Top European hospitals deny providing data for retracted COVID-19 research

Scores of leading European hospitals have denied providing data to Surgisphere, the company behind two now-retracted papers that were published in eminent journals.

One of these studies — which linked an anti-malarial drug to increased deaths in COVID-19 patients — was published in the renowned medical journal the Lancet in May. This came after U.S. President Donald Trump revealed in May that he was taking the controversial drug hydroxychloroquine as a prophylactic against coronavirus, despite scant evidence of its efficacy. The release of the study initially led the World Health Organization to pause its own trial of the drug.

Yet as questions emerged over the quality of the data used to support the study’s conclusions, the authors withdrew their support. The journal then took the rare step of retracting the paper.

Speaking to POLITICO, the Lancet’s editor Richard Horton questioned whether the lack of transparency around the data rendered the study a “monumental fabrication” and suggested the results were the result of “research misconduct.” He denied that the publication had been the result of a malfunction in the peer review process.

Dozens of hospitals, national health systems and ministries have denied providing the company with data or any other links.

At the same time, the prominent New England Journal of Medicine also retracted a paper that had used data from the same company. The NEJM study examined the use of blood pressure drugs called ACE inhibitors in coronavirus patients.

These retractions have since led to bigger questions about the use of large troves of health data in medical studies.

No paper trail

U.S.-based Surgisphere claimed to have gleaned data from around 96,000 patients in 700 hospitals across six continents, 50 of which were based in Europe, for the Lancet study.

For the NEJM study, the company said it had secured data from 169 hospitals, of which seven were in Spain; eight in Italy; five in Germany; five in France and seven in the U.K.

When challenged about the quality of the data, Surgisphere refused to open up its databases for audit, citing privacy and confidentiality agreements with the hospitals. It declined to even release a list of the institutions involved.

Surgisphere’s CEO and a co-author of both studies, Sapan Desai, didn’t respond to requests from POLITICO for comment, and the company’s website has been taken down. But in an interview conducted on Turkish state news channel TRT World just over a week before the study was retracted, Desai called the data “high quality evidence.”

“A study of this scale and quality is only possible with the kind of technology we’ve created at Surgisphere through a cloud-based data analytics system,” he said. “The real question here really is, with data like this, do we even need a randomized controlled trial?”

But dozens of hospitals, national health systems and ministries that responded to questions by POLITICO have denied providing the company with data or any other links — raising questions about the source of the information used in the study.

They also called into question Surgisphere’s claims that confidentiality agreements barred it from having its databases audited and revealing the sources of the information.

“When hospitals take part in clinical trials on medicinal products as trial sites, they are required to be listed in the clinical study report, which has to be prepared at the end of the trial,” said a spokesperson for Germany’s Federal Institute for Drugs and Medical Devices.

They noted that for “retrospective studies” — like the one in the Lancet paper — there are no particular legal obligations in place in Germany. There also are no rules stopping participating hospitals from being reported in the study.

Berlin’s Charité hospital went further, saying it is unusual for hospitals to insist on remaining confidential in studies. “As a public research organization of excellence, Charité seeks to be transparent about its processes as much as possible without endangering patient privacy rights,” it said.

POLITICO contacted around five leading hospitals or hospital groups per country in the U.K., Germany, France, Italy, Belgium, Spain and the Netherlands, asking if they had ever provided data to Surgisphere and whether it was common for hospitals to insist on confidentiality. Hospitals were selected based on rankings by the Newsweek magazine or, where that wasn’t available, on search engine results for a country’s major hospitals.

Other key public health authorities and ministries in these countries were also contacted.

More than 50 hospitals across Europe responded by categorically denying any links to Surgisphere. National bodies that spoke on behalf of hundreds of hospitals in their countries said they weren’t aware of any data-sharing deals with the company, but could not rule it out.

The NHS Greater Glasgow and Clyde, a U.K. National Health Service board in central Scotland, said it had had no interaction with the U.S. company, and that at its request, Surgisphere had taken a picture of one its hospitals off the company’s website. Also in the U.K., the Guys and St. Thomas’ NHS trust, which runs three London hospitals, said it hadn’t provided any data to the company.

Questions about why the co-authors would put their names to papers with suspect data have also loomed large.

In France, the Assistance publique – Hôpitaux de Paris, which represents 44 hospitals in the country — said it didn’t work with Surgisphere and had never provided data to the company.

Besides Charité hospital in Berlin, German hospitals Universitätsklinikum Hamburg-Eppendorf and the Medizinische Hochschule Hannover responded that they hadn’t contributed data to the Lancet study.

In Italy, hospitals that said they had no links to Surgisphere were Milan’s Niguarda and San Raffaele hospitals.

The NVZ Dutch Hospitals Association said that “to its knowledge” none of its 107 health care institutions had shared data with the company, but cautioned it could not be completely certain. Meanwhile, the Spanish health ministry said it was “not aware” of any Spanish hospital sharing data with Surgisphere but didn’t rule out the possibility.

Big data

An NEJM spokesperson also denied that any steps were skipped in the rush to get research out during the pandemic and said that it followed “the standards that the field has for peer review of large datasets.”

“This experience gives us the opportunity to think about what new information we need to establish confidence in large data sets,” the spokesperson added.

Questions about why the co-authors would put their names to papers with suspect data have also loomed large. For the most part, the co-authors have said that they didn’t actually see the raw data.

Universitätsspital Zürich, whose researcher Frank Ruschitzka put his name to the Lancet study, said he was “not involved in data acquisition,” while a spokesperson for Harvard’s Mandeep Mehra, who authored both the Lancet and the NEJM study, said he did not have access to the primary data before conducting either study.

“Data were shared with Dr. Mehra in the aggregate only, without patient, or hospital specific data. When asked, [Desai] declined to provide specific information on individual hospitals to Dr. Mehra and the remaining co-authors, citing Surgisphere’s legal contracts and limitations,” said the spokesperson, adding that Mehra withdrew the paper after Surgisphere refused to open up the data for audit.

Want more analysis from POLITICO? POLITICO Pro is our premium intelligence service for professionals. From financial services to trade, technology, cybersecurity and more, Pro delivers real time intelligence, deep insight and breaking scoops you need to keep one step ahead. Email [email protected] to request a complimentary trial.