week ending 2022-09-08 General

Unwanted Side Effects

This paper from Paris found that, while infant head trauma (which is usually caused by child abuse) did not go up in 2020, it doubled in 2021. 😬


Good news on the mucosal antibodies front! (Reminder: you want antibodies in the mucous because COVID-19 hits the mucous membranes of the respiratory or gastrointestinal system first, as this article explains.)

  • This article says that China has approved an inhaled version of Sinovac. The preprint giving the results says that it worked much better than injected Sinovac, but unfortunately still waned significantly over time. Full data has not been published yet.
  • This article says that India has approved a nasal vaccine from Bharat Biosciences. The data has not been published yet.

This paper says that COVID-trained T-cells aren’t present in nose mucous in people who are vaccinated, but are present in people who have had COVID-19 breakthrough infections. Furthermore, it says that they don’t show significant signs of waning.

I think this is really good news: I think it means that much better vaccines are possible. Like the inhaled vaccine from China and the nasal vaccine from India, perhaps!

I was reminded by this tweet thread that vaccine effectiveness dropping is partly illusory: the difference between outcomes for vaccinated vs. unvaccinated is going to be lower if the unvaccinated have immunity from, y’know, getting infected.

Last week, I showed a graph for the US which seems to show that the first booster didn’t do much but the second booster did:

This made no sense to me. The data appears to come from the US CDC numbers, and they also have cases by vax status:

This also looks strange: it looks like getting a first booster makes your risk go up but getting a second booster makes your risk go down. ??? Maybe it’s that the people who got one booster were those who were more vulnerable, and then the second booster helped a lot? I am really puzzled by these numbers.

This Letter to the Editor says that boosters antibody effectiveness in test tubes wanes at the rate of ~17-20% per month (depending on the strain) for those who did not have a COVID-19 infection, and at ~12-17% for people who had had an infection.


This paper from Germany says that after 274–383 days, 57% of the people in their study (people who had recovered from COVID-19) had signs of heart problems.

This preprint from the USA says that 21.5% of people surveyed between June 30 and July 2, 2022 (so likely BA.2 or BA.4/5 infections) had Long COVID symptoms after four weeks.

This article says that some billionaires started a research initiative with $15M dollars to study Long COVID. (I know $15M sounds like a lot, but it’s really not much for such important work. Still, it’s a nice start.)


This preprint is mostly about a new breath sampling device, but they also found that quiet breathing at rest (called “tidal breathing”) usually had no or very little detectable COVID-19 virus in it. They did find a lot of COVID-19 in the breath of people who were singing, however.

This preprint from the USA found that 17.3% of people had had COVID-19 between June 30 and July 2, 2022. As it was a survey, that’s likely to be an undercount.

This report from StatCan says that about 19.5% of Canadians said they had tested positive for COVID-19 (data collected through March). Another 8.1% thought they’d had it but didn’t have a confirmatory test. This is very encouragingly close to the 28.7% which the COVID-19 Immunity Task Force estimated in their report for March results. (Meanwhile, the updated COVID-19 Immunity Task Force web page estimates that 58.6% had been infected by the end of June 2022.)

This article reports on this paper, saying that they found weather patterns which correlated with COVID-19 spread, with UV being the biggest determinant of spread (and negatively correlated). The article makes it sound like it was really iffy, though: the study used three different ways of modelling and frequently the models would not agree on how a variable affected the result. 😬 I would also like to note that equatorial countries tend to have much younger demographics.


It is not clear which variant will hit us next. BA.2.75 is the one which most people have their eye on, but it seems underwhelming. Don’t get me wrong, it’s worse than BA.5, but not that much worse (and maybe not worse enough to cause another wave).

This preprint says that BA.4.6, BA.4.7, and BA.5.9 show heightened resistance to various monoclonal antibodies. Bebtelovimab is the only monoclonal antibody which is effective against all the Omicrons, and it is only available in the USA.

By the way, I have said (maybe in person, maybe here, I don’t remember) that it didn’t matter that much if our government got BA.1 or BA.5 boosters because the next strain was probably going to be BA.2 derived, and BA.1 was closer to BA.2 than BA.5 was. WRONG. BA.5 is closer to BA.2 than BA.1 is.

Does it matter? Probably not. The difference in effectiveness between the two against BA.1, BA.2, BA.4, and BA.5 is not huge; we don’t know what the next scariant is going to be. Maybe it’ll be Delta-derived, who knows? The difference between one dose and two doses is much higher than the difference between bivalent BA.1 and bivalent BA.5 vaccines. (I mostly want to make sure I issue a correction when I am wrong.)


This letter to the editor says that 2.3% of patients in the study who got Paxlovid had viral rebound, while 1.7% of the people who did not get Paxlovid rebounded.


As I said in today’s BC post, part of the rationale for waiting a little bit to roll out the boosters to have the best protection during the worst of respiratory virus season, i.e. Nov/Dec. I am utterly unconvinced that the worst months for flu will be Nov/Dec. The Canada Flu Watch shows that the flu peaked already — during week 19 (the week of 9 May 2022) — and is going down.

The BC flu report has a graph, but it lumps all the influenza-like illnesses — COVID, flu, RSV, rhinoviruses, adenoviruses, whatever. It also indicates that our respiratory infections have already peaked.

The BC flu report also says “several surveillance indicators suggest the earlier-than-expected influenza A(H3N2) activity this season may have peaked or plateaued in Australia” though they caution that the 2019 season also peaked early but then lingered. From the Australian flu report, it looks like it was high but also short.

However, Australia was probably doing a lot more testing than they did pre-pandemic, if their testing pattern was similar to the US/Canada. Also note that the 5-year mean (in yellow) has been dragged down by 2020 and 2021 having essentially zero (flu because of anti-COVID measures). Instead of dividing by five years, they should probably divide by three years, so that yellow line really ought to be 66% higher, which would have it peak at ~12K instead of ~7K (with higher testing).

We can also look at hospitalizations from flu in Australia, which won’t be affected by the testing rate. Hospitalizations this season have been lower than some pre-pandemic years:

Again, the 5-year mean line probably should be 66% higher, meaning that its peak would be ~160 instead of ~100, meaning that 2022 isn’t actually that much worse than normal.

The absolute amount of flu has been low in North America this year. From the US CDC, because they have nicer graphs than Canada does:

Recommended Reading

If the lipid nanoparticles used to sneak mRNA vax into your cells are your thing, this article talks about work being done to improve them.