|Home | About | Journals | Submit | Contact Us | Français|
We are, of course, flattered by this amount of interest in our article1 and would like to respond as follows to the multitude of interesting arguments. Spinal manipulation was first described in 1895 by the `magnetic healer' D D Palmer as a treatment of `subluxations' of the spine and other joints. Early chiropractors believed that `subluxations' were the cause of all diseases—to quote Palmer: `95% of all diseases are caused by displaced vertebrae, the remainder by luxations of other joints'.2 Today, 89.8% of (USA) chiropractors feel that spinal manipulation should not be limited to musculoskeletal conditions.3
It is thus not `methodologically unsound', as D Byfield and P McCarthy assume, but necessary to conduct a health technology assessment of spinal manipulations for the full range of conditions for which adequate data are available. Similarly, global assessments exist also in the chiropractic literature and are acclaimed by chiropractors—as long as they are not truly critical of their practice.4
Of course, Byfield and McCarthy are right, the majority of chiropractic patients suffer from musculoskeletal problems, but are they suggesting one must not ask questions about the rest? And, of course, the `straight' chiropractors adhering to Palmer's gospel are in the minority; but, in the UK, the influence of those `vitalists', who insist spinal manipulation is a panacea, is growing.5
It is not correct that we have `aggregated' different conditions. In fact, we assessed systematic reviews pertaining to different conditions quite separately.1 A systematic review is an accepted method for minimizing bias, the argument that our article maximized bias seems therefore illogical and has no basis. In this context it is worth noting that most of the commentators are affiliated with chiropractic or osteopathic organizations, while neither of us is on the payroll of an interested party. Byfield and McCarthy's claim that our approach `lacks statistical validity' is embarrassing—we did not use any statistics in this paper.
Systematic reviews inevitably require some inclusion/exclusion criteria. Thus, some articles will always be omitted which others would have liked to include—perhaps because of their favourable results. It is, however, misleading to imply that we systematically excluded studies of `manipulation as used in practice'. Most of the 16 evaluated reviews included such trials.
B J Lewis and G Carruthers are mistaken when stating that four of the systematic reviews included were `reviews of reviews'. In fact, they all were reviews of controlled clinical trials. Similarly, it is disingenuous to imply that we merely evaluated reviews of our `own opinion'. The fact that four of the 16 included articles were our own simply shows that we are research-active in this area. To exclude one's own work in systematic reviews would be woefully unscientific.
Several comments note that our conclusions are not in line with current guidelines. We also make this point in our article and suggest `... that these guideline be reconsidered in the light of the best available data'.1 Surely this is sensible? Yet Breen et al. categorically state `... there is enough evidence about manipulation in the back pain area', providing no reference in support of this statement. One could therefore be forgiven for concluding that it is more the result of wishful thinking than of critical evaluation.