Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
פרופ' אייל שחר Profile picture
פרופ' אייל שחר
@prof_shahar
פרופסור אמריטוס (בריאות הציבור, אוניברסיטת אריזונה), ד"ר לרפואה (אוניברסיטת תל אביב), מוסמך באפידמיולוגיה (אוניברסיטת מינסוטה)
Andrew Lowy Profile picture UnknownUnknowns Profile picture 2 subscribed
Jan 9, 2022 • 14 tweets • 5 min read
1/
 Thread.

Cumulative excess deaths in Europe.
EuroMOMO data (accessed Jan 6), recreated by “flu years” (Oct-Sep) instead of calendar years, by @OS51388957

All ages & by age group.

What may be learned about the pandemic?
A lot.

1)Magnitude
2)The price of panic 2/
 First, why flu years (Oct-Sep) instead of calendar years?
To avoid splitting winter mortality.
Explained here.
shahar-26393.medium.com/not-a-shred-of…
Nov 14, 2021 • 12 tweets • 3 min read
1/
 Thread.

Does waning protection end in null effect, or in VED?

Are we looking at evidence of VED in Qatar data? (buried in the Appendix, Table S11)

Note: highly technical thread, but don’t miss the red flags at the end.

nejm.org/doi/full/10.10… 2/
 We see the same alarming results for symptomatic infection.
Nov 7, 2021 • 13 tweets • 4 min read
1/
 Vaccine protection against NON-COVID death is “pseudo protection”.
It is evidence that VE against COVID death is over-estimated.

Solidifying the evidence for pseudo protection: “person-time”-based analysis of NON-COVID death.
England data, weeks 1-38

Graphs by @OS51388957 2/
 We recently documented lower non-CVOID mortality in vaccinated than unvaccinated (England).
Implies: vaccinated are healthier (on average).
Implies: rates of COVID mortality in vaccinated are lower, in part, because they are healthier.
Nov 4, 2021 • 10 tweets • 3 min read
1/
 Thread:
England data show COVID vaccine “protection” against NON-COVID death.

That’s evidence that VE against COVID death is confounded & over-estimated.

Likely substantially.

Graphs by @OS51388957 2/
 ONS published data on COVID death and all-cause death by V-status and age:
10-59
60-69
70-79
80+
Age range too wide for <60.
Restrict to 60+ (most deaths)
ons.gov.uk/peoplepopulati…
Nov 3, 2021 • 6 tweets • 2 min read
This is an interesting study, which raises a few methodological questions.

>>>

jamanetwork.com/journals/jama/… “We analyzed the odds of receiving a COVID-19 vaccine in the 28 days prior to spontaneous abortion compared with the odds of receiving a COVID-19 vaccine in the 28 days prior to index dates for ongoing pregnancies”

Why define the hazard period as 28 days and not longer?
>>>
Nov 2, 2021 • 6 tweets • 2 min read
Correct interpretation of this study:

All participants were vaccinated. So nothing can be inferred about the effect of vaccination.

By analogy, imagine a study of (only) low-dose aspirin users. Can you say anything about effects of aspirin?

No.

>>> Graphs show the effect of prior infection (vs. no-prior infection) on having a second (or first) infection in vaccinated.

Risk of a second infection much lower than risk of first infection (in vaccinated).

That's the effect of naturally-acquired immunity (in vaccinated)
>>>
Oct 27, 2021 • 14 tweets • 3 min read
1/
 Thread.
Why VE (COVID death) is far from 90-95% in elderly, and why we’ll never have a good estimate.

Bottom line: because of residual confounding bias and poorly designed trials. 2/
 What is confounding bias?

Two ways to explain:

Easy: vaccinated & unvaccinated are not exchangeable. Their risk of death is different, even if vaccinated had not been vaccinated.

Complicated: Causal diagram
Oct 23, 2021 • 6 tweets • 2 min read
1/
 Interesting analysis of non-COVID death, but incorrect interpretation by the authors.
Reported:
“This cohort study found lower rates of non–COVID-19 mortality among vaccinated persons compared with unvaccinated persons...”

cdc.gov/mmwr/volumes/7… 2/
 Risk of non-COVID death was lower by ~60%, comparing mRNA-V to non-V.

But no one assume benefit against non-COVID death! At best, expect null effect (or worse).

So why do we observe a biased estimate (VE= ~60%, against non-COVID death)? Image
Oct 21, 2021 • 5 tweets • 1 min read
Are the data compatible with 95% VE (death) in elderly?
More. From Scotland.
No age-specific data for deaths. Use hospitalized as a proxy.
Age 60+: true VE about 60%?
No.


