Journal of Manipulative and Physiological Therapeutics
Volume 29, Issue 7 , Pages 597-598, September 2006

Adverse reactions to chiropractic care in the UCLA Neck Pain Study

John A. Barnes School of Medicine, Department of Public Health Sciences, University of Hawaii, Manoa, Honolulu, HI

Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, MI

Article Outline

 

To the Editor:

The lack of statistical significance is Dr Rosner's first and apparently primary concern about our study.1 He writes that the odds ratios “do not approach statistical significance,” the odds ratios “lack statistical significance,” and the data “do not reach statistical significance.” We addressed this issue previously in our responses to letters to the editor.2, 3 Briefly, dichotomizing treatment effects as “significant” (P < .05) or “nonsignificant” (P > .05) has no scientific merit and is strongly discouraged by quantitatively sophisticated researchers, including epidemiologists and statisticians.4, 5

Dr Rosner cites the similarity of the estimated effects of manipulation vs mobilization on adverse outcomes occurring within 24 hours of treatment and later as evidence against Hill's “positive gradient” criterion for causal inference. In fact, the gradient criterion refers to the dose or duration of exposure, not to the elapsed time between treatment and outcome events. Perhaps Dr Rosner is mixing the temporality criterion with the gradient criterion. Nevertheless, the adverse effects occurred after treatment (satisfying the temporality criterion), and the similarity of effects for reactions occurring close to and further away from treatment is not an argument for or against the gradient criterion, which in any case is not necessary for inferring causation.

Dr Rosner “cherry picks” 1 estimate to support his contention that adverse reactions may not specifically be the result of cervical spine manipulation, but may be associated with other procedures such as electrical muscle stimulation (EMS) (not “electromagnetic stimulation”). In fact, in the table cited by Dr Rosner, 75% of the manipulation vs mobilization estimates are stronger than the respective EMS vs no EMS estimates, and 83% are stronger than the respective heat vs no heat estimates. Clearly, adverse reactions may be associated with treatments other than manipulation; however, when considering all the evidence, associations appear stronger and more consistent with manipulation vs mobilization than with the other treatments.

Dr Rosner is concerned that the manipulation vs mobilization effect may have been confounded by the other treatments, and that the distribution of predisposing conditions may have varied between the manipulation and mobilization groups, resulting in biased estimates. The treatment groups were assigned randomly, and because of the balanced factorial design, assignment to manipulation vs mobilization was not associated with assignment to EMS vs no EMS or to heat vs no heat.1 As stated in the text and tables,1 treatment-group assignment was included in all regression models, differences between treatment groups on baseline factors were small and clinically insignificant, and there did not appear to be any systematic differences between groups in expected prognosis or risk of adverse reactions.

Dr Rosner compares the risks of manipulation with the risks after other medical procedures, and he states that “chiropractic has been shown to be many orders of magnitude safer than medication or surgery.” Chiropractic may in fact be safer, but the data cited by Dr Rosner cannot be used to support this position. As we stated, complication rates from surgical and pharmaceutical treatments are likely higher than those from chiropractic interventions; however, the differences between medical/surgical and chiropractic patients must be taken into account to compute valid relative measures of safety. Prognostic differences may largely explain differences in complication rates. Furthermore, data cited by Dr Rosner and more recent findings6 on complications from nonsteroidal antiinflammatory drugs are mostly from patients with chronic pain who take large doses for long durations. These doses and durations may differ markedly from those of patients with neck pain seeking chiropractic care.

Finally, Dr Rosner questions our study because of the lack of details on the treatments and techniques. The specific treatment protocols are clearly described.1 Study participants had on average 2 to 3 treatments over the 2-week reporting period, and we and others have found no association between manipulation frequency and risk of side effects.7 Dr Rosner writes that the benefits of cervical manipulation have been “amply” shown, and he concludes that manipulation is “superior” to mobilization. Of the 30 articles cited, 29 are at least 5 years old, and 14 are at least a decade old. Dr Rosner misinterprets the 1 most recent and relevant study,8 and fails to cite other studies that do not support his preconceived beliefs.9, 10

An unbiased look at the most recent available evidence indicates that (1) manipulation is probably not more effective than mobilization for patients with neck pain,8, 9 (2) adverse reactions to manual therapy appear to be more common and more frequent after cervical spine manipulation than after mobilization or manipulation to other spinal regions,1, 7, 11 and (3) adverse reactions may impair prognosis and reduce patient satisfaction.12

Back to Article Outline

References 

  1. Hurwitz EL, Morgenstern H, Vassilaki M, Chiang L-M. Frequency and clinical predictors of adverse reactions to chiropractic care in the UCLA Neck Pain Study. Spine. 2005;30:1477–1484
  2. Hurwitz EL, Morgenstern H. Frequency and clinical predictors of adverse reactions to chiropractic care in the UCLA Neck Pain Study: letter. Spine. 2006;31:253–254
  3. Hurwitz EL, Morgenstern H. Frequency and clinical predictors of adverse reactions to chiropractic care in the UCLA Neck Pain Study: letter. Spine. 2006;31:254–255
  4. Rothman KJ, Greenland S. Modern epidemiology, 2nd ed. Philadelphia: Lippincott-Raven Publishers. 1998, pp. 64, 186-8, 660-1.
  5. Jewell NP. In: Statistics for epidemiology. Boca Raton: Chapman & Hall/CRC; 2004;p. 61–62
  6. Tannenbaum H, Bombardier C, Davis P, Russell AS. An evidence-based approach to prescribing nonsteroidal anti-inflammatory drugs: Third Canadian Consensus Conference. J Rheumatol. 2006;33:140–157
  7. Cagnie B, Vinck E, Beernaert A, Cambier D. How common are side effects of spinal manipulation and can these side effects be predicted?. Man Ther. 2004;9:151–156
  8. Gross AR, Hoving JL, Haines TA, Goldsmith CH, Kay T, Aker P. Cervical Overview Group. A Cochrane review of manipulation and mobilization for mechanical neck disorders. Spine. 2004;29:1541–1548
  9. Hurwitz EL, Morgenstern H, Harber P, Kominski GF, Yu F, Adams AH. A randomized trial of chiropractic manipulation and mobilization for patients with neck pain: clinical outcomes from the UCLA Neck-Pain Study. Am J Public Health. 2002;92:1634–1641
  10. Evans R, Bronfort G, Nelson B, Goldsmith CH. Two-year follow-up of a randomized clinical trial of spinal manipulation and two types of exercise for patients with chronic neck pain. Spine. 2002;27:2383–2389
  11. Senstad O, Leboeuf-Yde C, Borchgrevink CF. Frequency and characteristics of side effects of spinal manipulative therapy. Spine. 1997;22:435–441
  12. Hurwitz EL, Morgenstern H, Vassilaki M, Chiang L-M. Adverse reactions to chiropractic treatment and their effects on satisfaction and clinical outcomes among patients enrolled in the UCLA Neck Pain Study. J Manipulative Physiol Ther. 2004;27:16–25

PII: S0161-4754(06)00183-7

doi:10.1016/j.jmpt.2006.07.002

Journal of Manipulative and Physiological Therapeutics
Volume 29, Issue 7 , Pages 597-598, September 2006

Article Tools