By Olavo Amaral
How to deal with data that is too good to be true?
The choir performs early treatment of a Covid-19 took a hit they would have two weeks with the news that an Egyptian clinical trial demonstrating the efficacy of an ivermectin against the disease was taken from a platform on Analysis Rectangle preprints. The study showed some reduction on 19% in mortality for patients with severe disease compared to a group that had received hydroxychloroquine.
Many zero people had The serious work was taken there at the beginning, either because it came from obscure researchers, because it was written in ghoulish English, or because it produced a result too spectacular to be true. Zilch, however, prevented it from being included in several meta-analysis defending ivermectin, being responsible for a good part of the positive effect observed in them.
The article h was withdrawn from circulation after a journalist English Jack Lawrence decided to investigate it when he noticed signs of plagiarism. Some diesel versions perform articles include a hyperlink to operating system original data with password-protected electronic pay access. In a fluke, Lawrence kicked a little creative 1234 electronic saw the perform Excel spreadsheet with operating system raw data I learned materialize on his screen.
Da na Onwards, a police perform work on Chip Dark brown data showed zero inconsistencies, however strong evidence on fraud: several patients appeared to be clones created by copy-paste, with some data modified to disguise. As a result, the platform removed the article electronic operating system authors have not yet manifested.
The illustrative story to analyze another case that has been gaining space in the Brazilian media. In March, a team of researchers led by endocrinologist Flvio Cadegiani announced at a press conference spectacular results of proxalutamide, an antiandrogenic drug originally developed to treat prostate cancer, which would have led to some reduction in 19% in inpatient mortality with Covid-19.
The degree of success logo design drew attention to critics, which one pointed out how unlikely. Also contributing to the suspicions were the delay in the publication of 2 data (which did not appear as a preprint more than three months later), the high mortality in the placebo group, the meteorological recruitment of more than 600 patients in less than one master of science electronics, evidence of deviations from the protocol approved by the ethics committee.
Part performs skepticism, however, sony ericsson owes the factors zero related to the data. Since the beginning of the pandemic, Cadegiani had claimed positive results for early treatment with hydroxychloroquine j, ivermectin, nitazoxanide electronic dutasteride, in addition to proxalutamide itself in outpatients, some sequence of at least unlikely successes. His collaborator Ricardo Zimerman was a guest of the government bench at the CPI of the electronic pandemic, he became an electronic influencer in the right-wing electronic media social networks, making his presence felt in channels such as Osmar Terra. It is also worth mentioning that the repeated menes perform Presidente proxalutamida do not even work as an academic seal.
These would be good reasons, however, to close the eyes of a study that claims more about 90% about efficacy for some disease that causes millions of deaths? The motto of the Royal Modern society is, after all, nullius in verba (in no one’s words): scientific data should be more important to perform than whoever operating system presents.
Judging by the reception of the article, however, impersonality is on the low side. In the Technology magazine article, electronic cardiologist master of electronic medicine Eric Topol says that operating system therefore results too good to be true electronic that almost no intervention in the history of medicine with benefits of this size would be true. The same article mentions that the New Britain Record of Medication rejected the article on the grounds that operating system results therefore unexpectedly good, which would lead to the need to review 2 primary data that the journal claims it is unable to do.
After having its reputation exposed to zero scandal of a Surgisphere, it is understandable that a New Great Britain Journal would want to learn to take risks with articles that raise suspicions. Still, the decision’s heuristics seem unjustifiable, as well as the assertion that the world’s largest medical journal perform does not have the capacity to check operating system original perform study data, which Cadegiani claims to have offered to the editor.
That said, the offer doesn’t seem to be valid for the whole world. Even though the preprint states that operating system data is available upon justified request, my request to receive it bumped into the response that operating system authors prefer not to share them at this time with some false availability that echoes the password-protected hyperlink perform egyptian file . When questioned on Twitter, Cadegiani justified the refusal of the non-equity of treatment to different studies, suggesting that the fact that I did not request data from other works put my impartiality in question.
Amazingly, the refusal to provide original data on a study is some common reality in academic science. In the impossibility of accessing them, the belief in the statements of an article is almost always based on the word 2 authors. The words may sound like no one’s, but as a virus-like commercial from a few decades ago said, la guarantee soy yo. What makes the reputation of the speaker count, electronic a lot, to decide what to believe.
With this, an argument ends up migrating to investigative journalism or to seeing that social networks, where virulent advertisement hominem arguments about both operating system sides try to resolve some unsolvable issue while attacking reputations about critical electronic authors. Electronic as in any other topic, each group will end up finding the truth that suits them, leading to a polarization between lay electronic doctors.
The first step towards the solution performs obvious problem operating system anonymized data on a study they should be within reach of anyone who wants to analyze them. Although this data is normally requested by regulatory agencies, electronic that many articles claim that they can be obtained, they are rarely available on the fact that they are.
Even with open data, however, fraud better done perform than a rough copy-paste Egyptian perform can be hard on detecting article. Thus, it is necessary to evolve to auditing systems that allow checking whether what is written in an article reflects reality. In a where millions of therefore secret votes are counted in hours around the world, it shouldn’t be difficult to check whether people who took a drug in a study are alive or dead. Strangely, however, this does not seem to be a priority in academia, which is satisfied with a system based on trust that ends up sowing discord.
Eventually we will know if Cadegiani and his colleagues’ assertions are true proxalutamide was approved by Anvisa for new tests, and the Paraguayan government granted an emergency authorization for the use of the drug. Until then, however, we will spend several months harming thousands of people, either by depriving them of effective treatment or by selling the false hopes and side effects of an ineffective drug and its commercial analogues, which are already being prescribed off. -label in Brazil.
Both alternatives are inadmissible and attest to the failure of academic science to exercise the most basic level of quality control, namely whether published data is true. Something that should be a right for anyone, without the need for passwords, investigations or pleadings to the authors.
Olavo Amaral professor at the Leopoldo de Meis Institute of Medical Biochemistry at UFRJ and coordinator of the Brazilian Reproducibility Initiative.
Subscribe -sena Serrapilheira newsletter to follow more news from the institute and from the Cincia Fundamental blog.