As hard as it is for me to believe, I’ve been writing about homeopathy for more than a decade now. Regular readers, of course, know that homeopathy is quackery, utter pseudoscience based on prescientific vitalism based on two “laws”: the Law of Similars and the Law of Infinitesimals. The former states that, to relieve a symptom, you use a substance that causes that symptom in healthy people. There is, of course, no science or logic to support this as a general principle other than sympathetic magic. The latter Law states that to make a remedy stronger, you must dilute it. That in itself is silly enough and goes against much of what we know about how drugs work in the body, but it gets sillier. Homeopaths often dilute their remedies with serial dilutions of 100-fold each, usually designated as “C,” as in 1C = one 100-fold dilution. Between each dilution, a homeopath will tell you, you must vigorously shake (succuss) it in order to “potentize” it. The silliness gets even sillier when you consider that a typical homeopathic dilution of 30C = a 1 in 1060 dilution. Given that Avogadro’s number is only ~6 c 1023, which means that a 30C dilution is over 1036-fold greater than Avogadro’s number. Basically, any homeopathic remedy
stronger more dilute than 12C or so is very unlikely to have a single molecule of starting compound left. Most homeopathic remedies are water.
That’s why skeptics like myself like to use homeopathy as an example of quackery. Once you explain the laws of homeopathy, it’s very easy for most lay people to understand why homeopathy is quackery and why it basically can’t work. As I like to say, for homeopathy to work, scientists would have to be not just wrong, but spectacularly wrong, about some very fundamental laws of physics and chemistry. Homeopathy also is a very good “teachable” example to illustrate how randomized double-blind clinical trials (RCTs) can go wrong. To put it very simply, using Bayesian considerations, in which the plausibility of a hypothesis and hence its prior probability of producing a positive clinical trials are considered, testing treatments like homeopathy, which has about as low a pre-trial probability or plausibility as I can imagine, can be very good at producing false positive trials. Heck, Steve Novella and I even managed to publish a commentary on this in a scientific journal under the title, Clinical trials of integrative medicine: testing whether magic works?
All of this is just a bit of background I wanted to lay down for those who might be relatively new to this blog before I discuss what I learned about a clinical trial of homeopathic nosodes discussed in a STAT article by Helen Branswell, Should researchers study bunk science? Among respected scientists, a debate ensues. Basically, it’s asking the same question that Steve Novella and I answered two years ago with a resounding “No!” Depressingly, the arguments for testing homeopathic nosodes are basically the same naive or discredited arguments that Steve and I discussed then and have been discussing on and off in our respective blogs for years.
But first, what are homeopathic nosodes? Basically, nosodes are homeopathic remedies prepared from a diseased animal or person and can consist of saliva, pus, urine, blood, or diseased tissue, diluted to various degrees according to the principles of homeopathy. Many homeopaths sell nosodes as being the equivalent of vaccines. Take to a ridiculous extreme, some homeopaths were known to take blood or bodily fluids from Ebola victims to make, in essence, nosodes, an incredibly dangerous thing to do, not just for the patient being possibly exposed to Ebola but for the homeopath stupid enough to handle actual bodily fluids from Ebola victims without proper training and, one worries, proper protection.
The STAT story basically reports on the tension and disagreement between two Canadian academics, neither of whom is a quack or crank and both of whom are well respected, over whether doing a clinical trial of homeopathic nosodes can be justified:
A disagreement between two respected Canadian academics is raising some fundamental questions about when a disputed scientific issue has been studied long enough.
The debate centers on whether it’s still valid — or even ethical — to do research on products called nosodes, which are marketed as homeopathic “vaccines.”
And this is what it’s all about:
Dr. Mark Loeb, an infectious diseases researcher at McMaster University in Hamilton, Ontario, is seeking volunteers for a study that he thinks will show nosodes don’t activate an immune response and therefore cannot protect against diseases.
Tim Caulfield, a professor of health law and policy at the University of Alberta in Edmonton and a zealous debunker of quack science, argues there’s no need to run such a study. Science already knows the answer, said Caulfield, whose most recent book — on the impact of celebrity culture on health — is entitled “Is Gwyneth Paltrow Wrong About Everything?”
“It is com-pleeeeeeeete scientific nonsense,” Caulfield said of homeopathy, drawing out the word for added emphasis.
“There is no need to study it. … I don’t need to run a physics experiment to demonstrate that flying carpets don’t fly.”
