Estimating the vaccine effectiveness. Take 4.
A bit late to the party, after Clalit, Maccabi, and the MoH+Pfizer all show cased their results of the vaccine effectiveness. Yet, I am happy to present an updated version of my aggregate-level analysis.
A bit late to the party, after Clalit, Maccabi, and the MoH+Pfizer all show cased their results of the vaccine effectiveness. Yet, I am happy to present an updated version of my aggregate-level analysis.
1st draft preprint available on #medRxiv and hope to replace it with a new version soon.
medrxiv.org/content/10.110…
medrxiv.org/content/10.110…
The main difference is an improved formula to calculate expected daily incidence of cases. The problem is that cases are “eliminated” because of the vaccine, causing underestimation of expected daily cases. After hard work, with help from @geller_mic we can now correct for it. 
Another difference is that now I show beta values (sensitivity parameter) where there is no effect in days 0-13 +- 10%. The full methodology is available in my github (also all code and data available there) github.com/dviraran/covid…
The results are still based on the last update by the MOH from Feb 9, no new data yet and will update when available. Note that there is data on 14+ days after 2nd dose, but especially regarding severe cases, it is not yet reliable.
Positive cases: 72% for 60+, 78% for 60-. This is lower than the other studies, probably because other studies have information on symptomatic vs. asymptomatic. I only have them combined.
Hospitalizations: 83% for 60+. For 60- we see 83% in week 4 and 74% in week 5, but numbers are small, so large confidence intervals. Yet, one can say that young people after vaccine become less careful.
Severe cases: 86% for 60+. For 60- similar to hospitalizations - we see 85% in week 4 and 74% in week 3, but numbers are small, so large confidence intervals.
Still very low to no effect in week 3. As mentioned, it might be due to the UK variant, importance of 2nd dose, or a problem that 2 first weeks are combined.
Week 6-7 – data looks really good – 95%+. For positive cases, it shouldn’t change much. Why such increase between week 5 and 6? I don’t know. Hospitalizations and severe cases are expected to increase, but we will have to wait for the next update from the MOH.
Summary figure of the Vaccine Effectiveness estimations according to my analyses (as of February 9th)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
