Retracted Lancet study ‘a wake-up call’

‘From the medico’s point of view, you now think you can’t trust. Had it been any other journal, but The Lancet has got a hell of a reputation.’
 
That is Dr Roger Lord, a Research Fellow with The Prince Charles Hospital in Brisbane and lecturer in medical sciences at The Australian Catholic University.
 
He is referring to the retracted peer-reviewed study published in The Lancet in May examining the effects of hydroxychloroquine as a treatment for patients hospitalised with COVID-19, after discrepancies were noted within the data. Amid concerns it may have been falsified, an expression of concern was lodged by the journal as an investigation takes place.
 
The study analysed the data of 96,000 patients in hospitals around the world – data collected by Surgisphere, a US-based company owned by Dr Sapan Desai, a co-author of the paper.
 
It found that hydroxychloroquine was associated with heart problems and a higher risk of death, leading both the World Health Organization (WHO) and Peter Doherty Institute for Infection and Immunity, among others, to halt clinical trials on the anti-malaria drug. The trials have since recommenced.  
 
It is not the first time a journal has had to retract a paper. But with such glaringly discrepancies in the data, it is the calibre of the journal at the centre of the controversy that has scientists and healthcare workers in shock.  
 
‘Looking through the information that’s come out, some of it about the background of the people and all the rest, you wouldn’t know unless you did a search – and let’s face it, no one’s going to do that,’ Dr Lord told newsGP.
 
‘But things like the number of cases that they were claiming out of Australia when the Australian cases weren’t even that high at that time – well, that’s something that a really eminent journal like The Lancet should have been able to pick up on with their peer review process.
 
‘Why [didn’t they]? I don’t think they’re going to tell you.’
 
Experts have argued a rise in preliminary, unverified and misinterpreted research makes scientific scrutiny more difficult, especially in the current environment. That is where the peer-review process is meant to set studies apart as the gold standard, but this case seems to be proving otherwise.
 
‘Some journals are trying to prioritise publication to help inform the COVID-19 response. However, it is important that the research undertaken is methodologically sound and credible,’ Dr Jo-Anne Manski-Nankervis, an Associate Professor in General Practice at the University of Melbourne and Chair of the RACGP Expert Committee – Research (REC–R), told newsGP.
 
‘Publication processes rely on peer-review which is undertaken by volunteers, and it is not clear what issues those reviewers may have raised.
 
‘This is a complex case which demonstrates the importance of the integrity of researchers and the importance of having reliable datasets with clear governance structures available for use in research.’
 
Some experts have questioned why alarm bells did not go off to see that a little-known company, initially established to publish medical textbooks, was able to access extensive data and turn it around in such a short period of time.
 
‘It is a red flag that there were no statisticians that were authors on the study,’ Dr Manski-Nankervis said.
 
‘And it was surprising to see that data from Australian hospitals was said to be included in a dataset essentially owned by a private company in the US.’
 
The many unanswered questions regarding COVID-19 has seen a massive research response, with more than 23,000 papers published to date. That number is reportedly doubling every 20 days.
 
‘As an academic, trying to get through all the mountains of information that has been chucked out on COVID-19, it’s almost a full-time job,’ Dr Lord said.
 
‘So you would hope that when you pick up a copy of a journal like The Lancet that all that had work’s been done – you shouldn’t have to sort through and criticise.’


The retracted paper halted a number of hydroxychloroquine trials around the world. (Image: AAP)

But with the pandemic now such a focus, and countries like Australia staring down the barrel of a recession, Dr Lord is concerned that some may take advantage of the situation. 
 
‘If you’re a full-time researcher, unless you’re a really well-established one, you can probably see that the next couple of years are going to be really hard to get any money out of the government. So you better start switching things over to look at COVID-19,’ he said.
 
‘A classic example of this is to go on to look at the Australian New Zealand Clinical Trials Registry, and you look at some of the titles there. There was one – somebody who is obviously a psychologist – that was looking at the depression in patients that had bowel cancer, and then they decided to tack on to it depression in patients that had bowel cancer and who had also contracted COVID-19.
 
‘You could see it was like a “Me too, I’m jumping in. Some of my patients fit into this, so I better do that”.
 
‘We saw the same thing happen in Australia when they were struggling as to what sort of tests to use to pick up COVID. The kits that they were using, a number of them had all sorts of sensitivity issues and they were from companies that hadn’t made these sort of tests previously. 
 
‘So everyone wanted to jump on – “There’s a buck in it for me”.’
 
Dr Lord believes that could potentially go for journals, too.
 
‘They have been criticised in more recent times, and certainly during the pandemic, for publishing things a little bit more hastily and of poor quality,’ he said.
 
But could the fact that The Lancet study in question returned a negative result for the use of hydroxychloroquine have played a role? In the rush to get information out, Dr Lord said, that may have contributed to the oversight.
 
‘If they’re stopping people from using it, you could argue, “Well I’m not doing any harm, am I?” Whereas if it had been the other way around and the trial came out and said it’s doing wonderful things, it would have gone the other way. Suddenly the drug would flood into hospitals as a possible treatment,’ he said.
 
‘And then of course you’d have a lot of collateral damage in terms of patients not surviving the treatment.’
 
While it did not take long for independent statisticians and infectious disease specialists to query the paper, Dr Lord argues there have already been consequences.
 
‘It may have even scared a few patients who would have said they’d be in the trial, and now they won’t be,’ he said.
 
‘That’s important because if you slow down your recruitment, again you fall into the problem of not having enough patient numbers to actually get the statistical power to say yes, it is having an effect or no, it’s not. You’re still left in the not quite sure.’
 
Dr Manski-Nankervis agrees.
 
‘If the medication is shown to be of benefit, then this study may have delayed translation of a potentially life-saving medication,’ she said.
 
Neither has the retraction done any favours for the public’s trust in science, Dr Lord says.
 
‘All the knock-on effects that we have, even with the sociological components and people being isolated and whatnot, you want to get to something where people feel confident, safe again, to go out and know that it’s okay, there is some sort of treatment,’ he said. 
 
‘For those that have come across it, it’s yet more proof to say, “Oh look, see, even respected journals are publishing this sort of data.
 
‘If The Lancet publication does prove to contain fabricated data, this will be very disappointing both for the journal and the health and scientific community, which rely on the high standards of peer review provided by this publication.’
 
Dr Lord says, if anything, the situation will certainly be a ‘wake-up’ call.
 
‘Apart from getting egg on their faces, they’re going to be a hell of a lot more cautious about data coming their way,’ he said. ‘So maybe there’s a positive side to it in that sense.
 
‘But it shouldn’t have happened in the first place.’
 
Log in below to join the conversation.

coronavirus COVID-19 hydroxychloroquine peer-review