Research Review by Dr. Shawn Thistle©


Mar. 2007

Study Title:

Efficacy and safety of acupuncture for chronic uncomplicated neck pain: A randomized controlled study


Vas J et al.

Publication Information:

Pain 2006; 126: 245-255.


Chronic neck pain is very common, affecting up to 10% of men and 17% of women. Most cases of chronic neck pain are mechanical in nature, with causes including degenerative changes to facet joints, uncinate joints, and discs, or chronic soft tissue problems.

Conservative treatment is the recommended approach for uncomplicated neck pain, and can include manual therapies such as manipulation and mobilization, soft tissue therapies, electrotherapeutics, and rehabilitative exercise. Acupuncture is also frequently used for chronic neck pain, despite the lack of high quality studies demonstrating its efficacy.

In systematic reviews of acupuncture for neck pain, methodological deficiencies including small sample sizes, inadequate statistical power and analysis, and lack of appropriate "sham" or "placebo group" have been noted (even though this last short-coming is difficult to rectify in acupuncture trials, just as it is in many manual therapies).

This single-blind, randomized trial compared transcutaneous nerve stimulation-placebo (TENS-placebo) to a course of acupuncture treatment in patients with neck pain exceeding three months duration. Participants were outpatients aged 17 years and over who had received a diagnosis of uncomplicated mechanical neck pain (greater than 3 months) from their referring physician.

They had to have motion-related neck pain at the time of examination, exceeding 30mm on a Visual Analogue Scale, and had not received any treatment for their neck pain during the week preceding the trial.

Exclusion criteria were standard, and included:
  • previous treatment with acupuncture
  • inflammatory, malignant, psychiatric, infectious, or autoimmune disease
  • pain intensity less than 30mm on a VAS
  • cervical fracture, or previous surgery to the neck
  • occupation-related law suit
  • pregnancy
Patients were randomized to one of two groups:
  1. Acupuncture group (experimental) - consisted of 5 treatments in 3 weeks (2x in week 1, 2x in week 2, 1x in week 3) with dry needling for 30 minutes (manual stimulation applied every 10 minutes) at points determined by pain characteristics, and concurrent application of Vaccaria seeds to ear auricle points. Patients were then instructed to apply manual pressure to the ear points 10x/3x/day.
  2. TENS-placebo (control) - consisted of the same number of treatments performed with the patient in the prone position. Electrodes were placed on the acupuncture points for 30 minutes, but no current was passed through them.
Both groups were administered rescue medication (Diclophenac 50mg) once weekly that they could take at their own discretion. Unused medication was returned at trial completion and quantified.

The outcome measures used in this study include the pain level as measured by Visual Analogue Scale (primary outcome), the Northwick Park Neck Questionnaire (Spanish version, as the study was conducted in Spain), and the Short-Form-36 questionnaire. Patient perception of treatment efficacy was also measured with the Treatment Credibility Scale. Outcomes were measure at study inception, 3 weeks, and 6 months after the last treatment.

Pertinent Results:

  • 123 patients were randomized (average ~ 46yoa)- 45 were analyzed in total in the acupuncture group and 40 in the control group (others lost to follow-up and accounted for by appropriate statistical techniques)
  • baseline characteristics of the groups were not statistically different
  • changes in VAS pain levels were statistically better in the acupuncture group at study completion (3 weeks) - average change 44.1mm (average only 12.3mm in control group)
  • at 6 months, the changes in pain-VAS levels still remained significantly better in the acupuncture group (all other outcomes were no longer significant)
  • 70.5% of the patients in the acupuncture group took no rescue medication during the trial - compare this to only 17.7% in the control group
  • patient perception and confidence in the treatments was not statistically different
  • only mild adverse reactions to treatment were reported and affected the experimental and control groups to a similar degree

Conclusions & Practical Application:

This study had a reasonably sized study group and performed all appropriate statistical tests and measures (intention to treat analysis, multiple linear regression etc.). The results were not only statistically, but also clinically in favour of acupuncture for the treatment of chronic uncomplicated mechanical neck pain. The relative change in pain level in the acupuncture group was 62.2%, compared to only 20.4% in the control group. It is also important to note that rescue medication use was much lower in the acupuncture group.

One strength of this study is that the results were not clouded by a "sham" or "placebo" acupuncture technique. This has been one of the problems in studying acupuncture - it is very difficult to have true "sham" acupuncture, even when needles are inserted at "non-important" sites. Previous studies that have utilized this approach have shown treatment benefits even in the sham groups, suggesting a non-specific effect of acupuncture, but at the same time diminishing the strength of their results in the eyes of systematic reviewers.

This is one of the better studies to date on acupuncture for chronic mechanical neck pain. Since it was recently published, it is worthy to note that it was not included in the recent systematic review on this topic that was also published recently (Trihn K et al. Acupuncture for neck disorders. Spine 2007; 32: 236-243.).