Search

The Big-Data Mystery Behind Retracted Covid-19 Studies of Hydroxychloroquine, Other Drugs - The Wall Street Journal

krotoson.blogspot.com

A box of hydroxychloroquine at a pharmacy in Brazil: The lead author of a study that found risk of harm from taking the drug for Covid-19 has cast doubt on the existence of data that underpinned the study.

Photo: diego vara/Reuters

Surgisphere Corp., a little-known medical-data company, said it had a massive data set of patients that researchers could use to study the effect of blood-pressure and malaria drugs on Covid-19, leading to the publication of high-profile papers on the disease.

Now, the Harvard researcher who co-authored two articles on its research says he never saw the data and can’t verify its existence.

The New England Journal of Medicine, which retracted its article on their findings, says it didn’t sufficiently vet the underlying data and will revamp its review process.

And the Lancet, a British medical journal that also retracted an article based on Surgisphere statistics, says it is reviewing how it handles data sets while maintaining that the paper was peer-reviewed and edited according to its usual procedures.

“We have limited experience with reviewing or publishing studies like this one, which used a large database based on electronic medical records,” a New England Journal of Medicine spokeswoman said Tuesday in response to questions from The Wall Street Journal. “This is why, going forward, we will be sure to have such experts look at these studies as well.”

Lancet editors said they are “reviewing our requirements for data sharing and validation among authors, and data sharing following publication.”

Elsevier, publisher of the Lancet and other journals, said it was reviewing the approximately 20 articles in various publications that cite Surgisphere data.

The retractions and revelations about Surgisphere have raised questions about how medical journals decide which articles to publish and how authors vet raw data used to draw conclusions.

The NEJM study examined the effects of cardiovascular drugs and blood-pressure drugs in Covid-19 patients. The Lancet article found antimalarials, including hydroxychloroquine, provided no benefit as a treatment for Covid-19 infections, while raising risks of harm, including of death.

Both studies claimed to have analyzed thousands of patient data records supplied by Surgisphere.

Sapan Desai, Surgisphere’s owner, is a vascular surgeon who practiced until earlier this year at a midsize acute-care hospital in the Chicago suburbs. Created in 2008, Surgisphere’s business model has varied, from publishing a medical journal and a children’s book, “Bart from Iowa,” to providing health-care analytics.

VIDEO

As the coronavirus continues to spread around the globe, companies and academic labs are racing to develop a vaccine that would help society get back to normal. But there could also be costs to moving too quickly. WSJ’s Daniela Hernandez explains. Photo illustration: Laura Kammermann

Since publication, dozens of scientists around the world have asked how a private company could have amassed the data and why two of the world’s most prestigious medical journals could have published papers based on it without closely auditing the information or the company. The Guardian, a British newspaper, first reported these concerns last week. In recent days, The Wall Street Journal reached out to major hospitals and electronic health records providers; none said they had any data-sharing relationship with Surgisphere. And at least three other medical journals recently rejected papers that relied on Surgisphere data.

Since the retractions, Surgisphere has scaled down its website and deleted its social-media accounts. Dr. Desai didn’t reply to requests for comment.

The lead author on both papers, Mandeep Mehra,  is a well-known cardiologist at Brigham and Women’s Hospital in Boston with a professorship at Harvard University. He said Tuesday through a Brigham and Women’s Hospital spokesman that he never saw any of the data for either article. When Dr. Mehra asked Dr. Desai for various analyses, Dr. Desai would respond by providing tables or figures, the spokesman said.

“Each time Dr. Mehra presented the Surgisphere data team with a question, or a request for analysis, he was provided with appropriate and efficient responses to satisfy his, the peer reviewers’ and the journal editors’ queries,” the spokesman, said. He also said Dr. Desai had claimed he couldn’t share the data because of “client confidentiality contracts.”

Arthur Caplan, a bioethicist at New York University’s Grossman School of Medicine, said it is unusual that authors wouldn’t want to see the original data themselves, particularly data-analysis papers like the ones for which Dr. Mehra was lead author.

