OUP user menu

What Are the Clinical Criteria Justifying Spinal Manipulative Therapy for Neck Pain? A Systematic Review of Randomized Controlled Trials

Josephine Smith M Phty, Philip S. Bolton DC, PhD
DOI: http://dx.doi.org/10.1111/pme.12041 460-468 First published online: 1 April 2013


Objective Manipulation and mobilization are used to treat neck pain. However, little is known about the diagnostic criteria used to determine the need for manipulation in cases of neck pain. The primary aim of this study was to determine what diagnostic criteria are used to identify which neck pain sufferers should receive spinal manipulation or mobilization.

Design We systematically reviewed randomized controlled trials (RCT) involving mobilization or manipulation for neck pain. A data extraction pro forma was developed and trialled before two independent assessors extracted data sets from each RCT. A descriptive analysis was undertaken.

Results Thirty RCTs met the inclusion criteria. Acute and chronic “Mechanical” neck pain was the most common (43%) diagnosis at recruitment to the RCTs but some (10%) included patients with cervicogenic headache. Clinical criteria were used to determine the need for neck manipulation in over half (63%) of the RCTs. This usually involved exclusion of serious conditions, manual examination for tenderness on palpation, and/or altered vertebral motion in the neck or upper thoracic region which are known to lack validity. The remainder of the RCTs did not report a diagnostic strategy. All RCTs lacked detail descriptions of diagnostic criteria or interventions used.

Conclusions This systematic review highlights the absence of reliable and valid diagnostic protocols to determine the need for spinal manipulation in persons presenting with non-serious, idiopathic, or whiplash-associated (grade II) neck pain. Guidelines requiring the reporting of valid diagnostic criteria are needed to improve the quality of RCTs concerning manual therapy.

  • Manual Therapy
  • Neck Pain
  • Randomized Controlled Trial
  • Systematic Review
  • Spinal Manipulative Therapy


Cervical spine manipulation is one of several treatments available for conservative care of neck pain. Patients seen by pain physicians may have undergone manipulative therapy as part of their previous treatment, or pain physicians may consider manipulation as an option when invasive treatments are not indicated or prove ineffective. Some pain physicians may collaborate with physical therapists or chiropractors who provide cervical spine manipulation as part of a multidisciplinary program of management.

There is moderate to low-quality evidence that manipulation may be of benefit in the treatment of neck pain , but cervical spine manipulation is not without risks. Up to half of patients undergoing manipulation experience at least one adverse reaction, usually involving local discomfort typically lasting 24 hours , but vertebrobasilar stroke is a rare but catastrophic complication . This complication is thought by some to be a failure to diagnose an ischemic event presenting as neck pain rather than the manipulation being the cause . Consequently, it is important to have a clear diagnostic rational for considering a patient for cervical spine manipulation.

Pain physicians might trust that cervical spine manipulation is a specific treatment for a specific diagnosable entity, and that those who provide manipulation take steps to make specific diagnosis before undertaking this therapy. However, although this might be a respectful image of manipulative therapy, it is not necessarily the case. There are over 290 synonyms used to describe the entity or lesion that chiropractors and others manipulate,, but formal evidence of their ability to detect and treat these entities is lacking.

The present study was undertaken with the primary aim to determine what clinical criteria or characteristics are used to diagnose the need for cervical spine manipulation. A secondary aim was to consider the reliability and validity of these features. A clear understanding of these determinants is necessary if prudent referral and case management is to be achieved by pain physicians and others who might consider manipulation as a treatment for neck pain.


In order to determine the diagnostic criteria used by manual therapists, we undertook a systematic review of randomized controlled trials (RCTs) of spinal manipulation for neck pain. The review was limited to RCTs because they represent the highest grade of evidence regarding clinical studies and because these studies are expected to report the inclusion or eligibility criteria for the treatment being studied . The RCTs used were those included in the most recent Cochrane Collaboration on manipulation and mobilization , supplemented by an updated set of RCTs that satisfied the same inclusion criteria as used in the Cochrane review.

We obtained copies of the articles reporting the 27 RCTs that met the inclusion criteria (Table ). These were a subset of the 114 publications reporting data from 68 RCTs identified by the Cochrane review , supplemented by publications that appeared between July 2009 and December 2010 that were found using the same databases, the same search strategy, and the same inclusion criteria as that review.

