Jablonowski and Hooker conducted their analysis using a dataset provided by the Louisiana Department of Health. It included 1,775 children who died before turning 3 years old between 2013 and 2024 and had a record of being vaccinated. The preprint suggested children who received six recommended vaccinations in the second month of life were more likely to die in their third month compared to those who had not received the vaccinations.

Hooker has had two articles on vaccines retracted. One, published and retracted in 2014, suggested the measles-mumps-rubella (MMR) vaccine increased the risk of autism in African American boys. It was retracted for compromised peer review and concerns about the methods and analysis. The other, published in 2015 and retracted in 2017, focused on alleged conflicts of interest in autism research. It was retracted in part for the authors' failure to disclose conflicts of interest.

Loren Mell, a physician-scientist and clinical investigator at the University of California San Diego who reviewed the paper at our request, said the study has methodological shortcomings.

"There are numerous weaknesses to the scientific conduct of the study," which "eat into the validity of the conclusions," Mell said.

Mell took issue with the way the retrospective study selected the data. "There's a lot of assumptions that are baked into that that you would have to buy to be assured of the validity of their conclusions," he said. "There's too many flaws in that type of design."

"I think that the paper seemed to be more about advancing a predetermined position as opposed to a good faith scientific inquiry," Mell told us, referring to the paper as of "dubious scientific value."

"The way this particular article was written, you can tell right out of the gate that these authors have a vested stake in the outcome of their study and a prior bias," Mell said. "There's nothing quite as useless as a scientist who's already made up their mind about a subject."

After their latest preprint was withdrawn, Jablonowski and Hooker immediately reposted the paper on Zenodo, a repository run by the European Organization for Nuclear Research (CERN).