Surgisphere Founder’s Past Work Under Scrutiny

Two retractions in top-tier journals last week tied to a Surgisphere dataset have raised suspicions about other questionable projects and potential instances of fraud by company founder Sapan Desai, MD, PhD.

Many researchers watching the Surgisphere story unfold have questions about the validity of the so-called Surgical Outcomes Collaborative, a Surgisphere database that claimed to harness information from COVID-19 patients in 671 hospitals across 6 continents. A closer look into Desai’s past reveals Surgisphere is just one piece of a prolific resume he has built up over the past two decades, including some business ventures and research publications that appear duplicitous.

“The only explanation at this point that seems to hold any water is that he literally fabricated the data out of thin air,” said David Cohen, MD, MSc, former director of cardiovascular research at Saint Luke’s Mid America Heart Institute at the University of Missouri in Kansas City.

Universities including Stanford and Harvard, which were listed as research partners on the Surgisphere site, told The Guardian they were not aware of any such formal relationship with the company. More than a dozen of Desai’s colleagues, who served as mentors or co-authors on past publications, did not respond to requests for comment by MedPage Today.

Desai also managed the Vascular Outcomes Collaborative, a database used in a prior study on pneumatic compression therapy published in October 2019, and an abstract on dialysis graft management submitted to the Society for Clinical Vascular Surgery meeting in March, which was postponed.

The Surgical Outcomes Collaborative popped up in the New England Journal of Medicine and Lancet studies, but was not cited in any prior published research. Similarly, the Vascular Outcomes Collaborative leaves almost no trace.

The websites for both collaboratives have been removed, but an internet archive search found they were near-identical. Both listed Surgisphere as their parent company, and offered access to machine learning and big data to drive “high quality research” — all for an annual fee of $2,495.

Since the studies involving the Surgical Outcomes Collaborative were retracted from the NEJM and the Lancet — which have since initiated independent data reviews — another paper Desai published in 2005 that compared tissue from the inner ear structures of several rodents has also been scrutinized.

Elisabeth Bik, PhD, a microbiologist who pursues scientific fraud, analyzed the images used in the paper and said they appeared to have been digitally duplicated.

“You would not expect in a slice of tissue that certain cells would look exactly the same, yet this is what we are seeing here,” Bik told MedPage Today. “A couple of areas visible in the tissue slices are labeled to come from different animals, but that is biologically impossible to have features that are so alike.”

Those same images were republished again in a 2013 paper and presented as novel findings, without any mention of the 2005 paper, Bik said. “Research is supposed to be novel, so if you reuse an old paper, you should at least mention the old paper,” Bik said.

In a virtual presentation hosted by Trends Research on May 6, Desai said Surgisphere’s database stores fully de-identified patient data, making it “open for collaboration and research.” He added that he selected a machine learning model specifically because it had greater transparency that was “more suitable for scientific review.”

Yet when scientists demanded Desai reveal how the data was collected and analyzed in this system, their calls went unanswered.

Desai, 41, is an avid entrepreneur, with 11 trademarks listed in a 43-page resume obtained by MedPage Today. The surgeon also claimed more than 100 published journal articles.

From 2010 to 2013, Surgisphere published the Journal of Surgical Radiology, the website for which (since removed) stated it accrued nearly a million page views a month, despite the journal’s impact factor being very low.

One of Desai’s first ventures, Apex Testing LLC, offered exam preparation materials to medical students. Desai, either independently or through Surgisphere, currently has 30 textbooks for sale on Amazon, many of which are published through Catalyst Publishers and link readers to a site called ClinicalReview.com that no longer exists.

The Vascular Outcomes Collaborative also offered CME and medical education through ClinicalReview.com that cost $295 per year.

Desai, who disclosed that he was a consultant for W.L. Gore & Associates, received more than $40,500 from the company in 2018 related to a vascular graft device. He was the highest paid physician for the device listed in ProPublica’s Dollars for Docs database, earning three times the amount received by the second highest earner.

The surgeon is also involved in four medical malpractice lawsuits, and at least one involving a wrongful death, according to Illinois court records.

Bik said it would be useful to look into Desai’s other publications given the growing suspicion that a part of the dataset used in past work has been fabricated.

“It’s usually this process where you start to cheat a little bit and you get away with it, and then you do it a couple of more times,” Bik said. “You gradually grow the level of deceit over time.”

Last Updated June 11, 2020