JAMA Internal Medicine has rejected a petition by a group of clinicians calling for the journal to retract an article on oxycodone therapy for osteoarthritis. The request — which initially included two of the paper’s original authors — stems from 19-year-old court records revealing the sponsor of the study withheld evidence about opioid withdrawal that contradicted the paper’s conclusions.
Published in 2000 under the journal’s previous name, Archives of Internal Medicine, the study of 133 patients concluded that around-the-clock, controlled-release oxycodone was safe and effective for managing osteoarthritis pain and that withdrawal symptoms were rare. In 2007, documents made public during a federal criminal case against Purdue Frederick Company, the maker of OxyContin, showed the company concealed how many patients became dependent on the drug during the trial and omitted withdrawal data. Purdue pleaded guilty in 2007 to fraudulently marketing OxyContin with the intent to mislead, agreeing to pay the government more than $600 million in penalties.
Despite the withdrawal data being made public nearly 20 years ago, the study has remained in the scientific literature without changes, Andrew Kolodny told Retraction Watch. Kolodny is medical director for the Opioid Policy Research Collaborative at Brandeis University and president of Physicians for Responsible Opioid Prescribing. Kolodny has also testified before Congress about the opioid epidemic, assisted federal authorities with investigating the opioid industry and pain organizations, and acted as an expert witness for plaintiffs in litigation against opioid makers.
In late July, Kolodny and seven others, including two of the paper’s authors, called on JAMA Internal Medicine to retract the paper. The withheld data directly contradict the article’s central conclusion that controlled-release oxycodone can be abruptly discontinued without causing withdrawal, according to a July 29 letter sent by the group to Sharon K. Inouye, the editor in chief of the journal.
Kolodny had previously sent solo retraction requests in the fall and winter of 2025, which the journal declined.
The letter states “prompt action by the journal is especially important” because the article continues to influence the medical literature and clinical practice.
“It has been cited hundreds of times and, notably, appears in at least two clinical practice guidelines,” the letter reads. “The continued citation of a publication containing unreliable conclusions—including in contemporary documents intended to guide patient care—underscores the importance of correcting the scientific record.”
The paper has been cited 247 times, according to Clarivate Web of Science.
Jen Zeis, a spokesperson for JAMA Internal Medicine, told us the journal does not plan to retract the article.
“We have received a petition to retract this article and reviewed the request multiple times,” she said in an email. “We saw no evidence of fabrication or falsification in the published article. We have also discussed with two of the coauthors, including one who had cosigned the petition, who do not agree with this request for retraction.”
The article about the study disclosed funding from Purdue and identified coauthor Peter G. Lacouture as a company employee. The authors included two reports of withdrawal symptoms by patients who abruptly stopped taking the medication, but noted that “withdrawal syndrome was not reported as an adverse event for any patient during scheduled respites.”
Kolodny came across the 2007 court proceedings during his research for a federally funded project to identify articles in the literature with misinformation about opioids. According to the settlement documents, Purdue knew eight participants in the study had possible opioid withdrawal symptoms a year before the paper was published.
“Many in the medical community who didn’t pay attention to the criminal case against Purdue have no way of knowing that there is anything wrong with this paper,” Kolodny said. “Clinicians could very well wind up using this paper during their care of patients on oxycodone.”
Kolodny first wrote to JAMA Internal Medicine in October 2025 to request the paper be retracted. In November, an editor replied that the journal does not plan to retract the piece, according to correspondence we’ve seen. Kolodny wrote again in December with additional supporting information and legal documents, but the journal reiterated its decision.
“There is no evidence of fabrication or falsification in this article that warrants retraction, and we do not intend to take further action,” Annette Flanagin, executive managing editor for JAMA and the JAMA Network, wrote in a Dec. 19, 2025 email to Kolodny.
The signatories on the July letter originally included two authors of the paper, Roy Fleischmann, a rheumatologist and professor at University of Texas Southwestern Medical Center, and Barry Bockow, a rheumatologist and associate professor at the University of Washington. Bockow did not return messages seeking comment.
According to correspondence we’ve seen, Fleischmann told Kolodny that during the Purdue investigation, he testified before a grand jury that he was “not provided with all the relevant information” needed to “write the publication correctly.”
Fleischmann expressed his support for retraction in a July 14 email to Kolodny.
But Fleischmann has now changed his mind, telling us he “signed the letter in haste without looking at all the facts.” He has since reviewed the article in depth and said he agrees with the journal that “the manuscript was 100% accurate with no falsified data.”
“The possibility of dependence is discussed in the manuscript,” he told us. “For this reason, I changed my position and do not believe that this article should be retracted.”
The lead author of the study, rheumatologist Sanford H. Roth, died in 2024.
In an August 4 email to Kolodny, Flanagin wrote that “without agreement of all of the surviving coauthors to retract the article, we cannot move forward.”
The paper should have been retracted years ago, but “unfortunately no one has set the record straight,” said Caleb Alexander, an internist and epidemiologist who codirects the University of California, San Francisco-Johns Hopkins University Opioid Industry Documents Archive and a signatory of Kolodny’s letter to JAMA Internal Medicine. “These types of misleading articles have helped to fuel the opioid epidemic.”
