The Tooth Fairy
Another venerable scientific journal has fallen prey to “alternative” medicine research. The Journal of the American Dental Association (JADA) has joined the New England Journal of Medicine and Annals of Internal Medicine, among others, with its publication of “A pilot study of a chiropractic intervention for management of chronic myofascial temporomandibular disorder,” in the October, 2013, issue.
And what “chiropractic intervention” is this? The Activator Method Chiropractic Technique (AMCT) as taught by Activators Methods International (AMI), which also conveniently sells its own line of Activator Instruments. (Practice-building seminar DVDs sold separately.)
What is this Activator Method? In short, it is a method of detecting and correcting subluxations (the chiropractic version, not the medical one). Thus, the underlying premise of this study was that subluxations could cause myofascial temporomandibular disorder (TMD) pain and that finding and vanquishing these subluxations could bring about pain relief.
Now, as regular readers of this blog know, vertebral subluxations are a figment of the chiropractic imagination. They do not exist, as we have pointed out many, many, many times. And here is where alternative medicine has it all over conventional medicine and dentistry. In alternative medicine, once you invent a non-existent condition you can proceed to invent all sorts of diagnostic techniques and remedies to treat patients with this condition. The AMCT is a perfect example of this. Conventional medicine and dentistry, on the other hand, are saddled with the scientific method, biological plausibility, having to discard therapies when studies no longer support their use, and all sorts of other persnickety limitations.
The Activator Method Chiropractic Technique is a complete diagnostic and treatment package. The diagnostic method is laughably nonsensical. Here, we turn for an explanation to SBM contributor and chiropractor Sam Homola, writing for Chirobase:
The AM system is based on a concept of “pelvic deficiency (P.D.)”—also called “functional short leg”—which proponents define as an “apparent” difference in length, not an anatomical difference. To determine where the alleged problem is located, the practitioner holds the patient’s feet in various prescribed ways while the patient lies facedown on an examining table. Other parts of the body may be tested in various other ways.
AMI sells these shoes for patients to wear during their leg length analysis to “ensure accuracy.”
Leg-length Analysis Shoes
Now, to the treatment itself, which is equally nonsensical “chiropractic gimmickry,” again quoting from Chirobase:
If any inequality or “imbalance” is found, the practitioner taps various points along the spine, pelvis, and/or elsewhere with a handheld, spring-loaded mallet [called an Activator] until the legs appear to be equal in length . . . Slight variations of hip position or normal spinal muscle tension are probably responsible for the “imbalances.”
Ironically, the Activator design was derived from a dental impactor. Here’s the Activator IV, used in the JADA study:
The Activator
I’ve used a similar Activator, to great comic effect, in my talks to medical students on complementary and alternative medicine. They seem to enjoy passing it around and thumping each other.
Here’s the latest model, described on the website as “revolutionary not evolutionary.”
New, improved Activator
Climbing the stairway to FDA approval
Yet, the FDA has issued a 510(k) medical device clearance for the Activator Adjusting Instrument for “chiropractic adjustment of the spine and extremities.” How could that happen?
According to FDA records, Activator I (still for sale on the AMI website) was in use prior to the effective date of the federal law requiring pre-market approval. Such devices were grandfathered in under the law. After the law’s enactment, AMI applied for pre-market approval of its newer model, Activator II. It did so on the basis that II was “substantially equivalent” to I, as permitted by law. However, applicants must show that the device submitted for approval is similar in safety and effectiveness to the device already on the market. The required 510(k) summary submitted to the FDA cited a 1986 study published in the Journal of Manipulative and Physiological Therapeutics (JMPT): “Accuracy of Piezoelectric Accelerometers measuring Displacement of a Spinal Adjusting Instrument.” Unfortunately, I cannot access the full article or provide you with a link to this abstract, from which this quote is taken:
The impedance-head-equipped spring-loaded Activator chiropractic adjusting instrument had a low velocity when used on the patient and appeared to cause bone movement and a measurable EMG response.
According to the summary, this “demonstrate[s] the effectiveness of the instruments to cause vertebral movement for chiropractic adjustment at very low energy.” Note that there is no statement in the abstract about whether this “bone movement” actually did anything for the patient or even that its putative effect was measured at all.
But progress marches on, and eventually the Activator III appeared. Its premarket approval was based on substantial equivalency to Activator I and II, with no citation of additional studies of clinical effectiveness in the 510(k) summary. Activator IV apparently did not require FDA approval, or at least I didn’t find any evidence that AMI got it. That brings us to Activator V. This application was based on similarity to III and to two similar devices (FDA approved here and here) manufactured by other companies. Like the III application, the V didn’t cite any studies of clinical effectiveness, nor could it piggyback onto studies cited by the manufacturers of the two other devices, as they don’t cite any in their applications. There is now an Activator V-E, also approved, but unfortunately its summary does not appear on the FDA’s website. (V-E is not sold on AMI’s website.)
