A scientific paper linking an ingredient in vaccines to autism in mice has been discredited and a retraction is imminent, but Canadian researchers say the incident highlights a broader problem: flawed studies can continue to live online, even after a withdrawal notice.
Last month, researchers from the University of British Columbia asked to retract their paper reporting aluminum-triggered immune responses "consistent with those in autism." Editors of the journal that published the peer-reviewed study said they agreed to withdraw after finding "evidence of incorrect data."
"It attacks the credibility of science ... garbage science has an impact," said Jim Woodgett, investigator and director of research at the Lunenfeld-Tanenbaum Research Institute in Toronto, said of the soon-to-be-retracted paper.
Elsevier is one of the world's largest scientific publishing companies, overseeing the Journal of Inorganic Biochemistry. In a statement, it said the data in two of the paper's figures "are incorrectly presented."
The publisher also apologized to readers that those issues "were not detected" before publication.
Elsevier's policy on retraction says the electronic version of the journal will still link to the original paper, although it will be preceded by a retraction notice signed by the paper's authors and the journal's editors.
The study would only be completely stripped if it was defamatory, the subject of a court order, or deemed a "serious health risk" if someone were to act upon it.
Otherwise, the public will still be able to read and download the paper in its original form after clicking past the retraction notice — which Woodgett said is a problem.
"[Retraction] should be instantaneous," he said.
"The general public isn't typically well-versed in identifying what is scientific garbage, or pseudoscience, from what is real," he explained. "So, the damage is done in terms of, this paper is out there — people who are not well-versed in science can easily get misled by it."
Dr. Michael Gardam, an associate professor of medicine and infectious disease at the University of Toronto, said much of the same.
"There's actual harm that's happening as a result of these things," he said. "There will be other people that will glom onto this that, no matter the retraction, no matter what."
Woodgett, who started his lab 30 years ago, said retraction protocol can vary from journal to journal. The researcher said some publishers will "quietly and discreetly" pull articles down, without offering an explanation as to why. Other times, he said, they're simply slow to remove the publication — something he said "doesn't do anybody any good."
"Somebody may have read the paper, quoted it, and then three or four months later, it gets retracted … but it's still out there," Woodgett said. "If a journal says they're going to retract, then it should disappear … Unfortunately, that's not the case."
A digital version of the UBC study was still available to download as a PDF as of Friday afternoon, nearly a month the retraction was agreed upon.
Asked if he was concerned about the spread of allegedly falsified data, co-author Shaw said readers need to remember "this paper was done on mice" and take it with a grain of salt.
"A lot of people that have questions about vaccine safety were making more of this paper than was warranted," he told CBC News. "We try to caution people … don't make more of it than it is, because this is a model system where this data may or may not apply to humans."
UBC's vice president of research says she can't comment on specific cases, but that the university can investigate allegations of scholarly misconduct if they're warranted. Gail Murphy said faculty members can be fired if misconduct is proven.
The university has not published the paper on its own platforms.
Shaw's co-author, Lucija Tomljenovic, said she "had nothing to do either with collecting or analyzing any of the actual data" but that she agreed to the retraction.