This is more than a month old, but I missed it at the time.

It appeared that the authors had run entire paragraphs from press releases and websites about ivermectin and Covid-19 through a thesaurus to change key words. “Humorously, this led to them changing ‘severe acute respiratory syndrome’ to ‘extreme intense respiratory syndrome’ on one occasion,” Lawrence said.

Heh.

“The authors claimed to have done the study only on 18-80 year olds, but at least three patients in the dataset were under 18,” Lawrence said.

“The authors claimed they conducted the study between the 8th of June and 20th of September 2020, however most of the patients who died were admitted into hospital and died before the 8th of June according to the raw data. The data was also terribly formatted, and includes one patient who left hospital on the non-existent date of 31/06/2020.”

There were other concerns.

“In their paper, the authors claim that four out of 100 patients died in their standard treatment group for mild and moderate Covid-19,” Lawrence said. “According to the original data, the number was 0, the same as the ivermectin treatment group. In their ivermectin treatment group for severe Covid-19, the authors claim two patients died, but the number in their raw data is four.”

[...]

Brown created a comprehensive document uncovering numerous data errors, discrepancies and concerns, which he provided to the Guardian. According to his findings the authors had clearly repeated data between patients.

“The main error is that at least 79 of the patient records are obvious clones of other records,” Brown told the Guardian. “It’s certainly the hardest to explain away as innocent error, especially since the clones aren’t even pure copies. There are signs that they have tried to change one or two fields to make them look more natural.”

[...]

“Because the Elgazzar study is so large, and so massively positive – showing a 90% reduction in mortality – it hugely skews the evidence in favour of ivermectin,” Meyerowitz-Katz said.

“If you remove this one study from the scientific literature, suddenly there are very few positive randomised control trials of ivermectin for Covid-19. Indeed, if you get rid of just this research, most meta-analyses that have found positive results would have their conclusions entirely reversed.”

Kyle Sheldrick, a Sydney doctor and researcher, also independently raised concerns about the paper. He found numbers the authors provided for several standard deviations – a measure of variation in a group of data points – mentioned in tables in the paper were “mathematically impossible” given the range of numbers provided in the same table.

Sheldrick said the completeness of data was further evidence suggesting possible fabrication, noting that in real-world conditions, this was almost impossible. He also identified the duplication of patient deaths and data.

Not great.