Intended for healthcare professionals


Chiropractic for low back pain

BMJ 1998; 317 doi: (Published 18 July 1998) Cite this as: BMJ 1998;317:160

We don't know whether it does more good than harm 

  1. E Ernst, Professor,
  2. W J J Assendelft (E.Ernst{at}, Senior researcher
  1. Department of Complementary Medicine, Postgraduate Medical School, University of Exeter, Exeter EX2 4NT
  2. Institute for Research in Extramural Medicine, Vrije Universiteit Amsterdam, 1081 BT Amsterdam, Netherlands

    Chiropractic includes various techniques used in the hope of correcting vertebral disc displacements, freeing spinal joint adhesion, inhibiting nociceptive impulses, or correcting spinal misalignment. Several national guidelines on the treatment of low back pain recommend spinal manipulation, including chiropractic, as a symptomatic treatment for acute uncomplicated cases where pain fails to resolve spontaneously within the first months.1 How solidly are these recommendations based on evidence?

    There are many controlled trials of spinal manipulation and no fewer than 51 reviews.2 Surprisingly, in the review of Shekelle et al,3 which provided the basis for the recommendations mentioned above, the subset of randomised clinical trials on acute low back pain which generated these favourable recommendations did not contain one single trial of chiropractic. A recent systematic review restricted to chiropractic manipulation included only eight randomised controlled trials, all of which were methodologically flawed and “did not provide convincing evidence for the effectiveness of chiropractic for acute or chronic [low back pain].”4 Consequently, we can conclude only that the effectiveness of chiropractic as a treatment for low back pain has not been established beyond reasonable doubt.

    Is chiropractic safe? Cervical manipulations are burdened with severe adverse reactions, such as vertebrobasilar accidents and paralyses due to fractures.5 A literature review identified 165 vertebrobasilar accidents, including 29 deaths.5 Estimates of their incidence range from 1 per 200 000 to 1 per million cervical manipulations.5 A patient survey suggested that about 12% of users experience (mostly mild) adverse reactions.6 Mild adverse reactions were also reported after one third of all treatments in a prospective study.7 The risks of manipulating the lower spine seem to be lower, with fractures and cauda equina syndrome being the most serious reactions.5 Nevertheless, upper spinal manipulation is also occasionally performed in lower back pain. Finally, there may be important indirect risks associated with chiropractic. Potential overuse of radiographs by chiropractors is one example8; another is the negative attitude of some chiropractors towards immunisation.9 Thus, even if chiropractic manipulation were totally devoid of risks, the approach of chiropractors may not always be so.

    Lastly, does chiropractic save money for healthcare systems? There are few conclusive economic evaluations, but most of the rigorous studies do not suggest that chiropractic saves money. A study comparing the costs of care by chiropractors, primary care physicians, and orthopaedic surgeons in the United States indicated that the total direct outpatient cost per episode of low back pain was highest for urban chiropractors.10 One obvious reason is that, on average, chiropractors use more consultations per episode of back pain than other professionals.11 However, studies with other designs sometimes provide the opposite results.12 A review of workers' compensation studies concluded that “chiropractic cost-effectiveness is not yet convincingly proven.”13

    On the basis of current evidence, it seems uncertain whether chiropractic does more good than harm. More and better research is required.