Thread by @DiseaseEcology

Real-world Pfizer vaccine (& natural infection) efficacy against sars-cov-2 INFECTION
New Lancet paper posted today with fantastic data.
papers.ssrn.com/sol3/papers.cf…
Short Thread
tl;dr 1 dose reduces infection 72% on day 21; 7d post 2nd dose, 86%; previous infection 90%

Solid study design (for observational study)
Study of 23K health care workers in England, w/ PCR testing every 2 wks + rapid tests 2x/week & PCR confirmation of + rapid tests. 35% seropositive at start.

Vaccine hesitancy was higher in previously exposed, young, women, black (much lower), poorer.

For people w/out previous exposure ("negative cohort") vaccination reduced infection 72% (21d after 1st dose) or 86% (7d after 2nd dose).

Previous exposure reduced chance of infection 90%.

This is an observational study, so, despite authors best attempts to statistically control for differences, the unvaccinated and vaccinated groups may differ in ways that affect their exposure to sars-cov-2 in either direction & this could alter efficacy estimates.

Nonetheless, I think it provides solid evidence that vaccination reduces infection substantially (70-85%) & not just disease. I wish study had shown data on viral loads so we could assess effects on infectiousness as well.

Note: Estimate from this study (72%) is a bit lower after 1st dose than I estimated for Moderna (not Pfizer, but both mRNA) in this thread (90%):

https://twitter.com/DiseaseEcology/status/1359213735424872450

However, estimates are from diff days (21 vs 28d post 1st dose) & diff populations (HCW, gender, etc.).