A journal that last year corrected a paper claiming to show a homeopathic intervention improved quality of life and survival for people with advanced lung cancer has now retracted the article after the Committee on Publication Ethics got involved in the case.
The extensive correction and an accompanying editorial, published in September 2024 in The Oncologist, came two years after the Austrian Agency for Scientific Integrity asked the journal to retract the article due to concerns about manipulated data, we reported at the time.
The retraction notice, published November 24, acknowledged the watchdog agency’s retraction request. It also noted the previous corrections and expression of concern for the article, which originally appeared in October 2020.
“Subsequent to the two corrections, concerns have continued to be raised about the study,” the notice states. “In light of this continued uncertainty and the issues previously covered in the corrections, the journal no longer has confidence in the results and conclusions reported in the article and has decided to retract.”
Susan Bates, the journal’s editor-in-chief, told Retraction Watch she had no “further detail to provide” beyond what the retraction notice provides. As the journal changed publishers after the article first appeared, the version on the original publisher’s website remains unmarked.
Concerns about the study go back to shortly after the article was published. Norbert Aust, a retired mechanical engineer who started the Homeopathy Information Network in Germany, and Viktor Weisshäupl, a retired anesthesiologist now working on the Initiative for Scientific Medicine in Austria, first raised concerns in 2021 about how the study was conducted and reported, which led to the official investigation. They told us a formal complaint to COPE — submitted last November and signed by Harald Sitte, a pharmacologist at the Medical University of Vienna — was “the turning point” in getting the journal to act.
In February, COPE directed The Oncologist to engage with the Austrian Agency for Scientific Integrity and the Medical University of Vienna, where the lead author of the homeopathy study was affiliated. “This marked the first instance in which the journal demonstrated a willingness to participate in a professional dialogue with its critics,” Aust, Weisshäupl, and Sitte told us in a joint statement. Representatives of the journal, the university, and the watchdog agency held a video conference in June, the three critics said.
Following that meeting, Bates emailed Michael Frass, the lead author of the paper, a homeopathic practitioner who was working at the Medical University of Vienna, at the time the work was published. In the email dated July 29, which Frass shared with us and described on his website, Bates said the meeting had raised three questions for her, which she asked Frass to respond to by August 1. If he did not respond in time or could not “provide satisfactory explanations,” she wrote, “I now feel I will need to retract the paper.”
Bates asked Frass to address whether the randomized controlled trial included any participants who were originally enrolled in his single-arm study and to disclose the specific homeopathic compounds used in the study, along with their potencies and dosing regimens. Bates also wanted to know if Frass was recommending or prescribing the same compounds to patients in his private practice at the time the journal published the study.
Frass told Bates the randomized controlled trial did not include any participants who were originally enrolled in the single-arm trial, but also told us he did not understand what she meant, “as I always compare 2 groups as a clinician.”
In his response, Frass directed Bates to two tables in the paper for the homeopathic products used and their potencies. He also wrote, “It would go beyond the scope of this article if I were to list all the compounds of the medicines used.”
As for his private practice at the time the article was published, “of course” he was recommending and prescribing the same compounds to his patients, Frass wrote to Bates.
Bates did not respond, Frass said. On October 24, he received an email informing him of the decision to retract the paper, with a draft notice.
The draft notice included two specific reasons for retracting the article that weren’t in the final notice. First, because the homeopathic regimens were tailored to each patient and the paper did not report changes in dosing or potency made throughout the trial, the study’s findings would be “difficult to reproduce.” Second, the draft notice stated, Frass should have disclosed his homeopathic practice as a conflict of interest, as “the same homeopathic medicinal products were being marketed and prescribed” in his clinic during the trial.
In a response, Frass argued the trial, like others of homeopathy, was designed to assess “the technique itself, not efficacy of the individual medicines.” He maintained a trained homeopathic practitioner could replicate the study with the information provided. “Given that our study has replicated an existing prescribing technique, the argument that it could not be replicated again simply does not hold,” he wrote.
The point regarding his practice as an undisclosed conflict of interest was “unexpected” and “deeply confusing,” Frass wrote to the journal. “No marketing of medicines occurred and there was no conflict of interest of any kind.” Neither he nor any of his coauthors received any funding from the pharmacy that supplied the homeopathic preparations, he told us.
In a statement of disagreement with the retraction he also submitted, Frass maintained “the study was conducted ethically and rigorously, and the data presented are valid and accurately reported as already confirmed before.”
“We are surprised that a second attack was possible after the validity of the data had been confirmed” in the journal’s previous investigation, Frass told us. “In court proceedings, the following applies: if a defendant has been acquitted, the case cannot be reopened.”
Frass and eight other authors disagreed with the retraction, according to the notice. One author agreed, and the remaining five did not comment, according to the notice.
The retraction published in November “is a long-overdue and necessary step to uphold the integrity of scientific literature,” Aust, Weisshäupl, and Sitte said in their statement. Even still, they called the case “a textbook example of how journals should not deal with criticism from readers.”
In May 2021, Aust, Weisshäupl, and clinical oncologist Jutta Hübner of Jena University Hospital in Germany submitted a letter to the editor of The Oncologist detailing “some serious concerns that the results are a product of strong biases arising from modifications of the study parameters.” The journal never published the letter or responded to the concerns in it, according to the critics’ statement.