View this table:
Table 1

Inclusion and exclusion criteria used by Gross et al. to select the 27 randomized control trial included in their review of Manipulation or Mobilisation for Neck Pain and used in this systematic review

Randomized control trials and quasi-randomized control trialsNeck disorders caused by definite or possible long tract signs
No restrictions on methodological quality of randomized control trials or quasi-randomized control trialsPresence of other pathological entities
Manipulation or mobilization techniques used alone as a single-modal treatmentPresence of headache of non-cervical origin but associated with the neck
“Manipulation” defined as localized force of high velocity and low amplitude directed at specific spinal segments.Coexisting headache when either neck pain was not dominant or headache was not provoked by neck movements or by sustained postures
“Mobilization” defined as low grade/velocity, small or large amplitude passive movement techniques, or neuromuscular techniques within the patient's range of motion and within in their controlPresence of mixed headache which included more than one headache type
Could be applied to any region of vertebral column
Guiding principle was intention to treat neck pain
Adult (>18 year old) participants in study

A pro forma data extraction instrument was constructed to identify the criteria reported to determine the indication for manipulative therapy and the validity of these criteria (Table ). Two reviewers (one physiotherapist and one chiropractor) used the pro forma to extract data from each publication. A priori, it was agreed that disagreement regarding data extraction would result in it being reported separately.

View this table:
Table 2

Data extraction pro forma developed and used in this systematic review. Grey-shaded questions require data entry



The literature search identified five articles that were published since the Cochrane review (Table ). Two of the eligible articles were not included in the present study because we were not able to translate them satisfactorily. The final analysis was based on the 25 remaining articles (Table ). The two reviewers fully agreed on data extraction.

View this table:
Table 3

Table summarizing the inception diagnosis, the primary complaint, its duration and intervention site, and technique for each randomized control trial (RCT) included in this systematic review. Table 3A list all articles reporting RCTs reviewed by Gross et al. in July 2009 and Table 3B list eligible RCTs identified by us for the period July 2009 to December 2010. Full reference is identified in brackets “[ ].” Question mark “?” indicates not clearly defined in original article

Table 3A
Inception Diagnosis AuthorPrimary ComplaintDuration of ComplaintIntervention Site/Technique
Neck pain
Hurwitz et al. 2002
Neck pain and other cervicogenic headache radicular signs and symptomsAcute and chronicCervical or upper thoracic/manipulation or ? and mobilization
Neck pain
Wood et al. 2001
Neck painSubacuteCervical/manipulation
Neck pain
Yurkiw and Mior 1996
Neck painSubacute?
Insidious neck pain
Krauss et al. 2008
Posterior mid-cervical pain??/manipulation
Mechanical neck pain
Cassidy et al. 1992
Neck Pain and other degenerative changesAcute and chronicCervical/manipulation or mobilization
Mechanical neck pain
Cleland et al. 2005
Neck PainChronicThoracic/manipulation
Mechanical neck pain
Fernandez-de-las-Penas et al. 2004
Neck pain and other (whiplash-associated disorders)ChronicThoracic/manipulation
Mechanical neck pain
Gonzalez-Iglesias et al. 2009
Neck painAcuteThoracic/manipulation
Mechanical neck pain
Kanlayanaphotporn et al. 2009
Neck painChronicCervical/mobilization
Mechanical neck pain
Kanlayanaphotporn et al. 2009
Neck painChronicCervical/mobilization
Mechanical neck pain
Martinez-Segura et al. 2006
Neck pain and otherChronic?
Mechanical neck pain
Parkin-Smith and Penter 1998
Neck pain?Cervical or thoracic/manipulation
Mechanical neck pain
Strunk and Hondras 2008
Neck pain and otherChronic?/manipulation
Mechanical neck pain
van Schalkwyk and Parkin-Smith 2000
Neck pain?Cervical/manipulation
Mechanical neck pain
Vernon et al. 1990
Neck pain and otherChronicCervical/manipulation
Postural neck pain and/or whiplash-associated disorders
David et al. 1998
Neck pain and other (whiplash-associated disorders) degenerative changesAcute and chronicCervical/mobilization
Nonspecific neck pain or
cervical spondylosis
Sloop et al. 1982
Neck pain and other degenerative changesAcute and chronic?
Neck pain with arm and/or hand pain
Howe et al. 1983
Neck pain and other cervicogenic headache radicular signs and symptomsAcute and chronic? Lumbar/manipulation
Neck pain or neck and shoulder pain
Savolainen et al. 2004
Neck and/or shoulder pain?Thoracic/manipulation
Cervicobrachial and/or neurogenic pain
Coppieters et al. 2003
Neck pain and otherChronicCervical/mobilization
Cervical spondylosis
Egwu 2008
Neck pain and other degenerative changesAcute and chronicC5-C6/mobilization
Cervicogenic headache
Nilsson et al. 1997
Neck pain cervicogenic headacheChronicCervical/manipulation
Cervicogenic headache
Haas et al. 2004
Neck pain cervicogenic headacheChronic?/manipulation
Spinal pain
Giles and Muller 1999
Neck pain degenerative changesChronicFull spine/manipulation
Spinal pain
Muller and Giles 2005
Neck painChronic?/manipulation? mobilization
Bitterli et al. 1977 Excluded (see Methods)
Chen et al. 2007 Excluded (see Methods)
Table 3B
Inception Diagnosis
Primary ComplaintDuration of ComplaintIntervention
Neck pain
Leaver et al. 2010
Neck painAcuteCervical/manipulation or mobilization
Nonspecific neck pain
Gemmell and Miller 2010
Neck painAcuteCervical or thoracic/manipulation or mobilization
Mechanical neck pain
Lau et al. 2010
Neck painChronicThoracic/manipulation
Cervicogenic headache
Haas et al. 2010
Cervicogenic headacheChronicCervical or Thoracic/manipulation? mobilization
Whiplash-associated disorders (II)
Sterling et al. 2010
Whiplash-associated disordersChronicC5-C6/mobilization

