Roland Bryans, DC, Philip Decina, DC, Martin Descarreaux, DC, PhD,
Mireille Duranleau, DC, Henri Marcoux, DC, Brock Potter, BSc, DC,
Richard P. Ruegg, PhD, DCs, Lynn Shaw, PhD, OT,
Robert Watkin, BA, LLB, Eleanor White, MSc, DC
Guidelines Development Committee (GDC) Chairman;
OBJECTIVE: The purpose of this study was to develop evidence-based treatment recommendations for the treatment of nonspecific (mechanical) neck pain in adults.
METHODS: Systematic literature searches of controlled clinical trials published through December 2011 relevant to chiropractic practice were conducted using the databases MEDLINE, EMBASE, EMCARE, Index to Chiropractic Literature, and the Cochrane Library. The number, quality, and consistency of findings were considered to assign an overall strength of evidence (strong, moderate, weak, or conflicting) and to formulate treatment recommendations.
RESULTS: Forty-one randomized controlled trials meeting the inclusion criteria and scoring a low risk of bias were used to develop 11 treatment recommendations. Strong recommendations were made for the treatment of chronic neck pain with manipulation, manual therapy, and exercise in combination with other modalities. Strong recommendations were also made for the treatment of chronic neck pain with stretching, strengthening, and endurance exercises alone. Moderate recommendations were made for the treatment of acute neck pain with manipulation and mobilization in combination with other modalities. Moderate recommendations were made for the treatment of chronic neck pain with mobilization as well as massage in combination with other therapies. A weak recommendation was made for the treatment of acute neck pain with exercise alone and the treatment of chronic neck pain with manipulation alone. Thoracic manipulation and trigger point therapy could not be recommended for the treatment of acute neck pain. Transcutaneous nerve stimulation, thoracic manipulation, laser, and traction could not be recommended for the treatment of chronic neck pain.
CONCLUSIONS: Interventions commonly used in chiropractic care improve outcomes for the treatment of acute and chronic neck pain. Increased benefit has been shown in several instances where a multimodal approach to neck pain has been used.
The annual prevalence of nonspecific neck pain is estimated to range between 30% and 50%.  Persistent or recurrent neck pain continues to be reported by 50% to 85% of patients 1 to 5 years after initial onset.  Its course is usually episodic, and complete recovery is uncommon for most patients.  Twenty-seven percent of patients seeking chiropractic treatment report neck or cervical problems.  Thus, treatment of neck pain is an integral part of chiropractic practice.
Treatment modalities typically used by doctors of chiropractic (DCs) to care for patients with neck pain include spinal manipulation, mobilization, device-assisted spinal manipulation, education about modifiable lifestyle factors, physical therapy modalities, heat/ice, massage, soft tissue therapies such as trigger point therapy, and strengthening and stretching exercises. There is a growing expectation for DCs and other health professionals to adopt and use research-based knowledge, taking sufficient account of the quality of available research evidence to inform clinical practice. As a result, the purpose of the Canadian Chiropractic Association and the Federation Clinical Practice Guidelines Project is to develop evidence-based treatment guidelines. The clinical practice guideline (CPG) experience began in Canada with a consensus conference in April of 1993 that culminated with the publication of “Clinical Guidelines for Chiropractic Practice in Canada”  in 1994. Since then, the chiropractic profession in Canada has published 3 additional guidelines [6-8] that are intended to provide practitioners with the most current evidence for the treatment for patients in light of the clinician’s experience and the patient’s preferences.
The original Neck Pain Guideline  published in 2005 relied on studies that were drawn from the literature in a search conducted up to October 2004. The treatment recommendations developed at that time were supported largely by the expert opinion of the Guidelines Development Committee (GDC) in the absence of a solid, high-quality research base. Therefore, an update to the earlier neck pain guidelines that reflects evidence extracted from the published scientific literature about effective chiropractic treatment(s) for adult patients with nonspecific neck pain was needed. The purposes of this study were to develop evidence-based treatment recommendations for the treatment of nonspecific (mechanical) neck pain in adults and to present recommendations synthesized from this evidence and strength rating of each recommendation.
