Tuesday, January 04, 2005

FSU Chiropractic?

BoingBoing recently linked to a newspaper article that discusses an attempt by the Florida State Legislature to add a school of chiropractic to Florida State University. The article included a campus map drawn by some professors detailing a new school of pseudoscience (including such buildings as the "Bigfoot Institute" and "College of Dowsing"; the map is apparently what got BoingBoing's interest).

According to the article, many of the faculty at FSU are (understandably) outraged at the possibility of a chiropractic school, and a few have even threatened to resign if the school is approved. Apparently the legislature has already approved a budget of $9 million a year to fund this new school, even though the university's board of trustees has not approved the school.

Chiropractors are common, they appear to be doctors, and many people in the US go to them and leave satisfied, so why are professors threatening to resign if FSU gets a chiropractic school?

I've written about this topic before, and the reason why these professors are willing to resign is simple: there is very little evidence that chiropractic manipulation does anything that current medical science cannot, and there is little data available about the risks (especially long-term) of chiropractic treatments. Combine that with the observation that many chiropractic practices are laced with fraudulent and/or negligent activities, and this should be enough to stop the profession cold in its tracks (and also explains why no public university in the US currently has a chiropractic college associated with it).

The Florida state legislature commissioned a number of reports to provide information about the chiropractic field. The most interesting from a scientific perspective is "The Chiropractic Profession and Its Research and Education Programs" (PDF), which attempts to summarize the current body of scientific knowledge on chiropractic, but instead reads more like the advertising materials of your local chiropractor. Part of the reason for this might be that the report's authors appear to be distinctly lacking in clinical medical and scientific experience: the authors include one person with a Ph.D. in higher education finance and an MBA, one with a Ph.D. in economics, one with a Ph.D. in nursing, one with a doctorate in higher education, and one with a D.C. (doctor of chiropractic).

The report includes a lot of the typical arguments in favor of chiropractic, including the much-overused "The AMA hates chiropractic and is just trying to steal our customers" statement, along with other blather about how the vast majority of people are satisfied with their chiropractors (the vast majority of people were probably satisfied with their snake-oil salesmen, too).

After discussing how satisfied customers are with chiropractic, the report goes on to do a literature review. The report's authors focus a lot of their time on two of Shekelle et al.'s 1992 papers (a and b below), which are part of the oft-mentioned 1991 & 1992 "Rand Reports" on chiropractic. The peer-reviewed paper (Shekelle et al., 1992b) in this series is probably one of the papers most cited and most misrepresented by backers of chiropractic. It was a meta-study that reviewed published literature on chiropractic manipulations relating to lower back pain, and it came to the following conclusion:
"Spinal manipulation is of short-term benefit in some patients, particularly those with uncomplicated, acute low-back pain. Data are insufficient concerning the efficacy of spinal manipulation for chronic low-back pain." (Shekelle et al., 1992b)
In other words, the study looked solely at studies that evaluated lower back pain, and was only able to come to a conclusion with regard to patients with short-term (acute) lower back pain. This is not the ringing endorsement of chiropractic as a whole that much chiropractic literature makes it out to be. A more recent study (Assendelft et al., 2003, which I believe includes the lead author from the Shekelle et al., 1992 study), reviews even more literature on chiropractic manipulation related to lower back pain, and comes to a more concrete conclusion:
"Thirty-nine RCTs [randomized controlled trials] were identified. Meta-regression models were developed for acute or chronic pain and short-term and long-term pain and function. For patients with acute low back pain, spinal manipulative therapy was superior only to sham therapy or therapies judged to be ineffective or even harmful. Spinal manipulative therapy had no statistically or clinically significant advantage over general practitioner care, analgesics, physical therapy, exercises, or back school. Results for patients with chronic low back pain were similar. Radiation of pain, study quality, profession of manipulator, and use of manipulation alone or in combination with other therapies did not affect these results.

"Conclusions: There is no evidence that spinal manipulative therapy is superior to other standard treatments for patients with acute or chronic low back pain." (Assendelft et al., 2003)
Lest we bias ourselves by looking at just one meta-review paper, a separate paper (Cherkin et al., 2003) was also published in 2003 looking at the same topic:
"A meta-regression analysis of the results of 26 RCTs evaluating spinal manipulation for acute and chronic back pain reported that spinal manipulation was superior to sham therapies and therapies judged to have no evidence of a benefit but was not superior to effective conventional treatments." (Cherkin et al., 2003)
The Florida report for the state legislature says on numerous occasions that there is evidence that chiropractic manipulation is superior to standard medical treatment; these two reviews make it clear that when one analyzes the majority of the literature on the topic, chiropractic manipulation is not superior to standard treatment.

