US policymakers considering physical distancing measures to slow the spread of COVID-19 face a difficult trade-off between closing schools to reduce transmission and new cases, and potential health-care worker absenteeism due to additional childcare needs that could ultimately increase mortality from COVID-19, according to new modelling research published in The Lancet Public Health journal.
Using the latest data from the US Census Bureau's Current Population Survey to measure the childcare needs of health-care workers if schools are shut, researchers estimate that nationwide, at least one in seven medical workers may have to miss work to care for their children aged 3-12 years old, even after taking into account childcare provided by non-working adults and older siblings within the same household.
These additional childcare obligations could compromise the ability of the US healthcare system to respond to COVID-19 if alternative childcare arrangements are not made, researchers say.
However, the authors caution that the true impact of school closures on overall deaths from COVID-19 cannot be precisely predicted because of large uncertainties around estimates of transmission and infectivity, and to what extent a decline in the health-care workforce impacts the survival of patients with COVID-19.
"Closing schools comes with many trade-offs, and can create unintentional child-care shortages that put a strain on the health-care system", says Professor Eli Fenichel from Yale University in the USA who co-led the research. "Health-care workers spending less time providing patient care to look after their own children can directly influence the development of an epidemic and the survival of those patients. Understanding these trade-offs is vital when planning the public health response to COVID-19 because if the survival of infected patients is sufficiently sensitive to declines in the healthcare workforce, then school closures could potentially increase deaths from COVID-19." 
Support for mandatory school closures to reduce cases and mortality from COVID-19 comes from experience with influenza, or models that do not include the effect of school closure on the health-care workforce. Few studies have considered the trade-off between case reduction and disease burden and the potential loss of healthcare workers to childcare obligations.
In the study, researchers analysed data on more than 3 million individuals between January 2018 and January 2020 to assess family structure and probable within-household childcare options for health-care workers. They identified those most likely to require additional childcare for children aged 3-12 years old in the event of school closures by type of health-care occupation nationally and across different states, assuming that early childcare for children aged under 2 years remains open. They also modelled potential declines in the health-care workforce during school closures with estimates of case reductions from school closures to identify the point at which more lives are lost from school closures than are saved.
The analyses suggest that around 29% of US health-care workers need to provide care for children aged 3-12 years old. In households without a non-working adult or a sibling aged 13 years or older to provide care, the researchers estimate that 15% of health-care workers will require childcare--equivalent to around 2.3 million children nationwide--if schools close. However, the authors note that they were unable to account for health-care workers finding alternative methods of care for their children such as babysitters or friends.
School closures will be especially challenging for nurse practitioners (22% will need childcare), physician's assistants (21%), diagnostic technicians (19%), and physicians and surgeons (16%), as well as nearly 13% of the nursing and home health aids who are single parents and part of the group helping the elderly with infection control in nursing homes, researchers say.
The US states likely to have the greatest unmet childcare needs include South Dakota (21% of health-care workers will need childcare), Oregon (21%), and Missouri (21%). In contrast, Washington DC (9% health-care workers with unmet childcare needs), New Mexico (10%), and New Jersey (11%) are least likely to have health-care worker shortages if schools close.
Further analysis suggests that if the case fatality fraction (the share of people who die out of all those infected) rises from 2% to more than 2.4% when the health-care workforce declines by 15%, school closures could lead to a greater number of deaths than those they prevent. However, there is substantial variation across the country. For example, in South Dakota estimates suggest that the case fatality rate must not increase by more than 1.7% before school closures stop saving lives and start increasing overall mortality, whereas in Washington DC it is 4.1%--this is due to the low child care obligations in Washington DC relative to South Dakota.
"The US healthcare system appears disproportionately prone to labour shortages from school closures, particularly among those health-care workers providing infection control in nursing homes", says co-lead author Dr Jude Bayham from Colorado State University, USA. "These potential health-care workforce shortages should be a priority when assessing the potential benefits and costs of school closures, and alternative child care arrangements must be part of the school closure plan." 
According to Fenichel, "Closing schools and distancing in general is about bending the curve to stay below hospital capacity and reduce COVID-19 mortality, but how we distance in order to bend the curve can also influence the hospital capacity we need to stay below. We need to account for both."
The authors note some important limitations of the study, including that the authors informed their model based on the influenza virus, to which children are particularly vulnerable--however, early data on COVID-19 suggests children may be less vulnerable, so the benefits of school closures may be smaller than expected. On the other hand, the authors note that closing schools earlier in an outbreak could prevent more cases and lead to less health-care workers being infected and thus able to treat more patients. The study did not include mortality from other conditions that might occur if the health-care workforce declined, which should be taken into consideration when deciding about closing schools, the authors say.
NOTES TO EDITORS
It was conducted by researchers from Colorado State University, USA; and Yale University, USA.
The labels have been added to this press release as part of a project run by the Academy of Medical Sciences seeking to improve the communication of evidence. For more information, please see: http://www.sciencemediacentre.org/wp-content/uploads/2018/01/AMS-press-release-labelling-system-GUIDANCE.pdf if you have any questions or feedback, please contact The Lancet press office firstname.lastname@example.org
 Quotes direct from authors and cannot be found in the text of the Article.
Peer-reviewed / Modelling study / People
The Lancet Public Health