A Cochrane review included 51 studies with a total of 2920 subjects. 18 trials were of manipulation/mobilisation versus control; 34 trials of manipulation/mobilisation versus another treatment, 1 trial had two comparisons.
Cervical manipulation versus inactive control: For subacute and chronic neck pain, a single manipulation (three trials, no meta-analysis, 154 participants) relieved pain at immediate- but not short-term follow-up.
Cervical manipulation versus another active treatment: For acute and chronic neck pain, multiple sessions of cervical manipulation (two trials, 446 participants) produced similar changes in pain, function, quality of life (QoL), global perceived effect (GPE) and patient satisfaction when compared with multiple sessions of cervical mobilisation at immediate-, short- and intermediate-term follow-up. For acute and subacute neck pain, multiple sessions of cervical manipulation were more effective than certain medications in improving pain and function at immediate- (one trial, 182 participants) and long-term follow-up (one trial, 181 participants). These findings are consistent for function at intermediate-term follow-up (one trial, 182 participants). For chronic CGH, multiple sessions of cervical manipulation (two trials, 125 participants) may be more effective than massage in improving pain and function at short/intermediate-term follow-up. Multiple sessions of cervical manipulation (one trial, 65 participants) may be favoured over transcutaneous electrical nerve stimulation (TENS) for pain reduction at short-term follow-up. For acute neck pain, multiple sessions of cervical manipulation (one trial, 20 participants) may be more effective than thoracic manipulation in improving pain and function at short/intermediate-term follow-up.
Thoracic manipulation versus inactive control: Three trials (150 participants) using a single session were assessed at immediate-, short- and intermediate-term follow-up. At short-term follow-up, manipulation improved pain in participants with acute and subacute neck pain (five trials, 346 participants, pooled SMD -1.26, 95% CI -1.86 to -0.66) and improved function (four trials, 258 participants, pooled SMD -1.40, 95% CI -2.24 to -0.55) in participants with acute and chronic neck pain.These findings were consistent at intermediate follow-up for pain/function/quality of life (one trial, 111 participants, low quality).
Thoracic manipulation versus another active treatment: No studies provided sufficient data for statistical analyses. A single session of thoracic manipulation (one trial, 100 participants, moderate quality) was comparable with thoracic mobilisation for pain relief at immediate-term follow-up for chronic neck pain.
Mobilisation versus inactive control: Mobilisation as a stand-alone intervention (two trials, 57 participants) may not reduce pain more than an inactive control.
Mobilisation versus another active treatment: For acute and subacute neck pain, anterior-posterior mobilisation (one trial, 95 participants) may favour pain reduction over rotatory or transverse mobilisations at immediate-term follow-up. For chronic CGH with temporomandibular joint (TMJ) dysfunction, multiple sessions of TMJ manual therapy (one trial, 38 participants) may be more effective than cervical mobilisation in improving pain/function at immediate- and intermediate-term follow-up. For subacute and chronic neck pain, cervical mobilisation alone (four trials, 165 participants) may not be different from ultrasound, TENS, acupuncture and massage in improving pain, function, QoL and participant satisfaction at immediate- and intermediate-term follow-up. Additionally, combining laser with manipulation may be superior to using manipulation or laser alone (one trial, 56 participants).
Cervical manipulation for subacute/chronic neck pain: Manipulation alone compared to a control provided short-term pain relief following one to four treatment sessions (SMD pooled -0.90, 95% CI -1.78 to -0.02, statistical heterogeneity I² = 80%; 3 studies, n=130) and 9 or 12 sessions were superior to 3 for pain and disability in cervicogenic headache (1 study, n=25). Manipulation and mobilisation produced similar effects on pain, function and patient satisfaction at short and intermediate-term follow-up.Thoracic manipulation for acute/chronic neck pain: Thoracic manipulation as an adjunct to electrothermal therapy reduced pain (NNT 7; 46.6% treatment advantage) and increased function (NNT 5; 40.6% treatment advantage) in acute pain in one study (n=45). A single session of thoracic manipulation compared to placebo for chronic neck pain reduced pain at immediate follow-up (NNT 5, 29% treatment advantage; 1 study, n=36).
Mobilisation for subacute/chronic neck pain: A combination of Maitland mobilisation techniques was similar to acupuncture for immediate pain relief and increased function (1 study, n=51). Neural dynamic mobilisations led to statistically insignificant, but clinically important pain reduction immediately post-treatment in participants with neck pain of mixed duration when compared to pulsed ultrasound (1 study, n=20). Certain mobilisation techniques were superior.
Adverse effects were reported in 8 of the 27 studies. Three out of those eight studies reported no side effects. Five studies reported minor and temporary side effects including headache, pain, stiffness, minor discomfort, and dizziness. Rare but serious adverse events, such as stoke or serious neurological deficits, were not reported in any of the trials.Comment: The quality of evidence is downgraded by study quality (inadequate or unclear allocation concealment and lack of blinding), by inconsistency (heterogeneity in interventions and outcomes and variability in results across studies), and by imprecise results (limited study size for each comparison).