In this guideline, recommendations have been developed that updates the body of evidence supporting chiropractic treatment of neck pain. These recommendations offer a broad range of evidence-based treatment options for practitioners to use in patient-centered care. The development of these recommendations reflects the most recent evidence (2004 or later), which is limited to low-risk-of-bias studies. Wherever possible, recommendations were made for each of the treatment modalities identified as relevant to common chiropractic practice and for which current evidence was available. Limitations in the current evidence are described and used in making suggestions for advancing the quality of future research.
During review of the materials, a generalizable weakness of the studies was noted including the heterogeneity of treatment protocols (ie, the use of a primary intervention in combination with other therapeutic treatments). For example, many of the studies on manipulation were pragmatic and therefore included exercises, advice, and soft tissue work, thus making it difficult or impossible to isolate the therapeutic effect as a “stand-alone” intervention. When therapies are combined, for example, the use of manipulation with electrotherapy or exercise, it was sometimes possible to address making recommendations for the particular intervention “when provided in combination with.” In other instances, interventions are provided in combination with so many other treatment modalities, for example, manipulation with exercise, advice, stretching, and pulsed shortwave diathermy, that a recommendation can only be structured for a “multimodal” form of intervention. In developing treatment recommendations for multimodal interventions, the GDC considered the manner in which practitioners would apply them. We believe that, in many instances, the practitioner uses more than 1 treatment modality in the management of patients with nonspecific neck pain. All studies in which participants received more than 1 intervention or interventions in addition to the primary intervention being investigated are noted, and the recommendation was referenced as multimodal.
Several of the treatment recommendations in this document are diminished by some of the studies that based findings on too few study participants. Specific studies of “low subject numbers” are identified and recorded in The Literature Summary (Table 4). Although this limitation was considered a contributing factor to the imprecision of results and, ultimately, clinical relevance, our recommendations would be fortified by greater participant numbers and clinical relevance.
The inclusion of participants with variable duration of symptoms in a study made it difficult to formulate recommendations. In some cases, it was impossible to determine whether the observed effects (or lack of effect) of an intervention was caused by its impact on participants with acute, subacute, or chronic neck pain. Valuable data may have been missed in excluding studies in which the chronicity of the pain among the participants could not be determined (see above). Despite the positive outcomes reported, no recommendations could be formulated for neck pain of variable duration for the manual therapy, [33, 41] TENS,  thoracic manipulation, [25, 26], or traction  interventions.
Developing treatment recommendations related to the diversity of interventions reported as exercise (stability, mobility, relaxation, rehabilitation, range of motion, strength and endurance exercises, as well as stretching) was challenging. Although few studies are directly comparable in terms of the form of exercise used as the intervention, all demonstrated a degree of benefit for the participant.
Similarly, the breadth, diversity, and understanding of the intervention described as patient education (advice, training, supervision, and instruction of any kind provided to the patient) were a challenge. Many of the studies reported the inclusion of patient education (either generally or very specifically). In this article, the 11 RCTs identified as patient education were allocated to the exercise category because they specifically dealt with patient education and exercise. All encounters between the patient and practitioner incorporate at least some form of education to the patient. This component of care is essential when directing a patient for the elements of active care (eg, exercise). In addition, patients receiving the described interventions of passive care (eg, manipulation, mobilization, massage, etc) are also educated with regard to diagnostic, investigative, and treatment procedures; anticipated outcomes; potential adverse events; informed consent, and so on. Whenever the author(s) of a study has included an element of patient education as part of the treatment protocol, it has been included as part of the recommendation.
Comparison with SRs
As a result of the search and screening process,  current (2005 or later) SRs were identified that assessed the literature with regard to therapeutic benefit for the 10 treatment modalities reviewed in this guideline (Table 6). Although the SRs are considered current, the literature that they assess included studies that are sometimes much older. By contrast, the studies assessed in this guideline were limited to much more recent publications (2005 or later) and generally reflect a higher quality (low risk of bias). A number of SRs (N = 13) assessed the literature for more than 1 treatment modality and, of these, 7 identified interventions that were delivered in combination with other therapies (multimodal).
In general, the individual SR findings within an intervention category remained fairly consistent. For example, within the category of manipulation, 11 of 12 SRs identified by the search suggested some degree of therapeutic benefit from the intervention. Similarly, of the 13 SRs for exercise, all but 1 concluded that therapeutic benefit had been evidenced. Eleven SRs assessed the evidence for only 1 intervention.