Thus, it appears that the evidence of clinical effectiveness of the Activators submitted by AMI to the FDA is less than convincing, even though the FDA requires it for approval. (21 CFR Sec. 807.92) Interestingly, by the time V was submitted for approval in 2007, there was a 2005 article in the JMPT reviewing the studies of clinical effectiveness of the Activator Method and validity of leg-length analysis. The authors concluded that studies of AMCT diagnostic measures “did not permit assertions concerning the validity of AMCT assessment procedures for the detection of supposed joint lesions [one of the many names chiropractors have invented for the invented subluxation] or targets for adjustive intervention.” Clinical trials of the Activator method itself led them to the conclusion that it “may be a clinically useful tool” although they admit that the handful of studies they rely on have insufficient controls (e.g., lack of a placebo, not blinded) and not enough participants to demonstrate much of anything. And the funny thing is one of the authors of this article is Arlen W. Fuhr, D.C., co-creator of the Activator Method, President of Activator Methods International and the contact person on each of the pre-market approval applications filed with the FDA. Why wasn’t his article cited to the FDA?
The Tooth Fairy visits the Journal of the American Dental Association
Of course, all of this is simply Tooth Fairy Science in the first place, that wonderful term coined by SBM’s own Harriet Hall. “Tooth Fairy science seeks explanations for things before establishing that those things actually exist,” according to the Skeptic’s Dictionary. Dr. Hall offers this example:
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.
Speaking of tooth fairy science, let’s turn to the JADA article. The rationale for the study will sound familiar:
TMD can become a chronic problem lasting several years, and patients receive little help from traditional forms of treatment. Consequently, some patients may seek complementary and alternative (CAM) approaches to manage their TMD-related pain.
So, what to do? Should we expose chiropractors for ripping people off by telling them they have subluxations which may be causing their TMD pain and proceeding to “correct” those subluxations with little spring-loaded mallets? No, we study it because CAM is popular. (Actually, it isn’t.) And we study it with taxpayer money, in this case provided by NCCAM. No surprise there.
Do the authors cite any evidence that the Activator method is nevertheless effective for TMD in an attempt to justify this study, despite the lack of plausibility (which they never admit, of course)? Well, they cite a single case report and a prospective case series of 8 patients who were assessed at 6 weeks with no further follow-up. There was no blinding and no placebo control. I don’t know if this would normally justify a “prospective randomized controlled pilot trial” (as this study is described) of a medical or dental intervention, but perhaps some of you could comment.
Fifty-two (out of an original 80) completed the JADA study. Participants were divided into 4 groups which numbered between 11 and 14 in each group. They compared self-care and self-care plus either AMCT, sham AMCT or reversible interocclusal splint therapy (RIST). Self-care consisted of someone talking to the patient about TMD and suggesting such things as using OTC pain medications, jaw relaxation stretches, and reducing their stress levels.
Before I get into how AMCT was described, let me repeat Sam Homola’s explanation, because it should be fresh in your mind when you read what follows:
The AM system is based on a concept of “pelvic deficiency (P.D.)”—also called “functional short leg”—which proponents define as an “apparent” difference in length, not an anatomical difference. To determine where the alleged problem is located, the practitioner holds the patient’s feet [perhaps shod in AMI shoes] in various prescribed ways while the patient lies facedown on an examining table. Other parts of the body may be tested in various other ways. . . . If any inequality or “imbalance” is found, the practitioner taps various points along the spine, pelvis, and/or elsewhere with a handheld, spring-loaded mallet until the legs appear to be equal in length . . .
Of course, anyone with a background in medicine or dentistry would have his or her suspicions instantly aroused by such an explanation and likely seek further information. Or perhaps even toss the proposed article into the wastebasket. But let’s look at how the authors describe AMCT:
The AMCT protocol is a structured method of chiropractic treatment that involves the use of a series of biomechanical tests in determining how, where and when (or when not) to perform a mechanically assisted manipulation. These biomechanical tests involve well-defined joint or joint complex (motor unit) movements involving the area of examination. Each specific manipulation—called an “adjustment”— is given with a hand-held, springloaded instrument (Activator IV, Activator Methods International) that delivers a quick, shallow thrust. The AMCT protocol can include the entire spine and extremities, as well as the temporomandibular joint. The entire AMCT protocol has been described in detail elsewhere.
A “series of biomechanical tests?” Where is the description of functional leg-length analysis? Or the subluxations it supposedly uncovers? How about an explanation of the subluxations’ putative effect on TMD? Or what it is that this “quick, shallow thrust” is supposed to do?
And that description “in detail elsewhere?” The authors cite a chiropractic textbook written by the aforementioned Arlen W. Fuhr, a text which I expect resides exclusively in chiropractic school libraries and practitioner offices. A less obfuscatory explanation of ACMT can in fact be found in a source more readily available via medical libraries – the same 2005 article in the Journal of Manipulative and Physiological Therapeutics not cited to the FDA by AMI. (JMPT is indexed in PubMed.) Also not cited was a 2012 systematic review in the Journal of the Canadian Chiropractic Association (also indexed in PubMed), which found no difference in results between the Activator Method and manual therapy. Since none of the 8 studies were randomized clinical trials, all lacked a no-treatment or sham treatment group and were unblinded, no results could reliably be attributed to either treatment.
