[ literature review ]

KEVIN M. CROSS, PT, PhD, ATC1 • CHRIS KUENZE, MA, ATC2 • TERRY GRINDSTAFF, PT, PhD3 • JAY HERTEL, PhD, ATC4

Thoracic Spine Thrust Manipulation

Improves Pain, Range of Motion,
and Self-Reported Function in Patients
With Mechanical Neck Pain:
Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

A Systematic Review

N
Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

eck pain is a common musculoskeletal condition, with a neck pain, which includes
12-month prevalence among the general and work force patients without an identi-
fiable pathoanatomic cause
populations of 30% to 50%.19 As a consequence, neck pain is
and excludes patients with
responsible for a large proportion of physical therapy visits.3,26 neurological deficits, cervi-
A common general classification in clinical studies is mechanical cogenic headache, and systemic inflam-
matory conditions. 3,23 The operational
definition of mechanical neck pain most
TTSTUDY DESIGN: Systematic review. function at all stated time intervals.
frequently requires that the pain be ex-
TTBACKGROUND: Neck pain is a common diagno- TTRESULTS: Effect size point estimates for the
Journal of Orthopaedic & Sports Physical Therapy®

acerbated by motion. However, there

sis in the physical therapy setting, yet there is no pain change scores were significant for global is variability among studies, as patients
gold standard for treatment. This study is part of assessment across all studies (range, 0.38-4.03)
a growing body of literature on the use of thoracic with radiculopathy or whiplash mecha-
but not conclusively significant at the end range of
spine thrust manipulation for the treatment of nisms, for example, have been exclud-
active rotation (range, 0.02-1.79). Effect size point
individuals with mechanical neck pain. ed.10,17,18,29 While treatment for patients
estimates were large among all range-of-motion
TTOBJECTIVE: The purpose of this systematic re- change measures (range, 1.40-3.52), and the effect classified with mechanical neck pain has
view was to determine the effects of thoracic spine size point estimates of the change scores among been investigated, there is no consensus
thrust manipulation on pain, range of motion, and the functional questionnaires (range, 0.47-3.64) within the literature on a gold standard
self-reported function in patients with mechanical for treatment.3,6,19,28
also indicated a significant treatment effect.
neck pain.
TTCONCLUSIONS: Thoracic spine thrust manipula- One approach to conservative treat-
TTMETHODS: Six online databases were compre- ment of neck pain includes cervical
tion may provide short-term improvement in
hensively searched from their respective inception
patients with acute or subacute mechanical neck mobilization and thrust manipulation.
to October 2010. The primary search terms
included “thoracic mobilization,” “thoracic spine pain. However, the body of literature is weak, and The potential for complications caused
mobilization,” “thoracic manipulation,” and “tho- these results may not be generalizable. by thrust manipulation of the cervical
racic spine manipulation.” Of the 44 studies as- TTLEVEL OF EVIDENCE: Therapy, level 1b–. spine, such as vertebrobasilar artery in-
sessed for inclusion, 6 randomized controlled trials J Orthop Sports Phys Ther 2011;41(9):633-642. jury, has been extensively discussed in
were retained. Between-group mean differences the literature.2,14,20,21,34 Although the risk
doi:10.2519/jospt.2011.3670
and effect sizes for pretreatment-to-posttreatment
TTKEY WORDS: cervical spine, manipulative
of vertebrobasilar artery dissection has
change scores, using Cohen’s d formula, were cal-
culated for pain, range of motion, and subjective therapy, manual therapy been reported to be very low,14,25 screen-
ing tools to identify patients who are at

1
Physical Therapist, UVA Healthsouth, Charlottesville, VA. 2Doctoral Candidate, University of Virginia, Charlottesville, VA. 3Assistant Professor, Creighton University, Omaha, NE.
4
Joe H. Gieck Professor of Sports Medicine, University of Virginia, Charlottesville, VA. Address correspondence to Kevin Cross, 5004 Madison Court, Charlottesville, VA 22911.
E-mail: kmc7e@virginia.edu

journal of orthopaedic & sports physical therapy | volume 41 | number 9 | september 2011 | 633

41-09 Cross.indd 633 8/16/2011 6:28:33 PM

 [ literature review ]
risk of adverse effects from the thrust “neck pain,” which were also combined tigators (K.M.C. and C.K.) assessed the
manipulation have been proposed and with the Boolean operator “OR.” We then articles for inclusion into the review.
their use recommended, despite a lack of combined searches 1 and 2 using the
evidence supporting their validity.24,27,37 Boolean operator “AND.” We placed no Quality Assessment
Moreover, the literature has recommend- restrictions on the publication date; the Selected studies were evaluated using
ed avoiding manual therapies at terminal only limitation was that the language had the Physiotherapy Evidence Database
ranges of motion4 and cautioned against to be English. The reference lists of viable (PEDro) scale, which utilizes a 10-point
the use of cervical thrust manipulation studies were cross-referenced to identify scoring system to assess the methodologi-
due to the perceived risk of serious vas- additional articles not detected in the cal quality of randomized controlled tri-
cular complications, especially in specific original medical database searches. als.31 Two investigators (K.M.C. and C.K.)
subgroups of the population.27,32,34 Alter- independently assessed each study. If
natively, thoracic spine thrust manipula- Study Selection a disagreement in scores occurred, the
tion may effectively address mechanical Following the search in each database, investigators discussed the study’s qual-
neck pain.4,32 2 of the authors (K.M.C. and C.K.) in- ity to determine consensus on the final
Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

