The study authors, one published in The Lancet and the other in The New England Journal of Medicine, requested that the studies be withdrawn because the independent auditors were unable to access all of the information necessary to verify the data. Both studies used data from the data analysis company Surgisphere Corporation.
The Lancet retracted study, published on May 22, found that Covid-19 patients treated with hydroxychloroquine and chloroquine were more likely to die or suffer from dangerous side effects.
The study provided a counterpoint to President Trump, who called hydroxychloroquine a "game changer" for Covid-19. Several countries and the World Health Organization have suspended ongoing studies on the effectiveness of drugs based on The Lancet study, although WHO resumed its study on Wednesday.
However, other studies have also found that Covid-19 patients have not received treatment with hydroxychloroquine, and they may have experienced serious side effects. Other studies on hydroxychloroquine for Covid-19 are underway.
Efforts to verify the dataset
In their retraction, Drs. Mandeep Mehra, Frank Ruschitzka and Amit Patel wrote that after concerns were expressed about the data and analyzes conducted by Surgisphere and its founder, Sapan Desai, co-author of the study, they launched a review by third party peers with Desai's consent. They aimed to confirm "the completeness of the database and to reproduce the analyzes presented in the document".
"Our independent peer reviewers have informed us that Surgisphere would not transfer the full data set, customer contracts and the full ISO audit report to their servers for analysis, as this transfer would violate customer agreements and confidentiality requirements. As a result, our examiners were unable to conduct an independent and private peer review and therefore informed us of their withdrawal from the peer review process, "wrote the three researchers.
"Due to this unfortunate development, the authors are requesting that the document be retracted," they said. "We all entered into this collaboration to contribute in good faith and at a time of great need during the COVID-19 pandemic. We sincerely apologize to you, the editors and the readership of the journal for any embarrassment or inconvenience this may have caused. "
In a statement, The Lancet said it "takes questions of scientific integrity very seriously, and that there are many unanswered questions regarding Surgisphere and the data that would have been included in this study". He added: "Institutional reviews of Surgisphere's research collaborations are urgent."
The second retracted study, published in the New England Journal of Medicine, found that certain heart disease drugs, including ACE inhibitors, did not increase the risk of death for patients with coronavirus. The authors included Desai of Surgisphere and Mehra and Patel, who were study authors in The Lancet, as well as Drs. SreyRam Kuy and Timothy Henry.
"Because not all authors were granted access to the raw data and the raw data could not be made available to a third-party auditor, we are unable to validate the main sources of data underlying our article ", wrote the authors in the withdrawal.
"We apologize to the editors and readers of the Journal for the inconvenience this has caused."
Earlier this week, The Lancet and The New England Journal of Medicine expressed concerns about the Surgisphere data used in the two studies.
In a statement posted on its website after expressions of concern, Surgisphere said that "our multinational study of the sighting register published in the Lancet Medical Journal has elicited both praise and skepticism from the scientific community and global institutions.
"The Surgisphere registry is an aggregation of unidentified electronic health records for clients of QuartzClinical, Surgisphere's machine learning program and data analysis platform," he added. Surgisphere said it had detected a problem with a hospital in its database. "This hospital has been correctly reclassified in our database. The document's conclusions are not affected by this update," the statement said.
CNN contacted Surgisphere for comments Thursday.
"While it is important to hide individual identifiable patient data, there are other ways to verify the integrity of the data," said Stephen Evans, professor of pharmacoepidemiology at the London School of Hygiene and Tropical Medicine who was not involved in either study. from the Lancet study.
"For example, data providers should be able to confirm that they have provided data with at least the approximate number of patients involved. It is okay to withdraw the document under these circumstances."