“He did something very unusual but I can’t say unprecedented,” Dr. Caplan said.

Asked whether this handling of the data was in line with research norms, Mr. Murphy said matters regarding research conducted at Brigham & Women’s “are confidentially assessed per hospital and federal regulations.”

After criticism of the data arose, Dr. Mehra and co-authors on the NEJM article hired a firm to conduct an audit. They said Surgisphere said it couldn’t share access to the data, leading Dr. Mehra and the other co-authors to ask for the retractions.

The New England Journal of Medicine said that before publication, the paper was sent to four external experts and a statistical reviewer, none of whom had access to the underlying data. “We recognized in our internal review that the peer reviewers we had chosen for this article were subject-matter experts but not experts in big data studies like this one,” the spokeswoman said, adding the publication would correct that in the future.

The paper was published on an accelerated schedule, a step many medical journals have taken with Covid-19 research.

Meantime, almost immediately after publication of the Lancet article, researchers began to challenge its findings.

They pointed out that data for certain countries exceeded what the countries had reported. Few scientists had heard of Surgisphere, despite its claim to have accumulated data of millions of patients from hundreds of hospitals across several continents.

In retrospect, there were inconsistencies in the Lancet article that should have been caught, said Jerome Kassirer, a former editor in chief of the New England Journal of Medicine who wasn’t involved in the studies. “In this case, you have to look at the fact that as soon as the paper was published, other people who looked at it said, ‘Whoa, this is not plausible,’ ” he said.

Brigham & Women’s directed questions on that to the journal publishers. The NEJM spokeswoman disagreed the review process failed, saying the problem stemmed from the lack of big-data expertise.

A paper under consideration for publication in medical journals is sent to peer reviewers, who, experts say, are responsible for determining whether it should be printed and whether there are flaws in approach or methodology. The process is dictated by each journal, and the reviewers are typically anonymous and unpaid. “It’s quite apparent that this was a serious failure of peer review,” Dr. Kassirer said of the retractions.

In recent weeks, at least three medical journals, including JAMA Internal Medicine, rejected articles submitted for publication that cited Surgisphere data, the Journal found.

In light of the Surgisphere issues, the Journal of Vascular Surgery returned a paper relying on Surgisphere data recently submitted by Dr. Desai, plus another where he was a co-author, said Dr. Peter Gloviczki, the editor in chief. He said a third paper, on which Dr. Desai was a co-author, was accepted but put on hold.

Two leading U.S. providers of electronic health record data, Epic Systems Corp. and Cerner Corp., said they had no knowledge of Surgisphere or any deals to provide it with data.

Since the retractions and questions arose about Surgisphere, no hospital has said it had a data-sharing relationship with the company. NYU Langone Health told the Journal it was approached by Surgisphere and declined to participate in a data-collection initiative.

Aneesh Chopra, president of health-care analytics firm CareJourney and former U.S. chief technology officer under President Obama, said he doubted Dr. Desai was able to build this kind of database. “At best, I would presume he has tapped other networks at an aggregate level to generate reports without the ability to validate his results, or allow others to do so,” he said.

Write to Jared S. Hopkins at jared.hopkins@wsj.com and Russell Gold at russell.gold@wsj.com

Copyright ©2020 Dow Jones & Company, Inc. All Rights Reserved. 87990cbe856818d5eddac44c7b1cdeb8

Let's block ads! (Why?)



"behind" - Google News
June 11, 2020 at 05:26PM
https://ift.tt/3feEyCi

The Big-Data Mystery Behind Retracted Covid-19 Studies of Hydroxychloroquine, Other Drugs - The Wall Street Journal
"behind" - Google News
https://ift.tt/2YqUhZP
https://ift.tt/2yko4c8

Bagikan Berita Ini

Related Posts :

0 Response to "The Big-Data Mystery Behind Retracted Covid-19 Studies of Hydroxychloroquine, Other Drugs - The Wall Street Journal"

Post a Comment

Powered by Blogger.