Limited transmission of COVID-19 from open schools but teachers were affected: Swedish study

On March 18, 2020, Swedish upper-secondary schools moved to online instruction while lower-secondary schools remained open. This facilitates a comparison of infections and disease between groups that are comparable in other regards. In the study, all PCR-confirmed cases of SARS-CoV-2 and all healthcare registered cases of COVID-19 until the summer break are linked to register data on families and teachers in lower and upper-secondary schools.

Since the age of the student is likely to correlate with the severity of symptoms, student infectiousness and various types of risk behaviour, it is crucial to compare parents to children close in age. According to the study, the risk of infection was 17 percent higher among parents whose youngest child studied at the final year of lower-secondary rather than the first year of upper-secondary school. Had lower-secondary schools moved online, the estimates correspond to 500 fewer detected cases among a total of 450,000 lower-secondary parents (4.5 percent of the population). This can be compared to 53,000 PCR-confirmed cases in the total population during until the summer break in mid-June.

When comparing lower to upper-secondary teachers, we find that the risk for both PCR-confirmed infection and healthcare treatment due to COVID-19 doubled by keeping schools open. Among 124 occupations, upper-secondary teachers had a median risk of infection while lower-secondary were the 7th most affected. This comparison excludes healthcare workers who had markedly different access to PCR-testing. By the end of June, 79 out of 39 500 lower-secondary teachers had been hospitalized due to COVID-19, one of whom deceased. According to the study, this number had been down to 46 if lower-secondary schools had closed.

It is well-known that SARS-CoV-2 is transmitted within households. The study finds that the risk of a positive PCR-test was 30 percent higher among partners of lower-secondary teachers than among their upper-secondary counterparts. The estimates for more serious cases of COVID-19 are somewhat lower than for PCR-tests but -- just as for parents -- these estimates are imprecise.

Closing the schools is a costly measure with potentially long-run detrimental effects for students. The results for parents are in line with theoretical models predicting a limited impact of school closures on the general transmission of SARS-CoV-2. In an international comparison, the precautionary measures undertaken in Swedish schools are best described as mild. Thus, strict measures within open schools cannot explain the relatively minor impact on the overall rate of transmission. The results for teachers suggest that further precautionary measures could be considered.

The study does not analyse the impact of school closures for virus transmission among students. We note, however, that there are few cases of serious illness among the young. In particular, zero deaths from COVID-19 had been recorded among 2-19 year olds in Sweden until mid-summer 2020.