Coronavirus study that found US school closures cut life expectancy criticised by epidemiologist | Medical research

A study that found US school closures during the Covid-19 pandemic cut the life expectancy of each child in primary school by an average of three months contains “critically flawed assumptions” and “clear mistakes in study design”, according to a rebuttal led by an Australian epidemiologist.

The study, published in the Journal of the American Medical Association on 12 November, was widely shared on social media including by scientists, doctors and policymakers and was covered in dozens of news stories.

The researchers calculated the years of life lost that might be attributable to school closures, and compared that to the years of life lost attributed to the first months of the pandemic.

The study found a total of 5.5m life years would be lost due to school closures, higher than that directly attributable to Covid-19. This was determined by calculating the impact of missing days of school on educational attainment. The researchers linked this to an estimate of a 25% reduction in relative risk of death for each year of schooling, because education is associated with better health outcomes. The researchers then used these assumptions to estimate years of life lost that would be associated with school closures.

The researchers concluded there was “a 98.1% probability that school opening would have been associated with a lower total YLL [years of life lost] than school closure”.

An epidemiologist with the University of Wollongong, Gideon Meyerowitz-Katz, and a demographer with the University of Southern Denmark, Dr Ilya Kashnitsky, last week published a response to the study, describing it as “filled with errors”.

“This study has received enormous public attention, and its results immediately appeared in discussions of public health policies around schools worldwide,” they wrote.

“The authors suggested a causal chain that consists of two highly questionable links: (1), missing school is linked to overall educational attainment, and, (2), attainment is then linked to the length of life. The first link relies entirely on a single Argentinian study … of the long-term effects of teacher strikes on educational attainment of children who attended school during this time.

“In addition to the inappropriate assumptions that present day US children are directly comparable with Argentinian children from the 1970s and 80s and that teacher strikes have equal effect as remote learning during the lockdown … the second link relies on the unwarranted assumption that lost years of schooling can be directly translated into lost years of life.”

Meyerowitz-Katz (who is also a Guardian contributor) told Guardian Australia there were legitimate concerns about school closures. School closures have been a key issue during the pandemic in the US in particular, as they have put pressure on parents, including essential health workers, to provide childcare, and on teachers and children to adapt to online teaching. The role of children in transmission of the virus is not yet fully understood, but World Health Organization advice states evidence to date suggests transmission in educational settings is limited.

“My co-author and I are not saying that school closures are necessarily a good thing,” Meyerowitz-Katz said. “What we’re saying is that this study can’t provide evidence [that they are harmful]. The numbers that they’ve presented in part of their model are mathematically impossible. Given the procedure they claim to have used, they’re also wrong.”

He and Kashnitsky have called on the journal to correct or retract the paper.

Scientists have warned about an influx of scientific papers throughout the Covid-19 pandemic, as medical journals rush to publish new information about the virus.

The lead author of the JAMA schools paper, Dr Dimitri Christakis, is director of the Seattle Children’s Research Institute’s centre for child health in the US and told Guardian Australia his paper had been “through rigorous peer review,” pointing out that the criticisms of the study had not. He also referred to the disclaimer on the site on which Meyerowitz-Katz and Kashnitsky had posted their criticisms, which states: “Caution: Preprints are preliminary reports of work that have not been certified by peer review. They should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.”

Christakis did not address the criticisms of the paper, but said Meyerowitz-Katz and Kashnitsky “have so far refused to engage in the normal process of scientific discourse which is to submit a letter/comment to the journal for a response, or to conduct their own peer-reviewed work”.

The Journal of the American Medical Association did not respond to a request from Guardian Australia for comment. However, Meyerowitz-Katz said the journal editors told him that any concerns could be outlined by leaving a comment on the Christakis paper in the journal online.

Meyerowitz-Katz said he had resorted to publishing a comprehensive paper outlining his concerns, because the issues with the paper could not be addressed in a 600-word comment. In a further email from the JAMA editors, Meyerowitz-Katz and Kashnitsky were asked to also include a link to their full response, which has been made public, if they commented under the journal article.

Source link

Leave a Reply

Your email address will not be published.