The title of one of Edzard Ernst’s most recent articles, “Deaths after chiropractic: A review of published cases”,1
seems to have wasted no time in creating a worst-case scenario, frightening the reader from a profession which has successfully treated patients for 115 years and for which evidence supporting its effectiveness and safety is abundant.2-6
Ernst has declared an opposing view with copious publications for at least the past 16 years.7-14
Unfortunately, many of Ernst’s assertions regarding spinal manipulation
and chiropractic have been discredited in the past15,16
and, in several instances, found to be blatantly misleading.17
The current article1
is no exception.
- The problem in perspective:
Ernst reports a total of 26 deaths in 75 years of chiropractic practice. That would average to less than 1/3 death per year. Given the facts that there are an estimated 250M chiropractic visits per year in the United States,18 13.4M chiropractic visits per year in Canada,19 and 2.25M visits per year in the United Kingdom,20 the death rates in each country after chiropractic according to Ernst’s figures calculate to be 0.000000132%, 0.00000246%, and 0.0000132%, respectively.
These figures are absurdly infinitesimal when compared to the 230,000-280,000 deaths caused by iatrogenesis in medical treatment as reported by the Institute of Medicine,21 or from just the use of NSAID medications, producing 10,000-20,000 fatalities from multiple organ systems adversely affected—in the United States alone.22-24 Even what has been regarded as the more relatively benign COX-2 inhibitors and acetaminophen medications have been described to generate serious GI, cardiovascular, and hepatic problems at rates an order of magnitude greater than the side-effects attributed to spinal manipulation.25-29
A more complete accounting of the extremely low relative risks of spinal manipulation compared to medical interventions for back and neck pain has been presented elsewhere.30-32 Given far too little consideration by Ernst is the fact that the majority of vertebral artery dissections have been found to be spontaneous, possibly exacerbated by high levels of endogenous homocysteine.32,33 Final credibility to this argument is the fact that there was no evidence of excess risk of vertebrobasilar artery stroke associated with chiropractic care as compared to the primary care administered by allopathic physicians.34 This fact alone should do much to dismiss the unmistakable implications from Ernst’s argument that at least a large proportion of the deaths listed could be attributed to chiropractic care.
- Lack of risk-benefit analysis:
A balanced view of any healthcare intervention needs to make use of risk-benefit ratios in order that a fair and balanced analysis be presented. There is no such mention of any benefits in Ernst’s current publication,1 as is the case for any of his other papers as cited.7-14 This is in spite of the fact that the clear benefits of spinal manipulation as administered by chiropractors have been already discussed.2-6 Simple mathematics dictates that a denominator [benefits] which is set to zero will yield a risk-benefit ratio of infinity, regardless as to how infinitesimal the numerator [risks] may be.
- Incompleteness and carelessness of review:
In addition to the substantial lack of considerations pointed out above, it is peculiar to hear of Ernst’s assertion that his review is “systematic” when in fact through 1985 it accounts for just 16 deaths, whereas 24 through the same period were carefully described in a review of 257 cases that Ernst himself cited in his bibliography in discussing a different subject.35 Even more worrisome is the fact that in the same reference Terrett clearly points out that the majority of the cases of vertebrobasilar artery accidents were mistakenly—often deliberately—attributed to chiropractors instead of the myriad of other practitioners responsible for the adverse events.35 This point was not only ignored by Ernst but deliberately distorted to suggest that more chiropractors (17) were responsible for complications than medical practitioners (9), naturopaths (1), or physiotherapists (0).1 To make matters worse, there is more recent data which clearly demonstrates that vertebrobasilar artery accidents are more likely to occur in the hands of nonchiropractors rather than actual chiropractors.36 Finally, there is an inaccuracy in this actual citation itself35, in which the entity responsible for its publication is listed in Ernst’s article as “JCMIC” instead of “NCMIC.” All of these indications suggest that the article by Ernst falls far short of what would normally be considered as a carefully prepared, scholarly systematic review worthy of publication in the peer-reviewed, scientific literature.
What is peculiar is an apparent disclaimer by Ernst himself, in which he states in his article that “Obviously, the present article is not aimed at providing incidence figures; this would require a different methodology entirely. To date, no reliable incidence data are available.” Under these circumstances, one has to question how Ernst ever was able to invoke the term “systematic” in the first place.
- Signs of bias and blanket statements:
Statements to the effect that “when carefully evaluating the known facts, one does arrive at the conclusion that causality [of arterial dissection] is at least likely.” Unfortunately, we are never privy to what the “facts” may be, in addition to being given a statement that is contradicted by both the mechanistic32,33 and actual patient data34 supporting spontaneity discussed earlier. Finally, Ernst’s assertion that “the risks of chiropractic neck manipulations far outweigh their benefits” is clearly unsupported by the data in the literature with the complete ignorance of any benefits or risk-benefit ratios as argued earlier.
For these reasons, one must greet Ernst’s current article with extreme skepticism and hope that the host of its inaccuracies does not misguide the journal’s readership. It is certainly regrettable to contemplate the prospect of having this error-prone paper emulated in future citations in the scientific literature for years to come.