4 An Overview on the Efficacy of Manual Therapy ( Manipulations and Mobilisations ) on Nonspecific Cervical Pain : A Systematic Review in Adults

1. "spinal manipulation", which is defined as low-amplitude, high-speed manual operations that are short, precise and selective for one vertebral segment; these manipulations are applied until the normal physiological range of motion is exceeded but without reaching the anatomical limit1; 2. "spinal mobilisation”, which is defined as passive, low-speed movements of the vertebral segments within anatomical limits2.

, the Cochrane Musculoskeletal Review Group Specialised Trials Database (April 2004), and the Web of Science (1991).
A manual search was also conducted in the Journal of Manipulative and Physiological Therapeutics, Manual Therapy, Physiotherapy, Spine and Rehabilitation (Madrid).The following keywords were used: neck pain (cervicalgia), cervical spine (columna cervical), manual therapy (terapia manual), manipulation (manipulación), mobilisation (movilización), manipulation/mobilisation, and cervical manipulation versus mobilisation in adults.The first author of each study was used in subsequent searches to avoid missing relevant studies.

Selection of studies
We selected only meta-analyses and systematic reviews of RCTs that investigated the use of mobilisations and manipulations as treatments for nonspecific mechanical neck disorders.At least one of the following parameters was measured: pain, range of movement, pain on palpation, and overall or functional improvement.We excluded RCTs that analysed cervical pain with other aetiologies.The levels of evidence were classified in various ways by the authors 4 , as shown in Table 1.The definition of RCT quality ranks the level of evidence as low, medium or high for scores below 25%, between 25% and 50%, and higher than 50%, respectively, of the total maximum 5  Table 1.Levels of evidence 4 .

Results
The role of manual therapy in nonspecific mechanical cervical pain was determined by searching the literature and examining the results by year of publication.
Five low-quality RCTs were obtained from 1991 and earlier; therefore, it was not possible to draw conclusions, and further work to produce higher-quality studies is needed 6 .
From 1992 to 1996, 24 RCTs met the selection criteria; they were categorised by the type of intervention used (12, physical medicine; 9, manual therapy; 4, more than one form of intervention; 4, drug treatment; and 3, educational).We concluded that the various treatment techniques have not been studied in sufficient detail to properly allow for an assessment of their efficacy, and that the results were contradictory 7 .
From 1997, we identified 14 RCTs totalling 892 patients.In these studies, we found the most explicit systematic reviews on the distinctions between mobilisations and manipulations and among acute, subacute and chronic pain.There were no RCTs on manipulation and only 3 low-quality RCTs on mobilisation (two of which dealt with cervical whiplash) 8 .
From 1998 to 2002, 20 medium-quality RCTs were found; these RCTs showed better results for manual therapy and exercise (manipulation or mobilisation, manipulation and mobilisation or massage) than for the control groups (waiting list or placebo).There was no evidence that treatment by manipulation was better than the control 9 .Among the 33 selected RCTs from 2003, 42% ranked as high quality 10 .There was no evidence that treatment by manipulation was better than the control treatment 11 .
From 2004 to 2010, 12 RCTs met the selection criteria.Using the criteria developed by Koes et al. 6 (and later adapted by Sarigiovannis and Hollins 5 ), the RCTs had quality scores between 25 and 67 (out of a maximum of 100).Eight were medium-low quality, of which 6 reported positive results 12,[13][14][15][16][17] and 2 reported negative results 18,19 , and 4 were high quality, of which 2 reported positive results 20 21 and 2 reported negative results 22 *,23*,24 (* are from the same study; see Table 2).Thus, the evidence for the efficacy of manual cervical spine therapy remains inconclusive 5 .To evaluate the evidence for manual therapy, an analysis of the various processes yielded the following results.
Positive: A statistically significant difference was observed in the efficacy of manual therapy (manipulation/mobilisation) and other treatments.
Negative: No statistically significant differences were observed in the efficacy of manual therapy or other treatments.
Table 2.The methodological scores of the RCTs 4 according to the adopted criteria 10 .

