Manual Therapy 15 (2010) 334e354

Contents lists available at ScienceDirect

Manual Therapy
journal homepage: www.elsevier.com/math

Systematic review

Manual therapy and exercise for neck pain: A systematic review

Jordan Miller a, Anita Gross a, b, *, Jonathan D'Sylva a, Stephen J. Burnie c, Charles H. Goldsmith b,
Nadine Graham a, Ted Haines b, Gert Brønfort d, Jan L. Hoving e
a
School of Rehabilitation Science, McMaster University, Hamilton, Canada
b
Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Canada
c
Chiropractic Clinician and Lecturer, Canadian Memorial Chiropractic College, Toronto, Canada
d
Wolfe-Harris Center for Clinical Studies, Northwestern Health Sciences University, Bloomington, MN, USA
e
Coronel Institute of Occupational Health and Research Centre for Insurance Medicine, Academic Medical Centre, Universiteit van Amsterdam, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: Manual therapy is often used with exercise to treat neck pain. This cervical overview group systematic
Received 24 November 2009 review update assesses if manual therapy, including manipulation or mobilisation, combined with
Received in revised form exercise improves pain, function/disability, quality of life, global perceived effect, and patient satisfaction
26 January 2010
for adults with neck pain with or without cervicogenic headache or radiculopathy. Computerized
Accepted 8 February 2010
searches were performed to July 2009. Two or more authors independently selected studies, abstracted
data, and assessed methodological quality. Pooled relative risk (pRR) and standardized mean differences
Keywords:
(pSMD) were calculated. Of 17 randomized controlled trials included, 29% had a low risk of bias. Low
Manual therapy
Exercise
quality evidence suggests clinically important long-term improvements in pain (pSMD-0.87(95% CI:
Neck pain
1.69,
0.06)), function/disability, and global perceived effect when manual therapy and exercise are
compared to no treatment. High quality evidence suggests greater short-term pain relief [pSMD-0.50
(95% CI:
0.76,
0.24)] than exercise alone, but no long-term differences across multiple outcomes for
(sub)acute/chronic neck pain with or without cervicogenic headache. Moderate quality evidence
supports this treatment combination for pain reduction and improved quality of life over manual therapy
alone for chronic neck pain; and suggests greater short-term pain reduction when compared to tradi-
tional care for acute whiplash. Evidence regarding radiculopathy was sparse. Specific research recom-
mendations are made.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Background pain reduction, global perceived effect and patient satisfaction in

Neck pain is a frequent impairment associated with disability
and substantive health care costs (Côté et al., 1998; Linton et al.,
1998; Borghouts et al., 1999; Hogg-Johnson et al., 2008). Manipu-
lation, mobilisation, or exercise applied as single-modal treatment
approaches for neck pain have gained some support in Cochrane
reviews (Gross et al., 2010 found earlier in this issue of Manual
Therapy; Kay et al., 2009). Many practitioners believe that solo-care
approaches do not accurately represent clinical practice or best-
practice for individual patients. In our previous reviews (Gross
et al., 1996, 2003, 2004, 2007), results supported the use of
combined mobilisation, manipulation and exercise for short-term
acute and chronic neck pain with or without cervicogenic head-
ache. Other quality reviews (Spitzer et al., 1995; Bogduk, 2001;
Magee et al., 2000; Hoving et al., 2001; Peeters et al., 2001; Vernon
et al., 2006, 2007; Hurwitz et al., 2008) agreed with these findings.
Given that results were inconclusive for: 1) neck pain with radi-
culopathy; 2) additional outcomes including function and quality of
life; and 3) all outcomes at long-term follow-up, a systematic
review update was warranted. The Cervical Overview Group update
for other single- or multi-modal manual therapy approaches are
reported elsewhere in this issue of Manual Therapy (Gross et al.,
2010; D’Sylva et al., in this issue) (see Fig. 1).
2. Objectives

* Corresponding author. School of Rehabilitation Science, McMaster University,

Our systematic review update assesses the effectiveness of
Hamilton, Canada. manual therapy and exercise for neck pain with or without radic-
E-mail address: grossa@mcmaster.ca (A. Gross). ular symptoms or cervicogenic headache on pain, function/

1356-689X/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.math.2010.02.007
 J. Miller et al. / Manual Therapy 15 (2010) 334e354
Fig. 1. Flow diagram of study selection.
disability, quality of life, global perceived effect, and patient
satisfaction.
3. Methods
3.1. Study selection
Abbreviated inclusion criteria follow; see Gross et al., 2010
earlier in this issue for detailed definitions.
3.1.1. Types of studies
Randomized controlled trial (RCT) or quasi-RCT.
3.1.2. Types of participants
Adults with acute (<1 month) to chronic (>3 months) neck pain
with or without radiculopathy or cervicogenic headache.
3.1.3. Types of interventions
Manual therapy, including manipulation or mobilisation tech-
niques, combined with exercise compared to: a placebo; a wait list/
no treatment control; an adjunct treatment (for example: mobi-
lisation and exercise plus ultrasound versus ultrasound); or another
treatment.
335
3.1.4. Type of outcome
Pain, function/disability, quality of life, global perceived effect,
and patient satisfaction for short-term (closest to 4 weeks) to long-
term (closest to 12 months) follow-up.
3.2. Search methods
Computerized searches by a research librarian were updated to
July 2009 and included bibliographic databases without language
restrictions for medical, chiropractic, and allied health literature.
The search strategies were detailed in our Cochrane review (Gross
et al., 2010 found earlier in this issue).
3.3. Data collection and analysis
At least two reviewers independently conducted citation iden-
tification, study selection, data abstraction, and risk of bias
assessment according to Cochrane methodology detailed in Gross
et al., (2010) earlier in this issue. Agreement was assessed for
study selection using the quadratic weighted Kappa statistic (Kw);
Cicchetti weights (Cicchetti, 1976). Characteristics of included
studies can be found in Table 1.
336
Table 1
Characteristics of the included studies.
Study/Participants
MethodseJadad scale
1a 1b 1c 2a 2b 2c 3
Allison et al. (2002)
Chronic neck pain (cervical
brachial pain syndrome)
n(A/R) 36/40
1000000
Total Jadad score: 1/5
Total van Tulder Score: 4/11
Total risk of bias score: NC
Bonk et al. (2000)
Acute WAD
N(A/R): 97/97
1000001
Total Jadad score: 2/5
Total van Tulder Score: 2/11
Total risk of bias score: NC
Brodin (1984, 1985)
Chronic neck pain disorder
with some 25% having
radicular findings or lower
cervical degenerative
changes
n(A/R) 63/71
1100001
Total Jadad score: 3/5
Total van Tulder Score: 5/11
Total risk of bias score: NC
Bronfort et al. (2001);
Evans et al. (2002)
Chronic neck pain
n(A/R) 158e160/191
1100001
Total Jadad score: 3/5
Total van Tulder score: 8/11
Total risk of bias score: NC
J. Miller et al. / Manual Therapy 15 (2010) 334e354
Interventions
Index treatment
Neural Treatment (NT): technique: neural tissue techniques,
mobilisation, neuromuscular techniques, home mobilisation
exercise techniques; frequency: NR; dose: 10 repetitions, 1e3
times/day
Comparison treatments
Articular treatment (AT): thoracic and glenohumeral mobilisation,
home exercise (stretches, theraband strengthening)
Control group (CG): no treatment, allowed to seek treatment from
a non-physiotherapy health care provider
Co-intervention: NR
Duration of treatment: 8 weeks, number of sessions NR
Duration of follow-up: 0 days
Index treatment
Activity Therapy (AT): technique: a. mobilisation (passive),
b. strengthening: isometric exercises, inter scapular muscle
strengthening, c. postural exercise, d. advice; frequency:
3 sessions week 1, 2 sessions week 2 and 3; dose: NR;
route: cervical spine
Comparison treatments
Collar Therapy (CT): collar worn during day for 3 weeks;
no physiotherapy, activity, exercise or mobilisation
Co-intervention: analgesic, anti-inflammatory
Duration of treatment: 3 weeks, 7 sessions
Duration of follow-up: 12 weeks
Index treatment
Group 3 (G3): technique: passive mobilisation as described by
Stoddard, massage, manual traction, superficial heat, analgesics,
education (neck school including exercise); frequency:
3 sessions/week; dose: NR; route: cervical spine
Comparison treatments
Group 1(G1): analgesic
Group 2 (G2): technique: mock therapy including superficial
massage, manual traction, electrical stimulation, analgesics,
education (neck school including exercise); frequency:
3 sessions/week; dose: NR
Co-intervention: NR
Duration of treatment: 3 weeks, 9 sessions
Duration of follow-up: 1 week
Index treatments
Spinal manipulation and low-technology exercises (SMT/Ex):
teqhnique: a. chiropractic: manipulation, massage, described by
Frymoyer, b. cardiovascular exercises: warm-up on stationary bike,
c. stretching: light stretches as warm-up, upper body strengthening
exercises, d. progressive resisted exercises, strengthening of neck
and shoulders described by Dyrssen et al: push-ups, dumbbell
shoulder exercises; dynamic neck extension, flexion and rotation
with variable weight attachment pulley system; e. sham: sham
microcurrent therapy; frequency: 20 one hour sessions over 11
weeks; dose: manipulation/massage 15 min, microcurrent 45 min;
route: cervical spine
Comparison treatment:
Spinal manipulation alone (SMT): teqhnique: a. chiropractic:
manipulation, massage, described by Frymoyer, b. sham: sham
microcurrent therapy; frequency: 20 one hour sessions over 3
months; dose: manipulation/massage 15 min, microcurrent 45 min;
route: cervical spine
High tech MedX and Rehabilitation Exercise (MedX): technique: a.
cardiovascular exercises: warm-up on dual action stationary bike, b.
stretching: light stretches as warm-up, c. strengthening of neck and
shoulders: using variable resistance equipment; MedX equipment
resistance for neck extension and rotation to fatigue; frequency ¼
20 one hour sessions over 11 weeks; dose: 20 repetition max
Co-intervention: home exercises including resisted rubber tubing for
Outcomes
Pain (VAS, 0 to 10)
Baseline median: NT 4.6, AT 5.1, CG 3.3
End of study median: NT 2.1, AT 3.4, CG 3.8
Absolute Benefit: NT 2.5, AT 1.7, CG
0.4
Reported Results: significant favoring NT
SMD(NT v CG):
0.71 (95% CI:
1.52 to 0.09) [power 56%]
SMD(NT v AT):
0.63 (95% CI:
1.46 to 0.20) [power 65%]
Function (NPQ, 0 to 36)
Baseline median: NT 12, AT 12.5, CG 12.5
End of study median: NT 9.5, AT 11.0, CG 11.5
Absolute benefit: NT 2.5, AT 1.5, CG 1.0
Reported results: not significant
SMD(NT v CG):
0.34 (95% CI:
1.12 to 0.45) [power 69%]
SMD(NT v AT):
0.24 (95%CI:
1.56 to 0.57)[power 67%]
QoL: NR
GPE: NR
Patient satisfaction: NR
Side effect: NR
Cost of care: NR
Pain (9 point linear scale):
Baseline: NR
Reported results: significant difference favoring AT
RR: 0.13 (95%CI: 0.02 to 1.02) [power 34%]
Function: NR
QoL: NR
GPE: NR
Patient satisfaction: NR
Side effects: NR
Cost of care: NR
Pain (9 point linear scale):
Baseline: NR
Reported results:
RR(3 v o-cntl): 0.67 (95% CI: 0.43 to 1.04) [power 18%]
RR(3 v 2): 0.59 (95% CI: 9.39 to 0.91)
Function: NR
QoL: NR
GPE: NR
Patient satisfaction: NR
Side effects: 10 in o-cntl; G2 reported discomfort, usually small
complaints; RR: 9.22 (95% CI: 0.61 to 14.30); note one subject
dropped out due to acute cerebral disease (n ¼ 1), others
dropped out for the following reasons: acute abdominal pain
(n ¼ 1); vacation and infection (n ¼ 1); acute pain in several
joints (n ¼ 1); incapable of following planned treatment (n ¼ 4)
Cost of care: NR
Cumulative advantage (six patient-oriented outcomes)
Reported results: favors SMT/Ex over SMT; MONOVA significant
[Wilk's Lambda ¼ 0.85, (F(12, 302) ¼ 2.2, p < 0.01)]
Pain (11-box scale, 0 to 10)
Baseline mean: SMT 56.6, MedX 57.1, SMT/Ex 56.0
End of study mean: SMT 36.5, MedX 29.8, SMT/Ex 31.1
Absolute benefit: SMT 20.1, MedX 27.3, SMT/Ex 24.9
Reported results: group difference in patient-rated pain ANOVA
[F(2,156) ¼ 4.2, p ¼ 0.02] favors the two exercise groups
SMD(SMT v MedX): 0.31 (95%CI:
0.08 to 0.70) [power 29%]
SMD(SMT v SMT/Ex): 0.24 (95%CI:
0.14 to 0.61) [power 28%]
SMD(SMT/Ex v MedX): 0.06 (95%CI:
0.33 to 0.44) [power 28%]
Function (Neck Disability Index, 0 to 50)
Baseline mean: SMT/Ex 27.2, SMT 27.6, MedX 28.1
End of study mean: SMT/Ex 16.1, SMT 19.9, MedX 15.6
Absolute benefit: SMT/Ex 11.1. SMT 7.7, MedX 12.5
Reported results: no significant group differences were found
ANOVA: F[2, 156] ¼ 2.04, p ¼ 0.13
SMD(SMT v MedX): 0.33 (95% CI:
0.06 to 0.71) [power 23%]
SMD(SMT v SMT/Ex): 0.31 (95% CI:
0.06 to 0.68) [power 28%]
SMD(SMT/Ex v MedX): 0.31 (95% CI:
0.06 to 0.68) [power 25%]
QoL (SF36 0 to 100)
Baseline mean: SMT/Ex 71.7 MedX 69.0, SMT 69.1
End of study mean: SMT/Ex 76.6, MedX 78.0, SMT 74.3
Absolute benefit: SMT/Ex 4.5, MedX 5, SMT 5.2

