Closer Look
After reviewers suggested rejecting it, a paper won acclaim and grant money — all before retraction
Adobe, Alex Hogan/STAT
It was a bold claim on a life-and-death subject. A 2017 paper published in Nature Human Behaviour claimed a combination of brain scans and machine learning algorithms could identify people at risk for suicide with 91% accuracy. Based on the work, two of its authors a won a five-year, $3.8 million grant from the NIH to study the method further. But the paper has been retracted, six years after the initial peer reviewers raised questions about the manuscript and suggested rejecting it.
The authors admitted the work was flawed and retracted their article in April, but not before it had been cited 134 times in the scientific literature and ranked in the top 5% of research tracked by Altmetric. All this could have been avoided if the journal had followed the advice of its own reviewers, according to records of the peer review process obtained by Retraction Watch, Ellie Kincaid writes. Read more.
coronavirus
Study finds a link between taking a diabetes drug early in infection and lower risk of long Covid later
Long Covid is a baffling condition with no solution. A new study offers some preliminary hope: People who took the diabetes drug metformin within three days of testing positive for the virus were 41.3% less likely to develop long Covid than those who didn't get the drug. Researchers believe theirs is the first randomized trial to show a treatment that works when given at the time of acute infection, they write in The Lancet: Infectious Diseases.
The absolute risk difference shows 6.3% of people (35 of 564) who took metformin developed long Covid versus 10.4% who didn't (58 of 562) take the drug. There are caveats: The 1,126 trial participants all had overweight or obesity. "An alternative explanation … may be that [metformin] is addressing a tendency to hyperglycemia, which is considered a risk factor for adverse outcomes following infection with Covid," Stephen Lawrence of the University of Warwick, who was not involved in the study, said in a statement.
global health
Newborns are dying of sepsis after antibiotic resistance, especially in lower-income countries
Antibiotic resistance is a worldwide problem whose toll in newborn babies is drawing alarm. A study in PLOS Medicine tracking sepsis in 19 hospitals in 11 countries found that many newborns are dying because antibiotics to treat the bloodstream infection are losing their effectiveness. Almost 1 in 5 infants died of sepsis during the study, conducted from 2018 to 2020. Most of the deaths occurred in low- and middle-income countries, where sepsis has become resistant to antibiotics. The mortality rate among countries ranged from 1.6% to 27.3%, with much higher rates where incomes are lower.
Another worrying statistic: Hospitals in the study collectively used more than 200 different antibiotic combinations, frequently switching when treatments met high resistance. Antibiotics classified by the WHO as "last-line" were often used, despite being recommended only for specific, limited indications to preserve their power. But they were often the only antibiotics available, the authors note.
No comments