The RCTs enrolled patients presenting with neck pain, associated in some instances with other features (Table ). In most studies, the primary diagnosis rendered was mechanical neck pain (N = 12), neck pain (N = 4), nonspecific neck pain (N = 2), insidious neck pain (N = 1), postural neck pain (N = 1), and whiplash-associated disorder grade II (N = 1). In 11 of the 12 studies, mechanical neck pain was diagnosed by detecting reduced motion or so-called dysfunction in a cervical joint, cervical paraspinal tenderness, or aggravation of pain by neck movement or a particular posture. One study did not report a primary diagnosis other than neck pain . The remaining studies reported a primary diagnosis other than neck pain, such as cervicogenic headache (N = 3), spinal pain (N = 2), cervical spondylosis (N = 1), and cervicobrachial or shoulder pain (N = 1). The diagnosis of cervicogenic headache was said to be made on the basis of the criteria prescribed by the International Headache Society .

A total of 20% of the publications reported that their participants had headache as well as neck pain, diagnosed as cervicogenic headache in 10%, and non-defined headache in the other 10%. In addition, 16% of the studies had participants with shoulder pain, 10% with whiplash-associated disorders grade I to III, 6% with degenerative changes, 6% with pain in the arm or hand, and 3% with low back pain.

Two studies applied treatment only at the C5-6 spinal level, and 14 applied treatment to the cervical spine without stipulating the levels treated. Six studies treated the thoracic spine alone, whereas four studies treated the cervical and/or thoracic spine. One study included treatment of the lumbar spine, and two involved treatment of the entire spine. Two studies did not indicate what part of the spine was treated.

In 18 reports, the intervention was mobilization alone or manipulation alone . Four studies used mobilization or manipulation but not both, whereas two studies used a combination of these interventions. For six of the reports it was not possible to determine if mobilization or manipulation was use alone or in combination.

More than half (18/30) of the articles did not provide any information in their introduction that explained why manipulation was used as the treatment. Of the articles that referred to a manipulative lesion most (10/12) used a name or phrase to refer to it (Table ). Three articles did not provide a name. Two reports provided a label for the manipulable lesion but did not otherwise define the entity to which they were referring.

View this table:
Table 4

List of the phrases used to describe the manipulable lesion. The phrase used to describe the manipulable lesion (Name of Lesion) has been listed opposite to the name of the first author of the article reporting the randomized control trial (Study) with the full reference identified in brackets “[ ]”

Name of LesionStudy
Cervical segment motion restrictionCoppieters et al. 2003
Cervical spondylosisEgwu 2008
Thoracic joint dysfunctionFernandez-de-las-Penas et al. 2004
Mechanical neck pain with motion restrictionKanlayanaphotporn et al. 2009
Abnormal segmental motionKanlayanaphotporn et al. 2009
Joint dysfunctionKrauss et al.
Reduced mobility intervertebral joint dysfunctionMartinez-Segura et al. 2006
Mechanical joint dysfunctionMuller and Giles 2005
Joint fixationParkin-Smith and Penter 1998
Lateral flexion fixationvan Schalkwyk and Parkin-Smith 2000
Joint dysfunction/fixationVernon et al. 1990
Dysfunctional cervical spinal motion unitsWood et al. 2001

Although 63% of the articles specified clinical criteria used to provide evidence of a manipulable lesion (Table ), 30% did not do so. Three articles did not specify the criteria but reported that they were determined by the treating clinician. In one study , the only reference to criteria regarding the manipulable lesion was that the manipulation was repeated if the initial manipulation did not result in a popping sound.