Table 1 (continued )
Study/Participants
MethodseJadad scale
1a 1b 1c 2a 2b 2c 3
Giebel et al. (1997)
Acute neck pain with
cervicogenic headache, WAD
n(A/R) 97/103
1000001
Total Jadad score: 2/5
Total van Tulder Score: 5/11
Total risk of bias score: NC
Hoving et al. (2002);
Hoving et al. (2006);
Korthals -de Bos et al. (2001);
Korthals-de Bos et al. (2003)
Acute, subacute, chronic
neck pain with and without
radicular findings, or
Cervicogenic headache
n(A/R) 178/183
1100001
Total Jadad score: 3/5
Total van Tulder Score: 9/11
Total risk of bias score: NC
J. Miller et al. / Manual Therapy 15 (2010) 334e354
Interventions
rotation and flexion; no concurrent treatment for neck pain by other
health care providers
Duration of treatment: 11 weeks, 20 sessions
Duration of follow-up: 52 weeks
Index treatment
Group 1 (G1): technique: mobilisation (passive, neuromuscular): for
individual segments active-assisted followed by passive
movements under light traction; PNF: active, hold-relax/contract-
relax technique to scaleni, levator scapula and trapezius in sitting;
light traction with mobilisation; exercise; analgesic; anti-
inflammatory; frequency: NR; dose: NR; route: cervical spine
Comparison treatment
Group 2 (G2): collar: worn continuously, take off at night, advised
no exercise; analgesic; anti-inflammatory
Co-intervention: analgesics and antiinflammatories allowed in both
groups
Duration of treatment: 3 weeks, sessions NR
Duration of follow-up: 9 weeks
Index treatment
Manual Therapy (MT): technique: muscular and articular
mobilisation techniques, coordination and stabilization techniques;
low velocity passive movements within or at the limit of joint range;
excluded manipulation; frequency: one session/week; dose: 45 min
sessions; route: cervical spine
Comparison treatments
Physical Therapy (PT): technique: active exercise therapies:
strengthening, stretching (ROM), postural/relaxation/functional
exercise; optional modalities: manual traction, massage,
interferential, heat; excluded specific mobilisations techniques;
frequency: one session/week; route: cervical spine
Continued Care by General Practitioner (GP): type: advice on
prognosis, psychosocial issues, self care (heat, home exercise),
ergonomics (pillow, work position), await further recovery; booklet
(ergonomics, home exercise); medication: paracetamol, NSAID;
frequency:follow-up every 2 weeks was optional; dose: 10 min
sessions.
Co-intervention: analgesics and antiinflammatories allowed in both
groups, home exercise for all three groups
Duration of treatment: 6 weeks, median 6 sessions for MT, 9 sessions
PT, 2 sessions GP.
Duration of follow-up: 52 weeks
337
Outcomes
Reported results: SMT/Ex was superior to both MedX and SMT
SMD (SMT/Ex v MedX): 0.10(95% CI:
0.28 to 0.48)
SMD (SMT/Ex v SMT):
0.14(95% CI:
0.52 to 0.23)
GPE improvement (1e9)
End of Study Mean: SMT/Ex 78.6, MedX 78.2, SMT 91.9
Reported Results:
SMD (SMT/Ex v MedX): 0.01(95% CI:
0.39 to 0.41)
SMD (SMT/Ex v SMT):
0.23(95% CI:
0.62 to 0.61)
Patient satisfaction (1e7, satisfied to dissatisfied)
Reported results: A clinically worthwhile cumulative advantage
favoring manipulation/exercise [low tech] group over exercise
[hightech] & manipulation ANOVA: F[2, 158] ¼ 6.7, p ¼ 0.002
SMD(SMT v MedX): 0.26 (95% CI:
0.13 to 0.65) [power 49%]
SMD(SMT v SMT/Ex): 0.71 (95%CI: 0.33 to 1.10)
SMD(SMT/Ex v MedX):
0.44 (95% CI:
0.83 to
0.05)
Side effects: increase neck or headache pain 8 SMT/Ex, 9 MedX, 6
SMT; increased radicular pain 1 SMT/Ex; severe thoracic pain 1
SMT; all cases self-limiting and no permanent injuries;
RR(SMT/Ex v MedX): 0.81 (95% CI: 0.23 to 1.55)
RR(SMT v MedX): 0.61 (95% CI: 0.23 to 1.55)
Cost of care: NR
Pain (NRS101, 0 to 100)
Baseline mean: G1 46.70, G2 49.20
End of study mean: G1 0.64, G2 3.39
Absolute benefit: G1 46.6, G2 45.81
Reported results: significant favor G1
SMD@2w treatment:
1.04 (95% CI:
1.46 to
0.61)
SMD@3w treatment þ 9w follow-up:
4.88(95% CI:
5.68 to

