attribute death to the evil coronavirus?

(Summary: Swedes automatically count anyone who has tested positive for COVID-19 as a COVID-19 death; the Norwegians require that a doctor assess that an ancient person has specifically died because of COVID-19, rather than one of 3 pre-existing conditions, and then take time from his/her/zer/their day to report that to the authorities.)

From the article:

There are also reasons to think that Sweden is doing better than these comparisons suggest. Many countries don’t count COVID-19 deaths outside of hospitals. When people die at home, in nursing homes, or in prisons, they don’t show up in the coronavirus death count.

In the Stockholm region of Sweden, 42 percent of deaths took place in nursing homes for the elderly. In many countries, and some U.S. states, those deaths would not show up in the data.

According to Johns Hopkins University, Belgium has twice as many COVID-19 deaths per capita as the Netherlands. But in Belgium, almost half of those deaths are from nursing homes, while testing is more rare in Dutch nursing homes so fewer deaths there are attributed to the disease.

After France started to include nursing homes in the statistics, the total number of French COVID-19 deaths jumped by almost a third.

During the present pandemic, Sweden systematically checks the list of people who have tested positive for the virus against the population register. Every time the government discovers that someone who had the virus has died, that person is registered as a COVID-19 death if it happened within 30 days of the diagnosis—even if the cause of death was cancer or a heart attack.

It means that Sweden reports the number of people who die with COVID-19, not of COVID-19.

Even in a culturally and geographically similar country like Norway—celebrated for its low death rate—they do things differently. The Norwegians only count something as a COVID-19 death if a doctor concludes that someone was killed by the disease and decides to report it to the country’s public health authority.

The article confirms what I posted earlier, i.e., that Swedes are not running out of ICU capacity:

The Swedes who have died from the coronavirus did not die due to lack of hospital beds or ventilators. Thanks to a rapid increase in intensive care unit capacity, 20 percent of Sweden’s ICUs are unoccupied. Stockholm has built a new field hospital, already equipped to receive hundreds of COVID-19 patients, including 30 ICU beds. So far it has not had to open. The average age of the dead has been 81, which is close to our average life expectancy.

Why didn’t Swedes drop dead like the models said they would?

For example, the influential Imperial College model estimates a higher reproduction rate of the disease in Sweden than in other countries, “not because the mortality trends are significantly different from any other country, but as an artefact of our model…because no full lockdown has been ordered.”

In other words, the model could only handle two scenarios: an enforced national lockdown or zero change in behavior. It had no way of computing Swedes who decided to socially distance voluntarily.

[Believers in the Church of Shutdown, of course, will say that Swedes are completely different from Americans (note that 25 percent of people living in Sweden have no genetic, cultural, or linguistic connection to stuff we might consider “Swedish”; one quarter of the population was born somewhere else or has two parents born somewhere other than Sweden; Swedish 15-year-olds actually scored slightly lower than Americans in the science section of the PISA test (though they did a lot better in math)). Had at-risk Americans not been able to tap into strong leadership (from epidemiology professor Donald Trump in the White House!), they would have read media reports of mass deaths and not changed their behavior in any way. No American would have switched to work-from-home. No American would have decided to cook at home rather than spend an hour in a jammed restaurant. No American would have invested in a mask or Clorox wipes. Due to universal stupidity among Americans (or at least the nearly half who voted for Donald Trump), a Swedish approach of shutting down mass gatherings and trying to isolate the vulnerable could never have changed behavior or epidemic velocity in the U.S.]

Of course, we probably won’t be able to evaluate the success or failure of any country’s policy until early 2021 (otherwise we risk celebrating a country for preventing deaths when all that happened was that the country either postponed the death or classified it differently). But I think it is interesting that already we’re getting a glimpse into why apparently similar countries should have such different death rates.


16 thoughts on “Sweden may be recording COVID-19 deaths differently than other countries

  1. I live in Germany and if you do the math to adjust the numbers to be comparable, Sweden has about the same ratio if positively tested people (a few less, which is nice), but a lot higher fatality rate. That is, although in Germany people dying in nursery homes etc are counted in the statistics, so the way of counting is somewhat similar. In a few counties we are even doing forensic postmortem examinations to clarify the case. Experts also admit, that we are counting a few dearhs to much, because we are not differentiating between death by and death with Covid19.

    One thing, Sweden benefits from is the low density of population compared to
    Germany. And you have to keep in mind, that thei are about 2 weeks behind Germany which itself is about 2 weeks behind Austria in the pandemic timeline.

    • Hello Andrej!
      No, you are actually wrong. The Stockholm region, not Sweden as a whole, is one to two weeks behind the Italian timeline. Stockholm had their winterbreak for schools about at the same time as the footballmatch (soccer) between Atalanta and Valencia. That game is concidered as (one of) the crucial outbreak point for BOTH Italy and Spain…

      The game was played in Milano and thousands of Swedes (skiers in the Italian Alps) travelled by air via Linate Airport (and other airports in northern Italy/France/Austria) during that week. A week and a half later Austrias ski-resort Ischgl came into focus. Several hundred Swedes had been there. But THEN Sweden ordered THEM into isolation. So the “Austrian spread” has not been that bad, compared to the “Italian spread”.

      What I am trying to say is this: Sweden (the expanded Stockholm region) is, compared to many other regions/countries, almost in the same timeline as Italy/Spain.
      If you look at the bigger picture… Well, we are doing somewhat good. The problem is the homes for our elderly – we have not shielded our old folks good enough.

  2. Alphabet corporation was well ahead of the curve, but our company waited until the government finally started shutting down the streets to finally let us use our laptops remotely instead of bringing them to the city every day.