Acute neck pain
There were no RCTs for the treatment of acute neck pain by vertebral manipulation 25 (an absence of evidence) 4 .There were 3 low-quality RCTs for mobilisation 25 (moderate evidence) 4 and two for cervical whiplash 26,27 .In a randomised group of patients with acute neck pain (all of whom were previously treated with collars and analgesics), there were no differences among the mobilisation, TENS or control groups evaluated at 1, 6 and 12 weeks 19 .There was no evidence supporting the use of spinal manipulation 25 , and there was limited evidence against passive spinal mobilisation for acute neck pain 25 .

Cervical whiplash
For cervical whiplash, there were better pain reduction and mobility-recovery results at 8 weeks in the group treated with early active mobilisation than in those treated with conventional therapy (analgesics, advice and home exercise) 26 .There was less pain after 2 years with early active mobilisation compared with physiotherapy (cold or shortwave) or collars 27 .There was moderate evidence in favour of early active mobilisation 26,27 and no evidence supporting the use of spinal manipulation in whiplash 25 .

Chronic subacute neck pain
There were 2 RCTs 12, 28 (moderate evidence) 4 comparing manual therapy with mobilisation or spinal manipulation 7,25 .No differences were observed in short-term pain and range of movement in patients with chronic neck pain who were randomised to treatment groups for manipulation or mobilisation 28 .This result differs from other studies that showed better short-term results with manipulation 12 .Thus, there is unclear evidence on the difference in the efficacy between manipulation and spinal mobilisation 5,25 .
Four RCTs 17,31,32,33 (moderate evidence) 4 compared manipulation and/or mobilisation with other treatments 25 .There was less pain in the patients treated with mobilisation and salicylates (compared with massage and salicylates or traction/electrical stimulation and salicylates) at 3 months, but not over the long term 17 .
There were better initial results with manual therapy (manipulation and mobilisation) than with physiotherapy (short wave, electrotherapy and ultrasound), conventional therapy (analgesics, home exercises and advice) or placebo, but there were no differences at 3 or 12 months 31 .Increased muscle relaxation was achieved by manipulation, but this effect was not significant 31,32,33 (TE Global 0.42 (95% confidence interval, -0.005 to 0.85).Therefore, there is inconclusive evidence regarding the role of manual therapy (manipulation and mobilisation) in treating chronic neck pain 5,25 .
There was a better outcome in the manual therapy group (manipulation and/or mobilisation) than in the physical therapy and general medical treatment groups (analgesics, education and advice) at 7 and 26 weeks, but not after 1 year 20,29 .There was less pain, disability and drug consumption at 4 and 12 months after treatment, but with no differences between the treatment groups (physiotherapy and mobilisation, manipulation or intensive training) 18,29 .Mobilisation and strengthening exercises (isometrics with elastic bands) and mobilisation and resistance exercises (cephalic elevations in the prone and supine positions) produced better results than did the control treatment (recreational activity) in terms of reducing pain and increasing functional recovery after a year 30 .
When randomised to three treatment groups (i.e., manipulation (I), manipulation plus conventional exercise (II), or high-technology exercises, developed by MedX corporation, in addition to cervical extension isokinetic exercises (III)), no difference was observed among the groups at 3 months 22,29 .However, there was higher satisfaction at the end of 3 months in Group II 22,29 , better results and higher satisfaction at 12 months in Groups II and III 22,29 , better results at 24 months in Groups II and III 10,23 and higher satisfaction at 24 months in Group II 10,23 .
There was moderate short-and long-term evidence in favour of stretching programmes plus strengthening for chronic mechanical neck pain with or without headache 29 , and for stretching programmes for patients with chronic mechanical neck pain 10 .There was inconclusive evidence supporting the relative benefits of a programme of stretching plus strengthening exercises compared to manual therapy (mobilisation and manipulation) or to other therapeutic approaches 21,25,26,29  There was strong evidence against manipulation 12,33 and manipulation plus mobilisation 31,34,35 in isolation compared with controls (placebo/waiting list) for function and the general perceived effect in subacute or chronic mechanical neck disorders with or without headaches.
There was strong short-35-37* and long-term 35 ** evidence in favour of multimodal treatment (manipulation/mobilisation + exercise) compared with controls in subacute or chronic mechanical neck disorders with or without headaches for the following measures:

www.intechopen.com
There was inconclusive short-and long-term evidence on the effect of strengthening exercises for the relief of chronic mechanical neck pain 10 and on the role of manual therapy in chronic neck pain 5,10,25,29 .There were no clear differences between exercises and manual techniques or other physical therapies, or between strengthening and resistance exercises 11 .It was thus not possible to determine which technique or dosage is most effective or whether certain groups benefit more from a given form of therapy 10,25 .
Regarding external validity, there have been comparative metaanalysis reviews of treatment by manipulation and mobilisation in mechanical neck disorders.These studies have used the resulting pain after treatment as a measure 10,21 , and the most significant data are shown in Table 3.These studies demonstrate that there is inconclusive evidence for the efficacy of manual therapy (manipulation and mobilisation) in chronic neck pain 5, 10, 25,28 .

Radiating neck pain
There was greater improvement in function and pain with manual therapy (manipulation/mobilisation) directly on the cervical spine and indirectly on the shoulder and dorsal spine than without treatment 37 .We found limited evidence in favour of exercise and manual therapy (mobilisation/manipulation) in radiating chronic neck pain cases 9,10,25,37 .There was no evidence for the role of manual therapy (manipulation and mobilisation) in radicular cervical cases 10 .There was moderate evidence against stretching programmes plus strengthening for myofascial pain in the neck and shoulder 5,29,31 .Consequently, the evidence for the efficacy of manual therapy (manipulation and mobilisation) was inconclusive 5,29,31 .

Cervicogenic headaches
At both 7 weeks and after 1 year, the intensity and frequency of pain decreased more with manipulation, exercise, and manipulation with exercise than it did with no treatment 35 , and combining the treatments (manipulation and exercise) did not change the results 35 .There was strong short-and long-term evidence in favour of multimodal treatments that included exercise and mobilisations in subacute or chronic mechanical neck pain with headache, as assessed by pain reduction, improved function and general perceived effect 29 .There was moderate short-and long-term evidence in favour of strengthening exercises alone or with other treatments for pain, function and general perceived effect on chronic neck pain with headache 35 .There was also evidence against the efficacy of manipulation and/or mobilisation alone in the treatment of cervicogenic headaches 10,11 .Thus, there is inconclusive evidence regarding the efficacy of manual therapy (isolated manipulation and mobilisation) in the treatment of cervicogenic headaches 10,11 .