There is a recent but growing body of dividually selected the studies to be in- score. If consensus could not be attained,
literature evaluating the clinical effective- cluded in the systematic review. First, the then a third investigator (J.H.), blinded
ness of thoracic spine thrust manipula- context of each study’s title was screened to the previous assessment scores, re-
tion for patients with mechanical neck for relevance to the systematic review’s solved the disagreement.
pain. Therefore, the purpose of this sys- purpose. The abstracts of those studies
Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

tematic review was to evaluate the effects with relevant titles were then reviewed Data Extraction
of thoracic spine thrust manipulation on for pertinence to the topic. If an abstract Two investigators (K.M.C. and C.K.) in-
pain in patients with mechanical neck suggested that the manuscript provided dependently extracted data from the se-
pain, and the effects of thoracic spine information regarding the effect of tho- lected studies using standardized forms.
thrust manipulation on cervical range of racic spine thrust manipulation on cervi- Specifically, each investigator docu-
motion (ROM) and patient self-reported cal or neck pain, the article was read and mented the study design, study purpose,
function in that same population. thoroughly assessed for the inclusion or experimental and comparison interven-
exclusion criteria. Because the primary tions, number of subjects in each condi-
METHODS outcome for data analysis was neck pain, tion, follow-up intervals (as indicated by
Journal of Orthopaedic & Sports Physical Therapy®

the study had to report pain data that the duration from treatment initiation),

T
o provide a thorough system- permitted a preintervention-to-postint- outcome variables, and study quality
atic review, we referred to the PRIS- ervention analysis of changes in pain score (PEDro). The authors compared
MA Explanation and Elaboration scores. The 2 authors then compared their standardized forms for each study
document.35 The PRISMA checklist was their search results to create 1 compre- to verify the accuracy of data collection.
used in attempt to provide comprehen- hensive list of studies for inclusion. The outcomes of interest were pain,
sive and high-quality research to assess Studies were included in this review cervical spine ROM, and self-reported
the effectiveness of thoracic spine thrust if they satisfied the following criteria: function. To compare the effectiveness of
manipulation in patients with mechani- (1) the treatment group received thrust the experimental and comparison inter-
cal neck pain. manipulation to the thoracic spine; (2) ventions, we analyzed the change scores
only patients diagnosed with mechanical from the initial visit to each follow-up
Search Strategy neck pain were included in the study; (3) interval between each group.
Six online databases (CINAHL, Cochrane pain needed to be reported as an outcome
Library, PubMed, PEDro, Sport Discus, measure; (4) detailed eligibility criteria Data Analysis
and Web of Science) were comprehen- for the patients participating in the study Although all studies included in this re-
sively searched from their respective were provided.31 view assessed pain as one of the outcome
inception to October 2010. The search Studies were excluded based on the measures, the pain and self-reported
1 query included the terms “thoracic following criteria: (1) absence of a com- function parameters were frequently
spine manipulation,” “thoracic spine parison group; (2) treatment group measured using different instruments.
mobilization,” “thoracic manipulation,” received a thrust manipulation to the The pain measurements, specifically,
and “thoracic mobilization,” which were cervical spine; and (3) patients were were performed in varying contexts and
all combined with the Boolean operator diagnosed with neurological deficits, during different activities, such as pain
“OR.” The search 2 query used the terms cervicogenic headaches, or multiple di- at rest versus pain associated with end
“cervical spine pain,” “cervical pain,” and agnoses. Based on these criteria, 2 inves- range of cervical rotation. Most notably,

634 | september 2011 | volume 41 | number 9 | journal of orthopaedic & sports physical therapy

41-09 Cross.indd 634 8/16/2011 6:28:34 PM

 TABLE 1 Included Study Details

Authors Cleland et al7 Krauss et al29 Cleland et al10 Gonzalez-Igles et al17 Gonzalez-Igles et al18 Cleland et al11
Year of publication 2005 2009 2007 2009 2009 2010
Study design RCT RCT RCT RCT RCT RCT
PEDro score 7 6 7 7 7 7
Follow-up time Immediate Immediate 2-4 d 4 wk after initiation 3, 5, and 7 wk after 1 wk, 4 wk, and 6 mo
of treatment initiation of treatment after initiation of
treatment
Outcome measures VAS FPS, ROM NPRS, NDI NPRS, NPQ, ROM VAS, NPQ, ROM NPRS, NDI
Intervention group Supine thrust, n = 19 Supine thrust, n = 22 Supine thrust and cervi- Seated manipulation, Seated manipulation, Week 1: seated/supine
cal mobility, n = 30 1 × 3 wk; heat/TENS, 1 × 3 wk; heat/TENS, manipulations and
Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

3 × 3 wk; n = 23 3 × 3 wk; n = 23 cervical mobility

exercise, 2 × wk;
Weeks 2-4: cervical
mobility and strength
exercise, 1 × wk;
Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

n = 70
Control group Placebo, n = 17 Rest, n =10 Prone thoracic mobiliza- Heat/TENS, 3 × 3 wk; Heat/TENS, 3 × 3 wk; Cervical mobility and
tion and cervical n = 22 n = 22 strength exercise:
mobility exercise, week 1, 2 sessions;
n = 30 weeks 2-4, 1 × wk;
n = 70
Abbreviations: FPS, faces pain scale; NDI, neck disability index; NPQ, Northwick Park Neck Pain Questionnaire; NPRS, numeric pain rating scale; RCT,
randomized controlled trial; ROM, range of motion; TENS, transcutaneous electrical nerve stimulation; VAS, visual analogue scale.
Journal of Orthopaedic & Sports Physical Therapy®

all outcome assessments occurred at dif- was negative if the comparison group had study design or not utilizing a compari-
ferent time intervals following the inter- a larger treatment effect. The strength of son group (3 studies),9,13,15 and the use of
vention. Consequently, data could not be the effect size was determined as trivial both cervical and thoracic spine thrust
collapsed across studies for meta-anal- (<0.2), small (0.2-0.39), moderate (0.4- manipulations as interventions (2 stud-
ysis. In each study, the between-group 0.7), or large (>0.7).12 All nontrivial effect ies)36,38 (FIGURE 1).
mean difference (95% confidence inter- sizes (small to large) with 95% CIs that Each of the included studies reported
val [CI]) and effect size (95% CI) for did not include zero were considered to the use of 1 or more nonspecific thoracic
the preintervention-to-postintervention represent a significant treatment effect.22 spine thrust manipulation techniques,
change scores, using Cohen’s d formula, performed in either a supine or sitting
were calculated for pain, ROM, and self- RESULTS position (FIGURE 2). The use of addition-
reported function at each of the follow-up al modalities and therapeutic exercises
assessments. The effect size was calculat- Study Descriptions and Methodological varied. No comparison groups received
ed as d = xmanipulation – xnonmanipulation/SDpooled, Quality cervical or thoracic spine thrust ma-

