I identify as a recovering epidemiologist. Unfortunately for this study, I still remember how the scientific method works.
Last month, a client texted me to ask whether she should “go on the ivermectin protocol” for her joint pain. When I asked where she’d heard ivermectin could help with that, she told me everyone in her homeschool group was talking about the study showing an “84% benefit” in cancer patients. And if it could do that for cancer, she figured, surely it could help with inflammation too.
A number - 84 percent - detached from the thing it was supposedly measuring, floating around group chats and inspiring people to find more reasons to put pharmaceutical products - each with their own risks - into their bodies.
A group of authors associated with The Wellness Company published a study about the association of a hodge-podge of custom ivermectin-mebendazole protocols with a hodge-podge of individual self-reported cancer outcomes. In a peer reviewed journal, Anticancer Research.
All of us in the medical freedom movement - physicians, epidemiologists, naturopaths, homeopaths, activists, parents - have built reputations on distrusting sloppy institutional science.
We’ve endlessly mocked The Authorities for cutting scientific corners. Which means we’ll rightly be held to even higher standards than ever when we play in their sandbox.
The Problem Is, I Like These People
This isn’t a hit piece on the people involved. The author list includes physicians and researchers I profoundly admire - people who have put their licenses, their reputations, and their livelihoods on the line to say true things that were unpopular. Which is why this particular study does more damage to our movement than a hundred sloppy papers out of pharma-funded academic mills ever could. When The Authorities publish garbage, it’s business as usual. When our team does it, it’s a gift to everyone who wants more excuses to dismiss us and our credibility.
About That “84%”
The paper is referred to by The Wellness Company’s press release as a “groundbreaking… study.” And “The Largest Peer-Reviewed Study on Off-label Ivermectin + Mebendazole in Cancer Treatment Published in Anticancer Research Journal.” That sounds like legit science.
But in the actual paper, the published version in the journal, the authors more accurately call it a “prospective observational cohort.” And their findings? “The first real-world therapeutic signal.” “Hypothesis-generating” and calls for randomized clinical trials (RCTs). Of course, only a well-designed, large RCT can actually demonstrate clinical effectiveness of a therapeutic protocol when compared to current standard of care or placebo. The authors acknowledge that.
Not all observational studies are inherently terrible. (Just most.) But a useful one has to make a serious effort to identify and account for other factors that could explain the results - the pesky things epidemiologists call “confounders.” Which our intrepid author list didn’t do.
Here’s what actually happened, according to the paper: 197 patients who were already customers of The Wellness Company, already taking the compounded ivermectin-mebendazole capsules they’d bought from TWC’s telehealth platform, were invited to fill out a digital survey about how they were doing. Of those 197, only 122 (61.9%) completed the six-month follow-up survey. The other 75 - about 38% of the starting group - didn’t come back to complete the next survey. We don’t know what happened to any of them - whether they got better and stopped caring, got worse and stopped answering emails, or died. The paper doesn’t say, and by its own design, it can’t.
Of the 122 who stuck around, 84.4% reported “clinical benefit” - a composite the authors call the Clinical Benefit Ratio, combining self-reported “no evidence of disease,” tumor shrinkage, and stable disease into one headline figure. That’s the number that made it into every press release and Telegram forward. Remember - the inputs (cancer + drug combo) and outcomes were all self-reported. As in, no scans, no labs and no pathology. The paper said the patients started with “confirmed cancer” but no methods for confirming that cancer except for asking the patient are described. The dose wasn’t standardized either - patients and prescribers picked how many capsules to take, “most commonly one or two,” and reordered the drugs “as desired.” And during the six months being measured, 27.9% of participants were also on chemotherapy, 21.3% were getting radiation, 19.7% had surgery, and roughly half were taking other supplements while over a third had changed their diet. Forty percent of the patients were taking “other,” unspecified treatments.
If they had surgery, that likely meant that a tumor was removed or as much of it removed as the surgeon could get. Which means that the patient would, of course, be able to report “tumor shrinkage” on the basis of surgery alone. But these patients appear to have been included in the 84 percent reporting positive results associated with the ivermectin/mebendazole combo. The drugs are potentially getting credit here for cutting the tumor out of the bodies, even though it was a surgeon who did that.
There’s no version of the scientific method, ancient or modern, mainstream or dissident, in which you can hand people with an un-verified disease state an unstandardized dose of two drugs, let a bunch of them also do chemo or radiation or surgery or both, lose track of 38% of your subjects, ask the survivors to self-report their own cancer status with no verification, and then report a single clean percentage as if you’ve isolated what the drugs did. What got measured here was noise. What got reported was signal.
The Authors Know All This
To be completely fair: the paper’s own discussion section says almost all of this itself. Quoting directly: “Limitations are inherent to the observational, self-reported nature of the data. Outcomes were not clinically adjudicated or radiographically confirmed, no control group was available, and confounding from concurrent conventional therapies, supplements, and lifestyle changes cannot be excluded.” They go on to say the results “should therefore be regarded as hypothesis-generating” and call for “rigorous randomized, double blind, placebo-controlled trials.” If the press release had led with that paragraph instead of giant-font “84.4% reported clinical benefit,” we wouldn’t be having this conversation, and my client wouldn’t have texted me about her knees.
