FCER Response to Ernst's Systematic Review of Spinal Manipulation

Norwalk, Iowa — In the April 1, 2006 issue of the Journal of the Royal Society of Medicine, Ernst and Canter summarized 16 systematic reviews published between 2000 and May 2005 on the effectiveness of spinal manipulation. The Foundation for Chiropractic Education and Research (FCER) is always open to results of research if the research proves to be of good quality and lacking bias. The biases in this study are painfully apparent, rendering the sweepingly negative “findings” little more than the authors’ personal opinions. Ernst and Canter concluded that these data do not demonstrate that spinal manipulation is an effective intervention for any condition.” According to them, the data fail to demonstrate that spinal manipulation is effective for treating back pain, neck pain, dysmenorrhea, infantile colic, asthma, allergy, cervicogenic dizziness, and any medical complaint. The exception was for back pain, where spinal manipulation may be superior to sham manipulation but not to conventional interventions. Considering the possibility of adverse events associated with spinal manipulation, the authors determined that their review "does not suggest that spinal manipulation is a recommendable treatment." This study is so far from what would be considered a methodical and robust systematic review without bias as to render it highly suspect if not meaningless. Its methods of analyses have not been validated but rather reveal the authors' own carelessness — if not outright distortion — of the literature which it cites. Many of its glaring defects are ones that the authors attempt to criticize in other work. It fails to grasp how the hypotheses and methods of analysis in the reviews that it cites are bound to deliver differing conclusions, such that Ernst and Canter go out of their way to criticize the positive findings of a single chiropractic author while overlooking their own consistently negative findings for chiropractic which appear in no less than 25% of the reviews that they include in their discussion. Finally, the authors fail to recognize the major flaws in several of the primary sources of data which comprise the systematic reviews under scrutiny in this research. There are issues surrounding the Ernst and Canter review that substantially undercut any prompt and uncritical acceptance of its conclusions. These include (1) the failure of systematic reviews and meta-analyses to consider subgroups of patients receiving treatment, (2) the design flaws of a large number of the randomized controlled trials which comprised the systematic reviews addressed in this report, (3) the failure to consider that modern evidence-based medicine is based upon clinical observation as well as randomized clinical trials, (4) the failure to adequately address the relative risks of other treatments available in conventional medicine for the conditions discussed in this review, and (5) numerous revelations of bias of one of the authors (Ernst) which have been amply demonstrated and refuted elsewhere. Under these circumstances, Ernst and Canter's study can be greeted only with the most extreme skepticism. In lacking many of the elements required for a meaningful presentation of the evidence required for supporting treatment alternatives, this current report should not be considered worthy of guiding a clinical decision. As such, the glaring weaknesses of the report only serve to undermine the public's confidence in science as a means to inform health policy.