July 8, 2020
Was the Surgisphere case a one-off? Or does it highlight the bigger systemic problem of research fraud?
By Natalie Rhodes, Policy Officer
If you search for scientific research articles with COVID-19 in the title, you’ll see more than 17,000 articles published since the start of 2020[i], but this vital research is being undermined by weak or even fraudulent research practices. Perhaps the highest profile example so far is the Surgisphere case which saw a small US company seemingly fabricate a database, the data for which was purportedly from the medical records of nearly 100,000 COVID-19 patients treated in 167 hospitals[ii]. This database was then analysed and published in two of the world’s most influential medical journals[iii] [iv]. Both papers have since been redacted by the journals, but what damage has already been done? And is this a one-off incident or a reflection of the fraud that plagues academic research?
The study published in The Lancet purported to show that Hydroxychloroquine, the anti-malarial drug now infamously associated with the US president[v], was dangerous to COVID-19 patients. (Since then a separate UK study has demonstrated the drug is not an effective treatment and the UK government have suspended trials of the drug[vi]). However, as seen with other redacted papers its possible it will continue to be cited in future research papers[vii]. The most notorious example of the impact that falsified data can have even after its redaction is the paper, also published in The Lancet, incorrectly linking the MMR vaccine with autism, which is said to be a key underlying driver of the anti-vax movement[viii].
In addition to the studies in The Lancet and NEJM, a third study which analysed Surgisphere data was uploaded onto a pre-print server which allows researchers to publish their research without having to wait for publication in a journal[ix]. This paper reported that a drug called Ivermectin can be effective in treating COVID-19. As a result, the drug has now been included in official treatment guidelines in both Peru[x] and Bolivia[xi] alongside increased use in other Latin American countries. As highlighted by researchers at ISGlobal[xii], this poses risks such as a diversion of drug supplies away from the patients (who require this drug for its original use as an antiparasitic), unforeseen side effects in patients and reports of doctors administrating veterinary versions of the drug with harmful side effects after the human formulation had ran out[xiii].
Whilst this isn’t reason enough to begin accusing all medical journals or academics of research fraud, it still is a phenomenon which has yet to be taken as seriously as it should be. Nearly 1 in 50 scientists report having falsified or fabricated their data, with up to 1/3 utilising questionable research practices. This goes up to around 7 in 50 and nearly ¾ respectively when researchers were asked about the research practices of their colleagues[xiv]. Compare this to the fact that only 0.017% of scientific papers on PubMED (a popular research paper database) have been retracted, this indicates that a significant amount of research containing fraudulent data is circulating in academia and is further evidence of journals’ reticence to identify and call out poor science.
The question then arises; how is this allowed to happen? Prior to publication in a scientific journal a paper will undergo the peer review process which is supposedly used to ensure the research is scientifically robust. However, whilst peer reviewers are encouraged to scrutinise the raw data behind a paper, whey are not necessarily required or mandated to. (In the Surgisphere case the peer reviewers did not have access to the raw data). So as explained by the Editor-in-Chief of the Lancet, Richard Horton, the peer review process is not designed to capture research misconducts[xv]. This is just one of the many criticisms of the peer review process[xvi].
Tackling these cases of fraud within academic research is just as important as the research itself. Not only can cases such as Surgisphere have serious impacts on patient safety, but they also undermine trust in scientific research – something we need now more than ever. Corrupt researchers who fabricate or falsify data are obviously at the heart of fraudulent research. But many other players – research and scientific publishers, peer reviewers, funders, ethics committees to name a few – also have a role to play to deliver on a collective responsibility to protect patients and ensure the integrity of scientific research. It is in all our interests that all those involved in COVID-related research redouble their commitment and effort to undertake and report ‘clean’ science.