4.07)
NOTE that the large effect estimate for pain intensity in Giebel's
trial is an artifact of both groups markedly improving from
baseline to almost no pain. Clinically this benefit translates to
a 5.5% treatment advantage for the multimodal treatment.
Function [household activity, physical activity, activity of daily
living, social activity, neck mobility (11 point scale of MOPO
Fragenbogens)]
Baseline: NR
Reported Results: significant favor G1
SMD: 0.23 (95% CI:
0.17 to 0.63) (a positive sign denotes
advantage of the first group in the contrast) [power 100%]
QoL: NR
GPE: NR
Patient satisfaction: NR
Side effect: NR
Cost of care: G1 treatment economically favored
Direct care: G1 155DEM (78USD), G2 113DEM (57USD)
Sick days [number patients  days off work]: G1 187, G2 330
Pain (NRS, 0 to 10)
Baseline mean: MT 5.9, PT 5.7, GP 6.3
End of study mean: MT 1.7, PT 2.6, GP 2.2
Absolute benefit: MT 4.2, PT 3.1, GP 4.1
Reported results: significant favoring MT over PT
SMD (MT v PT):
0.41 (95% CI:
0.78 to
0.04)
SMD (MT v GP):
0.04 (95% CI:
0.40 to
0.32)
Function (Neck Disability Index, 0 to 50)
Baseline: MT 13.6, PT 13.9, GP 15.9
End of study mean: MT 6.4, PT 7.6, GP 7.4
Absolute benefit: MT 7.2, PT 6.3, GP 8.5
Reported results: significant favoring MT over PT
SMD (MT v GP): 0.17 (95% CI:
0.19 to 0.58) [power 22%]
SMD (MT v PT): 0.12 (95% CI:
0.48 to 0.25) [power 17%]
QoL (EuroQ, 0 to 100):
Baseline mean: MT 69.3, PT 75.3, GP 66.1
End of study mean: MT 73.5, PT 78.4, GP 70.2
Absolute benefit: MT 4.2, PT 3.1, GP 4.1
Reported results:
RR (MT v GP): 0.65 (95% CI: 0.40 to 1.06)
RR (MT v PT): 0.76 (95% CI: 0.45 to 1.28)
GPE (perceived recovery, 0e100%)
Reported Results: significant favoring MT over PT and GP
RR (MT v GP): 0.65 (95% CI: 0.40 to 1.06) [power 15%]
RR (MT v PT): 0.76 (95% CI: 0.45 to 1.28) [power 9%]
(continued on next page)
338
Table 1 (continued )
Study/Participants
MethodseJadad scale
1a 1b 1c 2a 2b 2c 3
Jull et al. (2002)
Chronic neck pain with
cervicogenic headache
n(A/R) 193/200
1100001
Total Jadad score: 3/5
Total van Tulder score: 8/11
Total risk of bias score: NC
Jull et al. (2007)
Chronic neck pain WAD II
n(A/R) 69/71
1100001
Total Jadad score: 3/5
Total van Tulder Score: 6/11
Total risk of bias score: NC
Karlberg et al. (1996)
Subacute neck pain with
cervicogenic headache
n(A/R) 17/17
1000001
Total Jadad score: 2/5
Total van Tulder Score: 5/11
Total risk of bias score: NC
McKinney et al. (1989);
McKinney et al. (1994)
J. Miller et al. / Manual Therapy 15 (2010) 334e354
Interventions
Index treatment
Manipulative Therapy (MT): technique: manipulation: high
velocity, low-amplitude manipulation described by Maitland;
mobilisation (low velocity); frequency: 2 sessions/week; dose: 30
min session duration; route: cervical spine
Combined Therapy (MT/ExT): technique: manipulation,
mobilisation, exercise; frequency: 2 sessions/week; dose: 30 min
session duration; route: cervical spine
Comparison treatment
Exercise Therapy (ExT): technique: therapeutic low load exercise to
cervical-scapular region: craniocervical flexor training with
pressure biofeedback, scapular muscle training, postural correction,
exercise performed throughout the day, isometric strengthening
with co contraction of neck flexion and extension, stretching as
needed; frequency: 2 sessions/w; dose: 30 min session duration
Control Group: no treatment
Co-intervention: NR
Duration of treatment: 6 weeks, 8 to 12 sessions
Duration of follow-up: 52 weeks
Index treatment
Multimodal physical therapy (MPT): exercise, mobilisations,
education and assurance
Exercise: technique: low load exercise to re-educate flexors,
extensors, and scapular stabilizers, exercises to retrain kinesthetic
sense; frequency: 10e15 sessions in 10 weeks; dose: low load
exercises; route: exercise advice and use of exercise diary
Mobilisations: technique: low velocity mobilisations; frequency:
10e15 sessions in 10 weeks; dose: low velocity mobilisations; route
cervical spine
Education and assurance: type: education and assurance provided
regarding ergonomics of activity of daily living and work practices;
frequency: recommended 2 times/day; dose: NR; route: education
and advice
Comparison treatment
Self management program (SMP): technique: education regarding
exercise, staying active and recovery process following a WAD,
ergonomic advice same in both arms; frequency: exercises
recommended 2 times/day; dose: NR; route: advice and education
Co-intervention: NR
Duration of treatment: 10 weeks, 10e15 sessions
Duration of follow-up: none
Index treatment
Physiotherapy Group (PT): technique: mobilisation (passive,
neuromuscular) as described by Kaltenborn and Lewit, soft tissue
treatment, physiotherapy treatment included exercise: stabilization
exercise described by Feldenkrais, relaxation techniques described
by Jacobson, non-steroidal anti-inflammatory, education;
frequency: median 13 sessions/9 weeks; dose: NR; route: cervical
spine
Comparison treatment
Delayed Treatment Group (D): wait period: 8 weeks without
treatment
Co-intervention: NR
Duration of treatment: median 8e9 weeks, 13 sessions
Duration of follow-up: none
Index treatment
Group 2 (G2): technique: mobilisation (passive, active) e active and
Outcomes
Patient satisfaction: NR
Side effect: benign and transient (increased neck pain > 2 days,
increased headache, arm pain/pins&needles, dizziness
Cost of care: total costs: not significant; total direct costs:
significant favors MT v PT; total indirect costs: significant favor
MT v GP; duration off work: significant favor MT v GP
Pain (headache intensity change score, VAS, 0 to 10)
Baseline mean: MT 4.8, ExT 5.4, MT/ExT 5.1, Cntl 5.3
Absolute benefit: MT 2.3, ExT 2.8, MT/ExT 2.7, Cntl 1.3
Reported results: significant favoring MT and ExT
SMD(MT v Cntl):
0.37 (95% CI:
0.78 to 0.04) [power 96%]
SMD(MT v ExT): 0.21 (95% CI:
0.18 to 0.61) [power 96%]
SMD(MT/ExT v ExT): 0.06 (95% CI:
0.35 to 0.46) [power 98%]
SMD(MT/ExT v Cntl):
0.58 (95% CI:
1.00 to
0.17)
Function (Northwick Park Neck Pain Questionnaire change
score, 0 to 36)
Baseline mean: MT 27.5, ExT 29.6, MT/ExT 29.7, Control 30.7
Absolute benefit: MT 11.2, ExT 15.7, MT/ExT 14.2, Control 6.4
Reported results: significant favoring MT or MT/ExT over
control; no significant difference between MT, ExT and MT/ExT
SMD(MT v Cntl):
0.39 (95% CI:
0.79 to 0.02) [power 100%]
SMD(MT v ExT): 0.32 (95% CI:
0.08 to 0.72) [power 100%]
SMD(MT/ExT v ExT): 0.11 (95% CI:
0.29 to 0.50) [power 59%]
SMD(MT/ExT v Cntl):
0.64 (95% CI:
1.06 to
0.23)
QoL: NR
GPE (participant perceived effect, VAS, 0 to 100)
Reported results: significant favoring MT and MT/ExT over
control, not significant for MT or MT/ExT when compared to ExT
SMD(MT v Cntl):
2.36 (95% CI:
2.89 to
1.83)
SMD(MT v ExT): 0.29 (95% CI:
0.10 to 0.69) [power 81%]
SMD(MT/ExT v ExT): 0.01 (95% CI:
0.38 to 0.40) [power 59%]
SMD(MT/ExT v Cntl):
2.73 (95% CI:
3.30 to
2.16)
Patient satisfaction: NR
Side effect: minor and temporary, 6.7% provoked by treatment
Cost of care: NR
Pain: NR
Function (Northwick Park Neck Pain Index, 0 to 36)
Baseline mean: MPT 37.7, SMP 38.4
End of treatment change score: MPT
10.4, SMP
4.6
Reported Results: significant
SMD(MPT v SMP):
0.49 (95% CI:
0.97 to
0.01)
QoL: NR
GPE perceived benefit (VAS, 0e10)
Mean perceived benefit: MPT 7.3, SMP-4.2
Reported results: significant
SMD(MPT v SMP) 1.32: (95% CI: 0.80 to 1.85)
GPE perceived symptom relief (VAS, 0e10)
Mean perceived symptom relief: MPT 6.9, SMP-4.2
Reported results: significant
SMD(MPT v SMP): 1.15 (95% CI: 0.64 to 1.66)
Patient satisfaction: NR
Side effect: NR
Cost of care: NR
Pain (headache intensity, VAS, 0 to 100)
Baseline mean: PT 54, D 56
End of study mean: PT 31, D 55
Absolute benefit: PT 23, D 1
Reported results: significant favoring PT
SMD:
1.47(95% CI:
2.58 to
0.36)
Function: NR
QoL: NR
GPE: NR
Patient satisfaction: NR
Side effect: NR
Cost of care: NR
Pain (VAS, 0 to 10)
Baseline median: G1 5.6, G2 5.3, G3 5.3

