Thread by @FaceDeBengals

🚨 New study (Safety Science, 2025): COVID spread in a restaurant wasn’t mainly aerosols, but larger droplets up to 40µm. That flips much of the early pandemic narrative. Let’s break it down 🧵 (1/15)

https://twitter.com/facedebengals/status/1965543122210885661

Case studied: Guangzhou, Jan 2020. 9 people infected in one restaurant. Researchers re-ran the outbreak using CFD airflow modeling + Wells-Riley infection theory. (2/15)

They tested two scenarios:
1️⃣ Aerosols (~1µm)
2️⃣ Multi-size droplets (up to 40µm). (3/15)

Results:
•Aerosol model → ~32 infections predicted.
•Multi-size droplet model → ~11 infections predicted.
Reality: 9 infections, clustered around the index case. (4/15)

Conclusion: larger droplets match the outbreak pattern better. Aerosols would have dispersed across the room, but the real infections stayed local. (5/15)

But here’s the kicker: the droplet model only worked if the infected person emitted 2520 quanta/hour 🤯. Typical “superspreader” values are below 1000. (6/15)

This means the model is sensitive to assumptions: deposition speed, surface stickiness, HVAC filtration. Still, it shows droplets can dominate in crowded, recirculated spaces. (7/15)

Policy insights:
•Boosting ventilation alone isn’t enough. In some spots, more airflow worsened exposure.
•Partitions + shorter meals (30min) reduced risk far more effectively. (8/15)

Masks for staff, shorter exposure times, and regular cleaning matter more than blind trust in ACH numbers. The study even suggests surfaces got most of the viral load. (9/15)

Takeaway: don’t worship a single dogma. “Only aerosols” was as simplistic as “only droplets.” Reality is mixed, localized, and context-specific. (10/15)

COVID science too often chose certainty over nuance. Instead of admitting limits, many experts pushed black-and-white narratives that later collapsed. (11/15)

Global management was driven by a “sanitarian” mindset: shut everything down, universal rules, little attention to context (crowded restaurants ≠ open parks). (12/15)

Experiments and models were rushed, assumptions rarely questioned, and dissenting voices painted as dangerous. That’s not science, that’s moral panic dressed up as expertise. (13/15)

Result: contradictory advice, population fatigue, distrust of public health. When you tell people “trust the science” while the science itself is shaky, you fuel skepticism. (14/15)

Lesson: next time, resist the sanitarian reflex of totalizing control. Accept uncertainty, be transparent about trade-offs, and target measures where risk is real. That’s how you keep both liberty and public health. (15/15)

16/15

This study shows why much of “Covid science” felt shaky: models were often run on oversimplified assumptions (like aerosols only), then presented to the public as settled truth.

17/15

This study shows why much of “Covid science” felt shaky: models were often run on oversimplified assumptions (like aerosols only), then presented to the public as settled truth.

18/15

The result? Policy built on fragile foundations. If droplets dominated in real-world outbreaks, then endless ventilation upgrades and outdoor mask mandates were misdirected energy.

19/15

This isn’t just about errors — it’s about the culture of certainty. Early Covid science too often acted like doctrine, not inquiry. “Trust the science” replaced “question the data.”

What this study debunks: the illusion that we had clear, singular answers. Real science is messy, contested, and conditional. Covid science forgot that — and the public paid the price.

Wang, Z., Jia, F., Galea, E. R., Ewer, J., & Grandison, A. (2025). Transmission of SARS-CoV-2 via larger respiratory droplets in a restaurant environment. Safety Science, 191, 106938. doi.org/10.1016/j.ssci…