Health Record: How many scientific frauds are there in health?

In 2006, American researchers published in the journal nature An important scientific article about Alzheimer's disease that links this disease to an abnormal protein, beta-amyloid. This article has since been cited as a reference in over 2,500 other publications – a huge number and a testament to the impact it has had. Except that it is on the verge of disappearing from the scientific literature, the authors have admitted that it contains manipulated images.

This phenomenon, that is, retraction of articles, is becoming more and more frequent, and is potentially fraught with consequences: publishing articles that then have to be retracted contributes to directing research in the wrong direction and can have repercussions for health-related decisions. One of the worst examples in this field is what was published in 1998 by The scalpelfrom a fraudulent article linking the vaccine to autism, retracted in 2010, which fueled the anti-vaccine movement whose effects are still being felt.

last year, nature king estimated The retraction rate for peer-reviewed journals is 0.2%, or 1 article out of every 500 articles published, across all scientific fields and in all languages. The Kingdom of Saudi Arabia, Pakistan, Russia, and China topped the list with the highest percentage of declines.

In the news Stady, the researchers looked at biomedical publications by at least one European author. Between 2000 and 2021, they calculated, the retraction rate for such publications quadrupled, from 11 per 100,000 articles published to 45 per 100,000. They then analyzed the 2,069 retractions in detail to classify the reasons. In a third of the cases, these were honest errors due to which the researchers themselves requested to retract their article. But two out of three times, the authors committed at least one “scientific misconduct,” such as fabricating data, manipulating images, making false statements about conflicts of interest, or plagiarism. Duplication—publishing the same article in more than one journal—is the most common case of misconduct, accounting for 22% of cases.

Copying an article, even if reprehensible, is still not as serious as fabricating data or hiding a conflict of interest! In the face of such outcomes, we can therefore see the glass as half full or as half empty. There is also a lot of work to be done so that we can better gauge the confidence we can have in published studies. Car on none of the above if the number of abstracts enlarges the part of the articles that have been published – notation with the outils of comparison of the images that are not present in the previous years – or part of the scientific advances, etc. In any way.

data

52,500

That, at least, is the number of premature deaths that might have been caused by fine particles emitted by California wildfires from 2008 to 2018, according to modeling conducted by public health researchers in Los Angeles. Scientists have succeeded in distinguishing the impact of the fires from other sources of PM2.5 particles (less than 2.5 micrometers in diameter), such as car pollution, using precise geolocation data. in PublishingThey also estimate that the health impacts from wildfires represent an economic burden of at least $432 billion (in 2015 US dollars) over these 11 years. According to the researchers, these numbers show that wildfires have more serious long-term health effects than previously thought.

If you liked this article, why not subscribe to our health newsletter? Every Tuesday, you'll first read the clear, detailed and always rigorous explanations from our team of journalists and health professionals. Simply enter your email address below. 👇