Table 1 (continued )
Study/Participants
MethodseJadad scale
1a 1b 1c 2a 2b 2c 3
Acute WAD
n(A/R) 170/247
1100001
Total Jadad score: 3/5
Total van Tulder Score: 5/11
Total risk of bias score: NC
Mealy et al. (1986)
Acute WAD
n(A/R) 51/61
1101001
Total Jadad score: 4/5
Total van Tulder Score: 6/11
Total risk of bias score: NC
Palmgren et al. (2006)
Chronic neck pain; radicular
signs and symptoms: NR
n(A/R): 36/41
1000001
Total Jadad score: 2/5
Total van Tulder score: 3/11
Total risk of bias score: NC
Persson et al. (2001);
Persson et al. (1994);
Persson et al. (1996);
Persson et al. (1997)
Chronic neck pain with
radicular findings
n(A/R) 79/81
1100001
Total Jadad score: 3/5
Total van Tulder score: 6/11
Total risk of bias score: NC
Provinciali et al. (1996)
Acute, subacute neck pain
with headache (cervicoen
-cephalic syndrome ¼
J. Miller et al. / Manual Therapy 15 (2010) 334e354 339
Interventions Outcomes
passive repetitive movements using principles of Maitland and End of study median: G1 3.0, G2 1.9, G3 1.8
McKenzie, heat/cold application, short wave diathermy, Absolute benefit: G1 2.6, G2 3.4, G3 3.5
hydrotherapy, traction, “the full gamut of physiotherapeutic aids Reported results: significantly better than rest (G1)
was available as deemed appropriate”, education e posture and SMD(2 v 1):
0.48 (95% CI:
0.90 to
0.06)
exercise to perform at home, standard analgesic, collar e fitted with SMD(2 v 3): 0.50 (95% CI:
0.28 to 0.39) [power 16%]
soft collar (intermittent use), frequency: 3 sessions/week; dose: 40- Function: NR
min sessions; route: cervical spine QoL: NR
Comparison treatment GPE: NR
Group 1 (G1): education: mobilisation after an initial 10e14 day rest Patient satisfaction: NR
period, general advice; analgesics; collar: fitted with soft collar Side effect: NR
(continuous use) Cost of care: NR
Group 3 (G3): education: posture correction, use of analgesics, use
of collar (restricted to very short periods in situations where their
neck was vulnerable to sudden jolting, if collar worn exercise should
be performed immediately after), use of heat sources, muscle
relaxation, encouraged to perform demonstrated mobilisation
exercises; analgesics
Co-intervention: NR
Duration of treatment: 6 weeks, 24 sessions
Duration of follow-up: 2 weeks
Index treatment Pain (pain intensity, linear analogue scale, 0 to 10)
Active Group (A): technique: mobilisation (passive) as described by Baseline mean: A 5.71, S 6.44
Maitland, exercise within the limits of pain, heat, ice, analgesics; End of study mean: A 1.69, S 3.94
frequency: daily, every hour at home; dose: NR; route: cervical Absolute benefit: A 4.02, S 2.50
spine Reported results: significant favoring active group
Comparison treatment: SMD:
0.86 (95% CI:
1.44 to
0.29)
Standard Group (S): soft cervical collar, worn for two weeks; rest for Function: NR
two weeks before beginning gradual mobilisation, analgesics QoL: NR
Co-intervention: NR GPE: NR
Duration of treatment: 8 weeks Patient satisfaction: NR
Duration of follow-up: none Side effect: NR
Cost of care: NR
Index treatment: (pragmatic, tailored to patient) Pain intensity (VAS, 0e100 mm)
Chiropractic care (chiro): technique: education, manipulation, Baseline mean: chiro 47.9, advice 42.2
myofascial technique, exercise (spine stabilizing for cervical region End of study mean: chiro 18.9, advice 45.3
and cervicothoracic junction); frequency: 3 to 5 sessions/week; Absolute benefit: chiro 29.0, advice -3.1
dose: NR; route: cervical spine Reported results: significant favoring chiropractic care
Comparison treatment: SMD:
1.56 (95% CI:
2.31 to
0.82)
Advice: advice given on simple regular exercise, done at own Function: NR
volition over 5 weeks GPE: NR
Co-intervention: information on anatomy, physiology of spine, QoL: NR
ergonomic principles, instruction on exercise and coping with pain, Patient satisfaction: NR
explanation of future outlook Side effects: NR
Duration of treatment: 5 weeks, 15 to 25 sessions Cost of care: NR
Duration of follow-up: none
Index treatment Pain intensity (VAS, 0 to 100)
PT Group: technique: physiotherapy decided by the physiotherapist Baseline mean: surgery 47, PT 50, collar 49
according to patient's symptoms and individual preferences End of study mean: surgery 30, PT 39, collar 35
[manual therapies (massage, manual traction, gentle mobilisation); Absolute benefit: surgery 17, PT 11, collar 14
modalities for pain relief like transcuatneous electrical nerve Reported results: not significant
stimulation, application of heat or cold (moist, ultrasound); exercise SMD(PT v collar): 0.16 (95% CI:
0.38 to 0.70) [power 82%]
(relaxation exercises; active stretching, strengthening, endurance SMD(PT v surgery): 0.33 (95% CI:
0.21 to 0.87) [power 76%]
exercises, postural correction); ergonomic instruction]; frequency: Worst pain (VAS, 0 to 100)
15 sessions/12 weeks; dose: 30e45 min sessions Baseline mean: surgery 72, PT 70, collar 68
Comparison treatments: End of study mean: surgery 42, PT 53, collar 52
Surgery Group: surgery [anterior cervical discectomy technique Absolute benefit: surgery 20, PT 17, collar 16
described by Cloward (1958); mobilisation on the 1st postoperative Reported results: not significant
day; cervical collar use for 1e2 days post-operatively ] SMD(PT v collar): 0.04 (95% CI:
0.50 to 0.57)
Collar Group: cervical collar (rigid collars during day; soft collar at SMD(PT v surgery): 0.28 (95% CI:
0.27 to 0.82)
night) Function: NR
Co-intervention: QoL: NR
Surgery group: 8 patients had 2nd operation, 11 patients received GPE: NR
physiotherapy Patient satisfaction: NR
PT group: 1 patient had surgery Side effects: NR
Collar group: 5 patients had surgery, 12 patients received Cost of care: NR
physiotherapy
Duration of treatment: 12 weeks, 15 sessions
Duration of follow-up: 56 weeks
Index treatment Pain intensity (neck pain intensity, VAS, 0 to 10)
Group A: technique: mobilisation (passive) as described by Mealy, Baseline median: A 6.8, B 7.4
massage as described by Mealy, exercise (eye fixation) as described End of study median: A 4.8 B 2.0
by Shutty to alter dizziness, neck school described by Sweeney, Absolute benefit: A 2.0, B 5.4
(continued on next page)
340
Table 1 (continued )
Study/Participants
MethodseJadad scale
1a 1b 1c 2a 2b 2c 3
fatigue, dizziness, poor
concentration,
accommodation and
adaptation to light intensity),
WAD
n(A/R) 60/60
1000001
Total Jadad score: 2/5
Total van Tulder score: 5/11
Total risk of bias score: NC
Skargren et al. (1998);
Skargren et al. (1997)
Chronic neck pain without
radicular findings
n(A/R) 317/323
1000001
Total Jadad score: 2/5
Total van Tulder score: 3/11
Total risk of bias score: NC
Vasseljen et al. (1995)
Chronic neck pain
n(A/R) 24/24
1000001
Total Jadad score: 2/5
Total van Tulder score: 4/11
Total risk of bias score: NC
Walker et al. (2008)
Chronic non-specific neck
pain
n(A/R) 94/98
1100001
Total Jadad score: 3/5
Total van Tulder score: 4/11
Total risk of bias score: 5/12
J. Miller et al. / Manual Therapy 15 (2010) 334e354
Interventions Outcomes
relaxation training based on diaphragmatic breathing in supine Reported results: significant favoring group A
position according to Shutty, active reduction of cervical and lumbar SMD:
0.79 (95% CI:
1.32 to
0.26)
lordosis based on suggestion provided by Neck School according to Function (Return to Work)
Sweeney, psychological support to reduce anxiety and limit Baseline: NR
emotional influence described by Radanov; frequency: 5 sessions/ Reported Results: significant favoring group A
week; dose: 1-h sessions SMD:
1.05 (95% CI:
1.59 to
0.26)
Comparison treatment: QoL: NR
Group B: transcutaneous electrical nerve stimulation, pulsed GPE (self assessment of outcome, ordinal scale
3 to þ3)
electromagnetic field, ultrasoundS, 10 1-h sessions/2 week Reported result: significant favoring group A, p < 0.001
Co-intervention: NR Patient satisfaction: NR
Duration of treatment: 2 weeks, 10 sessions Side effects: NR
Duration of follow-up: 24 weeks Cost of care:
Return to work: significant difference favors Group A,
a treatment advantage of 16 days; SMD:
1.05(95% CI:
1.59
to
0.51);
Sick days saved: 143 days saved favoring Group A
Index treatment Pain intensity (neck pain intensity change scores, VAS, 0 to 100)
Physiotherapy (PT): technique: 1% manipulation, 25% mobilisation, Baseline mean: Chiro 52, PT 61
15% traction, 25% soft tissue treatment, 33% McKenzie treatment, Absolute benefit: Chiro 16, PT 33
21% individual training, 15% transcutaneous electrical nerve Results: significant favoring PT
stimulation/ultrasound/cold, 15% individual program, 6% relaxation SMD (PT v Chiro):
0.66 (95% CI:
1.16 to
0.16)
training, 4% acupuncture, 1% instruction on individual training; Function (Oswestry Questionnaire, 0 to 100)
frequency: mean 7.5 sessions over mean 6.4 weeks Baseline mean: Chiro 25, PT 27
Comparison treatment Absolute benefit: Chiro 8, PT 12
Chiropractic (chiro): technique: 97% manipulation, 11% Results: not significant
mobilisation, 2% traction, 2% soft tissue treatment, 1% individual SMD(PT v Chiro):
0.32 (95% CI:
0.81 to 0.17)[power 100%]
training, frequency: mean 5.6 sessions over mean 4.9 weeks; dose: QoL (VAS 0 to 100):
NR; route: cervical spine Baseline mean: Chiro 38, PT 37
Co-intervention: 0e6 months of both chiropractic and End of study mean: Chiro 43.0, PT 46.1
physiotherapy treatment: Chiro 5.2%, PT 6.7% Absolute benefit: Chiro 5.0, PT 9.1
Duration of treatment: 5e6 weeks, 6 to 8 sessions Reported results:
Duration of follow-up: 52 weeks SMD (PT v Chiro):
0.22 (95% CI:
0.71 to 0.27)
GPE: NR
Patient satisfaction: NR
Side effects: NR
Cost of care:
Direct costs (for both neck and lumbar): significant favors PT,
SMD(PT v Chiro):
0.28 (95% CI:
0.50 to
0.05);
Indirect costs (for both neck and lumbar; of employed subjects);
a) median cost: not significant; SMD(PT v Chiro):
0.02(95%CI:

0.25 to 0.22);
b) sick leave: not significant, RR(PT v Chiro): 1.08 (95% CI: 0.75
to 1.54);
c) number of days off work: not significant; SMD(PT v Chiro):
0.06 (95% CI:
0.18 to 0.30)
Index treatment Pain intensity (neck pain intensity, VAS, 0 to10)
Group 1 (G1): technique: mobilisation (passive) e provided when Baseline mean: G14.2, G2 4.2
indicated, massage, exercise e strength on weight training End of study mean: G1 2.2, G2 2.1
apparatus, education e ergonomic principles, postural control, Absolute benefit: G1 2.0, G2 2.1
strength and flexibility training of shoulder and neck region; Reported result: not significant
frequency: 2 sessions/week; dose: 1 h sessions, 5e10 min of SMD: 0.09 (95% CI:
0.71 to 0.89) [power 9%]
massage, 5e10 min of exercise, and 3e4 min of stretching RR: 0.29 (95% CI: 0.07 to 1.10)
Comparison treatment Function: NR
Group 2 (G2): type: exercise e adopted from Dyrssen, 1.1 kg QoL: NR
dumbbells in both hands, 4 arm exercises each performed 10 times, GPE: NR
cycle repeated 3 times; load adjusted for 10 repetitions, abdominal Patient satisfaction: NR
and back exercises; breathing techniques; 5 min stretching exercise Side effects: NR
to shoulder/neck; education: same as Group 1; frequency: 3 Cost of care: NR
sessions/week; dose: 30 min sessions;
Co-intervention: NR
Duration of treatment: 5e6 weeks, 10 to 18 sessions
Duration of follow-up: 24 weeks; mailed questionnaire
Index treatment Pain intensity (cervical, VAS, 0 to 100)
Manual therapy and home exercise (MTE): technique: one to three Baseline mean: MTE 53.7, MIN 51.1
manual therapy techniques including thrust and non-thrust End of study mean: MTE 17.7, MIN 24.5
mobilisations, muscle energy and stretching techniques, home Absolute benefit: MTE 36.0, MIN 46.6
exercise including cervical retraction, deep neck flexors Reported results: significant at 3 week and 6 week follow-ups
strengthening, cervical rotation exercises; frequency: 2 sessions/ but not at 52 weeks
week; dose: NR; duration: 3 weeks; route: cervical spine 49 week follow-up SMD (MTE v MIN):
0.29 (95% CI:
0.71 to
Comparison treatment 0.13)
Minimal Intervention (MIN): technique: GP advice on posture, Function (NDI 0 to 50)
maintaining activity, range of motion exercises and medication use, Baseline mean: MTE 15.5, MIN 17.0
and sub-therapeutic ultrasound; frequency: 2 sessions/week; dose: End of study mean: MTE 5.5, MIN 10.6

