Releasing the results of the Recovery trial quickly, may have saved lives. But without the data available yet, the results cannot be scrutinised. Matt Morgan considers this complex decision

The authors were left with an impossible choice. Staring at the bold statistics showing that a simple, cheap drug may save tens of thousands of lives, what else could they do? One day of hesitation could translate into actual lives lost. A dad, a mum, a son. 

The response to the news that dexamethasone may reduce covid-19 deaths by as much as a third was as predictable as the rain when lockdown was eased. A chorus of cheer was followed by delicate, tentative questions about “the data”, then a loud roar of criticism. As yet, there is no pre-print of the findings and the researchers have said that the full results will be published shortly. 

Although the statistics are clear, that is all that we have. Until the results are posted on a pre-print server, or published in a peer reviewed journal, we cannot scrutinise the findings more closely. Instead, potentially the biggest medical discovery of our generation was announced by email in just 234 words, and that included “Yours sincerely.” Some say that covid-19 has murdered evidence based medicine. 

History is written by the winners and it’s too early to count the casualties. Yet nuance is still needed, even in a crisis. It is possible to be right in the moment, yet wrong in time. Many decisions in the pandemic may prove “right for covid” but “wrong for all”. Our prolonged lockdown in Wales is undoubtedly “right for covid”, by preventing more covid related deaths. But the dead don’t care why they died and longer lockdown may have many other consequences, such as undiagnosed cancers, or delays to treatment. This may even be the cause of more deaths overall when the accountancy is complete. Changing global medical practice through a press release may be “right for covid”, potentially saving tens of thousands of lives, by expediting what could be a life saving treatment. But putting the frail process of evidence based medicine into intensive care, with uncertain chances of survival, may ultimately prove “wrong for all”. Future standards may crumble, public belief in press-release medicine grows, transparency and peer review die. 

Covid-19 has shaken up life’s snow globe. The scene left when the flakes have fallen will be different from before. Different in many ways, and some will be good. Perhaps now is the time to resuscitate the ageing body of evidence based medicine. Resection of the maligned peer review process may give room for a life saving transplant. Bayesian platform adaptive trials have arrived, throwing p values under the bus. The discovery of the Higgs boson was announced globally through iterative, collaborative, mathematical automated process rather than by a man wearing a tie on an expensive podium to a bunch of his friends. Medical publishing is already learning from our colleagues in the STEM world by adopting pre-print servers. But the manner of this announcement shows there is still some way to go. 

Matt Morgan, Honorary senior research fellow at Cardiff University, consultant in intensive care medicine, research and development lead in critical care at University Hospital of Wales, and an editor of BMJ OnExamination. 

Twitter: @dr_mattmorgan

Competing interests: None declared