Last September, the journal invited Sitte to submit a commentary on the matter, which he did in January. The journal sent it out for review but never shared the reviews with Sitte, the critics said. On December 4, the journal rejected the commentary “since the paper has already been retracted,” Weisshäupl told us. He was “not surprised at all” by the journal’s decision, he said.
Even the corrections didn’t really address the issues the Austrian Agency for Scientific Integrity or the critics raised, their statement to us said. “We got the impression that the journal simply published the lead author’s explanations uncritically.”
The Oncologist, a publication of the Society for Translational Oncology, switched publishers from Wiley to Oxford University Press in January 2022, after Frass’ article and its first correction were published. The subsequent expression of concern and retraction notice have not been added to Wiley’s original webpage for the article.
When we previously asked Wiley if the publisher would update the page, a spokesperson for the publisher said their current policy for journals that had moved to another publisher was to “rely on the current publisher to carry any post-publication amendments,” but the company was planning to review the process.
After the retraction, we again asked Wiley about updating the original page. A spokesperson said the company now follows best practices from the National Information Standards Organisation, which recommend a journal’s current publisher work with its previous publisher to clearly label the article’s status, “especially in instances where the content may be published on both the previous and new publisher sites.” (Disclosure: Our Ivan Oransky was on the working group for those guidelines.)
Wiley received an update about the article from Oxford University Press on December 1, the spokesperson said, and will post a link to the retraction. At the time of posting, Wiley’s version of the article did not include the retraction notice.
Update, Dec. 16, 2025: This story previously said the article on Wiley’s website did not link to the editorial or previous corrections; it does. We have corrected this above. Wiley also confirmed the publisher will post a link to the retraction.
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].
Almost all homeopathic medics have been diluted to the point that they are placebos. So even a double-blind test is testing one placebo against another. If the cohort is large enough to dismiss ties, the homeopathic placebo should win half the time. The homeopaths then can tout their “successes.”
FYI: I do not know if it is still posted, but the center for quackery NCCAM posted an Orwellian statement that homeopathic treatment often beat placebos.
Homeopathy is to active ingredients as La Croix is to fruit.
100X in homeopathy is diluted to 1 in 10^100. There are only 10^81 atoms in the observable universe. La Croix is not that dluted.
This means that the retired engineer and his friends (including Jütta) are all part of an anti-homeopathy lobby that was unable to invalidate the data. And since they couldn’t do that, they put pressure on a professional at the University of Vienna. It’s ridiculous to retract an article just because the lead author prescribes individualized homeopathic medicines that are not patented. This once again shows how corruption prevails and the desperation of the so-called skeptics.
“part of an anti-homeopathy lobby”.
That would be the science community.
I recommend that you read the letter to the editor sent to the journal four years ago, which is linked in the article. Date falsification in the study protocol, which was only uploaded after the study had ended, and changes to the exclusion criteria from 1 to about 20 after study completion are incompatible with good scientific practice. And homeopathically diluted X-rays or table salt as a remedy against advanced lung cancer in the list of homeopathic remedies used in the study speaks for itself.
I have seen homeopathic “research” in action. Runaway grad student got enthralled by it. I attended meeting with her, the prof, and the prof’s toady. They were trying to “magnetize” water. The idea that the sample and the control should be treated the same way seemed new to them. Any difference between the control and the sample was a positive result. Later with zero effect predicted by modern science, proper controls, and unknown sign, 4-2 for one sign was a “positive” result. 4-2 and 2-4 come up more often together than 3-3 tie.
Outlandish conclusion require very heavy evidence. That certainly applies to homeopathy.
As I already stated, homeopathy beats the official placebo half the time because both are equally worthless.
Did you ever try to put pressure on a university? The biggest one in the country? Or on one of its officials?
How would you set about?
The only thing that would put pressure on any research institute would be the reduction of the funding. How would you achieve this?
The authors response to the retraction seems to be “We did what we wanted, so you shouldn’t retract it”.
There was no explanation or validation. No surprise that it was retracted – it should never have made it into publication in the first place.
Fractional amplitutdes are the next big thing in quantum computing, anathema in medicine.
As homeopathy has no scientific justification, it should not be tested in clinical trials. When the trials do show significant differences between the treatments, there are only three possible reasons, which are Type 1 error, an error in the conduct of the trial or fraud. Type 1 errors happen, I’ve seen the pointless comparisons at baseline of patient characteristics for randomised trials yield p-values of 0.001. The randomisation hasn’t failed it is just chance, as the distribution of p-values under the null is either uniform or approximately uniform. It would be interesting to check the clinical trials registry in the period after registration was required for homeopathic trials that have simply disappeared.
The registry should NOT allow trials to be taken down! A vast number of trials taken down and the ones with positive results combined indicates p-hacking.
Also one can always get very nice p values by combining data after the fact in an arcane way.
I meant that they were never reported, and I would expect that nobody is going to say what the results are. In other words, they have disappeared from the scientific record although they are in the trials registry. There are some nice tests for meta-analysis that detect missing studies. It would be worthwhile applying them to the accumulation of studies on homeopathy.
Correct. Meta-analysis does catch missing or culled data. I do not expect that homeopathic “researchers” would release their culled data. I just was trying to say that the registry could “lock” postings so they cannot “disappear.” They registry could also require more detail in posting to prevent target switching after the data were already collected. Stat methods for testing multiple hypotheses at the same time are well known.