S
where x represents the average change ix randomized controlled tri- nipulation, and only 1 study used grade
score and SD represents the pooled stan- als (RCTs) met our eligibility cri- 3 or 4 mobilizations to the thoracic
dard deviation between manipulation teria. Their median PEDro score spine as a comparative intervention.10
and comparison groups. was 7, with scores ranging from 6 to 7 In the RCTs with comprehensive sub-
The effect size was positive if the treat- (TABLE 1). Common items deducted from ject characteristics, the average duration
ment group that included the thoracic the scores involved lack of blinding of the of symptom onset was 3 months or less,
spine thrust manipulation had a larger subject and the therapist. Seven articles indicating that the condition was acute
treatment effect than the comparison were excluded due to the inclusion of or subacute.7,10,11,17,18 All studies had vari-
group, as indicated by a larger preinter- patients with nonmechanical neck pain able timelines for outcome assessments
vention-to-postintervention change score (2 studies),1,39 the use of methods below and ranged from immediately follow-
on the outcome variable. The effect size the minimum established experimental ing thrust manipulation to 6 months

journal of orthopaedic & sports physical therapy | volume 41 | number 9 | september 2011 | 635

41-09 Cross.indd 635 8/16/2011 6:28:35 PM

 [ literature review ]
after the beginning of the intervention.
TABLE 1 provides the information neces- Initial search results: n = 534
sary to interpret the results, based on the
individual study characteristics.
Excluded based on title: n = 490
Outcome Assessment
With the exception of the study by Cle-
land et al,11 the average change score and Abstracts reviewed: n = 44
either the SD7,10 or the 95% CI17,18,29 for
the change score were provided for the
variables in all of the included studies. In Excluded based on abstract: n = 31
both studies by Gonzalez-Iglesias et al17,18
and the study by Krauss et al,29 the SD
for the change scores was calculated from Manuscripts read: n = 13
Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

the equation for the 95% CI. The SD of

the change score was needed to calculate
Excluded based on content:
effect sizes. For the most recent article by
• Nonmechanical neck pain: n = 21,39
Cleland et al,11 the author provided, upon • Lack of comparison group:
request, the mean and SD of the change n = 39,13,15
Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

scores for the variables of interest. • Treatment included cervical

manipulation: n = 236,38
Pain
All included studies reported preinter-
vention and postintervention pain val- Included in review: n = 6
ues using either the visual analog scale,7,18
faces pain scale,29 or numeric pain rating
FIGURE 1. PRISMA flowchart for study selection.
scale.10,11,17 There was considerable vari-
ability among studies regarding pain
Journal of Orthopaedic & Sports Physical Therapy®

measurement and time of follow-up, so tion. Two separate studies by Gonzalez- naire.17,18 Effect size point estimates for
the pain data were not collapsed for anal- Iglesias et al17,18 reported cervical ROM in change scores among the functional
ysis. The effect size point estimates for the all planes, following a 3-week course of questionnaires were moderate to large
change scores of global pain ranged from electrothermal treatment and a thoracic and varied from 0.47 to 3.64. FIGURE 5
small to large across all studies (0.38 to spine thrust manipulation performed shows the effect sizes and 95% CIs for
4.03). FIGURE 3 shows the effect sizes and once per week. The ROM measures the self-reported functional outcome
95% CIs for the pain scores in each study. were taken at 3, 4, and 5 weeks follow- measures.
In contrast to a global pain assess- ing the treatment initiation. The average
ment, Krauss et al29 reported pain in mean improvement for cervical flexion Adverse Events
patients at the end of active left and and extension ranged from 8.1° to 12.0° Only 2 of the included studies presented
right cervical rotation as appropriate to and 7.0° to 11.4°, respectively, while the complications or adverse events as a re-
cause symptoms. The effect size point mean cervical rotation improvements sult of the interventions. Cleland et al10
estimates were generally smaller than varied from 7.7° to 12.5°. The effect size reported no significant differences in
those for pain at rest, ranging from 0.02 point estimates for ROM change scores the number of side effects experienced
to 1.79; but the 95% CI included zero for were large, varying from 1.39 to 3.23. by individuals in the thrust manipula-
most subgroups, indicating that conclu- FIGURE 4 presents the mean differences tion versus nonthrust group. Specifi-
sive treatment effects on pain at the end of the change scores between the thrust cally, aggravation of symptoms (n = 2),
range of cervical rotation were not pres- manipulation and comparison groups muscle spasm (n = 1), neck stiffness (n
ent (FIGURE 3). and 95% CIs for ROM measures. = 2), headache (n = 2), and radiating
symptoms (n = 2) were reported in the
Range of Motion Activity- and Disability-Related Outcomes nonthrust group, while aggravation of
Krauss et al29 assessed the immediate Functional outcome measures included symptoms (n = 8), muscle spasm (n = 1),
changes in active cervical rotation ROM the Neck Disability Index10,11 and the and headache (n = 1) were reported in the
following thoracic spine thrust manipula- Northwick Park Neck Pain Question- thrust manipulation group. The onset of

636 | september 2011 | volume 41 | number 9 | journal of orthopaedic & sports physical therapy

41-09 Cross.indd 636 8/16/2011 6:28:36 PM

 in the management of acute or subacute
mechanical neck pain to reduce pain,
improve cervical ROM, and improve
function. Positive effects were shown to
occur immediately following the first in-
tervention and continued up to 6 months
after participants were discharged from a
3-week program with repeated thoracic
spine thrust manipulations. These re-
sults, however, must be interpreted with
caution. Although the individual RCTs
were of high quality, as graded by the
PEDro scale, the overall body of litera-
ture had significant flaws that limit the
Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

generalizability of the results.