In comparing the treatment recommendations of this guideline with the findings of the relevant SRs, there would appear to be a general agreement. However, inconsistency within the SR findings or a paucity of high-quality evidence precludes complete agreement in the cases of massage, traction, and trigger point therapy. In these 3 instances, the SRs predate the studies used in developing the recommendations.
There were no serious adverse events reported in any of the citations used in developing these treatment recommendations. A summary of the adverse event reporting from the literature summary (Table 4) is shown in Table 7. Of the 43 studies included in this summary, 14 made no mention of adverse events. Of the remaining , all studies reported either none or only minor adverse events from a total of 1682 study participants and several treatment sessions (on average) per participant.
Considerations for Future Research
Since our original neck pain guideline published in 2005,  the number and quality of clinical trials in chiropractic care have increased significantly. Nonetheless, as a result of our experience in developing these practice guidelines, we would suggest the following be considered to help guide future studies.
We suggest the investigation of treatment interventions on a stand-alone basis that will allow the treatment outcomes to be evaluated without the influence of other forms of care. For example, when manipulative therapy is provided in combination with exercise, heat, cold, and so on, the benefit of the intervention becomes difficult to interpret, especially when the auxiliary therapies have also been shown to be of benefit.
The use of placebo, control, or sham comparators (whenever ethical) to determine the efficacy of a stand-alone treatment intervention is suggested. When comparing the outcomes of 2 or more interventions, it becomes increasingly difficult to establish if any of the treatment modalities provides anything more than placebo effect or the natural history of recovery, especially in instances of acute neck pain. In several instances, improvements that were identified in patient outcomes were frequently seen as “no better than” or “as good as” 2 or more interventions. Typically, no references are made to the natural history or progression of the condition.
A more thorough reporting of adverse events in the course of conducting a study for the balancing of benefit against risk when considering treatment options is needed. Although some studies do report that adverse events were queried and tracked by the researchers/clinicians, they were frequently reported as “none” or “minor,” with no additional information being provided. In other instances, there was simply no mention of adverse events whatsoever.
We suggest that authors clearly define and identify the composition of the participant pool in terms of the duration of symptoms (acute, subacute, and chronic) and that the reporting of results (outcomes) be separated for each “duration of symptoms” group. The results of some studies were reported for groups that included a mix of participants with acute, subacute, and chronic symptoms. Consequently, it was not possible to determine if one group fared better than another or if the response was truly shared. It appears that the focus of neck pain research remains on the chronic condition.
In summary, researchers are encouraged to use suitable controls as experimental comparators. We also suggest a clear separation of participants with acute and chronic symptoms within studies as well as a more thorough reporting of the occurrence or absence of adverse events. The investigation of treatment modalities on a stand-alone basis is needed.
The limitations of this study are consistent with those of SRs and clinical guidelines development. Although we made every attempt to include all relevant studies, it is possible that other relevant literature was missed. This study is limited in that literature was searched through December 2011; therefore, more recent literature studies in the publication process were not included in the recommendations. Thus, best judgement should be used to incorporate new high-quality evidence.
Although the focus of the guideline development was on chiropractic treatments, other stakeholders or contributions to what DCs do in practice could have been missed. The literature searched may have included procedures that DCs perform, but the research did not include practicing DCs and thus was omitted from our study. As with any use of the literature, we are limited by what has been published. Thus, publication bias may have an influence in the types of studies or topics included in our searches.
There are inherent limitations in guideline development. Expert opinion and interpretation are necessary procedures for guideline development. Thus, some subjectivity in judgments is present when assessing the strength of the evidence. Also, when evidence is lacking, expert opinion is required.
The studies included in this guideline indicate that cervical manipulation, mobilization, manual therapy, exercise, and massage can be recommended for the chiropractic treatment of nonspecific, mechanical neck pain. The strongest recommendations are typically made for the primary intervention in combination with another intervention, usually exercise and/or patient education. Owing to conflicting findings in the literature, no recommendation could be made for laser, TENS, or thoracic manipulation in the treatment of chronic neck pain or for the use of thoracic manipulation in the treatment of acute neck pain. There is a lack of evidence to support the use of laser, trigger point therapy, or traction for nonspecific, mechanical neck pain in adults.