|Conservative interventions provide short-term relief for non-specific neck pain: a systematic review
|Leaver AM, Refshauge KM, Maher CG, McAuley JH
This review found that some conservative interventions for non-specific neck pain improved pain or disability in the short term, but not in the long term. These conclusions appear to be reliable, but the review was limited by its failure to measure and explore statistical variation between the included trials.
To evaluate the effectiveness of interventions for non-specific neck pain in reducing pain and disability.
MEDLINE, CINAHL, EMBASE, PEDro and Cochrane Central Register of Controlled Trials (CENTRAL) were searched for articles from their inception to February 2008. Search terms were reported in an online appendix. The reference lists of included studies and relevant systematic reviews were checked. The search was limited to studies in English or those that had a suitable translation in English.
Eligible studies were randomised controlled trials (RCTs) of adults (aged over 18 years) with non-specific neck pain, in which any intervention was compared with placebo, a sham intervention, minimal intervention (such as waiting list) or no intervention (such as advice). Patients with specific diagnoses that were not confirmed by diagnostic tests were considered to have non-specific neck pain. The authors excluded trials of neck-pain prevention; those that did not specify neck pain in their inclusion criteria; those without separate outcomes for participants with neck pain; and those of participants with whiplash or trauma-associated neck pain. The outcomes had to be acute or chronic neck pain (assessed on a numeric scale) or disability (assessed on a multi-item scale).
Most participants in the included trials had neck pain of chronic or mixed duration. Interventions included medication, relaxation, acupuncture, exercise, manual therapy, multiple modes, and electrotherapy. No trials used surgery, injections or radiofrequency neurotomy. Controls received a sham physical intervention, minimal or no intervention, or placebo medication. Outcome measures included a numeric pain scale, the pain improvement rate, and a wide variety of disability scales.
Articles retrieved by the search were initially screened for relevance by one author. Relevant articles were screened independently by two reviewers who selected trials for inclusion. Disagreements were resolved by discussion.
Assessment of study quality
Trial quality was assessed using the PEDro scale. Items covered randomisation, allocation concealment, baseline similarity of groups, blinding, drop-out rate (15% or more), intention-to-treat analysis, and reporting of between-group differences and measures of variation. Trials were given up to 10 points for quality.
Where PEDro quality scores were available, for trials in the PEDro database, these were used. Other trials were assessed independently by two trained assessors.
Relative risks were extracted or calculated for dichotomous outcomes, and means and standard deviations were extracted for continuous outcomes, for each study, with 95% confidence intervals. If necessary, means and standard deviations were imputed, using published methods. The outcomes were categorised as short term (at or near the end of treatment), medium term (between three and nine months) or long term (over nine months). Pain and disability scores were converted to a scale from zero to 100 before calculating the effect size.
The data were extracted independently by two reviewers.
Methods of synthesis
Where appropriate, the data were pooled in a random-effects meta-analysis to calculate the pooled risk ratios and weighted mean differences, with 95% confidence intervals.
Results of the review
Thirty-three RCTs were included in the review. Trial quality was high, with 60% of trials (20 out of 33) scoring at least seven points out of 10 on the PEDro scale. The quality criteria were met for all 33 trials for randomisation, 18 trials for allocation concealment, 28 trials for baseline similarity, 16 trials for participant blinding, seven trials for therapist blinding, 21 trials for assessor blinding, 25 trials for drop-out rates, 11 trials for intention-to-treat analysis, and 30 to 33 trials for outcome reporting.
Pain outcomes: Interventions with a significant short-term benefit were manipulation (MD -22, 95% CI -32 to -11, three RCTs), multi-modal physical therapy with spinal manual therapy (MD -21, 95% CI -34 to -7, one RCT), specific exercise (MD -12, 95% CI -22 to -2, three RCTs), orphenadrine with paracetamol (MD -17, 95% CI -32 to -2, one RCT), and manual therapy (MD -12, 95% CI -16 to-7, two RCTs). Laser therapy improved medium-term outcomes (MD -20, 95% CI -33 to -7, five RCTs). No significant benefit was found from other interventions for the short, medium or long term (one to six trials; details were reported).
Disability outcomes: Interventions with a significant short-term benefit were acupuncture (MD -8, 95% CI -13 to -2, five RCTs) and manual therapy (MD -6, 95% CI -11 to -2, three RCTs). Significant medium-term benefit was found from laser therapy (MD -10, 95% CI -15 to -6, three RCTs; corrected data, see Other Publications of Related Interest). No significant benefit was found from other interventions for the short, medium or long term (one to three trials; details were reported).
Other findings were reported.
Some conservative interventions for non-specific neck pain improved pain or disability in the short term, but not in the long term.
The objectives and inclusion criteria were clear and relevant sources were searched for trials. It was unclear whether the search was limited by publication status and no attempts were made to retrieve unpublished trials, so the review may have been subject to publication bias. Limiting inclusion to translatable studies did not seem to exclude any trials from the review. It was unclear whether sufficient steps were taken to minimise the risk of reviewer bias and error in the initial stages of study selection.
Appropriate methods were used to assess study quality, and overall it was high. The statistical methods used to combine the data appear to have been appropriate, except that heterogeneity between the trials was not assessed and not explored. Visual assessment of the forest plots suggested that some of the pooled analyses had high heterogeneity and that the data might not have been suitable for pooling.
The authors' conclusions appear to be reliable, but the review was limited by its failure to measure and explore statistical variation between the included trials.
Implications of the review for practice and research
Practice: The authors stated that their findings supported combined manual exercise and physical therapy for non-specific neck pain, in the short term. Manual exercise could include neck manipulation, thoracic manipulation or neck mobilisation. Exercise that targeted specific impairments, such as head repositioning, seemed effective. General strength and conditioning programmes did not seem effective.
Research: The authors stated that trials were needed on laser therapy (in view of the inconsistency of evidence at different time points) and surgery (in view of the high potential for serious adverse events and high cost), and that there was a need for greater consistency in classifying neck pain and associated conditions. Trials should include a placebo or minimal or no intervention group. Reassurance, self-care advice and simple analgesics should be investigated as additions to high-quality care.
Funding received from the University of Sydney, and the National Health and Medical Research Council.
Leaver AM, Refshauge KM, Maher CG, McAuley JH. Conservative interventions provide short-term relief for non-specific neck pain: a systematic review. Journal of Physiotherapy 2010; 56(2): 73-85
Other publications of related interest
Erratum in: Leaver AM, Refshauge KM, Maher CG, McAuley JH. Conservative interventions provide short-term relief for non-specific neck pain: a systematic review. Journal of Physiotherapy 2010; 56(4): 222.
Subject indexing assigned by NLM
Acupuncture Therapy; Adult; Disability Evaluation; Electric Stimulation Therapy; Exercise Therapy; Humans; Laser Therapy; Magnetics; Neck Pain /therapy; Physical Therapy Modalities; Randomized Controlled Trials as Topic; Relaxation Therapy
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.