Table 1 (continued )
Study/Participants
MethodseJadad scale
1a 1b 1c 2a 2b 2c 3
Ylinen et al. (2003)
Chronic neck pain
n(A/R) 179/180
1000001
Total Jadad score: 2/5
Total van Tulder score: 6/11
Total risk of bias score: NC
J. Miller et al. / Manual Therapy 15 (2010) 334e354 341
Interventions Outcomes
10% duty cycle, 0.1 W/cm2, 10 min; duration: 3 weeks Absolute benefit: MTE 10, MIN 6.4
Co-intervention: comparable between groups Reported results: significant at all three time points
Duration of treatment: 3 weeks, 6 session 49 week follow-up SMD (MTE v MIN):
0.68 (95% CI:
1.11 to
Duration of follow-up: 49 weeks
0.25)
QoL: NR
GPE (Global Rating of Change scale,
7 to 7)
49week follow-up mean: MTE 4.5, MIN 2.6
Reported results: significant at all three time points
49week follow-up SMD (MTE v MIN): 0.46 (95% CI: 0.03 to 0.88)
Patient satisfaction: NR
Side effects: NR
Cost of care: NR
Index treatment Pain intensity (neck, VAS, 0 to 100)
Endurance (E) Group: Baseline median: E 57, S 58, C 58
a. 12 day institutional rehabilitation program b. exercises [neck End of study median: E 14, S 12, C 19
flexor muscles by lifting the head up from the supine position in 3 Absolute benefit: E 8, S 9, C 3
series of 20 repetitions; 5 sessions per week; 45 min for 12 days Reported results: pain was at the same level in the 3 groups
b. dynamic exercises for the shoulders and upper extremities by SMD(E v C):
0.74 (95% CI:
1.12 to
0.37)
doing dumbbell shrugs, presses, curls, bent-over rows, flyes, and SMD(S v C):
0.85 (95% CI:
1.23 to
0.48)
pullovers performing 3 sets of 20 repetitions for each exercise with Function (NDI, 0 to 50)
a pair of dumbbells each weighing 2 kg. Baseline median: E 22, S 21, C 22
c. exercises for the trunk and leg muscles against their individual Reported results: disability index was the same level in the 3
body weights by doing a single series of squats, sit-ups, and back groups
extension exercises SMD(E v C):
0.50 (95% CI:
0.87 to
0.13)
d. each training session concluded with stretching exercises for the SMD(S v C):
0.60 (95% CI:
0.96 to
0.23)
neck, shoulder, and upper limb muscles for 20 min QoL: NR
e. also advised to perform aerobic exercise 3 times/week for 30 min GPE: NR
f. received written information about the exercises to be practiced at Patient satisfaction: NR
home 3 times/week Side effects: NR
g. multimodal rehabilitation program, including aspects commonly Cost of care: NR
associated with traditional treatment: relaxation training, aerobic
training, behavioral support to reduce fear of pain and improve
exercise motivation, and lectures and practical exercises in
ergonomics
h. during the rehabilitation course, each patient received 4 sessions
of physical therapy, which consisted mainly of massage and
mobilisation to alleviate neck pain and to enable those with severe
neck pain to perform active physical exercises
Strength (S) Group:
a. 12 day institutional rehabilitation program
b. exercise used elastic rubber band to train the neck flexor muscles
in each session performed in sitting, a single series of 15 repetitions
directly forward, obliquely toward right and left, and directly
backward
c. aim to maintain the level of resistance at 80% of the participant's
maximum isometric strength re coded at the baseline and at follow-
up visits
d. load was checked with a handheld isometric strength testing
device during the training sessions
e. dynamic exercises for the shoulders and upper extremities by
doing dumbbell shrugs, presses, curls, bent-over rows, flyes, and
pullovers with individually adjusted single dumbbell, 1 set for each
exercise with the highest load possible to perform 15 repetitions f.
dynamic exercises for the shoulders and upper extremities by doing
dumbbell shrugs, presses, curls, bent-over rows, flyes, and pullovers
performing 3 sets of 20 repetitions for each exercise with a pair of
dumbbells each weighing 2 kg.
g. exercises for the trunk and leg muscles against their individual
body weights by doing a single series of squats, sit-ups, and back
extension exercises
h. each training session concluded with stretching exercises for the
neck, shoulder, and upper limb muscles for 20 min
i. also advised to perform aerobic exercise 3 times/week for 30 min
j. received written information about the exercises to be practiced at
home 3 times/week
k. multimodal rehabilitation program, including aspects commonly
associated with traditional treatment: relaxation training, aerobic
training, behavioral support to reduce fear of pain and improve
exercise motivation, lectures and practical exercises in ergonomics
l. during the rehabilitation course, each patient received 4 sessions
of physical therapy, which consisted mainly of massage and
mobilisation to alleviate neck pain and to enable those with severe
neck pain to perform active physical exercises
(continued on next page)
342 J. Miller et al. / Manual Therapy 15 (2010) 334e354

Table 1 (continued )

Study/Participants Interventions Outcomes

MethodseJadad scale
1a 1b 1c 2a 2b 2c 3
Comparison treatment
3. Control (C) Group:
a. 3 days at rehabilitation centre for tests and recreational activities
b. advised to perform aerobic exercise 3 times/week, 30 min
c. written information about stretching exercises to practice at
home 3 times/week for 20 min per day
Co-intervention: analgesics; 80% of Control group, 49% of Endurance
group, and 57% of the Strength group used additional therapists as
follows: massage and stretching about 65%; hot and ice packs,
electrotherapy, acupuncture, traction, and zone therapy from
between 5 and 7%
Duration of treatment: 12 session, 3 weeks
Duration of follow-up: 12 months

KEY: Methodological Quality Rating using the Jadad et al., 1996 Criteria and scores: 1a. Was the study described as randomized? (Score 1 if yes); 1b and c. Was the method of
randomization described and appropriate to conceal allocation (Score 1 if appropriate and
1 if not appropriate); 2a. Was the study described as double-blinded? (Score 1 if
yes); 2b and c. Was the method of double blinding described and appropriate to maintain double blinding (Score 1 if appropriate and
1 if not appropriate); 3 Was there
a description of how withdrawals and dropouts were handled? (Score 1 if yes).
n (A/R) e sample number analyzed/randomized; WAD e whiplash associated disorder; I e Index treatment; C e Comparison or Control treatment; v e versus; NC e not
calculated; NR e not reported; VAS e visual analogue scale; NRS e numeric rating scale; SMD e standard mean difference, RR e relative risk; CI e confidence interval;
p e probability value; MONOVA e multiple analysis of variance.

We calculated standardized mean differences with 95% confi-
dence intervals (SMD; 95% CI) and relative risk (RR) with associated
number-needed-to-treat (NNT) and treatment advantage (%) (see
Table 2 and Table 3 for definitions). To facilitate analysis, we used
data imputation rules (see Appendix 1). Power analyses were
conducted for articles reporting non-significant findings (Dupont
and Plummer, 1990).
Prior to calculation of a pooled effect measure (pSMD or pRR),
we assessed the reasonableness of pooling on clinical grounds (See
Gross et al. Part 1 for details). We reported the statistical hetero-
geneity (I2 > 40%, p < 0.05) between studies when pooled results
were found to be heterogeneous.
 the adapted Cochrane ‘Risk of Bias’ method (maximum score
12, high score greater than five) (Furlon et al., 2009); Specific
criteria for each of the items within these methodological
assessment tools are described in our Cochrane review (Gross
et al., 2010 earlier in this issue). The Cervical Overview Group
uses a calibrated team of interdisciplinary assessors of senior
and junior methodologists making it not feasible to upgrade
all former trials in the our series to the new Cochrane “Risk of
Bias” system in this update.
3.5. Qualitative analysis of trial results

3.4. Quality assessment
Methodological quality was judged using:
 the Jadad et al., 1996 criteria (maximum score five, high score
greater than two);
 the Cochrane Back Review Group criteria (van Tulder et al.,
2003) (maximum score 11, high score greater than five); and
We assessed the quality of the body of the evidence using the
adopted GRADE approach (see Fig. 2 and Gross et al., 2010
eAppendix 4 earlier in this issue for GRADE domains) (The
Cochrane Collaboration, 2008; Furlan et al., 2007);
 High quality of evidence: Further research is unlikely to change
our confidence in the estimate of effect. There are consistent
findings among 75% of RCTs with low risk of bias that can be

Table 2
Calculations for treatment advantage and number-needed-to-treat.

Term
Percent treatment advantage (%)
Number-Needed-to-Treat (NNT)
Definiton
Calculation of the clinically important difference or change on a percent scale was estimated as follows.
Karlberg et al., 1996 data are used in this example:
The assumption made was that a positive mean/median value is improvement and a negative is deterioration.
Treatment control
Mean/Median Mean/Median Mean/Median Mean/Median
Baseline [SD] Final [SD] Baseline [SD] Final [SD]
54[23] 31[10] 56[15] 55[20]
% Improvement [treatment] equals the difference between the change in the treatment group [23] divided by the treatment
baseline [54] which equals 42.6%.
% Improvement [control] equals the difference between the change in the control group [1] divided by the control
baseline [56] which equals 1.8%.
The treatment advantage equals 42.6% minus 1.8% which equals 40.8%.
The number of patients a clinician needs to treat in order to achieve a clinically important improvement in one. Assuming
the minimal clinically important difference to be 10%, the baseline of the experimental mean of 54 minus 10 units change
on a 100 point scale equals 44 and for the control 56 minus 10 units change equals 46.
For experimental group: 44 minus 31 divided by 10 equals z ¼ 1.3 which gives an area under the normal curve of 0.9032.
For control group: 46 minus 55 divided by 20 equals z ¼
0.5 which gives an area under the normal curve of 0.3085.
Then 0.9032 minus 0.3085 equals 0.5947.
NNT equals 1 divided by 0.5947 which equals 1.66 or 2 when rounded.
 J. Miller et al. / Manual Therapy 15 (2010) 334e354 343

Table 3
NNT & treatment advantage.

