Open questions of the pandemic, part 1

As we open another year of living with Covid-19 in the U.S., it's high time we asked the difficult questions.

1. What does living with the virus mean?

The following chart shows the number of Covid-19 deaths in the U.S. for three months from August 20, 2021 to January 3, 2022:

The data come from the OurWorldinData website. For all I know, these numbers are most likely undercounts. But what do we see in the chart?

For over three consecutive months (136 days), no fewer than 1,000 Americans died each day from Covid-19 (ignoring the reporting delay in late November probably associated with Thanksgiving). It appears that the U.S. government is managing to a roughly 1,200 daily deaths run rate, as the line goal-seeks this number on the right of the chart. Twelve hundred daily deaths amount to 438,000 annual deaths.

In a bad year, the flu kills ~40,000. So this "flu" is still causing 10 times as many deaths... with almost 60% of the population fully vaccinated with two doses, some with three doses.

Is Covid-19 like the flu? That was the question from day 1 of the pandemic. Has Covid-19 become like the flu? That is the shocking claim of some experts.

2. Endemic versus pandemic

Two common headlines have circulated in the media recently. One claims that Covid-19 is becoming endemic, like the flu. Another argues that this Omicron surge is fantastic because it is "mild" and getting lots of Americans infected is a great thing as it induces immunity through natural infections. (The recent policy shift that shortens the isolation time for infected people seems to endorse such a viewpoint.)

Those two headlines cannot coexist. Endemic diseases like the flu do not have lasting immunity - the defining characteristic of an endemic disease is that one will get re-infected periodically. Separately, the pursuit of "herd immunity" and the associated use of SIR-type epidemiological models assume that once infected, someone cannot get reinfected, as infected people are removed from the susceptible population. (For endemic diseases, one uses SIS models in which previously infected people become susceptible again.)

If Covid-19 is really endemic, there is no point in hosting Covid-19 parties because getting infected once does not prevent future reinfections.

3. Zero Covid

In the West, our journalists love to point out that lockdowns don't work. When they do work, our journalists say people are upset, or they say zero-covid countries are a risk to the rest of the world (link), or that people are leaving in droves (the journalist obviously belongs to a certain class for which people can pay to obtain citizenship status anywhere they like), or that people are starving to death.

Maybe the only criterion these journalists use is economic (and even that may not be supported by data) because the following chart shows the cumulative death per million in selected countries from the start of the pandemic to now:

There are two clusters of countries. The top 4 (Brazil, US, UK, Sweden) are countries most often associated with anti-lockdown sentiments. The journalists in these countries argue that lockdowns have no benefits while destroying the economies, and worse, usurping the freedom of their citizens. Sweden, one should note, was initially no lockdown but subsequently changed strategy.

The other cluster (South Korea, Australia, Taiwan, Hong Kong, New Zealand) contains countries most associated with zero Covid policies, which typically means lockdowns whenever cases flare up. The "failure" of such policies has led to cumulative death rates that are at least 20 times lower than those experienced by the anti-lockdown countries.

***

The news on Covid-19 is confusing and frustrating. It won't get better until we clarify some of these fundamental questions - there can be two sides to each debate but what muddles things up is when people make arguments that require holding incompatible views simultaneously.

The next post will feature a few more open questions.

 

Comments

[Michael Droy]

In a bad year, the flu kills ~40,000. So this "flu" is still causing 10 times as many deaths... with almost 60% of the population fully vaccinated with two doses, some with three doses.

Is Covid-19 like the flu? That was the question from day 1 of the pandemic. Has Covid-19 become like the flu? That is the shocking claim of some experts.

Health analysis always used to be done using QALYs - quality adjusted life years. For 2 years the term QALY has disappeared from all conversations.
But any Flu vs Covid 19 comparison demands a QALY lost analysis not a death count. Flu certainly can kill otherwise healthy patients in a way Covid almost never dows.

Then you have the "death certificate issue" now as with or by Covid.
In US a bad year is, you say, 40k flu deaths. In the UK a bad year (1999) is 60,000 flu & pneumonia deaths (counted together, half that in a good year). 0.1% of the population.
Roughly 350k by US standards and close to the level of Covid last year. And yet the US in recent years has reported death by Flu & Pneumonia of 50-60k.

https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/datasets/deathsduetocovid19comparedwithdeathsfrominfluenzaandpneumonia
Table 5.
https://www.cdc.gov/nchs/pressroom/sosmap/flu_pneumonia_mortality/flu_pneumonia.htm
I summed up the deaths by years.

I'm not arguing that a 350k equivalent is roughly the same as 400k Covid deaths. But clearly there are different recording practices in flu deaths between US & UK and that recording practice have a huge impact on totals. It gives us a clue about the difference in reporting practices between flu and Covid inside the UK or inside US.
What ever a fair balance of deaths is, both practices and the impact on QALYs means your 10 to 1 ratio is clearly a double exaggeration.

Time to recognise that the real impact of Covid from 2022 onwards is going to be 1 to 2 times the impact of Flu, while we will benefit from reduced victims of flu.

BTW as a resident of UK I can assure you that most of the media follow the fear porn version of the covid story, and the libertarian version is very low key. Only recently is anyone talking about how most covid hospitalisations are of patients who entered hospital for non-covid reasons.

I do hope in Part 2 you will address the impact that addressing Covid full on has lead to other health services, health and deaths from traditional sources, and in particular the impact of health services over the past 30 years on the incredible reductions we have seen in deaths / increase in Life Expectancy.

I'm also very curious about any ideas of how we can differentiate going forward between excess deaths from Covid, reduced deaths because Covid victims died early, and increased deaths due to other Time of Covid factors (economic, social and health service).
It seems to me that

[Kaiser]

MD: No one can dispute your argument since it is based on unverifiable assumptions of what certain governments did or did not do relative to specific data series, including claims of massive counting errors on some data but not others. Yet in your comment on my other post, you also believe that a massive prediction error is nothing to sneeze at. If one has license to invent one's data, then all stories are possible.

[Michael Droy]

Ditto your claim (that I quoted) that
In a bad year, the flu kills ~40,000. So this "flu" is still causing 10 times as many deaths... with almost 60% of the population fully vaccinated with two doses, some with three doses.

But I dispute your claim that we cannot verify that different countries count differently. The Flu + pneumonia numbers make it very clear that they do count differently. In fact I'm not even sure anyone would argue that
And certainly there are counting errors but I'm not claiming that, I'm claiming massive differences in counting methods.

With the other post - you need to clarify what a mass prediction error is (upside and downside).

[John Birge]

Kaiser,
Interesting stuff. A question: in your view, in the U.S., has any state done clearly better than others in minimizing deaths per capita? (after adjusting for age, urbanization etc). Bill Maher espoused the opinion that essentially there is no rhyme or reason to state death rates per capita and therefore any mitigation measures taken were ineffective. What does the data really say?
Cheers!
John Birge