Around 400 U.S. patents contain at least one citation to a retracted scientific paper, a new study has found.
The analysis, posted last month to the preprint server SSRN, analyzed granted patents in the United States in PatentsView, a database developed by the United States Patent and Trademark Office (USPTO) using artificial intelligence algorithms, and the Retraction Watch Database.
Of the more than 13 million patents in the USPTO database, 136,162 contained citations to at least one scientific paper. Of these, 401 patents cited retracted papers: 195 cited the retracted paper before it was filed, 92 cited papers retracted while being examined, and 114 cited a paper that was retracted after the patent was granted, the study revealed.
“Science has correction mechanisms, but correction signals often do not travel into patent records,” Hana Kim, of the Korea Advanced Institute of Science and Technology in Daejeon, South Korea, who conducted the study, told Retraction Watch in an interview.
Kim said citing a retracted paper shouldn’t automatically invalidate a patent but argued patent offices and information platforms must be able to identify retracted papers easily. “That is a fixable infrastructure problem,” she said, noting that patent systems should be linked to metadata from the Retraction Watch Database and Crossref. Kim said patents that cite retracted papers should be corrected.
Of the 401 patents that cite retracted papers, Kim conducted a further audit of 39 high-impact filings — defined by how many citations the patent itself has received, among other reasons. Nearly half (19) relied substantially on the retracted papers they cited.
Those 39 patents include one on DNA microtechnologies and nanotechnologies, one on a novel way of engineering transgenic animal models, and one for delivering therapeutic proteins into cells using gene editing techniques. We reached out to the lead scientists named on all three patents to find out why their patents cite retracted papers or if they plan to do anything about it, but didn’t hear back.
Janet Freilich, a law professor at Boston University who was not involved with the new study but has written about inaccurate information in patents, agreed with Kim. “There are a lot of inaccuracies that are permissible in a patent that would not be permissible in a paper,” she told us. “The standards are simply different.”
“It’s an embarrassment that the patent office doesn’t pay attention and then act upon information about scientific papers being tainted by retraction,” said Michael Meurer, a law professor also based at Boston University who has written a book on failures in the patent system.
Freilich said she thinks the conclusions of Kim’s study are likely to be “broadly correct.” Patents should recognize when papers supporting them have been pulled, she said, because one purpose of issuing patents is to disseminate scientific information.
“The reason that many countries make patents public is because the patent system is supposed to be a repository of cutting-edge information that other scientists can then learn from and build on,” Freilich said. “That works a lot better if there’s a process for identifying and dealing with erroneous information, which otherwise will propagate and confuse readers.”
Rules for amending patents when mistakes or errors are found differ in different jurisdictions, noted Freilich, who has previously called for new mechanisms to retract patents. The United States has no process to formally retract patents, she said, but if a patent contains inaccurate information, a judge may choose to invalidate it in court, which would mean it can’t be enforced against other patents.
“That would happen if the informational problems are severe enough that the patent is then deemed not to meet the required disclosure standards but those required legal standards don’t fully match up with the scientific notion of retraction,” Frelich said. “That doesn’t really do anything for the information because the information is still out there.”
Although in some cases a certificate of correction can be appended to the patent, Freilich said, it does not change the text of the patent itself and is usually not easy to access. “It’s not as good as a journal that will put the red line through information,” she said.
Certificates of correction are also not common, Freilich said, and often don’t correct the record for initial patent filings, so earlier, incorrect versions could still be accessible. “It ought to be a more commonly used mechanism,” she said. “The patent office is not used to the idea of these kinds of corrections.”
But Freilich said “no new matter” is a long-standing rule in U.S. patent law that forbids adding new material to a patent that wasn’t in the originally filed document. “The concern is that you would try to broaden your patent or strengthen your patent in ways that would not be fair to either the public or other inventors,” she said.
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on X or Bluesky, like us on Facebook, follow us on LinkedIn, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].