Amongst the many changes to daily life during the early covid-19 pandemic, school closures may have been the most noticeable and most hotly debated. A study in JAMA sought to elicit the effects of school closures on incidence of covid-19 cases and deaths. By examining the timing of school closures across all 50 U.S. states, the study's authors hoped to determine if there was a temporal component to such an effect. In a time series analysis performed from March 9, 2020 and May 7, 2020 they found a decline in incidence of cases and mortality by 62 percent and 58 percent, respectively. Furthermore, the data revealed that states where closures occurred early, when cumulative incidence was low, had the largest reductions—72 percent for the states in the lowest quartile for incidence versus 49 percent for the highest. Based on these percentages, the authors' model predicted an overall decrease of 1.4 million cases over a 26-day period and 40,000 fewer deaths over a 16-day period.
Of course, time series (before-and-after) studies are fraught with limitations as a number of behavioral ("non-pharmaceutical") interventions were enacted in quick succession across many states. As states started to take the drastic measure of closing schools, they were also educating the public about improved hand hygiene, social distancing, and mask usage, not to mention restaurant, office and public space closures. While the authors made statistical adjustments for many of these factors, not all of them could be directly accounted for with data, making it difficult to accurately separate out school closures from many other simultaneous interventions. Another limitation included analysis at the state level, which can't account for differences amongst counties or individuals who traveled between states. Furthermore, as has been a centerpiece of all covid-19 related discussions, testing ability and reporting between states can vary greatly, possibly misrepresenting the number of actual cases in each state.
Ultimately, closing schools seemed an unavoidable step in the fight against the covid-19 pandemic and coincided with a significant decrease in disease burden and mortality, as this study shows. While adjusting for many other factors still resulted in a statistically significant difference, one would be hard pressed to take this data as decisive or definitive given the variety of measures each state enacted concurrently with school closures.
A new study summarizing a national survey conducted in Australia was posted on the preprint server MedRxIV yesterday. The results (from surveys taken in May) are concerning. Similar to a studies done in the United States and elsewhere, a majority (92 percent) of participants knew that handwashing can limit the spread of SARS-CoV-2. However, any number under 100 percent is alarming. Among participants, 42 percent had the mistaken belief that being unable to hold one's breath for 10 seconds without coughing implied coronavirus infection. Mask wearing was views were mixed, with only 23 percent understanding that surgical masks mainly prevent outward spread of an already infected individual. 33 percent either falsely believed or were unsure as to whether antibiotics would decrease the chance of infection. The survey excluded persons who had already been infected. Gender, age, and regional representation among the 1500 participants was appropriately balanced. However, only 1 percent of participants self-identified as Aboriginal, compared to 3.3 percent of Australia's population.