Surgisphere Corp | In the eye of the HCQ storm

The firm that gave data for hydroxychloroquine study is accused of fabrication

Until May, Surgisphere Corporation was a quotidian company in the world of medical services. Its forte was data analytics, particularly, as its website claims, in its use of art and artificial intelligence tools to help physicians and medical researchers use patient data related to a wide variety of conditions.

Once COVID-19 struck and upended the world economy, the singular focus of countries became all-things medical — diagnostics, reagents, vaccines and drugs. Scientists everywhere started to dust out their bottom cabinets for drugs, made for other purposes, that might have some role in inhibiting the virus. Thus began the meteoric rise of hydroxychloroquine (HCQ) , an anti-malarial and used in the treatment of lupus and rheumatoid arthritis. It was promoted by U.S. President Donald Trump and quite happily endorsed by Indian pharmaceutical companies too — they being the largest manufacturers of the drug.

While several studies started to appear showing that HCQ didn’t measurably improve clinical outcomes in patients, the decisive blow was dealt in a May 21 study published online in the reputed Lancet . “Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis” was published by Indian-origin cardiologist Dr. Mandeep Mehra of Harvard Medical School; Dr. Frank Ruschitzka, University of Zurich; and Dr. Amit Patel, University of Utah. It also included, as a co-author, Dr. Sapan Desai, the founder of the Chicago-based Surgisphere.

The study was premised on health indicators from about 96,000 patients — an astoundingly large data set — spanning 671 hospitals and six continents and concluded that the medicine didn’t improve mortality in COVID-19 patients and was linked to cardiac arrhythmia. Based on this, the World Health Organization announced a pause on the use of HCQ in its solidarity trial, pending a safety review and following suit, several governments in Europe too stopped the use of HCQ in trials.

Within days, however, questions started to emerge from scientists around the globe on the statistical inferences made. More seriously, there were questions on the data set. Mortality rates attributed to countries seemed higher than what was available in public records, health records from COVID-19 patients in Africa seemed disproportionate and it seemed impossible for researchers to scrutinise records from individual hospitals.

Questions on data

Peter Ellis, chief data scientist at the Australia-based Nous Group, penned an expose that accused Surgisphere of fabricating data. He argued that Surgisphere’s database couldn’t have been put together in two months, given that it can take years for electronic health records to be compiled into a machine-learning format. Moreover, the data involved was the personal health records of thousands and getting ethical approvals across continents for such private details would be globally controversial. Investigations by the Guardian Australia and an interview by The Scientist of Mr. Desai showed that no hospital, which was supposedly part of Mr. Desai’s database, was willing to confirm their association for the HCQ study.

Mr. Ellis and the Guardian also reported that the LinkedIn profiles of Surgisphere’s employees showed none of them with the demonstrated expertise of dealing with troves of patient records and clinical data.

For its part, Surgisphere’s website claims to be a “physician-led public service organisation”. Founded in 2008, it claimed to have “amassed data” on over 240 million patient encounters. The HCQ study was not funded by any drug company, private or public donor, or political organisation, it noted in a statement after the controversy broke out. The authors of the Lancet study, some of whom were also authors of a study that linked COVID-19, cardiovascular disease and the role of certain cardiac drugs in the New England Journal of Medicine , retracted their papers, saying that Surgisphere wasn’t able to provide the source data for analysis to independent reviewers, citing “client-confidentiality agreements”.

A day before, the WHO had lifted its ‘pause’. The Scientist reported that Mr. Desai is named in multiple medical malpractice lawsuits in the U.S.. His firm appeared once to be in the business of selling medical textbooks and publishing a journal, Surgical Radiology . In his interview to the publication, Mr. Desai said he’d be considering an independent audit of the HCQ study.