Regardless of intent, looks like a respected researcher taking homeopathy seriously. This can only help homeopathy. https://t.co/bbJtU7csyB
— Timothy Caulfield (@CaulfieldTim) October 3, 2016
One of these days Prof. Caulfield. and I need to meet. He’s showing some sweet, sweet Insolence, there, and I approve heartily. What Caulfield is saying is basically what Steve and I wrote two years ago when we compared a trial of this type to a clinical trial of magic. In fact, because homeopathy is basically very similar to the principles of sympathetic magic, I have a hard time thinking of a better example of exactly the sort of phenomenon that Steve and I were lamenting that this very clinical trial by Dr. Loeb. It is the very epitome of what Harriet Hall has labeled “Tooth Fairy science,” defined by her as seeking explanations for things before establishing that those things actually exist and frequently elaborates:
You could measure how much money the Tooth Fairy leaves under the pillow, whether she leaves more cash for the first or last tooth, whether the payoff is greater if you leave the tooth in a plastic baggie versus wrapped in Kleenex. You can get all kinds of good data that is reproducible and statistically significant. Yes, you have learned something. But you haven’t learned what you think you’ve learned, because you haven’t bothered to establish whether the Tooth Fairy really exists.
Here’s what I mean. The clinical trial being run by Dr. Loeb can be found on ClinicalTrials.gov (NCT02825368). It will consist of three groups:
- Controls, who will receive a calcium carbonate oral dose and two sterile saline injections (0.5 ml each, intramuscular) as placebo.
- Active Comparator: Conventional vaccine group. These will receive one dose of Tdap, one dose of MMR, and calcium carbonate pellets as a placebo.
- Experimental: Homeopathic vaccine group. This group will receive Diphtheria (Diphtherinum®), pertussis (Pertussinum®), tetanus (Tetanotxicum®), measles (Morbilinum®) and mumps (Ourlianum®) nosodes, as well as two sterile saline injections (0.5 ml each, intramuscular) as placebo.
Three weeks later, antibody levels for diphtheria, pertussis, tetanus, mumps, and measles will be measured in the subjects. There will be 150 subjects in the study, roughly 50 randomized to each group. Dr. Loeb will be looking at other measures of immunity as well.
Now, there’s nothing wrong with this study, as far as its design goes under the evidence-based medicine (EBM) paradigm. It’s a perfectly serviceable RCT looking at the response of a biomarker (in this case, measures of specific immunity) to an intervention. It’s just that Caulfield is correct. It’s a completely unnecessary clinical trial. Dr. Loeb has clearly fallen victim to methodolatry, which in medicine is the profane worship of the RCT as the only valid method of experimentation. Dr. Loeb states that his hypothesis is that “will be no different than placebo,” which is rather odd and makes me wonder how on earth he got funding for this study, as it’s uncommon for the hypothesis of a clinical trial to be that there will be no difference between the two interventions. After all, the purpose of clinical trials is usually to find a better treatment than standard of care or a treatment that is better than placebo.
Be that as it may, the second question I have is how this study got through the Canadian equivalent of the institutional review board (IRB); that is, the ethics committee overseeing clinical trials. The McMaster University IRB-equivalent truly messed up here when it gave its stamp of approval on this profoundly unethical clinical trial, as described in the STAT article:
Further, Caulfield argued that conducting the study may be unethical. His rationale: Research ethics require that there is what’s known as equipoise — scientific uncertainty — if a question is to be studied. There is no scientific uncertainty about nosodes, he insisted.
Loeb said he could find no previous study looking at whether nosodes triggered an immune response. And his arguments satisfied McMaster’s Institutional Review Board, which approved the study.
Dr. Ross Upshur, an ethicist and a professor at the University of Toronto’s Dalla Lana School of Public Health, disagreed with Caulfield on the issue of the study’s ethics — but agreed with him that regardless of the outcome of the study, it would not change the thinking of people who believe nosodes work.
Would he have approved the study if he had been on the McMaster review board? Upshur, who has conducted research with both Loeb and Caulfield in the past, said he could argue both sides of this debate, but said if pressed to make a call, he probably would have argued that the study wasn’t worth doing.
I argue more strongly than Caulfield: The study is unethical. Period. No doubt about it. Even Dr. Upshur seems to know that the McMaster IRB acceptance of the argument that there is no previous study looking at whether nosodes trigger an immune response was not appropriate, given that basic science alone is enough to show that homeopathic nosodes are utter nonsense. After all, he basically admitted that if he were on the McMaster IRB he would have argued that the study wasn’t worth doing. After all, there is a very small, but not zero, potential for subjects in the control group. Remember, this is a group whose members have, according to the trial protocol, already received their childhood primary DTaP and MMR vaccines and thus don’t need Tdap or MMR vaccine, to suffer a vaccine reaction. I suppose you could finesse their receiving the Tdap as their adult dose, but not the MMR. Then there’s also a small, but not zero, risk from receiving the homeopathic nosodes, which are not, strictly speaking, all water.
Basically, this is the very definition of an unethical clinical trial. There’s no expected benefit to the experimental group or, quite frankly, to the active comparator group, either. The control group is the only one without an expected risk is the control group. Even then, though, one can argue that receiving saline injections could result in a hematoma. That’s harm too. Granted, it’s small and not very likely, but it’s risk nonetheless. So, basically, 150 human research subjects are being subjected to varying degrees of risk for no benefit. Now, don’t get me wrong. Sometimes we do subject clinical trial subjects to low degrees of risk for biomarker studies, but we generally don’t do it when the hypothesis is that there is no difference between control and the experimental intervention. We do it for drugs for which we want to determine if they actually “hit” the molecular target, as determined by seeing the expected changes in biomarkers in either blood or tissue, the idea being to use this information to validate a promising drug and demonstrate its promise by proving that it does what it’s supposed to do on a molecular level before doing a large scale clinical trial.
