Researchers have uncovered an alleged instance of image manipulation within previous work published by the founder of Surgisphere, Buzzfeed News reports.
Two COVID-19 papers that relied on a Surgisphere database were retracted last week, one from the Lancet and one from the New England Journal of Medicine. While the database was said to include data on about 100,000 COVID-19 patients from 1,200 hospitals, the Guardian reported that there were discrepancies between data in the Surgisphere database and that from other sources like Johns Hopkins University. In the Lancet retraction notice, the authors said they sought an independent audit of the data, but that Surgisphere declined to provide their reviewers with the full dataset, citing contract and confidentiality agreements.
According to Buzzfeed News, Elizabeth Bik, a consultant on scientific image manipulation, took a deeper dive into work previously published by Sapan Desai, the founder of Surgisphere. Her analysis uncovered possible image manipulation within a 2004 study in Journal of Neurophysiology on the anatomy of the vestibular system, a finding Buzzfeed News says two other researchers concurred with. Bik posted her findings to both her blog and to PubPeer.
Buzzfeed News adds that the journal is looking into the matter.