Flawed data from an almost unknown U.S. health care analytics company caused several national governments and the World Health Organization (WHO) to change policies and treatments for COVID-19.
Surgisphere’s CEO, Sapan Desai, co-authored two studies published in the Lancet and the New England Journal of Medicine. But he failed to explain the methodology for collecting the data or the data, the Guardian reported.
The company claimed to have collected data from hospitals around the world. Scientific articles that caused changes in coronavirus treatment policies in Latin American counties were based on that flawed data.
A data scientist for an international corporation called the database a scam. More than 100 doctors dispute the Lancet study as well.
President Donald Trump has promoted the use of the antimalarial drug for COVID-19 patients, and said he took it himself. The WHO and research organizations worldwide stopped trials of hydroxychloroquine based on those articles. The WHO plans to resume trials of the drug, the Guardian reported.
The Lancet and the New England Journal of Medicine expressed concern about the published studies, the Guardian reported. Authors who aren’t affiliated with Surgisphere commissioned an independent audit to determine the validity and provenance of the database.
The WHO’s data safety monitoring committee said hydroxychloroquine taken by COVID-19 patients did not appear to increase their risk of death, so the public health agency will resume all parts of its Solidarity trial.
The peer-reviewed study published on May 22 by the Lancet linked the antimalarial drug with a higher mortality rate for COVID-19 patients, the Guardian reported. It also was said to increase heart problems.
The study backed its claims with data analyzed that Surgisphere said it collected from 1,200 hospitals and almost 15,000 patients who received the drug. That led the WHO to stop its hydroxychloroquine trials. Johns Hopkins University data showed COVID-19 deaths didn’t reach 73 until April 23, while the Surgisphere data claimed that figure as of April 21, the Guardian reported.
Five hospitals in Australia that would have been vital for Surgisphere’s database denied any relation or knowledge of the company, Guardian Australia reported.
Surgisphere started in 2008 publishing medical textbooks. Inexplicably it claimed to become the owner of an international database with access to data from 1,200 hospitals worldwide and 96,000 patients.
“It doesn’t make sense,” Dr. James Todaro told the Guardian. “It would require many more researchers than it claims to have for this expedient and [size] of multinational study to be possible.”
Todaro runs MedicineUncensored, where the results of hydroxychloroquine studies are posted.
“We use a great deal of artificial intelligence and machine learning to automate this process as much as possible, which is the only way a task like this is even possible,” Desai told the Guardian.
He said the research collaboration for his company’s QuartClinical content management system was started several years ago without listing a date.
Desai said the company’s health care partners convert the data to the format the database needs, de-identifying it in the process. But the QuartzClinical website said it does all the work.
A data scientist with Nous Group, which performs data integration for government agencies, told the Guardian that Surgisphere’s database was “almost certainly a scam.” Hospitals couldn’t do what Desai said.
“De-identifying is not just a matter of knocking off the patients’ names, it is a big and difficult process," Peter Ellis, the chief data scientist of Nous Group, told the Guardian. “I doubt hospitals even have capability to do it appropriately. It is the sort of thing national statistics agencies have whole teams working on, for years.”
The hospital names and information from the Surgisphere database has not been released despite sharing agreements by the Lancet.