Safety
Most sources indicated that the incidence of serious accidents during treatment by cervical spine manipulation is low (approximately 1 per million per year) 1,3 .The most commonly described injuries were Wallenberg's syndrome, dissection or thrombosis of the vertebral or carotid arteries and brainstem injury 1 .Adverse reactions were more likely to occur after manipulation than after cervical spinal mobilisation 3 .Therefore, iatrogenic sequelae may be reduced, outcomes may be improved, and satisfaction and security may be increased by using mobilisation 1 .Only one study reported adverse effects from manual therapy 21 .
Spinal manipulation and mobilisation are commonly used in the treatment of cervical spine disorders 35 .Their use has been associated with serious complications, including an increased incidence of cerebrovascular accidents (CVAs) [39][40][41][42] and minor side effects, such as headache, stiffness, and symptom worsening [43][44][45] .In a systematic review of the adverse effects of spinal manipulation, Ernst 41 suggested that spinal manipulation is associated with frequent mild and transient adverse effects, as well as more serious complications that can lead to permanent disability or death.The incidence of reported adverse effects has varied between studies (ranging, for example, between 1 per 50,000 manipulations 44 and 1 per 228,050 43,46 ).
A review of the literature related to cervical artery dysfunction and manual therapy suggests that due to reporting bias, inferences about the magnitude of the risks of manipulative therapy should be conservative in relation to the surveys 46 .Other authors have also stated that due to concerns about the validity of the calculations applied to these data, it is not currently possible to estimate the risk of complications after treatment in a meaningful way without reporting the incidence of risk for cervical manipulation 47 .
Ernst concluded that incidence figures cannot be reliably estimated at present, due to the lack of sufficiently broad and rigorous prospective studies 41 .Thiel and Bolton have suggested the need for a system to record adverse effects on a routine basis that is not based on the practitioner's subjective recall 48 .Several tests also exist to gauge the risk of adverse effects, with and without the use of mobilisation or high-velocity thrust (HVT) techniques, which have not been as widely reported in the literature 43 .
Beca (2002) reported a higher incidence of minor adverse reactions with the use of non-HVT techniques (27.5%) compared with HVT techniques (16.1%).Magarey reported a higher rate of adverse effects associated with the use of non-HVT techniques (1 out of 180 therapists per week of treatment) compared with HVT techniques (1 out of 177.5 therapists per week of treatment) 44 .Magarey also reported that adverse effects were caused by the particular test procedures, which involved rotation.In contrast, Hurwitz reported that patients who received spinal manipulation were more likely to experience adverse effects than were patients treated with mobilisation; however, his risk estimates were imprecise.The reported side effects associated with mobilisation included increased pain, headache and fatigue 49 .
The use of functional testing of the position of the cervical spine has been proposed as part of the evaluation of vertebrobasilar insufficiency (VBI) before the application of HVT and non-HVT techniques to the cervical spine 50 .However, functional position tests have been criticised for their "lack of sensitivity, specificity 51 and validity 42" .The poor validity of the functional position tests for the detection of alterations in blood flow has also been noted [52][53][54] .
It appears that the risk of adverse reactions is associated with the testing procedures themselves 39 coupled with the time consumed by the testing 44 , suggesting that the clinical utility of functional position testing is questionable.However, these tests are currently defended for VBI assessment as part of a comprehensive assessment protocol that also includes a detailed subjective evaluation and places special emphasis on the therapists' clinical reasoning in the process 50 .

Discussion
The definition and concept of manual therapy varies according to different authors.For example, the study with the highest-scoring methodology includes mobilisation of the spine and soft tissue through coordination and stabilisation exercises 20 .The wide variety of manipulative techniques used and the qualifications of the professionals involved make comparing studies difficult 20 .Virtually all authors have agreed on the need for high-quality and long-term RCTs to establish precisely the efficacy and safety of manual therapy 1,2 , thereby facilitating meta-analyses rather than only systematic reviews 5 .It is encouraging to note that the three papers with the highest scores for methodological quality were published after the year 2000 20,21,23 .However, none of these RCTs included pre-randomisation, and only one included post-randomisation for psychosocial assessment of the patients 21 .
Additionally, the qualifications or professional experience of the manual therapist were not considered; these qualifications are important for the proper indication and application of cervical spinal manipulations 5 .
It would also be desirable to implement placebo treatments that are as similar as possible to manual therapy techniques but without any specific activity 5 .However, the absence of evidence for the efficacy of physical treatment does not mean that such treatments are not effective (according to evidence based medicine), although the evidence does suggest that manual spinal therapy has a definite placebo effect 1,2 .
There is a need for higher-quality and longer-term RCTs to demonstrate the efficacy and safety of manual therapy in general, and of its main techniques (manipulation and mobilisation) for mechanical cervical spine disorders in particular 55 .There should be a national notification system for adverse effects, applied on a routine basis, that utilises a protocol for collecting the adverse effects associated with the use of these techniques and the therapist's VBI assessment"56".
There is no evidence to suggest that physiotherapists are better qualified and are more effective in the application of cervical spinal manipulations than are other healthcare professionals"57".
The populations with neck pain, with or without headaches, in the RCTs were quite homogeneous.
Howe"32" reported a rapid and significant improvement of symptoms in patients with a painful or rigid neck, pain or paresthesia in the shoulder, or pain or paresthesia in the hand.
The main weaknesses of this study include the following: im sub-optimal randomisation and a failure to mention drop-outs Bitterli"34" reported an improvement of 35% in the group receiving early active mobilisation, but found no improvement after spinal manipulation.This study has a high risk of bias due to the low quality of the methodological design (non-randomised trial, small sample).
Jull (2002) "35" reported a reduction of the frequency and intensity of headache and neck pain when using spinal manipulation, and the effect lasted until the 12-month follow-up.However, the inability to control the placebo effect could increase the risk of bias (see Tables 4, 5, and 6).This review has some limitations.Although we used broad search criteria, we cannot guarantee that we did not miss any relevant publications.Due to the number of RTCs reviewed, the total number of subjects, and the low design quality, it is difficult to draw clear conclusions.Although the study populations in the RCTs are quite homogenous, it is not possible to perform a meta-analysis.
This overeview had the advantages of spanning the available literature on nonspecific cervical pain, included only the highest-quality studies, and used recommended methods for systematic reviews.
In the future, studies evaluating the efficacy and safety of manual therapy should be designed according to the international CONSORT recommendations.Furthermore, investigators need be very careful when performing sample size calculations in order to avoid sources of bias.
The information in the studies should be sufficient to allow researchers to reproduce the results independently.The data could suggest a bias in favour of physiotherapists for the treatment of neck pain.However, this bias does not mean that physiotherapists are better qualified or that they are more effective in the application of cervical spinal manipulations compared with other healthcare professionals (see Tables 7, 8, and 9).