Author/Comparison NNT Advantage (%)

Allison et al. (2002) outcome: pain 3 [clinically important pain reduction] 69.4%
Brodin (1985) outcome: pain 4 [complete neck pain reduction] N/A
Bronfort et al. (2001) outcome: pain 10 [clinically important pain reduction] 12.5%
Giebel et al. (1997) outcome: pain 8 [complete neck pain reduction]9 [complete H/A reduction] 5.5%
Hoving et al. (2002) outcome: pain 20 [clinically important pain reduction] 5.0%
Jull et al. (2002) outcome: pain 5 [clinically important pain reduction] 27.1%
McKinney et al. (1989) outcome: pain 11 [clinically important pain reduction] 17.1%
Mealy et al. (1986) outcome: pain 6 [clinically important pain reduction] 40.8%
Palmgren et al. (2006) outcome: pain 3 [clinically important pain reduction] 67.9%
Provinciali et al. (1996) outcome: pain 6 [clinically important pain reduction] 36.9%
31 [complete pain reduction]
Skargren et al. (1998) outcome: pain 4 [clinically important pain reduction] 26.1%
Vasseljen et al. (1995) outcome: pain 11 [clinically important pain reduction]4 [substantive pain reduction] 11.9%

generalized to the population in question. There are sufficient
data, with narrow confidence intervals. There are no known
or suspected reporting biases. (All of the domains are met.)
 Moderate quality of evidence: Further research is likely to have
an important impact on our confidence in the estimate of
effect and may change the estimate. (One of the domains is
not met.)
 Low quality of evidence: Further research is very likely to have
an important impact on our confidence in the estimate of
effect and is likely to change the estimate. (Two of the
domains are not met.)
 Very low quality of evidence: We are very uncertain about the
estimate. (Three of the domains are not met.)
4. Results
4.1. Description of studies
We selected 17 trials representing 31 publications from 1820
citation postings (See Fig. 1):
 17 studied neck pain: acute (Mealy et al., 1986; McKinney
et al., 1989; Giebel et al., 1997; Bonk et al., 2000); subacute
(Karlberg et al., 1996); chronic (Brodin, 1985; Vasseljen et al.,
1995; Skargren and Oberg, 1998; Bronfort et al., 2001;
Persson and Lilja, 2001; Allison et al., 2002; Jull et al., 2002;
Ylinen et al., 2003; Palmgren et al., 2006; Walker et al.,
2008); and mixed duration (Provinciali et al., 1996; Hoving
et al., 2002)
 5 studied whiplash associated disorders (WAD I and II): acute
(Mealy et al., 1986; McKinney et al., 1989; Giebel et al., 1997;
Bonk et al., 2000); and mixed duration (Provinciali et al., 1996);
 1 studied degenerative changes: chronic (Brodin, 1985);
 5 studied cervicogenic headache: acute (Giebel et al., 1997);
subacute (Karlberg et al., 1996); chronic (Jull et al., 2002); and
mixed duration (Provinciali et al.,1996; Hoving et al., 2002); and
 3 studied neck disorders with some radicular signs and
symptoms including WAD III: chronic (Brodin, 1985; Persson
et al., 2001); mixed (Hoving et al., 2002).
Agreement between pairs of independent authors from diverse
professional backgrounds for manual therapy was Kw 0.83, SD 0.15.
We excluded 77 RCTs based on the type of participant, intervention,
outcome, or design. One Spanish RCT is awaiting additional data
and 17 RCTs were ongoing studies.
4.2. Quality assessment for included studies
Five trials (29%) had a low risk of bias (Mealy et al., 1986;
Bronfort et al., 2001; Persson et al., 2001; Hoving et al., 2002; Jull
et al., 2002) and 12 trials had high risk of bias (Brodin, 1985;
McKinney et al., 1989; Vasseljen et al., 1995; Karlberg et al., 1996;
Provinciali et al., 1996; Giebel et al., 1997; Skargren et al., 1998;
Bonk et al., 2000; Allison et al., 2002; Ylinen et al., 2003;
Palmgren et al., 2006; Walker et al., 2008). See Fig. 3 for
summary table of risk of bias findings. Methodological weakness
that we found in multiple trials included: failure to describe or use
appropriate concealment of allocation (53%, 9/17) and lack of
effective blinding procedures [observer 59% (10/17); patient 100%
(17/17); care provider 100% (17/17)]. We note two limitations in
applying the methodological criteria to our trials: 1) it is difficult to
blind the patient and impossible to blind the care provider in
manual treatments; and 2) when self-report measures are used, the
trials do not fulfill the observer blinding criteria. Only a few trials
avoided co-intervention (24%; 4/17) and acceptable compliance
was found in 24% (4/17) of trials.

4.3. Main results

Various combinations of manual therapy and exercise emerged

for neck pain. Our findings are first reported by outcome and then
listed by type of comparison. The quality of evidence is an integral
part of our summary of findings reported in Table 4.

4.3.1. Pain

Fig. 2. Depiction of GRADE domains and scoring. Six domains may result in (
1)  versus a mock therapy or no treatment control: We found
subtraction while three domains may result in (þ1) addition. evidence of long-term pain relief from four trials comparing
344 J. Miller et al. / Manual Therapy 15 (2010) 334e354

Fig. 3. Methodological quality summary: review of authors judgements about each methodological quality item for included studies.

manual therapy and exercise to a control for chronic neck
disorder (Allison et al., 2002); subacute and chronic neck
disorder with headache (pSMD
0.87; 95% CI:
1.69 to

0.06; Karlberg et al., 1996; Jull et al., 2002); and chronic neck
pain with or without radicular findings (Brodin, 1985). This
translates into an absolute benefit of 23e27 mm VAS units,
a treatment advantage as high as 69%, and an NNT varying
from three to five. See Fig. 4 for forest plots of all pain
comparisons.
 versus primarily exercise with or without modalities: Bronfort
et al. (2001), Hoving et al. (2002), Jull et al. (2002), and
Vasseljen et al. (1995) compared manipulation, mobilisation
and exercise to exercise in participants with neck pain of
mixed duration, with or without cervicogenic headache.
Since the interaction effect of manipulation, mobilisation and
exercise is unclear, we elected to present these data within
this review. Results from the study by Vasseljen et al. (1995)
could not be combined due to a dissimilar outcome
Table 4
Summary of findings across all outcomes and comparisons.

Quality assessment Summary of findings

Study disorder subtype Design Limitations Inconsistency Indirectness Imprecision No. of patients Effect
follow-up (generalizability; (sparce data;
Int'n Cntl Effect size Clinical impact Quality
period group size) group size)
Effect size (95% CI) Absolute benefit
or pooled effect treatment
Size (95% CI) advantage NNT
1. Manipulation, mobilisation þ exercise versus mock therapy or no treatment control
Pain
Allison et al. (2002) chronic rct-I High (
1) na (-1) (
1) 17 10 SMD
0.71 AB 25 mm; TA 69%; NNT 3 very low
cervicobrachial pain (
1.52, 0.09)
Karlberg et al. (1996) subacute to rct-I High (
1) I2 54% (
I) 9 8 pSMD
0.87 AB 23 mm; TA nc; NNT nc low
chronic neck pain with rct-LT Low 48 46 (
1.69,
0.06) AB 27 mm; TA 27%; NNT 5
cervicogenic headache
Brodin (1985) chronic neck rct-ST Low na (
1) (
1) 35 41 RR 0.67(0.43, 1.04) AB nc; TA nc; NNT 4 low
pain þ/
radiculopathy and
degenerative changes
Function/Disability
Jull et al. (2002) subacute to rct-LT Low na (
1) (
1) 48 46 SMD
0.64 AB 14 NPQ; TA 31%, NNT 6 low

J. Miller et al. / Manual Therapy 15 (2010) 334e354

chronic neck pain with (
1.06,
0.23)
cervicogenic headache
Allison et al. (2002) chronic rct-post High (
1) na (
1) (
1) 17 10 SMD
0.34 AB 25 NPQ; TA 13%; NNT 11 very low
cervicobrachial pain (
1.12, 0.45)
Global perceived effect
Jull et al. (2002) subacute and rct-LT Low na (
1) (
1) 48 46 SMD
2.73 AB 69%; TA nc; NNT nc very low
chronic neck pain with (
3.03,
2.16)
cervicogenic headache

2. Manipulation, mobilisation þ exercise versus traditional care or general practitioner care

Pain
Giebel et al. (1997) acute WAD rct-ST High (
1) 47 50 pSMD
0.97 AB 47 mm; TA 5%, NNT 8 moderate
Mealy et al. (1986) acute WAD rct-I Low 26 25 (
1.32,
0.63) AB 40 mm; TA 41%, NNT 6
McKinney et al. (1989) acute WAD rct-ST High (
1) na (
1) (
1) 54 26 RR 0.96(0.58, 1.61) AB 34 mm; TA 17%, NNT 11 very low
Hoving et al. (2002) acute to rct-LT Low I2 50% (
1) 58 61 pSMD
0.14 AB 42 mm; TA 5%, NNT 22 low
chronic neck pain þ/
radiculopathy (
0.42, 0.13)
Walker et al., 2008 chronic neck pain rct-LT High (
1) 46 42 AB 36 mm; TA nc; NNT nc46
Function
Giebel et al. (1997) acute WAD rct-ST High (
1) na (
1) (
1) 47 50 SMD 0.23 nc very low
(
0.17, 0.63)
Hoving et al. (2002) acute to rct-LT Low I2 87% (
1) 58 61 pSMD
0.28 AB 7 NDI ;TA 29%; NNT 7 low
chronic neck pain þ/
(
1.05, 0.49)
radiculopathy or
cervicogenic headache
Walker et al. (2008) chronic neck pain rct-LT High (
1) 46 42 AB 6 NDI; TA 27%; NNT 5
Global perceived effect
Hoving et al. (2002) acute to rct-LT Low na (
1) (
1) 58 61 RR 0.65 nc low
chronic neck pain þ/
(0.40, 1.06)
radiculopathy or
cervicogenic headache
Walker et al. (2008) chronic neck pain rct-LT High (
1) na (
1) (
1) 46 42 SMD
0.52 nc very low
(
0.94,
0.09)
Quality of life
Hoving et al. (2002) acute to rct-LT Low na (
1) (
1) 58 61 SMD
0.04 AB 12 EuroQ(0e100), low
chronic neck pain þ/
radiculopathy (
0.40, 0.32) TA 2%, NNT na
or cervicogenic headache