But TWC’s media team skipped straight past that paragraph, and so did every outlet that ran with the number. A Clinical Benefit Ratio of 84.4%, printed at the top of an abstract under the banner “Anti-Cancer Research Journal,” does a huge amount of work that the caveats three pages later can’t undo.
And, by the way, the drug everyone on Telegram thinks this study is about is fenbendazole, the veterinary drug from the Joe Tippens story that’s been floating around the health-freedom world for years. This paper tested mebendazole instead - a related but different, human-approved anthelmintic.
Has Hell Frozen Over?
You know it’s bad when two groups who agree on almost nothing arrive at identical take-downs within weeks of each other.
On one end, Dr. Tom Cowan - about as far from a status-quo defender as it gets - devoted a full podcast to walking through this study line by line, and landed on nearly every problem I just listed: the self-reported diagnoses, the unexplained 38 percent dropout, the absence of any control group, the patient-chosen dosing, the fact that many participants were also getting conventional treatment at the same time. His sharpest point, and I think it’s the right one: if you would never accept a patient’s unverified self-report of “I have diabetes” as grounds for publicly claiming your “banana diet” (or whatever) cures diabetes, you don’t get to accept unverified self-reported cancer status as grounds for publicly claiming a drug combination could be treating cancer.
On the other end, TechARP - a status-quo fact-checking outfit whose entire mission is to make all of us in the med-skeptic crowd look like dummies - ran its own line-by-line breakdown and rated the 84% claim “False.” Its critique overlapped with Cowan’s almost point for point: unverified self-report, no control group, a cohort that was pre-selected for having already bought and started the protocol (which makes the placebo effect more likely and selection bias enormous), and confounding from concurrent chemo, radiation, and surgery that the study design has no way to separate out. TechARP added one thing Cowan didn’t emphasize: most if not all every single one of the twelve listed authors is affiliated with, or draws salary from, The Wellness Company - the same company that sells the compounded ivermectin-mebendazole protocol being evaluated.
But I actually appreciate the honest conflict-of-interest disclosure - a helluva lot better than we get from sneaky pharma-funded authors in most peer-reviewed articles.
Still, when Dr. Cowan and a Fauci-worshiping, COVIDian fact-checking site independently produce the same list of methodological failures, it gets my attention. We can’t wave off the critics with the usual “well, of course the establishment hates it” or “of course the fringe are quacks.”
Pay no attention to those peer reviewers behind the curtain.
The journal’s own walk-back got lost in the outrage cycle over the summer.
Anticancer Research issued an Expression of Concern about the paper in June, and it initially raised questions that sounded ethical in nature - institutional review board approval, verification of participants’ baseline diagnoses, medical documentation. That’s the kind of language that implies someone did something dishonest.
Then, in a September update, the editorial board walked that specific framing back. They withdrew the audit request tied to IRB approval and diagnosis verification, and they said so explicitly: “No allegation of fabrication, falsification, plagiarism, image manipulation, or other research misconduct is made against any author.”
All of this ethical rigamarole was really just fig-leafing for the fact that the journal’s so-called “peer-review” process was a complete fail on the merits of the study. By dancing around the fact that they investigated and found no ethical failings - the authors didn’t lie or cheat - the journal delayed having to confront the much more embarrassing problem: its peer reviewers apparently thought this was good enough science to publish.
And when journals publish that kind of paper, the outcome is predictable. It produces the “Study finds…” headlines, social media frenzy and the patients like my client who are now more likely to purchase drugs with a known risk profile, from the company the authors own or affiliate with, with an unknown benefit for a disease they don’t even have.
For the record, I got nothin’ against Ivermectin
You could make an argument that the healers and researchers who are working outside the conventional sick-care system shouldn’t be trying to prove ourselves inside the academic literature clown car at all.
After all, one of the most foundational principles of my fields of naturopathy and homeopathy is bioindividuality - the idea that each person is an unrepeatable ecosystem responding to a custom set of material, energetic, and immaterial inputs all at once. That makes it very difficult to group patients with similar symptoms into a treatment group that receives a new protocol or into a control group that receives the standard of care or placebo.
I get it. Why play by the RCT rules when they assume a premise we reject?
But it’s not actually the scientific method we reject. Even without an RCT, as a recovering epidemiologist, I don’t dismiss all observational studies outright. There can be well-designed ones that do provide useful information and “hypothesis generation,” even if they don’t demonstrate causation.
I’m not asking for a Vinay-Prasad-approved double-blind trial before anyone’s allowed to make a clinical recommendation - plenty of good, honest care happens without one, and I practice plenty of it myself. But when you decide to publish something with “Anticancer Research” in the title and a quantitative finding in the headline, you hold yourself to a standard that would survive being read by someone who wants you to fail (which is most of the white coat religious cult). This one wouldn’t have survived a first-year biostatistics section, and, ironically, the people I would have considered the most qualified to have caught that before publication were the twelve authors on the byline.
Sources:
Original study: Real-world Clinical Outcomes of Ivermectin and Mebendazole in Cancer Patients (Anticancer Research, Vol. 46, Issue 6)
Editorial Board update: Update to the Expression of Concern (Anticancer Research, September 1, 2026)
TechARP fact-check: Ivermectin Cancer Study: 84% Fact-Checked
Dr. Tom Cowan podcast discussion (September 2, 2026)