Thread by @x2IndSpeculator

I've been trying to combine daily Israel data with the @IsraelMOH FDA presentation & the booster study from @NEJM. Extending their ideas, I created charts that provide the same valuable real-life information as the study but CONTINUOUSLY in REAL-TIME.
1/
nejm.org/doi/pdf/10.105…

MOH correctly understood even a seemingly modest decrease in vaccine effectiveness results in a significant increase in relative risk for the vaccinated, impacting the pandemic control.
97%>85% = 5-fold increase in relative risk 3%>15% (1:33>1:6.6)
2/
fda.gov/media/152205/d…

So, to quantify the protective effect of the additional booster dose, it's best to compare rate ratios & calculate x-fold changes in relative risk.
"Protection is given as a fold reduction in risk relative to people who received only two vaccine doses."
3/
gov.il/BlobFolder/rep…

"...susceptibility of a person who receives a booster dose would decrease to approximately 5% (i.e., 50% divided by 10) relative to that in an unvaccinated person & would bring the VE among booster recipients to approximately 95%, a value similar to the original VE."
1:2>1:20
4/

The Israeli study was also careful to exclude all possible transient behavioral change effects by comparing incidence rates to 4-6 days post-boost.
"...it is preferable to assess the effect of the booster only after a sufficient period has passed since its administration."
5/

"We compared rates at least 12 days after receipt of the booster with rates during days 4 to 6, when the booster effect was expected to be small & behavioral changes after vaccination were less marked... the rate of confirmed infection was lower by an estimated factor of 5.4."
6/

That's about half a decrease in relative risk compared to the main analysis in this table. My results end somewhere in between because, unlike this study, daily MOH datasets I use differentiate vaccinated with-booster from without-booster by the 7-days-post-the-3rd-dose mark.
7/

Since the definition of "fully vaccinated" became murky, without the MOH database at my disposal I can't know daily sizes of subgroups, which makes calculating attack rates & risk ratios on my own impossible.
I can only rely on the Israeli MOH-provided daily incidence rates.
8/

So, we get daily incidence rates of confirmed infection, (active & new) severe illness & deaths for two age & three vaccination subgroups. The only intervention I'll make is to average cases across 7-days, and new severe & deaths over 14-days (otherwise they look too erratic).
9/

Here's what comes out of it for the Over-60s.
RRR vs. infection for the original 2-dose vaccination lingers between 40-50%, while booster's RRR jumped above 90% to levels of the vaccine protection in Q1 vs. the #Alpha variant.
10/

The other way to look at this is to turn incidence rates into simple risk ratios (1:X), where X is a risk factor for the unvaccinated compared to boosted or non-boosted vaccinated.
Dividing those two ratios reveals a fold reduction in risk for boosted relative to non-boosted.
11/

In Over-60s, on September 24, incidences (per 100k) & the risk of infection among subgroups look like this.
Boosted : unvaccinated = 5.6 : 71.4 = 1 : 12.75
Non-boosted : unvaccinated = 41.9 : 71.4 = 1 : 1.7
1 : 1.7 > 1 : 12.75 = 7.5-fold reduction in relative risk (red line).
12/

Also, at the beginning of the year, for the 2-dose vaccinated (green line), we see the risk ratio jumping around 1:10, while now it's around 1:2. That's the 5-fold increase in relative risk mentioned before in the study, a.k.a. the effect of immunity waning vs. infection.
13/

The Under-60s are a larger subgroup with more infections, so calculations are steadier & more reliable. The effects are similar to the Over-60s.
Under-60s started boosting later, but RRR in the booster subgroup jumped above 90% rather quickly. A 2-dose recovered to about 50%.
14/

Looking at risk ratios, waning is visible again with the green line falling 1:12 > 1:2. But, also in this subgroup 1:2 risk for the 2-dose vaccination turned to 1:12 after the booster. That's a 6-fold reduction in risk and a complete restoration of protection vs. infection.
15/

Active severe illness showing the same trends in the Over-60s as infections, and with greater success. RRR vs. severe in vaccinated without booster settling at 80%, while boosted RRR climbed fast to 97%, similar to Q1 in 2-dose vaccinated.
16/

Risk factors in 2-dose vaccinated again show 5-fold waning of immunity from 1:25 to 1:5. That's reversed, even improved by boosters: almost a 10-fold reduction in relative risk from 1:5 to 1:45, above Q1 levels.
(Mid-year jumps should be ignored due to near-zero incidences.)
17/

Because of a high number of events, charts for infections in both age groups & active severe for 60+ are straightforward. The rest have a problem of daily incidences often being zero. I tried to correct it by averaging but, in the end, had to exclude those days. Thus, gaps.
18/

That's why the Israelis stopped at 60+. Since this is all a fast-moving situation, they didn't yet have enough events & dates for a scientific study. Since I don't have to worry about that, I make charts as soon as data are available. Even with gaps, trends are recognizable.
19/

Though there's a month less of data, similar trends in RRR vs. severe are seen among the Under-60s. They are holding a bit better with the original vaccination (above 80%), probably due to the later timing of their two doses, so boosters are for now having less impact.
20/

People opposed to boosting Under-60 looked only at this smaller reduction in risk: 2.7-fold for Under-60 vs. 9-fold in Over-60.
But, the effect vs. infections is overwhelming & it's too soon to judge if there really is less additional benefit vs. severe or it's due to timing.
21/

So until now, real-life data show additional booster effect:
7.5-fold reduction in risk of infection for Over-60
6-fold reduction in risk of infection for Under-60
9-fold reduction in risk of severe disease for Over-60
2.7-fold reduction in risk of severe disease for Under-60
22/

Final datasets cover new severe hospitalizations & deaths. Similar trends again: significant additional benefit of boosters vs. severe disease & a little less notable vs. deaths. Among the Over-60, 9-fold & 6-fold risk reductions, plus a complete restoration of effectiveness.
23/

In the Under-60 group, there are huge gaps in these two datasets due to zero daily events. It renders these charts (especially deaths) unusable for now until more time passes. But, I'll show them for comparison. Besides, previously mentioned patterns are still visible.
24/

Combining all RRRs & separating them by a booster status & age leads to these charts. Currently, in both age subgroups vaccinated WITHOUT a booster are showing RRR of circa 50% vs. infection & circa 80% vs. severe/fatal outcomes.
25/

Boosters restored RRR to circa 92% vs. infection in both age groups & 95%/97.5% vs. severe/fatal outcomes in Under-60/Over-60.
(RRR vs. deaths in Under-60 should be ignored for now because of too few events & too few dates.)
26/

Combining fold risk reductions (booster effect) shows the greatest impact where it's needed:
vs. severe in the Over-60s
vs. infection in the Under-60s
Sadly, relatively the smallest additional benefit boosters achieve vs. deaths as some people can't generate enough Abs.
27/

#Israel charts open up 4 questions determining the future of the #SARS2 pandemic in 2022:
1) Are current intramuscular mRNA vaccines 50%/80% or 92%/97% in the MEDIUM- to LONG-TERM?
2) What are all the factors that drove RRRs in originally vaccinated with 2-doses DOWN THEN UP?
28/

3) How big is an additional vaccine/booster benefit AFTER a BREAKTHROUGH INFECTION?
4) If vaccine effectiveness is fully restored (even surpassed some of the original levels) by the same boosters so quickly vs. #Delta, will we ever get an UPDATED variant-specific version?
29/29