School opening associated with lower test-adjusted COVID-19 case rates in children.

Background: There is conflicting evidence from prior studies on the relationship between in-person schooling and transmission of SARS-CoV-2 among children. This may be due to multiple confounders in estimating this relationship, including the decision to close schools, community rates of infection, and rates of testing.

Methods: Regression-based observational study to estimate the relationship between school openings and COVID-19 case rates among children, while accounting for potential confounders including community case rates, mitigations in schools, and rates of testing among schoolchildren. The setting is US school districts in the Fall of 2021, from 3 weeks prior through 12 weeks after school opening, using restricted data obtained from the Centers for Disease Control and Prevention. Data were available for school districts in 2592 counties, containing 86% of the US population.

Results: School openings were associated with a brief rise in cases among children relative to adults, with a peak of 39.3 [37.7, 40.9] additional cases per 100 000 per week. However, children were tested at higher rates when schools were in session. After adjusting for testing rates, case rates among children were significantly lower after schools reopened by 4.7 cases per 100 000 compared with over summer break.

Conclusion: School reopening in the USA in the 2021-22 academic year was accompanied by an increase in SARS-CoV-2 testing in children and a brief rise in pediatric cases. When testing rates are accounted for, school reopening was associated with a decrease in COVID-19 cases among children relative to adults. A lower threshold for testing in the school setting may be an important confounder in studies of SARS-CoV-2 transmission.