Had the JMPT article been cited, the peer reviewers would not only have discovered the obvious: the lack of validity, not to mention the underlying pseudoscience, of functional leg length analysis and lack of evidence of effectiveness (and, again, the pseudoscience) of the Activator Method. They would also have found a refreshingly honest admission by its authors: there is no evidence that the chiropractic subluxation exists in the first place.
Let’s imagine two scenarios here. First, the peer reviewers were misled by what was made to sound like a fairly innocuous procedure by using vague terms like “structured method of chiropractic treatment” and “biomechanical tests.” Part of me hopes this is the case, although I do think someone should have asked for more detail. And part of me doesn’t want to contemplate a second scenario: that more information was requested by the reviewers and a more straightforward explanation provided by the authors and that it informed them exactly what AMPT is.
And the results of the JADA study?
After going through an impressive amount of verbiage about subject selection, control groups, confidence intervals, baseline characteristics, and all the other accoutrements typical of Tooth Fairy Science, we learn there was no statistically significant difference between results in each group. But that’s ok because the study was simply looking at the “feasibility of conducting a full-scale RCT to evaluate the effectiveness of AMCT” although the next study, according to the authors, should be a comparative effectiveness study. An odd suggestion, as such studies are generally done to compare treatments that are effective. Otherwise, why would you do it?
However, I can’t say nothing useful was learned, although likely not what the researchers intended. First, for all of the hoopla we hear about patients preferring CAM practitioners for their more “holistic” approach, mean satisfaction with care was actually greater for the RIST and self-care-only groups than with the AMCT and sham-AMCT groups. Second, they report “subtle differences” between the chiropractor’s interaction with participants in the AMCT and sham AMCT groups. Actually, some of the differences (more fully explicated in a poster presentation by the authors) weren’t so subtle. The chiropractor talked more to one group than another, but which was which is not reported. And in the active AMCT group the median number of Activator device “clicks” was almost twice as many as in the sham group (42 clicks v. 22 clicks).
But what about the idea that CAM practices must be studied because, as the authors suggest, “some people may seek complementary and alternative approaches to manage their TMD-related pain.” True enough, “some people” do exactly that. “Some people” may seek crystal therapies or use biomats, or tape magnets to their jaws. Should we study those too? After all, they have no less plausibility than AMCT. Must we study every treatment anyone makes up? And does it matter why “some people” try these therapies? Does it matter that they seek them out because chiropractors are misleading them about AMCT’s effectiveness and taking their good money for a worthless treatment? Wouldn’t it be more useful to tell them this rather than spending money on implausible treatments that we know can’t possibly work? Do we think for one second that chiropractors will stop using AMCT because the studies don’t support its use? (Answer: they won’t.)
Which brings me to other troubling aspects of this study.
Why was this study approved by the University of Iowa’s Institutional Review Board (IRB)? I am certainly not an expert in bioethics, but it is my understanding that no trial should be approved that does not conform to the “World Medical Association Declaration of Helsinki – Ethical Principles for Medical Research Involving Human Subjects,” which states:
21. 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.
As well, ethical and legal principals require informed consent before a human subject agrees to participate in a trial. As Steve Novella said yesterday in his post on informed consent:
This principle is, in turn, based largely upon the principle of autonomy – people have the right to control their own lives, and one cannot have control without information.
The same is true for informed consent in the legal context, which is based historically on the tort of battery: no one can intentionally touch you without your consent. And to consent you have to know what you are consenting to.
So let’s review what we know about AMCT.
- It is based on the “theory” that subluxations (chiropractic version) of the spine affect human functioning and can cause ill health. As chiropractors themselves admit, there is no real support for this “theory.” And whatever chiropractors think, as any responsible scientist will tell you (and as one neuroscientist in fact told me) the whole idea is nonsense.
- These subluxations can be discovered by something called “functional leg length analysis,” which even a casual observer can see makes no sense (anatomical or otherwise) whatsoever.
- That thumping the vertebrae or other articulation with a small, spring-loaded mallet will reduce the subluxation by moving it back into its proper location, an obviously dubious concept.
- This “correction” of the subluxation will have a beneficial effect on the patient’s condition.
And how do we know this? Through information readily available via a search of literature indexed on PubMed. There are plenty of other journal articles with similar information, but articles on AMCT alone will tell you everything you need to know. As will Science-Based Medicine, for that matter.
Did the University of Iowa IRB have this information? I don’t know, but it was certainly easily available. Would this information be relevant to a person making a decision on whether to participate? Of course. Were the subjects told this? I doubt it. (And if anyone out there has the informed consent documents used in this trial, I would certainly love to see them.)
If I can figure all of this out with my 10th grade biology science education and an internet connection, why couldn’t the editors of JADA? Why couldn’t the University of Iowa IRB? Good questions.