The trial did not find a statistically significant personal protective effect from adding a mask recommendation; confidence intervals were compatible with meaningful benefit or harm. Limitations included adherence, missing data, self-reported tests, no blinding, and no assessment of source control.
Infection with SARS-CoV-2 occurred in 42 participants recommended masks (1.8%) and 53 control participants (2.1%). The between-group difference was −0.3 percentage point (95% CI, −1.2 to 0.4 percentage point; P = 0.38).
Abstract, Results
The trial did not find a statistically significant personal protective effect from adding a mask recommendation; confidence intervals were compatible with meaningful benefit or harm. Limitations included adherence, missing data, self-reported tests, no blinding, and no assessment of source control. Bearing on the claim: partly supports. Verification note: Authors explicitly warn against using the findings to conclude that community masking would not reduce infections; source control was not tested.