It is critical to remember that these studies are only looking at chiropractic manipulation relating to lower back pain; they say nothing about chiropractic treatment of other conditions, even though chiropractors regularly attempt to treat other conditions.

The Florida state report also uses some horrible analyses in their report, for instance,
"There have been at least 18 randomized trials for manipulation with head and neck pain complaints. Nine favored manipulation and eight found manipulation equal to other treatments, although conventional levels of statistical significance were reached for only some of the outcomes in some studies."
They start off by saying that nine of 18 studies favored manipulation, which makes it sound as though chiropractic has incredible effects, but only after that do they say that not all the results were statistically significant. For those not up on their statistics, common practice in science is to assume that if some treatment is not shown to be statistically significantly different from some other treatment, than those two treatments must be assumed to be the same. In other words, if there is no statistically significant difference you cannot say, "Treatment 1 was higher than treatment 2," because statistically they are the same. Thus the report's statement that nine trials favored manipulation is most likely deceptive and incorrect.

Unfortunately, I have been unable to determine the extent of the report's misrepresentation. Two papers are cited for neck and head results (Hurwitz, 1996; Vernon, 1999), but the Vernon (1999) reference is missing from the literature cited section (only Vernon, 1990 is listed), and while I can't get the full text of the other paper (Hurwitz, 1996), it appears from the abstract to base its conclusions primarily on three studies.

The Florida report concludes its research review with a section showing that chiropractic manipulations are safer than conventional medical treatments. Here are the data on chiropractic injury rates that it includes:
"5 to 10 per 10 million manipulations vertebrobasilar;
3 to 6 per 10 million for major impairment;
fewer than 3 deaths per 10 million manipulations; and
about 1 per 100 million manipulations for complications involving
canda equina (Hurwitz, 1996)."
Here's what some recently published literature has to say about the injury rates of chiropractic:
"The most valid studies suggest that about half of all patients will experience adverse events after chiropractic SM [spinal manipulation]. These events are usually mild and transient. No reliable data exist about the incidence of serious adverse events. These data indicate that mild and transient adverse events seem to be frequent. Serious adverse events are probably rare but their incidence can only be estimated at present. Further prospective investigations are needed to define their incidence more closely." (Ernst, 2001b)
Cervical spinal manipulation is generally considered to be more dangerous than other forms of spinal manipulation, and actual rates of injury from cervical spinal manipulations are currently unknown. While many chiropractors state that these manipulations are safe, there are suspicions that the injury rates are higher than currently believed. For instance, a group of neurosurgeons in Tulsa OK wrote a paper on a number of patients who had suffered neurological problems after cervical spinal manipulation (Malone et al., 2002); at the end of their paper they estimate the rate of injuries based on the population of their geographical area, the number of cases they've seen, and the approximate prevalence of chiropractic care:
"Based on these data, with our local 20 patients who experienced CSMT [cervical spinal manipulation therapy] related complication during the study period, and assuming that 17,000 patients had undergone cervical manipulation during the study period, then the complication rate was approximately one irreversible complication per every 850 patients undergoing a series of manipulations in the local region of the study. If the mean number of manipulations per patient is assumed to be 10, then the risk of complication in this study would be one in every 8500 cervical manipulations. Using other published data to estimate the number of cervical manipulations in this regional population, there could be up to 180,000 procedures annually, yielding a complication rate of one irreversible per 45,600 cervical manipulations. This does not take into account that our group represents only one third of the neurosurgeons in the geographical region around Tulsa." (emphasis mine, Malone et al., 2002)
Ernst, the author of the first paper on injury reports I quoted above, provides more details on evaluations of chiropractic injury rates in a comment published in Stroke (Ernst, 2001a):
"The best way to arrive at such information [figures on the incidence of major problems] is to prospectively study large samples of consecutive patients. Five such investigations have been published, and none reports a single case of a serious complication. This apparently confirms the assumption that complications are extreme rarities. Vis a vis the many thousand manipulations carried out daily, 200 or 300 complications in 5 years could be almost negligible. While we all hope that this is true, one must consider underreporting: if a patient suffers a serious complication after spinal manipulation, her chiropractor is unlikely to see her again, and the physicians who do might not think of a link between manipulation and the adverse effect. And even if they consider an association, are they likely to publish this as a case report? Moreover, none of the prospective studies available to date have enough power to detect events that occur less frequently than 1 in approximately 500 patients. Interestingly, most of these studies agree that mild, transient adverse effects (eg, local discomfort) are experienced by roughly every second patient who receives spinal manipulation.

