A variety of health care providers, including physicians, physical therapists, osteopathic physicians, and chiropractors, offer spinal manipulation in their practices. It is advocated as a means of reducing pain in the spine (and other joints) and increasing range of motion. It is thought that manipulation resets the position of the spinal cord and nerves, allowing the nervous system to function optimally and improve the body's biomechanical functioning. Randomized controlled trials have not shown the effectiveness of this form of therapy. Stevinson and Ernst systematically reviewed the literature to define the risks of spinal manipulation.

A computerized literature search using MEDLINE (via PubMed), EMBASE, and the Cochrane Library of the following terms—adverse effects, adverse events, spinal manipulation, stroke, vascular accident, risk, adverse effects, chiropractic, complications, manual therapy, osteopathy, and safety—yielded several case studies and retrospective surveys about the adverse effects of spinal manipulation. Approximately one half of patients who undergo spinal manipulation have mild to moderate undesirable effects; these include local discomfort (occurred in 53 percent of patients in one series of 4,712 patients); headache (12 percent); tiredness (11 percent); and radiating discomfort (10 percent). Most (74 percent) of these reactions resolved within one day. Similar results were found in other series.

Serious adverse consequences, including death, were reported in a number of cases. Cerebrovascular accidents, with permanent neurologic sequelae, were most often reported. It is difficult to gauge the incidence of serious adverse events, but other serious events occurring after spinal manipulation included dislocation, vertebral fracture, disk herniation progressing to cauda equina syndrome (occurring after lumbar manipulation), and vertebrobasilar accidents (occurring after rotational cervical treatment). It seems that there is approximately one event for every 1 to 2 million manipulations, although this is only a rough estimate. A population-based, case-control study including 582 cases of vertebrobasilar accidents estimated that there would be about 1.3 of these adverse events for every 100,000 treated patients (younger than 45 years) within one week of the treatment.

The authors conclude that the risk of serious adverse events with spinal manipulation is very low, although the risk of less serious events is more common.

editor's note: Clinical Evidence (published by BMJ, June 2001, 5th ed.), concluded that there was a slightly higher risk than was found in the Stevinson study: 1 per 20,000 to 100,000 patients with vertebrobasilar strokes. They conclude that spinal manipulation is “likely to be beneficial,” but as the authors of the preceding study note, effectiveness is dependent on a number of factors, including competence of the practitioner and limitations of the therapy itself. Just as it is important to inquire whether patients are using over-the-counter or alternative medications, it is also prudent to be aware that they are receiving manipulations, whether from other physicians, chiropractors, or physiotherapists.—g.b.h.