FIGURE 2. Common thoracic spine thrust manipulation techniques described in the literature. (A) Supine thrust
manipulation technique: the patient lies supine, with the arms wrapped across the chest. The therapist pulls the Evaluation of the Body of Literature
elbows of the patient toward the chest to induce passive cervical and thoracic flexion. Then, the therapist provides
The primary limitation of the current
a high-velocity, low-amplitude thrust through the arms of the patient toward the thoracic spine at a targeted level,
as dictated by the manipulative hand. (B) Seated thrust manipulation technique: the patient sits with the arms body of literature is the limited number
Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

crossing the chest. The patient leans back into the chest of the therapist at the midthoracic level. Passive thoracic of RCTs. Our literature search identi-
spine flexion is applied until the therapist perceives tension, then the therapist provides a distraction thrust in an fied 6 RCTs with 3 different lead au-
upward direction. thors.7,10,11,17,18,29 Although the subject
demographics for the 2 RCTs authored
by Gonzalez-Iglesias et al17,18 were very
similar, they were verified to be 2 com-
0.45 (0.08, 0.83), 6 mo11
2.58 (1.79, 3.38), 7 wk18
pletely different samples of subjects.
1.92 (1.21, 2.62), 5 wk18 While our search was extensive through
3.82 (2.84, 4.80), 4 wk17 6 databases, biases may exist within our
Journal of Orthopaedic & Sports Physical Therapy®

0.38 (0.04, 0.71), 4 wk11 search criteria. We performed literature

4.03 (3.01, 5.04), searches in medical databases that were
3 wk18
0.79 (0.45, 1.14), 1 wk11
accessible through the medical school or
1.58 (1.00, 2.16), 2-4 d10
accessible online. While these databases
1.75 (0.98, 2.53), immediate7
1.39 (–0.28, 3.05), immediate; B to L29*
capture the vast majority of physical ther-
1.79 (0.05, 3.53), immediate; B to R29* apy trials, it is possible that they might
0.02 (–1.31, 1.35), immediate; L29* have missed or failed to index certain tri-
0.74 (–0.48, 1.97), immediate; R29* als.33 Moreover, our decision to limit this
review to articles published only in the
–2 –1 0 1 2 3 4 5 6
English language might have excluded
Favors Comparison Favors Thoracic Spine Thrust Manipulation
pertinent articles in other languages.
FIGURE 3. Effect sizes (95% confidence interval) for pain relief following thoracic spine thrust manipulation.
Similarly, our decision to use only pub-
Abbreviations: L, pain and measurement with left cervical rotation; R, pain and measurement with right cervical lished studies might have introduced bias
rotation; B to R, pain with bilateral rotation and measurement with right rotation; B to L, pain with bilateral and inadvertently magnified the effects of
rotation and measurement with left rotation. All time measures are relative to the initiation of interventions. *Pain thoracic spine thrust manipulations. Nor
measured at end range of cervical rotation.
were experts of spinal thrust manipula-
tion and mechanical neck pain contacted
these side effects was not significantly dif- group throughout the trial were reported. to ensure that there were no unintention-
ferent between groups, with individuals ally overlooked articles within the inclu-
reporting an onset of side effects within DISCUSSION sion and exclusion criteria. Within the
24 hours of treatment, with a duration considerations for bias that are normal

T
of no greater than 24 hours, regardless he results of this systematic for systematic reviews, the minimal vari-
of group assignment. In a later study by review indicate that thoracic spine ability among the clinicians and patient
Cleland et al,11 no adverse events in either thrust manipulation may be utilized types for the studies included in this sys-

journal of orthopaedic & sports physical therapy | volume 41 | number 9 | september 2011 | 637

41-09 Cross.indd 637 8/16/2011 6:28:37 PM

 [ literature review ]
A B

8.10 (4.62, 11.58), 5 wk18

7.00 (4.27, 9.73), 5 wk18

10.70 (8.69, 12.71), 4 wk17

10.00 (8.14, 11.86), 4 wk17

12.0 (9.95, 14.04), 11.40 (9.20, 13.60),

3 wk18 3 wk18

0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
ROM Change Score Difference, deg ROM Change Score Difference, deg
Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

C D

12.10 (8.88, 15.32), 5 wk R18 8.70 (7.03, 10.37), 5 wk R18

8.50 (5.20, 11.80), 5 wk L18

5.90 (3.56, 8.23), 5 wk L18
9.60 (7.49, 11.71), 4 wk R 17
Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

9.60 (7.55, 11.65), 4 wk R17

8.30 (6.38, 10.22), 4 wk L17

12.50 (9.80, 15.20), 3 wk R18 8.10 (6.19, 10.00), 4 wk L17

11.20 (8.87, 13.53), 3 wk L18

9.30 (7.55, 11.05), 3 wk R18
8.33 (3.40, 13.26),
immediate R29
7.69 (3.58, 11.80), 10.80 (8.40, 13.20), 3 wk L18
immdediate L29

0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16
ROM Change Score Difference, deg ROM Change Score Difference, deg
Journal of Orthopaedic & Sports Physical Therapy®

FIGURE 4. Mean difference (95% confidence interval) in cervical ROM change scores between the thoracic spine thrust manipulation and comparison groups. Positive values
favor thoracic spine thrust manipulation. All time measures are relative to the initiation of interventions. (A) Cervical flexion. (B) Cervical extension. (C) Cervical rotation. (D)
Cervical lateral flexion. Abbreviations: L, left; R, right; ROM, range of motion.

tematic review diminished the external to provide direction for physical therapy generic locations of the midthoracic and
validity of the results. interventions,5 this systematic review of upper thoracic spine.10,11,17,18 Cleland et al10
Regarding cervical spine pathology, all the effectiveness of thoracic spine thrust extrapolated on the decision for using a
subjects were diagnosed with mechanical manipulation among a narrower scope generic location for thoracic spine thrust
neck pain, and most of them had an aver- of neck injuries (acute and subacute manipulation, recognizing that the tech-
age symptom onset duration of less than mechanical neck pain) provides support nique does not target specific vertebral
3 months.7,10,11,17,18 Krauss et al29 did not for its potential as an intervention. Due segments. In contrast, to simulate clinical
report the duration of symptoms of their to the relatively homogeneous patient practice, 2 studies7,29 attempted to specifi-
study participants. Very few participants sample in the included studies, the re- cally manipulate segments that were de-
with chronic neck pain were included in sults may not be generalizable across pa- termined to be hypomobile during joint
the selected trials, which limits the gen- tients with differing diagnoses or onset mobility testing. Nevertheless, outcomes
eralizability of the findings beyond pa- durations. among the current studies did not appear
tients with acute and subacute neck pain. The RTCs described 2 thoracic spine to be influenced by the specific technique
The broad definition of mechanical neck thrust manipulation techniques, yet none or rationale for application.
pain, from a pathological perspective, of the studies provided clinical reasoning Variation among the comparative
reflects the proposed treatment-based for the selection of a specific technique. interventions also blurred the effects
neck pain classification categories of The specific site to which the thrust ma- of thoracic spine thrust manipulation.
pain control and conditioning and mobil- nipulation was applied also varied among Comparison treatments included place-
ity.5 Considering that the purpose of the studies, with 4 RCTs attempting to pro- bo thrust manipulation,7 rest,29 thoracic
treatment-based classification system is vide the thrust manipulation at 1 or more spine mobilization and cervical mobility