Conclusions
There is no evidence to support the use of spinal manipulations for acute neck pain, and there is limited evidence against passive spinal mobilisation.However, there is strong evidence against manipulation alone or manipulation in addition to mobilisation in isolation compared with control groups (placebo/waiting list) in terms of improving function and the general perceived effect for the treatment of subacute or chronic mechanical neck disorders with or without headache.There is strong short-and long-term evidence in favour of multimodal treatments (manipulation/mobilisation plus exercise) compared with control groups for reducing pain, thereby improving function and the general perceived effect in subacute or chronic mechanical neck disorders with or without headaches.
There is no evidence to support the role of manual therapy in cervical radicular conditions.There is evidence against the efficacy of isolated manipulation and/or mobilisation in the treatment of cervicogenic headaches.There were no serious adverse effects associated with the use of HVT techniques; a number of minor adverse effects were reported.The adverse effects associated with the use of non-HVT techniques were more serious and included a transient ischaemic attack, a fall due to this attack, and a fainting episode.The adverse effects associated with the use of non-HVT techniques justify a specific investigation, especially in view of their widespread use on the upper cervical spine.
There was a low utilisation of VBI assessment protocols, and the questionable utility of VBI assessment protocols in clinical practice was highlighted in one study.Positional VBI tests cannot detect all of the patients at risk of adverse effects associated with the use of manual therapy.Additional large-scale studies are needed to investigate the risk of serious adverse reactions associated with the use of both HVT and non-HVT techniques.Ideally, this research should not depend solely on subjective information obtained from providers, as was the case in this study.Finally, a notification system for adverse effects should be used on a routine basis.This system should incorporate protocols for collecting the adverse effects associated with the use of these techniques and the therapist's VBI assessment.
EX = exercise; RCT = randomised clinical trial; SM = spinal manipulation, transcutaneous electrical nerve stimulation; VAS = visual analogue scale; -= inconclusive results.Table 4. Spinal manipulation for the treatment of cervicogenic headaches in three of the included RCTs low risk of bias; 0 = unclear risk of bias; < 0 = high risk of bias. .

Table 5 .
59ality Evaluation in three of the included RCTs (Jadad score)59

Table 6 .
59ality Evaluation in three of the included RCTs (Cochrane tool)59