Michael Hiltzik: How Retracted Research Papers Contaminated Global Coronavirus Research | Business

On June 4, Lancet, one of the world’s most prestigious scientific publications, the British medical journal, withdrew its paper, which was one of the most important in a new area of ​​coronavirus research.

Treatment of COVID-19 with the antimalarial drug chloroquine or hydroxychloroquine does not increase the risk of heart-related mortality in hospitalized COVID-19 patients, according to a peer-reviewed paper published by Lancet on May 22. It is said that.

The first author of this paper used a worldwide database of 96,000 patients from nearly 700 hospitals on six continents and was highly regarded as a Harvard Cardiologist.

The findings were very impressive and urged researchers in the United States and abroad, including the World Health Organization, to immediately discontinue their research on antimalarials because of the risk of heart problems.

But within hours of publishing the paper, questions were raised about the database. Immediately following its withdrawal, another paper withdrawn from the same database and published by the similarly prestigious New England Journal of Medicine was withdrawn.

The withdrawal is portrayed by defenders of the peer-review system of scientific publishing as evidence that the system is working — the papers have been published, they have been scrutinized through regular scientific give and take, and shortly after them. The drawbacks that were withdrawn were exposed.

Isn’t that how scientific methods are supposed to work?

Well, no. The whole point of peer review in which a submitted article is subject to academic review by a panel of subject matter experts is to identify any shortcomings before publication.

In this case, Mehreep R. Mehra, Harvard’s lead author, had no direct access to the patient database itself, but relied on an analysis by a strange Chicago company called Surgisphere, who owns the database. I admitted that. When I asked for access to the database after it was published, it was denied.

“We can no longer guarantee the authenticity of major data sources,” Mehra and two of his three co-authors said in a withdrawal statement. The name of Sapan S. Desai, CEO of Surgisphere, the third co-author, is not mentioned in the statement.

It is self-evident that neither Lancet nor NEJM saw the ball here.

“This is not a bad co-author’s story,” says Lee Turner, a bioethicist at the University of Minnesota. “There have been some rather serious breakdowns within the journal.”

Worse, the breakdown has led to politically-affected skepticism about coronavirus research in general and antimalarial drug research in particular. The main offender of this effort to undermine science is President Trump, who turned the use of antimalarial drugs as a cure for pandemics into a partisan cause.

Sure enough, Trump’s re-election campaign used the publishing blunder for its own purposes. The first study claimed by the campaign was “happily advertised by mainstream media in an attempt to undermine credit and attack President Trump.”

Campaigns were ongoing, noting that Trump was often accused of having engaged in a “war with science.” “The actual war on science was done by the people behind the study and Lancet.”

Poor peer review has been understood for some time. “Yes, this is a wake-up call,” says Ivan Olanski, whose website for Retraction Watch tracks the withdrawal of lancet and NEJM withdrawal research papers. “But we’ve had a wake-up call for years.”

Perhaps in the most notable reversal of the scientific paper, Lancet in 2010 withdrew a study by British researcher Andrew Wakefield, who claimed to associate autism with the measles, mumps, and rubella (MMR) vaccine.

But the withdrawal was only twelve years after Lancet published the study. In the interim, Wakefield’s work was under intense scientific attack.

Shortly after withdrawal, Wakefield lost his license to practice medical care in England. However, he remains a prominent figure in the anti-vaccine movement, and his uncovered papers have fueled anti-vaccine activity to this day.

Lancet also endorsed Paolo Macchialini, an Italian surgeon whose claims of developing an artificial trachea with stem cell engineering, including a spectacular special report from NBC News anchor Meredith Vieira. The claim was ultimately regarded as the product of unethical and deceptive work.

“Paolo Macchiarini is not guilty of scientific misconduct,” Lancet declared in 2015, citing findings from the Karolinska Institute, where Macchiarini worked as a researcher. “It couldn’t be defended from pulling the scientist’s professional reputation out of the stigma,” Lancet said.

Three years later, when Karolinska reversed, the lancet also reversed, with the heading “Paolo Macchialini’s Final Verdict: Guilty of Cheating.” The journal also withdrew two papers by Macchiarini published in 2011 and 2012.

By those standards, the retreat of coronavirus paper occurred at the speed of lightning. However, given the politically charged environment and global despair for the treatment of COVID-19, the original paper was particularly damaging to respect for the scientific process.

The withdrawal threatened to undermine the reliability of the first large-scale objective study of antimalarial drugs as a treatment for COVID-19. This is a study of 821 subjects who were exposed to the virus at home or at work as a health care worker or first responder, then treated with either hydroxychloroquine or placebo.

The study, led by the University of Minnesota and published in the New England Journal on June 3, found that the drug did not help prevent infection by the coronavirus.

This takes us to the strange side of the withdrawn research. Let’s take a look.

The research question first arose in Australia. There it turns out that the Lancet paper cites COVID-19 deaths in more countries than actually occurred during the period of the paper. Surgisphere explained that data from Asian hospitals were mistakenly added to the Australian figures. Lancet has made a simple fix.

However, this flaw prompted further investigation of Surgisphere and its database. Health officials wondered how a small company with around 12 employees could build a worldwide patient database containing tens of thousands of samples.

By compiling such data, confirming that individual IDs have been deleted, and then creating software tools that make the database useful, by the time we reach agreement with hundreds of hospitals that contribute to patient records. It takes several years and a large team work. It seemed far beyond the capabilities of Surgisphere.

Desai rejected the request to access raw data. He did not respond to our request for comment, but on May 30, British scientist The Scientist said his database was “audited by an outside third party, irrespective of us.” I’m willing to do that,” he said (helping to verify all this.)

But after the first question arose, Mehra and other co-authors launched their own “third-party review of Surgisphere by an independent third party…” to assess the origin of database elements, Database integrity and reproduction of the analysis presented in the paper. “They continued their setback.

Mehra of Harvard portrays himself not as a victim of a company that seems to be providing the exact data he needs and his ambition to answer important health questions .. He suggests that a desperate quest for information pushed him aside to his professional standards.

“I wanted to contribute to this research when I needed it very much, so it’s clear that it wasn’t enough to make sure the data source was suitable for this use,” he emailed. And told me. “So, and for all the confusion-either directly or indirectly-I’m really sorry.”

Is that enough? In the original paper, Mehra states that he and co-author Amit Patel had “full access to all the data in the study.” I say not now.

As previously observed, the treatment of COVID-19 or the frenzy of finding a treatment led to the removal of many barriers that have hindered or delayed the spread of past scientific research. That is especially true of the barriers created by the for-profit model of scientific publishing.

In many ways it’s good. Information sharing is especially important at this point, as pathogens that are still misunderstood threaten the world. But it can also lead to the dissemination of half-finished or unburned research. This is not necessarily a major issue as long as the “preprinted” papers that have not yet been peer reviewed are labeled.

But when half-finished research arrived with the unnaturalness of the world’s seemingly most credible journal, it’s time to wonder if dedication to scientific rigor came second after the desire for influence. ..

“These journals are major brands,” Turner told me. “This motivates the fence to be motivated in terms of dramatic publications.” Now, as a study definitively revealing a popular novel based on a huge sample of 100,000 patients There are few more dramatic papers in the atmosphere.

We need to find out more about this before the smoke goes out. Mehra and his colleagues need to explain how they came to rely on databases they couldn’t guarantee. Desai must explain the origin of the database, even if it exists.

Lancet and NEJM need to explain how to publish a paper that must be withdrawn within two weeks. Why would anyone have to trust their publications without it?

Photos (contact 312-222-4194 for image help):