The paper illustrates the broader pattern of the opioid industry using scientific evidence to support unsubstantiated claims to inflate the benefits of opioids and minimize the dangers, said Ravi Gupta, an internist and assistant professor at Johns Hopkins University School of Medicine who signed the letter. Gupta and several other signatories authored a 2024 paper about that pattern.
“It is imperative to retract papers that advanced unsubstantiated claims to maintain scientific validity, as well as for accountability and to minimize patient harm,” Gupta told us.
David Juurlink, professor of medicine and pediatrics at the University of Toronto, also signed the letter. He has been central in the investigation into a 20-year-old case report of a baby’s death from opioid poisoning, brought back into the spotlight earlier this year by an article in the New Yorker. He also coauthored an analysis finding a five-sentence letter in the New England Journal of Medicine was heavily cited for stating opioid addiction was rare. The letter prompted the journal to add an editor’s note to the original letter.
Kolodny said he has been surprised and disappointed by the journal’s repeated refusals to retract the paper. When he first encountered the misrepresented findings in the court record, it seemed like “very-low hanging fruit” to get the article retracted, compared with more complex retraction pursuits.
“If a journal is presented with credible information that a paper warrants retraction, it should retract the paper,” he told us. “The burden shouldn’t be on the person identifying the problem to reach out to other experts to get a group letter or to reach out to the co-authors. It should stand on its merit.”
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].
I remember writing a PubPeer entry for this article which doesn’t appear to have been approved- it’s been known for a long time that these conclusions in this paper are heavily overstated and the methods were questionable. Disappointing to see someone come with even more evidence that this paper is questionable and get nowhere.
Facts Only
* A study published in 2000 in the Archives of Internal Medicine examined 133 patients treated with around-the-clock, controlled-release oxycodone for osteoarthritis pain.
* The study concluded that controlled-release oxycodone was safe and effective for managing pain and that withdrawal symptoms were rare.
* In 2007, documents from a federal criminal case against Purdue Frederick Company showed the company concealed data on patient dependence and omitted withdrawal data during trials.
* Purdue pleaded guilty in 2007 to fraudulently marketing OxyContin with the intent to mislead.
* The authors reported two instances of withdrawal symptoms by patients who abruptly stopped medication but noted no withdrawal syndrome was reported during scheduled respites.
* A petition was sent to JAMA Internal Medicine calling for retraction based on the withheld withdrawal data contradicting the article’s conclusion regarding abrupt discontinuation.
* JAMA Internal Medicine declined the retraction request, stating there was no evidence of fabrication or falsification in the published article and noting disagreement from some coauthors.
* The study has been cited 247 times according to Clarivate Web of Science.
Executive Summary
A group of clinicians requested that JAMA Internal Medicine retract an article concerning oxycodone therapy for osteoarthritis due to withheld evidence regarding opioid withdrawal. The study, published in 2000, concluded that controlled-release oxycodone was safe and effective for pain management and that withdrawal symptoms were rare. In 2007, documents related to a federal criminal case against Purdue Frederick Company, the manufacturer of OxyContin, revealed the company concealed data on patient dependence and omitted withdrawal information during the trial. Despite this public information, the study remained in the literature without revision.
The requesting group argued that the withheld withdrawal data contradicted the article's central conclusion that abrupt discontinuation would not cause withdrawal. The journal declined the retraction request, citing no evidence of fabrication or falsification in the published work and noting disagreement from some coauthors. The author, Andrew Kolodny, contended that continued citation of findings with unreliable conclusions influences medical practice.
The decision to retract was complicated by evolving views among the original authors; one author later revised support for the request, stating they believed the manuscript was accurate without falsified data. Despite this, other stakeholders argued for retraction based on the omission of critical withdrawal data, which is framed as necessary for correcting the scientific record and minimizing patient harm.
Full Take
The conflict highlights a tension between the perceived scientific integrity of published literature and the accountability demanded by external legal and public health findings. The situation reveals how institutional gatekeeping—the journal's decision not to retract—can stabilize potentially flawed knowledge within clinical practice, especially when the weight of the evidence shifts from the publication itself to external legal proceedings. The trajectory suggests that scientific consensus can be decoupled from a rigorous application of all available contextual data, particularly when high-stakes industry actions overshadow corrections.
The shift in one author’s position, moving from supporting retraction to defending the manuscript's accuracy after reviewing the context, illustrates how personal conviction interacts with institutional inertia and external pressure. This process suggests that accountability is often contingent on consensus among parties rather than pure adherence to objective data. The pattern observed is an attempt to manage liability by compartmentalizing error, where a small subset of authors or editorial bodies refuse to acknowledge discrepancies found in broader investigations, allowing the literature to persist and influence care despite known flaws.
What are the systemic implications when external investigative findings related to corporate misconduct—which directly impact patient safety—are excluded from the review process? How does the weighting of established scientific citation versus newly revealed legal context shape future accountability mechanisms for medical publications? What role does skepticism play in ensuring that published consensus remains resilient against powerful economic interests?
Sentinel — Human
The text is strongly grounded in specific legal and academic events, exhibiting the complex, layered argumentation characteristic of investigative journalism, though it synthesizes these facts into a narrative structure.