"Where does this leave us when trying to critically evaluate the safety of spinal manipulation? We know that serious complications do exist. We also know of plausible explanations of how spinal manipulation might lead to serious adverse events; eg, sudden rotational and hyperextensive head movements can cause a traumatic dissection of the extracranial arteries. The incidence of life-threatening complications, however, is unknown, and previous estimates have all been based on assumptions which may or may not be true." (emphasis mine, Ernst, 2001a)
Thus chiropractic is a profession which has very little research behind it (relative to the medical profession), and what research it does have shows that it is not any more successful than other, far better-researched medical practices. Combine this lack of research regarding clinical usefulness with the lack of good data regarding the safety of chiropractic treatments (even for those treatments where there is some evidence of patient benefit), and this makes chiropractic unacceptable for medical care. Would you follow the treatment your doctor prescribed if she said, "We do not know the frequency of adverse effects of this medication"? I certainly wouldn't – that's why I go to a doctor, to get the best available treatment based on the most current research, including extensive research on the risks of the treatment (or, if the risks of the treatment are unknown, this is made clear).

Considering that the majority of chiropractic treatments have never been shown to have any significant benefits, what few treatments have been shown to have benefits with regard to placebos still do not improve on current medical practice, and that the risks of chiropractic treatments are unknown, it seems fair to say that virtually any chiropractor making a claim that they can safely provide treatment that is better than current medical practice is misrepresenting the truth.

Thus, that a state-sponsored school would consider opening up a college to train chiropractors is ridiculous; state schools should not be in the business of supporting such misguided professions, even if the public is satisfied with them.


Assendelft, W.J.J., S.C. Morton, E.I. Yu, M.J. Suttorp, and P.G. Shekelle. 2003. Spinal Manipulative Therapy for Low Back Pain: A Meta-Analysis of Effectiveness Relative to Other Therapies. Annals of Internal Medicine 138(11): 871-881. (abstract, full-text PDF is free)

Cherkin, D.C., K.J. Sherman, R.A. Deyo, and P.G. Shekelle. 2003. A Review of the Evidence for the Effectiveness, Safety, and Cost of Acupuncture, Massage Therapy, and Spinal Manipulation for Back Pain. Annals of Internal Medicine 138(11): 898-906. (abstract, full-text PDF is free)

Ernst, E. 2001a. Life-Threatening Complications of Spinal Manipulation. Stroke 32: 809. (full-text)

Ernst, E. 2001b. Prospective investigations into the safety of spinal manipulation. J Pain Symptom Manage: 21(3):238-42. (abstract)

Hurwitz, E.L., Aker, P., Adams, A.H., Meeker, W., Shekelle, P.G. 1996. Mobilization and manipulation of the cervical spine: a systematic review of the literature. Spine 21: 1746-60. (abstract)

Malone, D.G., N.G. Baldwin, F.J. Tomecek, C.M. Boxell, S.E. Gaede, C.G. Covington, AND K.K. Kugler. 2002. Complications of cervical spine manipulation therapy: 5-year retrospective study in a single-group practice. Neurosurgical Focus 13(6): 1-8. (full-text PDF)

Shekelle, P.G., A.H. Adams, M.R. Chassin, E.L. Hurwitz, R.E. Park, R.B. Phillips, and R.H. Brook. 1992a. The Appropriateness of Spinal Manipulation for Low-Back Pain. Indications and Ratings by an all Chiropractic Expert Panel. Santa Monica, CA: The RAND Corporation, R-4025/3-CCR/FCER. (list of Rand papers)

Shekelle, P.G., Adams, A.H., Chassin, M.R., Hurwitz, E.L., Brook, R.H. 1992b. Spinal Manipulation for Low-Back Pain. Annals of Internal Medicine 117(7):590-8. (abstract)

“The Chiropractic Profession and Its Research and Education Programs” December 15, 2000. Report for Florida State University by MGT of America, Tallahassee, FL. (PDF)

No comments: