Improvement with chiropractic treatment is not clinically important


A study published in Spine indicates that usual short-term chiropractic treatment provides greater improvements in spinal pain than sham treatment. However, these improvements are not clinical important and most patients in the study correctly identified the type of treatment they were receiving (sham or chiropractic).

Bruce Walker, School of Health Professions, Murdoch University, Perth, Australia, and others write that small treatment effects and methodological limitations in previous studies of chiropractic treatment for spinal have “raised uncertainty” about the value of the therapy. They add: “Our primary aim was to examine the comparative effectiveness of a brief chiropractic intervention compared to sham treatment for participants with spinal pain. The secondary aim was to explore the clinical and demographic variables associated with treatment satisfaction.”

In the study, patients with spinal pain (most of who had had their pain for more than three months) were randomised to receive usual chiropractic care or sham treatment. Sham treatment consisted of de-tuned ultrasound, an Activator instrument (a hand-held device that delivers a low impulse thrust delivered at its lowest output through a tongue depressor to disperse any remaining force), and a randomly placed hand on the spine while the de-tuned ultrasound head was placed on patient’s back/neck. According to Walker et al, all chiropractors performing sham treatment were “trained to show equal enthusiasm for the sham treatment as they would with usual care.”

Of 183 participants in the study, 91 received sham treatment and 92 received usual care. While patients receiving usual chiropractic care did experience greater improvement in pain intensity and physical function than the sham treatment patients, the differences between the groups were not clinically important. The authors also report that there were no significant differences between groups in the percentage of patients reaching the threshold of minimally acceptable outcome–34% for sham group vs. 29% of usual care (p=0.42).  

According to Walker et al, the small treatment effects observed in the study may be a result of chiropractors treating non-specific spinal pain as a homogenous condition rather than “a collection of conditions with different aetiologies”. “If non-specific spinal pain does consist of a heterogeneous conditions, then a specific intervention such as chiropractic treatment may be only effective in select cases,” they note. 

Although patients were supposed to be blinded to the treatment they received, 67% of patients in the sham group and 85% of patients in the usual care group correctly guessed the nature of the treatment they were receiving. Walker et al comment that the lack of success in attempting to blind patients to treatment allocation may be a result of the “inherent difficulties” in identifying a credible sham intervention for use in a chiropractic trial and because they did not ask participants if they had previously received chiropractic treatment. They explain: “If the sham group contained a reasonable number of non-naïve participants, then this may have impacted substantially on the success of the blinding procedure.”

Walker et al state that in patients who received usual care, correctly identifying their treatment allocation was a significant predictor of treatment satisfaction but this was not the case in the sham treatment group. Overall, patients in the usual care group reported greater treatment satisfaction than those in the sham treatment group (78% satisfied vs. 56% unsatisfied, respectively; p=0.004).

Further research in the use of chiropractic treatment for spinal pain should, the authors conclude, “Concentrate on identifying clinically relevant spinal pain subgroups to which chiropractic and other manual therapy treatments may be best suited.” They add that it would be “worthwhile to identify credible sham interventions and to conduct studies examining how non-specific effects, such as interpersonal attributes, contribute to treatment outcomes.”