Here’s a point Steve and I made in our article that’s relevant here:
Even so, the major assumption underlying EBM is that by the time an investigational treatment is ready for RCTs it has passed all preclinical tests and has thus demonstrated biological plausibility. Before, CAM or IM, treatments without biological plausibility and compelling evidence from preclinical studies and pilot clinical trials usually did not reach the stage of RCTs. Indeed, so integral to this process is biological plausibility based on preclinical data that the Declaration of Helsinki  states, ‘medical research involving human subjects must conform to generally accepted scientific principles, be based on a thorough knowledge of the scientific literature, other relevant sources of information, and adequate laboratory and, as appropriate, animal experimentation’.
Homeopathy fails that test. Miserably.
Steve Novella, I, and others like to refer to “science-based medicine” (SBM) instead of EBM, the difference being that SBM takes into account what EBM does not. EBM relegates basic science to the lowest rung on its evidence ladder or pyramid, which is not inappropriate when dealing with clinical trials of treatments that have some scientific plausibility. But when we’re talking about magic like homeopathy, with as close to zero prior probability and plausibility as I can imagine, what RCTs are testing is basically noise and bias in the clinical trial process. I realize that it goes against the grain of scientists like Dr. Loeb and many others, for whom the RCT is the be-all and end-all of clinical research, ever to conclude that there is no need for an RCT. They just can’t accept that basic science alone can be sufficient to rule out the need for an RCT, even in the case of homeopathy, which is to me one of the best examples, if not the best example, of an alternative medicine treatment that is so ridiculously, incredibly implausible that RCTs of homeopathy can pretty much never be justified. Nor is this a justification:
Dr. Peter Palese, a renowned influenza researcher at the Icahn School of Medicine at Mount Sinai in Manhattan, also understands Caulfield’s concern. He calls homeopathy “hogwash.” But Palese noted that Loeb conducts first-rate research, and there has been little of that done on homeopathy.
“If a guy like Mark Loeb does it, it is a service to the rest of us,” he said.
Loeb acknowledged the true believers will not be swayed by the findings of a clinical trial. But he said he hopes to generate evidence that will provoke regulators to take a harder line on nosodes.
As you can see, Dr. Loeb is far from alone. At one level, I can understand the naïveté and idealism of scientists like Dr. Loeb and Palese. What I can’t understand is how, even when someone like Dr. Loeb understands that a negative result will not sway true believers, he persists anyway. After all, given how good he is, what is cost of his doing this clinical trial? Leaving aside the issue that this trial is a complete waste of money, there is also the opportunity cost of having a good scientist devote his time and attention to designing, carrying out, and analyzing the results of a clinical trial like this, particularly when he expects it to be negative anyway and is only doing it in the naively misguided belief that one more negative trial of homeopathy will sway Canadian regulators. What far more worthy scientific questions could Dr Loeb have investigated with the time and resources he is wasting to ask this pointless question about homeopathic nosodes?
One wonders if Dr. Loeb has considered one final question. What will he do if, contrary to his expectations, he finds an increase in the specific immune response to the various antigens he’s testing for in response to the homeopathic nosodes? His hypothesis is that there won’t be such a finding, and he hopes that finding will provide ammunition for Canadian regulatory authorities to ban homeopathic nosodes. It probably won’t, because, as Steve Novella put it, if 200 years of science is not enough to convince someone that homeopathy is bunk, or to convince regulators, then one more study is not going to do it. Also, such a false positive result is not as implausible as one might think; Dr. Loeb should read some of John Ioannidis’ studies, if he doesn’t believe me. As I’ve explained time and time again, when you test treatments with very low pre-test probability and plausibility, false positives are far more common than you might think. Dr. Loeb “knows” that homeopathy is “bunk.” But what will happen if his study is one of those false positives?
As you might expect, I side with Caulfield, although I don’t do it for the main reason he objects, that doing studies of quackery like homeopathy somehow “legitimizes” it (although certainly that’s a significant concern). As a physician, I’m more concerned about the ethics of the clinical trial process and how doing such studies is profoundly unethical, as much as Dr. Loeb might delude himself otherwise. There are more than enough data to conclude conclusively that homeopathic nosodes are complete and utter nonsense. Further clinical trials are unnecessary. The trial that Dr. Loeb is doing is the very epitome of quackademic medicine. It’s also an excellent example of the pernicious influence of quackademic medicine, which can tempt an otherwise excellent researcher to believe that it is worth doing a clinical trial of magic like homeopathic nosodes.