Senior scientists have reported flaws in an influential World Health Organization-commissioned study into the risks of coronavirus infection and say it should not be used as evidence for relaxing the 2-metre physical distancing rule.
Critics of the distancing advice, which states that people should keep at least 2 metres apart, believe it is too cautious. They seized on the research commissioned by the WHO, which suggested a reduction from 2 metres to 1 would raise infection risk only marginally, from 1.3 per cent to 2.6 per cent.
But scientists who delved into the work found mistakes they believe undermine the findings to the point they cannot be relied upon when scientists and ministers are forming judgments about what constitutes safe physical distancing.
“The analysis of infection risk at 1 metre versus 2 metre should be treated with great caution,” said Prof David Spiegelhalter, a statistician at Cambridge University, who has participated in the British government’s Scientific Advisory Group for Emergencies . “I’m very suspicious of it.”
Prof Kevin McConway, an applied statistician at the Open University, went further and called the analysis inappropriate. He said the work “should not be used in arguments about how much greater the infection risk is at 1-metre minimum distance as opposed to 2 metres”.
The study, published in the Lancet, is the latest to come under fire from experts who fear that in the midst of the pandemic some research papers are being written, reviewed and published too fast for sufficient quality checks to be performed. Earlier this month, the Lancet and another elite publication, the New England Journal of Medicine, were forced to retract coronavirus studies after flaws in the papers emerged.
Led by researchers at McMaster University in Ontario, the report pooled data from previously published studies to estimate the risk of becoming infected with coronavirus at different distances. It also considered how face masks and eye protection might help prevent the spread of disease.
But in the analysis the authors assume the proportional impact on risk of moving from 2 metres to 1 metre is the same as moving from 1 metre to zero. “They are forcing the proportional fit to be the same,” Spiegelhalter told the Guardian.
McConway believes there is a more fundamental problem in the way the risks of infection at different distances are compared in the study. He said: “The method of comparing the different distances in the paper is inappropriate for telling you exactly how the risk at 2-metre minimum distance compares to a 1 metre minimum distance. It does not support, and should not be used in, arguments about how much greater the risk is with a 1 metre limit versus a 2-metre limit.”
Another scientist, Prof Ben Cowling at the WHO Collaborating Centre for Infectious Disease Epidemiology and Control at the University of Hong Kong, flagged further issues with the work. He tweeted that he was “not taking the whole paper very seriously” because it looked only at distance and not how long a person was exposed for.
McConway said he had raised questions about the analysis with the authors and was waiting to hear back. He believed peer review by the Lancet and the WHO should have spotted the problems. “I think they did it in such a rush – the authors, possibly the WHO, and the Lancet peer reviewers – that important things were missed,” he said.
“Everyone believes that the risk of infection at 1-metre is higher than at 2-metre and we need to know how much higher because there’s a trade-off between the increased risk and the gains from moving to 1-metre. But if you don’t know how the risks at 1 metre and 2 metres compare, how do you know how to trade it off? It’s finger in the air stuff,” McConway said.
In a statement, the WHO said it recommends keeping a distance of 1 metre or more. “The evidence used to inform this guidance was based on a systematic review of all available, relevant observational studies concerning protective measures to prevent transmission of the coronaviruses that cause Sars, Mers and Covid-19. After checking for relevance, 44 comparative studies done in health-care and non-health-care settings were included.
“The findings of this systematic review and meta-analysis support physical distancing of 1 metre or more, which is in line with the existing WHO recommendation that people should physically distance at least 1 metre,” the statement said.
The Lancet and the authors of the study have been contacted by the Guardian for comment. – Guardian