>>> That’s true VE, if risk of hospitalization for 72K unvaccinated is similar to risk of hospitalization in 1.4M vaccinated HAD THEY NOT BEEN VACCINATED.
Similarity is expected in a randomized trial. Not in observational data.
>>>
Oct 17, 2021 • 11 tweets • 3 min read
1/
 Thread: Are the data compatible with 90-95% VE against COVID death in elderly?
Bottom line: no

Israel:
Comparing CFR in recent wave (vaccinated country) to CFR last year summer/fall wave (unvaccinated). Indirect measure. 2/
 Used this table created by @OS51388957
2020 period (left) vs. 2021 period (right)
Sep 25, 2021 • 7 tweets • 3 min read
1/
 A thread on competing viruses/viral interference:
Rhinovirus & SARS-Cov-2. Optimistic signal for fall/winter?
(Rhinovirus is a common cause of the common cold) Image 2/
 Competing viruses is an old theory.
Likely explain the disappearance of flu & other respiratory viruses at the beginning of the pandemic. (No, it’s not masks/NPI).
In particular, interference was proposed between Rhinovirus & flu pandemic.
pnas.org/content/116/52…
Sep 24, 2021 • 6 tweets • 2 min read
1/
 Inspired by the analysis of @profnfenton of UK data (age-standardized ALL-CAUSE mortality by V status.)

I focused on NON-COVID mortality (ONS excel file, Table 5).

2/
 Examined the period April-June, a time of low background COVID deaths. (Non-COVID deaths dominated.)

Omitted deaths <21d of dose 1 (small numbers)

Figure shows clear rank order. Image
Sep 17, 2021 • 13 tweets • 4 min read
1/
 Israel data: NEJM article
Booster group vs. 2-dose group.

BIG questions on results & methods.

“We considered 12 days as the interval between the administration of a booster dose and its likely effect on the observed number of confirmed infections”
nejm.org/doi/full/10.10… 2/
 No. You can’t skip 12 days after booster, especially when there is evidence that the risk of infection is increased during the first 1-2 weeks post dose 1.

Reasons for increased early risk: immune suppression, infection at V sites.
Aug 8, 2021 • 6 tweets • 2 min read
1/6
 A scary thread on case-reports.
@barakshalev 2/6
 A case-report is part of medical research.
Its weaknesses: very small sample (n=1), no estimated effect.
Nonetheless, it is research.
How come?
It is an observational study and there is “sort of” control.
Aug 8, 2021 • 5 tweets • 1 min read
1/
 They are at least incompetent.
The analysis in the post is valid. What CDC is doing here corresponds to the proportional mortality ratio (PMR) in epidemiology. (You can read about it in the textbook "Modern Epidemiology").
>> 2/
 I will quote one sentence from that book:

"An implicit assumption of a proportional mortality study is that the overall death rate for categories other than the ones under study is not related to the exposure".
>>
Aug 7, 2021 • 4 tweets • 2 min read
1/
 Now that estimated VE for severe COVID is down to 70-80% from those false estimates of >95%, we are still left with cardinal questions:

1. Still biased?
2. What is the effect of V on all-cause hospitalization/death?

2/
 Still biased?

What are the characteristics of unvaccinated elderly?
Are there shared causes of V status and health outcomes?
Are those unvaccinated elderly (~10%) exchangeable with their vaccinated counterparts (90%) on key health-related variables?
Aug 6, 2021 • 5 tweets • 2 min read
1/
 Cancelled my subscription to Ha’aretz after they rejected my paper on Sweden (published on Medium). (They were scared.)

1. I/others analyzed published VE studies from Israel and pointed to evidence of biases & over-estimation

2. They claimed ~95% VE. False for all endpoints. 2/
 3. Little/no VE for infection.

4. And it is not just waning, if true. (Ab levels decline after natural infection, too, but protection is not lost. Memory cells play a role in re-infection.)

5. VE for severe disease is currently 70-80% at most (and still possibly biased).
Jul 22, 2021 • 7 tweets • 2 min read
1/
 They claimed that lockdown before the peak terminated the rise.

They were wrong. False claim.
Refutation: Sweden & South Dakota (for example).

Now they claim that mass vaccination "decoupled deaths from cases".

2/
 They are wrong again. False claim.
Refutation: Sweden & Netherlands (for example).

Let’s see what happened there in recent months.
Jul 13, 2021 • 4 tweets • 2 min read
1/
 Israel: another paper on VE
jamanetwork.com/journals/jama/…
Count of cases is very confusing:

2/
 Trying to split a total of 243 (eTable 1) into mutually exclusive categories of vaccinated and reconcile with a flow chart (Figure 2, with additions).

No clear understanding how the numbers are split.
Jul 9, 2021 • 11 tweets • 4 min read
1/
 Scared by Delta wave? Read this thread.
Is Delta endemic wave that would have gone unnoticed in normal times?

Projections (& some philosophy at the end)

Key points:
Most COVID deaths are 60+.
Trend in elderly is the key - not all ages, not younger.

2/
 Here is what we see:
Continued divergence between infected vulnerable and non-vulnerable (by age).
Was fairly parallel in winter.
This divergence is the most important difference between current (summer) wave and winter wave.

coronavirus.data.gov.uk/details/cases?…
Jul 8, 2021 • 4 tweets • 2 min read
1/
 Thread: Vaccination of recovered.
This line of research has been interpreted as: ”People who have had a PCR confirmed covid-19 infection may only require one dose“.
Is there another possible interpretation?
Recall: vaccinating recovered is newscience.
bmj.com/content/372/bm… 2/
 Most exposed clear up the virus without much trouble.
Likewise, if re-encountered through respiratory route.
Injection is not natural route for re-encounter.
“That offender, which I handled easily, shows up on my immune cells, far away!”
Is the immune system confused?