Thread by @DrScottBalsitis

CDC reports another study that fails to control for vaccination rate differences between mask-wearers (higher vax rate) and non-mask wearers (lower). Vaccines were highly protective vs infection (pre-Omicron), so of course the mask group has lower rates. The vaccines work.🧵

https://twitter.com/CDCgov/status/1489675180242870278

Using proper control groups is critical, and the most basic principle of good study design. If you don't control for obvious confounding variables (vaccination rate) that will impact the result, you can't assign the effect to the variable you're testing (masks).

Like many other studies, this one lacks a well-matched control group and instead "adjusts" the raw data after the fact, for multiple uncontrolled variables, to generate the numbers reported as the take-home message. These adjustments inject researcher assumptions into the data

Such studies should not be relied upon to make policy. Fortunately, there are multiple studies that do compare COVID rates between well-matched masked and unmasked student groups, without having to adjust the raw data. As a reminder, here they are.

The CDC's own study, from before vaccines were available. "The 21% lower incidence in schools that required mask use among students was not statistically significant compared with schools where
mask use was optional." cdc.gov/mmwr/volumes/7…

UK government study, published last month. It found an 11% lower rate of COVID absence with student face masks, but the difference was again not statistically significant. assets.publishing.service.gov.uk/government/upl…

A study of 1.6 million students (pre-vaccine so vaccines can't confound the result). "We do not find any correlations with mask mandates." Key data shown below. Blue = no student masks, green = student masks. medrxiv.org/content/10.110…

The next study is particularly powerful, because in-school transmission was carefully tracked by identification of every index case and testing of every contact. At the time, under-6s did not mask but universal masking was required for 6+ years old. journals.lww.com/pidj/Fulltext/…

The result: transmission rates were associated with age, but no difference seen when making the change from no-masks (blue background) to with-masks (green background).

The last one looked at mask use within student athletics specifically, and again found "There were no significant associations between COVID-19 incidence and face mask use."
meridian.allenpress.com/jat/article/57…

Study quality is far more important than quantity. It remains true that every well-controlled, real-world study of student masking fails to find a convincing benefit.

CDC studies of student masking also fail to measure possible downsides. Measuring both possible benefits and possible harms is an ethical necessity for any intervention, otherwise we can't possible evaluate the whole risk-benefit profile.

Fortunately, the UK did carefully evaluate possible harms. 80% of students reported masks made it more difficult to communicate, and 55% felt masks made learning more difficult.

Teachers felt the same: 94% thought masks made communication with students more difficult, with 59% saying masks made it a lot more difficult.

The UK also notes negative impacts on "understanding, engagement, and feelings of connection." Some students, such as those with hearing loss, are much more negatively impacted than others. They cite multiple studies quantifying these effects further.

If there is any benefit to student masking, it's too tiny to measure in any well-controlled real-world study to date, while downsides to masking are widely present. When is the right time to stop using something that is both ineffective and detrimental? Now.