Outcome	Number of studies	Overview of effect by study	Comparison used in each study	Effect direction per study (positive ▲; negative ▼; no change/mixed effects/conflicting findings ◀▶)
Outcome category: transmission‐related outcomes
Number or proportion of cases	2 modelling studies (Bershteyn 2020; Burns A 2020)	Bershteyn 2020: policies include daily symptom screening, and monthly or weekly testing of 10%, 20%, or 100% of attendees, with testing occurring either on the most optimal day (the first week day of a 5‐day work week, which is Monday for USA public schools) or the least optimal day (the last week day of a 5‐day week, which is Friday for USA public schools). Compared to no testing or isolation, a policy requiring index cases to self‐isolate if they develop symptoms, in‐school transmission is predicted to occur during presymptomatic infection (days 1 through 4) and asymptomatic infection (26% to 39% of index cases). In the absence of additional testing for asymptomatic individuals, this policy predictably reduced transmission by 34.8% to 41.8% relative to no isolation. The impact of weekly testing varied according to the day of the week in which testing was deployed, due to the lack of in‐school transmission over the two‐day weekend. The first week day (Monday) was the most optimal day for testing, while the last weekday (Friday) was the least optimal. Testing on Monday averted 27.1% to 34.0% more infections than testing on Friday, and could reduce transmission by 61.8% to 64.2% without symptom‐based isolation. The most effective testing and isolation strategy used a combination of testing 100% of attendees on the first week day together with symptom screening and isolation of all those who are symptomatic, for an overall transmission reduction of 68.6% to 71.1% relative to no testing or symptom‐based screening.	Full opening of schools with no measures in place	Positive ▲
		Burns A 2020: in the baseline scenario of no intervention, the study predicted a median attack rate of 0.79 (IQR 0.56 to 0.9). The estimated attack rates were 0.79 (IQR 0.56 to 0.9), 0.71 (IQR 0.43 to 0.86), and 0.72 (IQR 0.43 to 0.86) at 1 and 2 days of isolation following fever in the scenario of 50% fever detection. The effects varied according to the rate of detecting fever. Applying an 88% detection rate compared to a 50% detection rate, implementing a one fever‐free day predicts an 8% reduction in the attack rate. At this higher rate of symptom detection, increasing the isolation to 6 days predicts a 15% reduction in the median attack rate to 0.43 (0.03 to 0.82) compared to no policy.	Full opening of schools with no measures in place	Positive ▲
Shift in pandemic development	1 modelling study (Burns A 2020)	Burns A 2020: with no policy in place, the peak number of infected people is assumed to be 148 (IQR 82 to 213) and the interval between the first and last day with at least two cases would be 139 (IQR 120 to 154). Implementing a policy of two days of home isolation following the last episode of fever predicted a reduction in all outcome categories: peak number of infected people is predicted to sink to 124 (IQR 58 to 184). The interval between the first and last day with at least two cases would increase to 145 (IQR 127 to 157). The effects varied according to the rate of detecting fever. If the rate of detecting fever is a higher rate of 88%, implementing a 1 fever‐free day achieves a 20% reduction in the peak concurrently infected and a 7‐day increase in the interval between the first and last day with at least two cases.	Full opening of schools with no measures in place	Positive ▲