As an interesting aside to this, the recent study into vaccine (Pfizer) effectiveness out of Qatar showed a significantly negative effectiveness for asymptomatic covid.
Ie, they found a higher proportion of vaccinated individuals got asymptomatic covid compared with the unvaccinated population.
The actual impact of this is unclear, because the risk of transmission with asymptomatic covid hasn't really ever been quantified properly.
As an interesting aside to this, the recent study into vaccine (Pfizer) effectiveness out of Qatar showed a significantly negative effectiveness for asymptomatic covid.
Ie, they found a higher proportion of vaccinated individuals got asymptomatic covid compared with the unvaccinated population.
The actual impact of this is unclear, because the risk of transmission with asymptomatic covid hasn't really ever been quantified properly.
Except what you say just isn't true. Table 4 makes this quite clear.
Did you make this calculation yourself?
Please show your working out, so we can see.
@onion
Table 4 quite clearly shows negative vaccine effectiveness against asymptomatic infection.
There's no interpretation -- they give a set of numbers and the ones for the longest time since vaccination are negative.
I really don't understand how this can be interpreted in any other way -- what is your interpretation?
The paper is here: https://www.nejm.org/doi/pdf/10.1056/NEJMoa2114114, I'm referring to table 4 and their estimates are (for the last 3 data points in their table):
month 5 24.3%
month 6 -2.1%
month 7+ -33.3%
@onion
Table 4 quite clearly shows negative vaccine effectiveness against asymptomatic infection.
There's no interpretation -- they give a set of numbers and the ones for the longest time since vaccination are negative.
I really don't understand how this can be interpreted in any other way -- what is your interpretation?
The paper is here: https://www.nejm.org/doi/pdf/10.1056/NEJMoa2114114, I'm referring to table 4 and their estimates are (for the last 3 data points in their table):
month 5 24.3%
month 6 -2.1%
month 7+ -33.3%
What you describe here is accurate but your prior assertion didn't say this. You said vaccines had a negative effectiveness against symptomatic infection... Full stop.
The Qatar study (now peer reviewed and published in the NEJM) said nothing of the sort.
If you're referring to table 4 look at the sample size, 25 v 29 out of 26,000. Huge CIs. (−181.8 to 36.9)
Means the data is so noisy its impossible to make any claim off it.
I agree. There needs to be far more work done in this area -- the different estimates of vaccine effectiveness produced by different people at different times are highly concerning and possibly indicate that methodology issues might be present -- these really need to be resolved.
You can see an indication of the potential problems in that single Qatar paper.
- Their initial estimates of vaccine effectiveness using TNCC were negative after about 5 months (preprint). However, in the full release they removed some cases and ended up with positive effectiveness. There's no explanation anywhere about why these cases were removed.
- Their estimate of asymptomatic infection is negative after 5 months. Sure, these are low numbers, but that's what you'd expect (fewer people get tested without symptoms). I'd note that if the TNCC methodology is biased because of a 'different disease, similar symptoms' problem then you'd expect to see very different results for asymptomatic infections.
- Their estimate of vaccine effectiveness using multivariate regression (table S11), which should be less sensitive to the 'similar symptoms' problem inherent in TNCC also shows deeply negative vaccine effectiveness -- this is a real problem because all sufficiently powerful methods should give a similar answer for the vaccine effectiveness. Hence S11 is a red-flag that screams 'check everything now!', not something to shove into the tail end of the annex.
So, are the vaccines now offering negative effectiveness? Well, we can't be sure but there certainly are data that suggest it. This isn't a minor inconvenience, but an absolutely massive factor that would put the whole approach of mass vaccination into question.
The right thing to do here (actually, some time ago) would be to spend huge amounts of effort to make sure that the vaccines aren't making things worse -- instead, the solution appears to shout even louder about the subset of analyses that say things are fine, and suppress the voice of anyone that says different.
The Qatar study (now peer reviewed and published in the NEJM) said nothing of the sort.
If you're referring to table 4 look at the sample size, 25 v 29 out of 26,000. Huge CIs. (−181.8 to 36.9)
Means the data is so noisy its impossible to make any claim off it.
Agreed 6th and 7th month assessment of effectiveness can be ignored.
Prior months show good effectiveness, declining.
