Professor G Neil Martin, Research Integrity Advisor and psychologist, casts a critical eye over science’s response to COVID-19.
This article is part of our new series, Regent's Review: thoughts, research and academic discussion on the rapidly unfolding pandemic.
Summary
-
Scientists still know little about major patterns of behaviour change, but we are better than educated guesswork.
-
The ethics of rapid publishing is going to be tested; rushed bad research is worse than delayed good research.
-
Good science can, and will, lead to healthier people.
Does rapid research come at a hefty price?
Just one month ago, the shops were open, the bars were heaving, the theatres were entertaining and life seemed relatively normal.
The news from China, then Italy, about the increasing death toll resulting from COVID-19 induced some concern but little government action here or in other countries not yet badly affected.
Then, there followed the empty shelves, the business closures, University closures, the social distancing, the movement restriction, the Coronovirus Act 2020, and more deaths – this time in the UK and the US, a country that has – as of 30 March – reported the greatest number of COVID-related deaths in the world.
On 27 March, the Prime Minister of the UK, England’s Chief Medical Officer, and the UK’s Health and Social Care Secretary, were all confirmed positive with Coronavirus. All of this in the space of a month. How do we respond?
The increasing speed of research
This invisible killer needs an adversary and scientists across the world have been informing governments about the protective steps they need to take to safeguard their citizens, and are increasing their efforts to:
-
(i) understand this new strain of coronavirus,
-
(ii) develop diagnostic tests which provide rapid results, and
-
(iii) develop a vaccine and/or treatment
And Universities have not been slow to respond. Universities UK’s Twitter account (@UniversitiesUK) has been regularly tweeting some of the major new research projects underway in UK Universities including the University of Leicester’s effort to find a test for asymptomatic carriers, Glasgow University’s genome sequencing of the virus, the University of East Anglia’s 3D printing of personal protective equipment and ventilator parts, amongst tens of others. Oxford University has called for participants to take part in a study of a potential vaccine for COVID-19, requesting healthy volunteers between 18-55.
But as this monumental effort has been rightly accelerating (we need responses to some life-threatening issues), some concerns have been expressed at the speed – some would argue, haste – with which some research proposals and papers are being reviewed and approved.
Some researchers have questioned whether the rigour of review and the quality of proposals is being adequately managed. And there have been some very rapid processes in place for the peer-review of science that have been developed in the past ten days alone.
For example, the Royal Society’s Open Science portal promises rapid review of registered reports submitted (reviewed within 48 hours). The organisation’s first Stage-1 approval-in-principle decision has taken only six days. Over 300 scientists have already agreed to act as reviewers in a variety of disciplines and confirmed they can review within two days. I am one of them.
In psychology, the Psychological Science Accelerator has developed a similar, if less robust, process and within 10 days of launch has 139 collaborating laboratories.
The British Psychological Society has developed a research framework of its own but no plans for implementation.
The dangers of haste
Behavioural science has been a key player in the government’s response to COVID-19 because so much depends on people changing their behaviour to save themselves and others.
The behavioural science evidence used to inform government decision-making and released by the government on March 20 surprised many people as it was based on very little evidence.
I made the point at the time that if you cite seven references to support your decision-making (which is what one set of recommendations included) and none of those report empirical work, then you have a problem with your behavioural science: your decision-making is not robust and it is not informed.
Scientists still know little about major patterns of behaviour change, but we are better than educated guesswork.
The government claimed that blanket restriction of movement measures could not be implemented at once because of behavioural fatigue, and this surprised many colleagues – there appeared to be no evidence to support this.
A large group of psychologists and other academics wrote an open letter to the government asking it to publish this evidence.
The government’s Behavioural Insights Team has also been the subject of much – often accurate – criticism with its reliance on debunked studies or inaccurate interpretations of questionable studies. There is better psychology out there.
The rush to conduct studies under what is a unique set of circumstances can come with some pretty awful downsides. Some of these have been summarised in a thread by Laura Scherer on Twitter.
Rushed bad research is worse than delayed good research.
Some have criticised researchers for opportunism. Others have criticised the lack of robust scrutiny.
One notable case is the recent publication of a paper demonstrating ostensibly positive results of the effect of hydroxycholoroquine and azythromicin on COVID’s progress, published in the Elsevier journal, the International Journal of Antimicrobial Agents. It was reviewed in a day and written by the Editor-in-Chief. The forensic scientist Elizabeth Bik and others had spotted a number of serious methodological flaws in the paper (missing participants, confounds, lack of randomisation). It transpired that the paper was also under review at Pubpeer but had been published unaltered within a day by the journal.
Unusual times, unusual methods
The ethics of rapid publishing is going to be tested. But there are some things we should not do.
Publishing results that are questionable, give false hope, may not replicate and have not been adequately scrutinised helps no one.
The effort, imagination, alacrity and industry of our scientists – here and across the world – is testament to what we can do when we are up against a fatal virus which, at the time of writing, is still not giving up.
Good science can, and will, lead to healthier people.
