I have been really curious to see what a rigorous, randomized, large scale study on masks and COVID 19 would show, testing community transmission (i.e. source control) and we finally have one. 342,126 adults over 6 months in 600 Bangladesh villages.

Preregistered too. Some villages were not instructed on or provided masks, some were with cloth masks, and some with surgical. Those instructed with and given masks largely wore them, confirmed by >1 million observations(!). Symptoms were checked on 98% and many were tested.

Villages not given masks mostly didn't use them. The results:
0.76% of the not-masking villages got COVID-19
0.74% of the cloth-masking did
Cloth masks had no significant effect on community transmission (p=0.54 54% likely we'd see this if cloth masks made no difference at all)

Surgical masks however, showed a p=0.043 (so we are over 95% confident they helped) and had about 1/10 fewer cases. I feel rather vindicated, and wish N95/KN95 would have been tried too. But I am going to guess this will be ignored since it's inconvenient.


It's inconvenient for mask promoters since cloth didn't make a noticeable difference and surgical didn't make a big difference. It's inconvenient for mask skeptics since surgical did make a difference. Probably 74 out of the 100k didn't get sick and maybe a life was saved.

· · Web · 0 · 0 · 0
Sign in to participate in the conversation
Scriptjunkie Social

scriptjunkie's server