View this table:
Table 5

List of the clinical criteria or test used to identify the need for manipulation or mobilization. The articles referring to these test or criteria are listed opposite with the name of the first author of the article reporting the randomized control trial (Study) with the full reference identified in brackets “[ ]”

Clinical CriteriaStudy
Segmental mobility testingCleland et al. 2005
Segmental motion restrictionCoppieters et al. 2003 , Howe et al. 1983 , Hurwitz et al. 2002 , Krauss et al. 2008 , Yurkiw and Mior 1996
Neck range of motionDavid et al. 1998
Thoracic spine springness testFernandez-de-las-Penas et al. 2004
Thoracic flexion restrictionFernandez-de-las-Penas et al. 2004
Pain precipitated or aggravated by particular neck movementHaas et al. 2004
Resistance/limitation of passive neck movementHaas et al. 2004
Neck muscle tendernessHaas et al. 2004
Palpatory changes in neck muscleHaas et al. 2004
Palpable asymmetry of transverse of C1Howe et al. 1983
Hypomobile or painful segmentKanlayanaphotporn et al. 2009
Abnormal segment mobility (Maitland)Kanlayanaphotporn et al. 2009
Discomfort on Joint challenge/pressureKrauss et al. 2008
Lateral glide test of C3-C5Martinez-Segura et al. 2006
Motion palpation lateral flexion fixationvan Schalkwyk and Parkin-Smith 2000
Palpation findingsNilsson et al. 1997
Hypomobile diagnostic palpation techniquesStrunk and Hondras 2008 ,Vernon et al. 1990

In only 17% of the articles was it stated that participants were excluded from the study if they did not present with clinical evidence of a manipulable lesion. It was not possible in most of the reports to determine if participants were excluded at any time on the basis of the absence of a manipulable lesion.

A quarter of the articles identified the segmental level at which the manipulable lesion was considered to lie in the participants. Most of these (88%) involved multiple segments in the cervical or cervicothoracic (C5-T1) vertebral column, two of which identified thoracic or “other” locations were manipulable. One report identified the manipulable specifically at T4. The remainder of the reports did not specify which segments of the vertebral column were treated.

Only one study reported that examination was conducted after treatment to determine if the treatment had resolved the manipulable lesion and this was limited to assessment of localized pain . Although 87% of reports did not declare a measure of the success of treating a manipulable lesion, four indicated that success was supported by the detection of an audible sound during treatment. In one article , success (although not its prevalence) was recorded when “a release was produced.” Two other articles referred in the discussion section of the article to the occurrence of a crack when treating the manipulable lesion.


In the literature reviewed, it is conspicuous that the indications for manipulation of the cervical spine are neither systematic nor specific. The use of manipulation for neck pain is at best empirical.

Most of the studies reviewed performed manipulation simply on the basis of a complaint of neck pain. Most used synonyms for neck pain as the diagnosis, but no more specific diagnosis was formulated. Qualifying the diagnosis as one of mechanical neck pain does not tell us anything about the cause of pain, but simply reflects what aggravates it.

Although many studies did refer to various criteria which might serve as an indication for manipulation, they are only clinical phenomena as none of these has been shown to be reliable and valid. Signs such as joint fixation and joint dysfunction have been shown to have poor scores for interobserver reliability . An early study suggests that examination of passive accessory intersegment motion might be valid but a larger, more rigorous study found it not to be the case.

Some studies claimed to have diagnosed cervicogenic headache according to the criteria of the International Headache Society , but these criteria emphasize that any clinical tests used to satisfy the criteria must be both reliable and valid. In the absence of tests with these properties, as was the case in the reviewed studies, the criteria require positive responses to controlled diagnostic blocks, which none of the studies of manipulation used.

Studies of manipulation may not formally address the reliability and validity of diagnosis, leaving it implied that the therapist had the necessary skills to make the diagnosis as reported in three studies in this review . However, this is concerning as available literature casts doubt on both the reliability and validity of manual diagnosis. Alternatively, treatment may have been carried out on the basis of a set of principles rather than the identification of a specific manipulable lesion. Indeed, some advocates of mobilization apply their intervention on this basis , as was the case in one of the studies reviewed . However, those principles have not been validated.

Evidence of a manipulable lesion might be inferred if studies showed that diagnostic signs changed after treatment, but none of the studies reviewed provided such evidence. Implicitly, either the signs did not change or they were not tested.

There is no available evidence, based on this review, that manipulative therapists diagnose and successfully treat lesions that pain physicians cannot detect or treat.

It may be that there are specific entities for which manipulation might be indicated, but this still has to be shown. Future studies of treatment should use inclusion criteria that have been shown to be reliable and valid; and they should show that relief of pain is associated with normalization of the clinical signs used to detect the manipulable lesion. However, until that is done, referring patients for manipulative therapy does not serve to address an elusive cause of pain but rather is simply transferring the care of a person with neck pain.


  • No grants or support provided specifically for this review. The authors have no conflicts of interest.


View Abstract