    Not sure if the current protesters are business owners who want their employees in the office, employees desperate to support their employers to keep their jobs, or just protesters looking for something to protest.

    Swedes famously don’t have to work. Their socialist system covers all needs from heaven, so that would make it easier to voluntarily shut down businesses.

  3. Phil,

    Thank you for writing this blog, I was just rereading a post from December 22, 2018 On Natural Causes An Epidemic of Wellness…

    Great post!

  4. Phil, is it possible the swedes do not want to take everyone hostage due to the Stockholm syndrome? Stockholm after all is in Sweden.

  5. Phil,
    Thanks for another outstanding post and that link to the article in Reason. I have to say I was more of a disciple of the Church of the Shutdown, but you have opened my eyes and mind to other possibilities. It will be interesting, as you and Johan say, to look back and see how this plays out (assuming I am still around!)

    • Paul: I have never considered myself enough of an expert to be a priest in any of the current religions (there are at least three, right? (1) Test, contact trace, and quarantine (South Korea); (2) Shutdown and cower (U.S., Germany, Spain, et al.); (3) Adjust slightly and accept that this will be with us like the flu (Sweden)).

      I am subject to media-induced bias. I wasn’t that afraid at first, but then got exposed to a constant stream of media stories regarding young healthy people who were killed. (I’m not sure where they find these stories, actually. Today’s Massachusetts data show only 4 people under the age of 60 who were killed, all of them previously unhealthy, and a median age of “80s”. See https://www.mass.gov/doc/covid-19-cases-in-massachusetts-as-of-april-19-2020/download ) So I switched over to the Church of Shutdown mentally. But then I had doubts: if there is no treatment, won’t we suffer all off the same problems the day that we emerge from our bunkers? (The Imperial College folks had an answer to this: shut down for 80% of the time over the next couple of years! But then their dogma changed: shutdown for part of spring 2020 and ye shall mostly be saved.)

      Right now I feel that the Swedes have the most logically coherent story and also that Church 3 is the only church that hasn’t had to radically change its dogma from week to week (Imperial College: “500,000 will die” then “20,000 will die and it was always 20,000”).

      So I’m still a little afraid when I read the dramatic stories of coronasuffering and coronadeath (somehow our media never chooses to write about a non-famous patient in his/her/zir/their 80s or 90s dying; it is almost as if these deaths are irrelevant!), but I don’t believe that young people hiding in their crummy apartments are going to change my chance of dying.

      I do think it is possible that both Churches 2 and 3 are right: Church 3 is right for Sweden because they can treat a patient without investing several hundred hours in paperwork and bureaucracy. Church 2, on the other hand, is right for the U.S. because we aren’t organized enough to do Church 1 and our health care system can handle only a few patients at a time, having been optimized for extracting huge $$ from each of a handful of complex cases rather than being optimized as an assembly line for simple cases.

      The implication from the above paragraph is that Americans with money should try to relocate to Sweden for a couple of years until the U.S. figures out what to do.

    • @philg None of the options has really worked, all have their pros and cons:
      1. Korea, Japan and Singapore are now having second waves, probably because it is impossible to track asymptomatic cases, which have been proven to comprise quite a large fraction of the population.
      2. South Africa (the country in which I live) was exposed quite late, but did a prompt shutdown which restricted travel related cases (like Australia), and looked like it might fall into category #1. After a few weeks we find the worst of both scenarios, a damaged economy, with a protracted leaky shutdown and slow exponential growth.
      3. Voluntary self-distancing may work in countries (Sweden) with the right demographics, cultural norms, and optimum healthcare, but it is too soon to tell.
      4. Third world countries (India, Brazil, and most of Africa) will be very hard hit, with chaotic interventions, and ineffective healthcare.

    • Thanks, Phil, well articulated and concur on the three “religions”. What’s interesting as you point out, theoretically that until a vaccine appears, they all ultimately yield the same number of deaths (well I guess the later in time you push out people who catch it, the more chance there are therapeutics that can reduce the number of deaths). But a chart that has a BIG RED curve over time interval t and one that shows a green curve over a longer period of time, the red curve is alarming even though the deaths are the same!

      Gordon has articulated well the pros/cons and said what I was going to say about Sweden and their path forward. And it sounds like South Africa is much like the US.

      Finally, with some places opening up more (Florida has just started opening up some of it’s beaches), we will see if people can act more like the Swedes. I have my doubts..

  6. To all Trump supporters: RESIST the shutdown, get out there and mingle! Get together in as large numbers as possible, GO TO CHURCH, shake hands, hug each other, go back to work! Protest, claim your liberty back, take your guns!!! And please, make sure your dear leader joins in your rallies.

  7. Why are age groups the preferred method of presenting the data – just because that’s an easy number to procure? Wouldn’t it be a lot more illuminating for the independent variable to be comorbidity? As well as comforting to any otherwise healthy 70 year olds who might be around.

    • Mitch: Despite what must be closing in on $1 trillion spent on electronic health record systems, I don’t think it is generally possible for a U.S. hospital where a person is treated for COVID-19 to get usable records from institutions that previously treated the same person. So I don’t think we will ever have good data from the U.S. on the prior health status of COVID-19 victims.

  8. The Economist has some interesting data on excess mortality in specific regions and countries, which should deal with issues of under-testing or failing to report cause of death. For example, from March 19 to April 2, England and Wales had 7200 excess deaths compared to the historical average. For the period March 20 to April 3, New York City had 5100 excess deaths. For March, Jakarta had 1500 excess deaths. Doesn’t include data for Sweden and Norway yet.

Comments are closed.