638 | september 2011 | volume 41 | number 9 | journal of orthopaedic & sports physical therapy

41-09 Cross.indd 638 8/16/2011 6:28:38 PM

 those which compared the intervention
0.58 (0.20, 0.95), 6 mo11 with a comprehensive exercise program.11
This reduction in effect size for the addi-
2.06 (1.34, 2.79), 5 wk18
tion of the thrust manipulation interven-
tion highlights the therapeutic benefit of
3.51 (2.58, 4.45), 4 wk17
a methodically devised exercise program
0.47 (0.13, 0.80), 4 wk11 on managing cervical spine pain.3 While
there was clear variability in the control
3.64 (2.68, 4.59), 3 wk18 intervention among the studies (TABLE 1),
their results indicate that patients with
0.56 (0.22, 0.89), 1 wk11
mechanical neck pain who received tho-
racic spine thrust manipulation experi-
1.10 (0.56, 1.65), 2-4 d10
enced significant decreases in pain.
Pain at the end range of cervical ro-
Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

–1 0 1 2 3 4 5 6 tation, in contrast to global pain ratings,

Favors Comparison Favors Thoracic Spine Thrust Manipulation was the primary outcome measure in 1
trial.29 Only those patients who experi-
FIGURE 5. Effect sizes (95% confidence interval) for self-reported function following thoracic spine thrust enced pain with bilateral cervical rotation
manipulation. All time intervals are relative to the initiation of interventions.
reported decreased pain at the end range
Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

rotation, following the thrust manipula-

exercises,10 electrothermal therapies,17,18 investigated the clinical effectiveness of tion. The authors, by assessing pain at
and a comprehensive exercise program.11 a thoracic spine thrust manipulation as the end ranges of cervical rotation as op-
Only 2 studies7,10 used a comparison in- an isolated manual therapy technique for posed to at rest, have suggested that the
tervention directed at the thoracic spine, patients with neck pain. Since that time, source for mechanical pain may still be
and only 2 studies10,11 used a comparison there have been several investigations of present at the end range, even though
intervention that directly addressed the the clinical utility of thoracic spine thrust cervical spine ROM has improved.29 This
pathological area with therapeutic ex- manipulation in the management of me- would account for the lack of improve-
ercise. However, regardless of the com- chanical neck pain across various follow- ment in pain reported in this study as
Journal of Orthopaedic & Sports Physical Therapy®

parative interventions, the results of the up times and within diverse intervention compared to the other studies in this
statistical analysis suggest that treat- programs. Within these investigations, review.
ment programs which incorporate a the effectiveness of 2 distinct thoracic Patients who received thoracic spine
thoracic spine thrust manipulation have spine thrust manipulation techniques thrust manipulation alone or in combi-
larger treatment effects on the outcome have been reported (FIGURE 2), both of nation with ROM exercises or modalities
measures. which provided short-term improve- had increased cervical spine mobility. In
The follow-up intervals for all stud- ments in pain, cervical ROM, and self- all RCTs, thoracic spine thrust manipula-
ies included in this review were relatively reported function among patients with tion resulted in larger ROM changes and
short. The current literature has individ- acute or subacute mechanical neck pain. significant CIs. Each study that measured
ual reports for treatment effectiveness Thoracic spine thrust manipulation, cervical ROM used the cervical range-of-
immediately following the treatment7,29 performed by itself or in combination motion (CROM) device, which has an es-
and at 2 to 4 days,10 3 to 7 weeks,17,18 and 6 with other interventions, may decrease tablished minimal detectable change for
months11 after treatment initiation. There neck pain, with the decrease occurring patients with neck pain.16 In the majority
is no study, to our knowledge, which in- immediately after a single thrust manip- of follow-up intervals, the point estimate
vestigates evidence of treatment effec- ulation intervention and persisting up to of the within-group ROM change scores
tiveness beyond 6 months posttreatment 6 months. In each study, thoracic spine exceeded the minimal detectable change
initiation. Although various follow-up in- thrust manipulation was found to have a only in the thoracic spine thrust manipu-
tervals have been reported, they have not positive effect size when compared to the lation group (TABLE 2), and the between-
been validated by other research groups, control intervention. The range of effect group ROM change scores for all ROM
and long-term outcomes have not been sizes for change in pain following thrust were positive for the thrust manipulation
reported. manipulation intervention has been groups. However, there was variability in
wide. Studies using a control intervention the magnitude of the treatment effect
Clinical Implications of passive treatment or a placebo7,10,17,18 across time (FIGURE 4). Further research is
Prior to 2005, there were no RCTs that reported larger positive effect sizes than necessary to examine the long-term treat-

journal of orthopaedic & sports physical therapy | volume 41 | number 9 | september 2011 | 639

41-09 Cross.indd 639 8/16/2011 6:28:39 PM

 [ literature review ]
which used the Northwick Park Neck
Comparison of Within-Group ROM
Pain Questionnaire,17,18 the change score
Changes in the Thoracic Spine Thrust
TABLE 2 in only the thrust manipulation group
Manipulation Groups to the CROM
exceeded the minimal clinically impor-
MDCs for Patients With Neck Pain
tant difference at 3, 4, and 5 weeks af-
ter treatment initiation.40 Unfortunately,
Cervical ROM/Follow-up Within-Group Mean Difference (95% CI) MDC9016
global assessments of change scores were
Cervical flexion 9.6°
not provided for either group, which
3 wk18 12.4° (10.9°, 14.0°)
limits the interpretation of improve-
4 wk17 11.6° (10.0°, 13.2°)
ment by the Northwick Park Neck Pain
5 wk18 11.2° (8.8°, 13.5°)
Questionnaire.40
Cervical extension 7.0°
3 wk18 11.6° (10.3°, 13.6°)
CONCLUSION
4 wk17 11.2° (9.6°, 12.8°)
Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

I
5 wk18 6.4° (4.3°, 9.1°)
n the current literature, thorac-
Cervical right rotation 7.6°
ic spine thrust manipulation reduced
Immediate29 8.2° (4.9°, 11.5°)
pain and improved ROM among pa-
3 wk18 11.1° (9.1°, 12.3°)
tients with acute or subacute mechanical
4 wk17 9.8° (8.4°, 11.2°)
neck pain. Optimal treatment parameters
Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

5 wk18 7.8° (5.4°, 9.7°)

are not clear from the current literature.
Cervical left rotation 6.7°
It appears that thoracic spine thrust ma-
Immediate29 7.1° (4.5°, 9.7°)
nipulation can provide a positive treat-
3 wk18 10.8° (9.3°, 12.4°)
ment effect immediately following thrust
4 wk17 8.9° (7.4°, 10.4°)
manipulation, as well as short-term
5 wk18 6.4° (3.9°, 8.3°)
improvement in symptoms following a
Cervical right lateral flexion 5.9°
course of repeated thrust manipulations.
3 wk18 9.7° (9.1°, 11.6°)
However, there are several weaknesses
4 wk17 11.1° (9.4°, 12.8°)
in the body of literature, particularly in
Journal of Orthopaedic & Sports Physical Therapy®

5 wk18 9.5° (8.4°, 11.5°)

the limited number of RCTs and the con-
Cervical left lateral flexion 9.1°
sequent lack of variability among lead
3 wk18 10.0° (8.3°, 11.7°)
authors, which limits the generalizability
4 wk17 9.4° (8.3°, 11.7°)
of the results for multiple patient types
5 wk18 7.6° (5.6°, 9.7°)
and clinical environments. Therefore,
Abbreviations: CI, confidence interval; CROM, cervical range of motion device; MDC, minimal detect-
able change; ROM, range of motion. clinicians must apply these findings to
their clinical practice with caution. Tho-
racic spine thrust manipulation should
ment effect of thoracic spine thrust ma- various cervical pathologies,30 the only be considered when treating patients
nipulations on cervical spine ROM and study to report a Neck Disability Index with mechanical neck pain, especially
in comparison to a standardized physical change score conclusively greater than if cervical spine thrust manipulation is
therapy program. the minimal clinically important differ- contraindicated or the patient is averse
Based on this review, the effectiveness ence among patients with mechanical to cervical spine thrust manipulation.
of thoracic spine thrust manipulation, in neck pain8,41 was the Cleland et al11 study Future research should establish treat-
conjunction with exercises or modalities at the 4-week and 6-month follow-ups af- ment parameters that include short- and
on self-reported function in patients with ter treatment initiation. The significance long-term effects, as well as compare
mechanical neck pain, appears to depend of these changes in self-reported func- treatment effectiveness to interventions
on the duration of the follow-up interval. tion are further illustrated by the signifi- directed at the cervical spine. t
While only 4 studies10,11,17,18 included self- cantly greater number of patients in the
reported function as an outcome variable, thrust manipulation and exercise group,
all reported moderate to large effect sizes. as compared to the exercise-only group, REFERENCES
Although there is disparity in the litera- who reported a successful outcome at
1. Browder DA, Erhard RE, Piva SR. Intermittent
ture regarding the psychometric proper- the 4-week and 6-month follow-ups af-
cervical traction and thoracic manipulation
ties of the Neck Disability Index among ter treatment initiation.11 Of the studies

640 | september 2011 | volume 41 | number 9 | journal of orthopaedic & sports physical therapy

41-09 Cross.indd 640 8/16/2011 6:28:40 PM

 for management of mild cervical compressive 12. C  ohen J. Statistical Power Analysis for the 2002;136:713-722.
myelopathy attributed to cervical herniated Behavioral Sciences. 2nd ed. Hillsdale, NJ: Law- 24. H urley L, Yardley K, Gross AR, Hendry L,
disc: a case series. J Orthop Sports Phys Ther. rence Erlbaum Associates, Inc; 1988. McLaughlin L. A survey to examine attitudes
2004;34:701-712. 13. Costello M. Treatment of a patient with cervi- and patterns of practice of physiotherapists
2. Carlesso LC, Macdermid JC, Santaguida LP. cal radiculopathy using thoracic spine thrust who perform cervical spine manipulation. Man
Standardization of adverse event terminology manipulation, soft tissue mobilization, and Ther. 2002;7:10-18. http://dx.doi.org/10.1054/
and reporting in orthopaedic physical therapy: exercise. J Man Manip Ther. 2008;16:129-135. math.2001.0430
application to the cervical spine. J Orthop 14. Ernst E. Prospective investigations into the 25. Hurwitz EL, Morgenstern H, Vassilaki M, Chiang
Sports Phys Ther. 2010;40:455-463. http:// safety of spinal manipulation. J Pain Symptom LM. Adverse reactions to chiropractic treatment
dx.doi.org/10.2519/jospt.2010.3229 Manage. 2001;21:238-242. and their effects on satisfaction and clinical
3. Childs JD, Cleland JA, Elliott JM, et al. Neck 15. Fernandez-de-las-Penas C, Palomeque-del- outcomes among patients enrolled in the UCLA
pain: clinical practice guidelines linked to the Cerro L, Rodriguez-Blanco C, Gomez-Conesa A, Neck Pain Study. J Manipulative Physiol Ther.
International Classification of Functioning, Dis- Miangolarra-Page JC. Changes in neck pain and 2004;27:16-25. http://dx.doi.org/10.1016/j.
ability, and Health from the Orthopedic Section active range of motion after a single thoracic jmpt.2003.11.002
of the American Physical Therapy Association. spine manipulation in subjects presenting with 26. Jette AM, Smith K, Haley SM, Davis KD. Physical
J Orthop Sports Phys Ther. 2008;38:A1-A34. mechanical neck pain: a case series. J Manipu- therapy episodes of care for patients with low
http://dx.doi.org/10.2519/jospt.2008.0303 lative Physiol Ther. 2007;30:312-320. http:// back pain. Phys Ther. 1994;74:101-110; discus-
Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

4. Childs JD, Flynn TW, Fritz JM, et al. Screening dx.doi.org/10.1016/j.jmpt.2007.03.007 sion 110-105.
for vertebrobasilar insufficiency in patients with 16. Fletcher JP, Bandy WD. Intrarater reliability 27. Kerry R, Taylor AJ. Cervical arterial dysfunc-
neck pain: manual therapy decision-making in of CROM measurement of cervical spine ac- tion: knowledge and reasoning for manual
the presence of uncertainty. J Orthop Sports tive range of motion in persons with and physical therapists. J Orthop Sports Phys Ther.
Phys Ther. 2005;35:300-306. http://dx.doi.org/ without neck pain. J Orthop Sports Phys Ther. 2009;39:378-387. http://dx.doi.org/10.2519/
10.2519/jospt.2005.1312 2008;38:640-645. http://dx.doi.org/10.2519/ jospt.2009.2926
5. Childs JD, Fritz JM, Piva SR, Whitman JM. jospt.2008.2680 28. Korthals-de Bos IB, Hoving JL, van Tulder MW,
Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

Proposal of a classification system for patients 17. Gonzalez-Iglesias J, Fernandez-de-las-Penas C, et al. Cost effectiveness of physiotherapy,
with neck pain. J Orthop Sports Phys Ther. Cleland JA, Alburquerque-Sendin F, Palomeque- manual therapy, and general practitioner care
2004;34:686-696; discussion 697-700. del-Cerro L, Mendez-Sanchez R. Inclusion of for neck pain: economic evaluation alongside a
6. Cleland JA, Childs JD, Fritz JM, Whitman JM, Eb- thoracic spine thrust manipulation into an randomised controlled trial. BMJ. 2003;326:911.
erhart SL. Development of a clinical prediction electro-therapy/thermal program for the man- http://dx.doi.org/10.1136/bmj.326.7395.911
rule for guiding treatment of a subgroup of pa- agement of patients with acute mechanical 29. Krauss J, Creighton D, Ely JD, Podlewska-Ely J.
tients with neck pain: use of thoracic spine ma- neck pain: a randomized clinical trial. Man Ther. The immediate effects of upper thoracic trans-
nipulation, exercise, and patient education. Phys 2009;14:306-313. http://dx.doi.org/10.1016/j. latoric spinal manipulation on cervical pain and
Ther. 2007;87:9-23. http://dx.doi.org/10.2522/ math.2008.04.006 range of motion: a randomized clinical trial. J
ptj.20060155 18. Gonzalez-Iglesias J, Fernandez-de-las-Penas Man Manip Ther. 2008;16:93-99.
7. Cleland JA, Childs JD, McRae M, Palmer JA, C, Cleland JA, Gutierrez-Vega Mdel R. Thoracic 30. MacDermid JC, Walton DM, Avery S, et al. Mea-
Journal of Orthopaedic & Sports Physical Therapy®

Stowell T. Immediate effects of thoracic manipu- spine manipulation for the management of surement properties of the neck disability index:
lation in patients with neck pain: a randomized patients with neck pain: a randomized clinical a systematic review. J Orthop Sports Phys Ther.
clinical trial. Man Ther. 2005;10:127-135. http:// trial. J Orthop Sports Phys Ther. 2009;39:20-27. 2009;39:400-417. http://dx.doi.org/10.2519/
dx.doi.org/10.1016/j.math.2004.08.005 http://dx.doi.org/10.2519/jospt.2009.2914 jospt.2009.2930
8. Cleland JA, Childs JD, Whitman JM. Psycho- 19. Haldeman S, Carroll L, Cassidy JD. Find- 31. Maher CG, Sherrington C, Herbert RD, Moseley
metric properties of the Neck Disability Index ings from the bone and joint decade 2000 AM, Elkins M. Reliability of the PEDro scale for
and Numeric Pain Rating Scale in patients with to 2010 task force on neck pain and its as- rating quality of randomized controlled trials.
mechanical neck pain. Arch Phys Med Rehabil. sociated disorders. J Occup Environ Med. Phys Ther. 2003;83:713-721.
2008;89:69-74. http://dx.doi.org/10.1016/j. 2010;52:424-427. http://dx.doi.org/10.1097/ 32. Maigne J-Y. A Prevention of the Vertebrobasilar
apmr.2007.08.126 JOM.0b013e3181d44f3b Accidents Following Cervical Thrust Manipula-
9. Cleland JA, Flynn TW, Childs JD, Eberhart S. The 20. Haldeman S, Kohlbeck FJ, McGregor M. tions: Recommandations of the SOFMMOO.
audible pop from thoracic spine thrust manipu- Stroke, cerebral artery dissection, and cervi- Available at: http://www.sofmmoo.com/english_
lation and its relation to short-term outcomes cal spine manipulation therapy. J Neurol. section/2_neck_pain/Cervical%20Manip.%20
in patients with neck pain. J Man Manip Ther. 2002;249:1098-1104. http://dx.doi.org/10.1007/ Recommandation%20SOFMMOO.htm. 2003.
2007;15:143-154. s00415-002-0783-4 33. Michaleff ZA, Costa LO, Moseley AM, et al.
10. Cleland JA, Glynn P, Whitman JM, Eberhart 21. Haldeman S, Kohlbeck FJ, McGregor M. Un- CENTRAL, PEDro, PubMed, and EMBASE are
SL, MacDonald C, Childs JD. Short-term ef- predictability of cerebrovascular ischemia the most comprehensive databases indexing
fects of thrust versus nonthrust mobilization/ associated with cervical spine manipulation randomized controlled trials of physical therapy
manipulation directed at the thoracic spine in therapy: a review of sixty-four cases after cervi- interventions. Phys Ther. 2011;91:190-197. http://
patients with neck pain: a randomized clinical cal spine manipulation. Spine (Phila Pa 1976). dx.doi.org/10.2522/ptj.20100116
trial. Phys Ther. 2007;87:431-440. http://dx.doi. 2002;27:49-55. 34. Miley ML, Wellik KE, Wingerchuk DM, Demaer-
org/10.2522/ptj.20060217 22. Hopkins WG, Marshall SW, Batterham AM, Hanin schalk BM. Does cervical manipulative therapy
11. Cleland JA, Mintken PE, Carpenter K, et al. J. Progressive statistics for studies in sports cause vertebral artery dissection and stroke?
Examination of a clinical prediction rule to medicine and exercise science. Med Sci Sports Neurologist. 2008;14:66-73. http://dx.doi.
identify patients with neck pain likely to benefit Exerc. 2009;41:3-13. http://dx.doi.org/10.1249/ org/10.1097/NRL.0b013e318164e53d
from thoracic spine thrust manipulation and a MSS.0b013e31818cb278 35. Moher D, Liberati A, Tetzlaff J, Altman DG.
general cervical range of motion exercise: multi- 23. Hoving JL, Koes BW, de Vet HC, et al. Manual Reprint--preferred reporting items for system-

@ MORE INFORMATION
center randomized clinical trial. Phys Ther. therapy, physical therapy, or continued care by a atic reviews and meta-analyses: the PRISMA
2010;90:1239-1250. http://dx.doi.org/10.2522/ general practitioner for patients with neck pain. statement. Phys Ther. 2009;89:873-880.
ptj.20100123 A randomized, controlled trial. Ann Intern Med. 36. Parkin-Smith G, WWW.JOSPT.ORG
Penter C. A clinical trial investi-

journal of orthopaedic & sports physical therapy | volume 41 | number 9 | september 2011 | 641

41-09 Cross.indd 641 8/16/2011 6:28:41 PM

 [ literature review ]
gating the effect of two manipulative approaches Ther. 1997;20:80-85. 2006;22:820-826. http://dx.doi.org/10.1097/01.
in the treatment of mechanical neck pain: a 39. S avolainen A, Ahlberg J, Nummila H, Nis- ajp.0000210937.58439.39
pilot study. J Neuromusc Syst. 1998;6:6-16. sinen M. Active or passive treatment for neck- 41. Young BA, Walker MJ, Strunce JB, Boyles RE,
37. Rivett D, Shirley D, Magarey M, Refshauge K. shoulder pain in occupational health care? A Whitman JM, Childs JD. Responsiveness of the
Clinical Guidelines for Assessing Vertebrobasilar randomized controlled trial. Occup Med (Lond). Neck Disability Index in patients with mechani-
Insufficiency in the Management of Cervical 2004;54:422-424. http://dx.doi.org/10.1093/ cal neck disorders. Spine J. 2009;9:802-808.
Spine Disorders. Available at: http://www.phys- occmed/kqh070 http://dx.doi.org/10.1016/j.spinee.2009.06.002
iotherapy.asn.au/index.php/quality-practice/ 40. Sim J, Jordan K, Lewis M, Hill J, Hay EM,
guidelines. Accessed November 1, 2010. Dziedzic K. Sensitivity to change and internal

@ MORE INFORMATION
38. Rogers RG. The effects of spinal manipulation consistency of the Northwick Park Neck Pain
on cervical kinesthesia in patients with chronic Questionnaire and derivation of a minimal
neck pain: a pilot study. J Manipulative Physiol clinically important difference. Clin J Pain. WWW.JOSPT.ORG

PUBLISH Your Manuscript in a Journal With International Reach

Downloaded from www.jospt.org at on August 12, 2019. For personal use only. No other uses without permission.

JOSPT offers authors of accepted papers an international audience. The

Journal is currently distributed to the members of the following
organizations as a member benefit:

• APTA's Orthopaedic and Sports Physical Therapy Sections

Copyright © 2011 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

• Sports Physiotherapy Australia (SPA) Titled Members

• Physio Austria (PA) Sports Group
• Belgische Vereniging van Manueel Therapeuten-Association Belge des
Thérapeutes Manuels (BVMT-ABTM)
• Comitê de Fisioterapia Esportiva do Estado do Rio de Janeiro
(COFEERJ)
• Mais Fisio
• Canadian Physiotherapy Association (CPA) Orthopaedic Division
• Sociedad Chilena de Kinesiologia del Deporte (SOKIDE)
• Suomen Ortopedisen Manuaalisen Terapian Yhdistys ry (SOMTY)
Journal of Orthopaedic & Sports Physical Therapy®

• German Federal Association of Manual Therapists (DFAMT)

• Hellenic Scientific Society of Physiotherapy (HSSPT) Sports
Injury Section
• Chartered Physiotherapists in Sports and Exercise Medicine (CPSEM)
and Chartered Physiotherapists in Manipulative Therapy (CPMT)
of the Irish Society of Chartered Physiotherapists (ISCP)
• Israeli Physiotherapy Society (IPTS)
• Gruppo di Terapi Manuale (GTM), a special interest group
of Associazione Italiana Fisioterapisti (AIFI)
• New Zealand Sports and Orthopaedic Physiotherapy Association
• Norwegian Sport Physiotherapy Group of the Norwegian Physiotherapist
Association
• Portuguese Sports Physiotherapy Group (PSPG) of the Portuguese
Association of Physiotherapists
• Singapore Physiotherapy Association (SPA)
• Orthopaedic Manipulative Physiotherapy Group (OMPTG) of the
South African Society of Physiotherapy (SASP)
• Swiss Sports Physiotherapy Association (SSPA)
• Association of Turkish Sports Physiotherapists (ATSP)

In addition, JOSPT reaches students and faculty, physical therapists and

physicians at more than 1,400 institutions in the United States and around the
world. We invite you to review our Information for and Instructions
to Authors at www.jospt.org and submit your manuscript for peer review at
http://mc.manuscriptcentral.com/jospt.

642 | september 2011 | volume 41 | number 9 | journal of orthopaedic & sports physical therapy

41-09 Cross.indd 642 8/16/2011 6:28:42 PM