Johns Hopkins Study: [bolding by me] "By contrast, closing cafeterias and playgrounds and the use of desk shields are associated with lower risk reductions (or even risk increases); however, this may reflect saturation effects because these are typically reported along with a high number of other measures." In other words, the "odds ratio of COVID-19–related outcomes" for using desk shields was only increased "compared to the reduction resulting from a generic mitigation measure." It was not measured compared to NO MITIGATION whatsoever, which the summary insinuates.
MA Study: The plexiglass dividers were identified as one of three "possible risk factor" categories through "observation of smoke," not measurement. The other categories were "occasions of mask removal at distances less than 6', primarily for eating and drinking" and "high-traffic areas (especially in two-way entry/egress areas) and shared offices." Further, this was at ONE SCHOOL! "Because no additional cases were identified in screening among staff at other schools, or among middle school students, all other schools in the district continued hybrid or in-person learning." Of the three categories, I wouldn't put my money on the plexiglass dividers being the most likely contributing factor... but I'd also want to confirm with actual measurements across all of the categories, not just assumptions, guessing, and "observation of smoke."
GA Study: The summary already includes the qualifier "compared with ventilation improvements and masking," but doesn't include a comparison to NO MITIGATION. Looking at the study, classrooms with "desks or tables with barriers" for "all classrooms" had 2.92 cases per 500 students enrolled. That was better than the following mitigation characteristics that were also measured: Optional mask requirements for teachers and staff members (4.42), Optional mask requirements for students (3.81), Flexible medical leave policies for teachers not offered (3.98), No ventilation improvements (4.19), Unknown ventilation improvements (2.95), Desks or tables separated by greater than/equal to 6 ft in some/no classrooms (3.09), Desks or tables separated by greater than/equal to 6 ft in all classrooms (3.02), and desks or tables with barriers in some/no classrooms (3.13). Based on those numbers, it's pretty safe to assume that desks or tables with barriers for all classrooms also performed better than no mitigation whatsoever, especially when you look at the rates in Private/Parochial/Independent schools (4.05) and Nonmetropolitan schools (3.41) that likely had little/no mitigation.
Cubicle Study [1] [bolding by me]: The summary just plain got this WRONG! "The patient with the index case of TB was working as chief of staff in a large commercial business office [...] Screening was extended on two occasions to take in the whole ground floor, reception and mezzanine level, and to include staff who transited the area on a regular basis [...] The workplace had an open-plan design with low-profile cubicle dividers and closed air conditioning. The practice of 'hot-desking' (where most staff are not allocated a permanent desk) involved about 75%–80% of the staff in the main office area. The patient with the index case of TB had a permanent desk." In the discussion: "In our case, there were several factors in the workplace design that may have contributed to transmission including a closed air-conditioning system, modern open-plan office design with low profile design of cubicle dividers that allows workers to see and communicate directly with their colleagues without standing, and the practice of 'hot desking'." It was LOW dividers (essentially, NO DIVIDERS) that contributed to the spread, not HIGH dividers.
British Research: Meta analysis. They got the Lessler (Johns Hopkins) study wrong. Their take on the Johns Hopkins study completely ignores that it was a comparison to other mitigation, not a comparison to NO mitigation: "Analysis from a very large US online survey of self-reported school-based mitigations in the US suggests that desk screens are associated with an increase in COVID-19 risk (Lessler et al 2021)." That's just NOT what that study concludes. Still, the British research does state, "Screens and barriers are likely to have benefits in reducing the risk of exposure to larger aerosols and droplets from exhaled breath when people are face to face and close together (less than 2m) (high confidence)." Like a lot of this, the best part of this article is the last bullet: "There remains a need for further research to look at the effectiveness of screens and barriers in real-world settings both from the perspective of direct mitigation of the virus in exhaled breath or managing behaviour."
[1] Non-paywalled: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.mja.com.au%2Fjournal...