(continued on next page)

345
Table 4 (continued)

346
Quality assessment Summary of findings
Study disorder subtype Design Limitations Inconsistency Indirectness Imprecision No. of patients Effect
follow-up (generalizability; (sparce data; Int'n Cntl Effect size Clinical impact Quality
period group size) group size)
Effect size (95% CI) Absolute benefit
or pooled effect treatment
Size (95% CI) advantage NNT
3. Manipulation, mobilisation þ exercise versus advice (including exercise advice)
Pain
Ylinen et al. (2003) (endurance rct-LT High (
1) na (
1) (
1) 58 59 SMD
0.74 AB 35 mm; TA nc; very low
focus versus advice) chronic neck pain (
1.12,
0.37) NNT nc
Ylinen et al. (2003) (strength focus rct-LT High (
1) na (
1) (
1) 60 59 SMD
1.85 AB 40 mm; TA nc; NNT nc very low
versus advice) chronic neck pain (
1.23,
0.48)
Palmgren et al. (2006) chronic neck pain rct-post High (
1) na (
1) (
1) 18 19 SMD
1.56 AB 29 mm; TA 68%, very low
(
2.31,
0.82) NNT 3
Function
Ylinen et al. (2003) (strengthening rct-LT High (
1) na 60 59 SMD
0.50 AB 9 NDI; TA 30%; NNT 4 very low
focus versus advice) chronic neck pain (
0.87,
0.13)
Ylinen et al. (2003) (endurance focus rct-LT High (
1) na (
1) (
1) 58 59 SMD
0.60 AB 8 NDI; TA 23%; NNT 5 very low
versus advice) chronic neck pain (
0.96,
0.23)

J. Miller et al. / Manual Therapy 15 (2010) 334e354

4. Manipulation, mobilisation þ exercise versus other treatment
Pain
Provinciali et al. (1996) acute to rct-IT High (
1) na (
1) (
1) 30 30 SMD
0.79 AB 20 mm; TA 37%, NNT 6 very low
chronic neck pain with cervicogenic headache and WAD (
1.32,
0.26)
Persson et al. (2001) (manipulation, rct-LT Low na (
1) (
1) 27 27 SMD 0.33 AB 11 mm; low
mobilisation, exercise, versus (
0.21, 0.87) TA nc; NNT nc
sxercise) chronic neck pain
with radiculopathy
Persson et al. (2001) rct-LT Low na (
1) (
1) 27 27 SMD 0.16 AB 11 mm; TA nc; NNT nc low
(manipiulation, mobilisation, exercise versus (
0.38, 0.69)
collar) chronic neck pain
with radiculopathy
Bonk et al. (2000) acute WAD rct-post High (
1) na (
1) (
1) 47 50 RR 0.13 nc very low
(0.02, 1.02)
Brodin (1985) chronic neck pain rct-ST High (
1) na (
1) (
1) 23 23 RR 0.67 nc very low
with and without radiculopathy (0.43, 1.04)

5. Manipulation, mobilisation þ exercise versus primarily manipulation or mobilisation

Pain
Skargren et al. (1998) chronic neck pain rct-LT High (
1) 28 39 pSMD-0.48 AB 16 mm; moderate
(
0.78,
0.18) TA 25%, NNT 4
Bronfort et al. (2001) chronic neck pain rct-LT Low 55 56 AB 26 mm;
TA 9%, NNT 14
Function
Skargren et al. (1998) chronic neck pain rct-LT High (
1) I2 92% (
1) 28 39 pSMD-0.31 AB12 Oswestery Units; low
(
0.61,
0.02) TA 20%; NNT 8
Bronfort et al. (2001) chronic neck pain rct-LT Low 55 56 AB 11 NDI;
TA 13%; NNT 9
Global perceived effect
Bronfort et al. (2001) chronic neck pain rct-LT Low na (
1) (
1) 51 50 SMD-0.23 nc low
(
0.62, 0.16)
Patient satisfaction
Bronfort et al. (2001) chronic neck pain rct-LT Low na (
1) (
1) 51 50 SMD-0.38 nc low
(
0.76,
0.01)
Quality of life
Skargren et al. (1998) chronic neck pain rct-LT High (
1) 28 39 pSMD-0.48 AB 9 VAS (0e100), TA 12%; NNT 9 moderate
Bronfort et al. (2001) chronic neck pain rct-LT Low 55 56 (
0.78,
0.18) AB 5 SF36 (0e100), TA -1%; NNT 52
 6. Manipulation, mobilisation þ exercise versus exercise with or without modalities
Pain
Jull et al. (2002) subacute to chronic rct-LT Low 48 51 ST: pSMD
0.50 ST: AB 34 mm; TA 40%; NNT 6 high
neck pain with cervicogenic headache (
0.76,
0.24) LT: AB 27 mm; TA 0.3%; NNT 38
Bronfort et al. (2001) chronic neck pain rct-LT Low 55 49 LT: pSMD
0.10 ST: AB 33 mm;
(
0.42, 0.21) TA 14%; NNT 9
LT: AB 25 mm;
TA 3%, NNT 11
Hoving et al. (2002) acute to chronic rct-LT Low 58 59 ST: AB 35 mm;
neck pain þ/
radiculopathy or TA 10%; NNT 16
cervicogenic headache LT: AB 42 mm;
TA 29%, NNT 7
Vasseljen et al. (1995) chronic neck pain rct-ST High (
1) na (
1) (
1) 12 12 RR 0.67 AB 20 mm; Very low
(0.35, 1.28) TA 12%, NNT 11
Function
Jull et al. (2002) subacute to chronic rct-LT Low na 48 51 pSMD-0.00 AB 14 NPQ; TA 5% high
neck pain with cervicogenic headache (
0.22, 0.22) favors exercise;
NNT na
Bronfort et al. (2001) chronic neck pain rct-LT Low na 55 49 AB 11 NDI; TA -4%
favors exercise; NNT na
Hoving et al. (2002) acute to chronic neck rct-LT Low na 58 59 AB 7.2 NDI;
pain þ/
radiculopathy or cervicogenic headache TA 7%; NNT na

J. Miller et al. / Manual Therapy 15 (2010) 334e354

Global perceived effect
Jull et al. (2002) subacute to chronic rct-LT Low na (
1) (
1) 48 51 pSMD-0.14 nc low
neck pain with cervicogenic headache (
0.44, 0.15)
Bronfort (200) chronic neck pain rct-LT Low na (
1) (
1) 51 45 nc
Hoving et al. (2002) acute to chronic neck rct-LT Low na (
1) (
1) 58 59 RR 0.76 nc low
pain þ/
radiculopathy or cervicogenic headache (0.45, 1.28)
Patient satisfaction
Bronfort et al. (2001) chronic neck pain rct-LT Low na (
1) (
1) 51 45 SMD 0.06 nc low
(
0.33, 0.44)
Quality of life
Bronfort et al. (2001) chronic neck pain rct-IT Low I2 67% (
1) 55 49 pSMD-0.18 AB 5 SF36 moderate
(
0.64, 0.28) (0e100), TA -6%
favore exercise;
NNT na
Hoving et al. (2002) acute to chronic neck rct-IT Low 58 59 AB 12 EuroQ
pain þ/
radiculopathy or cervicogenic headache (0e100), TA 12%; NNT na

Key: N e number; rct e randomized controlled trial; na e not applicable; nc e not calculated data not available; WAD e whiplash; ST e short term; LT e long term; I2 e Iganen value; pSMD e pooled standard mean difference; RR
e Relative Risk; AB e absolute benefit; TA e treatment advantage; NNT e number needed to treat; Quality e Cochrane GRADE of high, moderate, low, or very low; NPQ e Northwick Park Neck Pain Questionnaire; NDI e Neck
Disability Index.

347
348 J. Miller et al. / Manual Therapy 15 (2010) 334e354

Fig. 4. Forest plot of pain comparison: Manipulation or mobilisation and exercise versus comparison group.

measure for pain. When the remaining data were pooled,
results favored manipulation, mobilisation and exercise over
exercise alone in the short-term [SMD pooled
0.50 (95% CI:

0.76 to
0.24); heterogeneity: p ¼ 0.25, I2 ¼ 27%], while
results were similar in the long-term [pSMD
0.10 (95% CI:

0.42 to 0.21); heterogeneity: p ¼ 0.13, I2 ¼ 50%]. Differences
in exercise treatment and study groups may explain some of
the heterogeneity in results.
 versus primarily mobilisation and manipulation: Two trials
(Skargren et al., 1998; Bronfort et al., 2001) compared
 J. Miller et al. / Manual Therapy 15 (2010) 334e354
manipulation, mobilisation and exercise to manipulation and
mobilisation alone for chronic neck pain. Since the interaction
effect of manipulation, mobilisation and exercise is unclear,
we again elected to present these data within this review.
Pooled data favored combining exercise with mobilisation
and manipulations for intermediate to long-term pain relief
[pSMD
0.48 (95% CI:
0.78 to
0.18)] with a 9e25% treat-
ment advantage and NNT from 4 to 14.
 versus traditional care: Five trials (Mealy et al., 1986; McKinney
et al., 1989; Giebel et al., 1997; Hoving et al., 2002; Walker
et al., 2008) compared manipulation or mobilisation and
exercise to traditional care (at least two of three interventions
included: collar, medication and advice). McKinney et al.
(1989) used an outcome that we were not able to pool.
Giebel et al. (1997) and Mealy et al. (1986) found greater pain
relief in participants with acute WAD at short-term follow-up
[pSMD
0.97 (95% CI:
1.32 to
0.63]. Hoving et al. (2002)
and Walker et al. (2008) found no long-term difference in
pain between groups for subjects with neck pain of chronic or
mixed duration [pSMD
0.14 (95% CI:
0.42 to 0.13)].
 versus advice (including exercise advice): Palmgren et al.
(2006) and Ylinen et al. (2003) compared manual therapy
and exercise to exercise advice in participants with chronic
neck pain. Ylinen et al. (2003) reported that mobilisation,
massage and exercise produced greater pain relief than
exercise advice one year after treatment [endurance focused
exercise SMD
0.74 (95% CI:
1.12 to
0.37); strength
focused exercise SMD
0.85 (95% CI:
1.23 to
0.48)].
Palmgren et al. (2006) reported greater pain reduction with
the combined treatment approach immediately following the
treatment period [SMD
1.56 (95% CI:
2.31 to
0.82)]. The
absolute benefit across these treatments varied from 29 to 40
mm on a pain scale 0e100 mm and translates into a treat-
ment advantage of 68% and NNT 3.
 versus other treatment: Four trials (Brodin, 1985; Provinciali
et al., 1996; Bonk et al., 2000; Persson et al., 2001) compared
manipulation, mobilisation and exercise to other interventions:
1. modalities: transcutaneous electrical nerve stimulation,
pulsed electromagnetic field therapy and ultrasound
(Provinciali et al., 1996);
2. surgery (Persson et al., 2001);
3. collar (Bonk et al., 2000; Persson et al., 2000); and
4. medication as an adjunct to both trial arms (Brodin, 1985).
Provinciali et al. (1996) and Bonk et al. (2000) reported signifi-
cant results favoring mobilisation and exercise over other inter-
ventions for acute or subacute WAD with or without headache.
Brodin (1985) and Persson et al. (2001) reported no significant
difference between the groups for chronic neck pain with or
without radicular findings.
4.3.2. Function and disability
 versus a mock therapy or no treatment control: When
compared to a control, there was evidence of functional
improvements immediately post treatment and after long-
term follow-up with this combined care approach for chronic
neck pain (Allison et al., 2002) and for subacute/chronic neck
pain with cervicogenic headache (Jull et al., 2002). Long-term
treatment advantage of 31% could be achieved for one in six
patients or a 14 point absolute benefit on the Northwick Park
Neck Pain Questionnaire (0e36 scale) (see Fig. 5 for forest
plots of all function comparisons).
 versus primarily exercise with or without modalities: Bronfort
et al. (2001), Hoving et al. (2002) and Jull et al. (2002)
349
compared manipulation, mobilisation and exercise to exercise
alone. When the data were pooled, we found no statistical
difference between groups at short-term and long-term follow-
up for mixed duration or chronic neck pain with or without
cervicogenic headache [pSMD 0.00 (95% CI:
0.22 to 0.22)].
 versus primarily manipulation or mobilisation: Bronfort et al.
(2001) and Skargren et al. (1998) compared manipulation,
mobilisation and exercise to manipulation or mobilisation
alone. Pooled data favored manual therapy and exercise
[pSMD
0.31 (95% CI:
0.61 to
0.02); heterogeneity: p ¼
0.04, I2 ¼ 0%]. This reflects and absolute benefit from 11 Neck
Disability Index units to 12 Oswestery units with a treatment
advantage of 13e20% and NNT of 8e9.
 versus traditional care: We found three trials (Giebel et al.,
1997; Hoving et al., 2002, and Walker et al., 2008) comparing
manipulation or mobilisation and exercise to traditional care.
Giebel et al. reported no significant difference in short-term
functional improvements for acute WAD. At long-term follow-
up, pooled data suggests no significant difference between
groups for participants with neck pain of chronic or mixed
duration [pSMD
0.28 (95% CI:
1.05 to 0.49); heterogeneity:
p ¼ 0.006, I2 ¼ 87%]. The differences in results between groups
may be explained by the differences in exercise protocol.
 versus advice including exercise advice: Ylinen et al. (2003)
reported that mobilisation, massage and exercise produced
greater improvements in function than exercise advice for
chronic neck pain at long-term follow-up [endurance focus
SMD
0.50 (95% CI:
0.87 to
0.13); strength focus SMD

0.60 (95% CI:
0.96 to
0.23)].
4.3.3. Quality of life
 versus primarily exercise with or without modalities: Two trials
(Bronfort et al., 2001; Hoving et al., 2002) showed no statis-
tically significant difference when we pooled data to compare
manipulation or mobilisation and exercise to exercise alone at
long-term follow-up for chronic or mixed duration neck pain
[pSMD
0.16 (95% CI:
0.67 to 0.35); heterogeneity: I2 ¼ 73%,
p ¼ 0.06]. Statistical differences in the results may be
explained by some differences in exercise treatments or
patient groups
 versus primarily manipulation or mobilisation: We identified
two trials investigating the effects of manipulation or mobi-
lisation and exercise to manipulation or mobilisation alone on
quality of life for chronic neck pain (Bronfort et al., 2001;
Skargren et al., 1998). Pooled data suggested manipulation
or mobilisation and exercise had a similar effect at both short
and long-term follow-up [pSMD
0.17 (95% CI:
0.47 to
0.12); heterogeneity I2 0%, p ¼ 0.08].
 versus general practitioner care: One trial showed no signifi-
cant difference when manual therapy and exercise were
compared to general practitioner care (Hoving et al., 2002).
4.3.4. Global perceived effect
 versus a mock therapy or no treatment control: There was
evidence of a long-term benefit favoring manual therapy and
exercise when compared to a wait-list control (for subacute
and chronic neck disorder with cervicogenic headache SMD

2.73; 95% CI:
3.30 to
2.16) (Jull et al., 2002). The treat-
ment advantage was 69%.
 versus primarily exercise with or without modalities: Bronfort
et al. (2001), Hoving et al. (2002) and Jull et al. (2002)
350 J. Miller et al. / Manual Therapy 15 (2010) 334e354
Fig. 5. Forest plot of function comparison: Manipulation or mobilisation and exercise versus comparison group.
reported no statistically significant difference in global
perceived effect between groups receiving manipulation,
mobilisation and exercise and groups receiving exercise alone
for chronic neck pain with or without cervicogenic headache
and neck pain of mixed duration at long-term follow-up.
The combination of manipulation and exercise also produced
similar results to manipulation alone [pSMD
0.14 (95%
CI:
0.44, 0.15); I2 10%, p ¼ 0.29; Bronfort, 2001; Jull, 2001].
 versus traditional care: We found equal or greater benefits
when this care approach was compared to traditional care for
global perceived effect at long-term follow-up for neck pain
of chronic or mixed duration (Hoving et al., 2002; Walker
et al., 2008).
4.3.5. Patient satisfaction
 versus primarily exercise with or without modalities: There was
no difference in patient satisfaction when manipulation and
exercise was compared to exercise alone (Bronfort et al., 2001).
 versus primarily manipulation and mobilisation: When
compared to manipulation alone, manipulation and exercise
resulted in greater patient satisfaction at long-term follow-up
for chronic neck pain (Bronfort et al., 2001).
4.3.6. Adverse events
Side effects were reported in 18% (3/17) of trials. All side effects
were benign and transient and included cervical pain, thoracic pain,
headache, radicular symptoms, and dizziness. The rate of rare but
serious adverse events such as stokes or serious neurological deficits
could not be established from our review. Adverse events are dis-
cussed further in another review in this series (Carlesso, in this issue).
4.3.7. Cost of care
There was moderate evidence favoring reduced costs for care
consisting of manual therapy and exercise for acute, subacute, and
chronic mechanical neck disorder with or without headache or
radicular findings (Giebel et al., 1997; Hoving et al., 2002;
Provinciali et al., 1996; Skargren et al., 1998).
 J. Miller et al. / Manual Therapy 15 (2010) 334e354
4.4. Summary of findings
The summary of findings table (Table 4) details the relative and
absolute treatment effect with related quality assessment.
We note low quality evidence (2 pooled trials, 111 participants)
suggesting manipulation, mobilisation and exercise produce
greater long-term pain reduction (absolute benefit 23e37 mm, NNT
5, treatment advantage 27%) when compared to no treatment for
chronic neck pain and subacute/chronic neck pain with cervico-
genic headache. Additionally, low quality evidence suggests an
improvement in function (absolute benefit 14 to 25 Northwick Park
Neck Pain Questionnaire; NNT 6, treatment advantage 31%), and
global perceived effect (treatment advantage 69%) for subacute/
chronic neck pain with cervicogenic headache. The evidence is less
clear for chronic cervicobrachial pain.
High quality evidence (3 pooled trials, 320 participants)
suggests that manipulation or mobilisation and exercise produce
greater short-term pain relief than exercise alone and similar
effects in long-term pain, function, global perceived effect, patient
satisfaction and quality of life to various forms of exercise alone for
subacute/chronic neck pain with or without cervicogenic headache
and neck pain of mixed duration.
We found moderate quality evidence (2 pooled trials, 178
participants) showing that the combination of manipulation or
mobilisation and exercise produces greater pain reduction and
quality of life than manipulation or mobilisation alone for chronic
neck pain. Additionally, low quality evidence (2 pooled trials, 178
participants) supports evidence of benefit for improved function
and patient satisfaction.
We determined moderate quality evidence supports the use of
manipulations, mobilisations and exercise over traditional care for
reduction in pain at short-term follow-up for acute WAD (2 pooled
trials, 141 participants), but there is low quality evidence that this is
not achieved in the long-term for neck pain of chronic or mixed
duration (2 pooled trials, 208 participants). There is low quality
evidence showing no difference in function at short-term follow-up
for acute WAD and no difference in function, global perceived effect
or quality of life at long-term follow-up for neck pain of chronic or
mixed duration.
Very low quality evidence suggests that manipulation, mobi-
lisation and exercise may be superior to treatments with primarily
advice on exercise.
There was sparse low or very low quality evidence of no
difference for chronic neck pain with radiculopathy when manip-
ulation, mobilisation and exercise are compared to collar use,
surgery and analgesic medication.
5. Discussion
In our previous systematic review:
 up to 1996 (Gross et al., 1996): results remained inconclusive
for mobilisation or manipulation as a single intervention and
suggested support for combined mobilisation, manipulation
and exercise for short-term pain reduction.
 up to 2003 (Gross et al., 2003): results showed no evidence in
support of manipulation or mobilisation alone but showed
further support to the use of combined mobilisation, manipu-
lation and exercise in achieving clinically important but modest
pain reduction, global perceived effect and patient satisfaction
in acute and chronic neck disorder with or without cervicogenic
headache. There was insufficient evidence available to draw
conclusions for neck disorder with radicular findings. Other
high quality reviews (Spitzer et al., 1995; Bronfort, 1997; Magee
et al., 2000; Bogduk, 2001) agreed with these findings.
351
 up to 2009: The benefits of combined mobilisation or
manipulation plus exercise, were reinforced with additional
trials, across multiple outcomes, and in the long-term. A
clinically important change across multiple outcomes was
noted in subacute/chronic neck pain with or without cervi-
cogenic headache. Manipulation or mobilisation added to
exercises alone provides some added short-term pain relief.
We are in concordance with other findings showing a dearth of
quality evidence for manual therapy and exercise for WAD
(Verhagen et al., 2007). We are in agreement with the best evidence
synthesis by Hurwitz et al., (2008) who supported the use of
manual therapy and exercises for neck pain in comparison to
alternative treatments and suggested a lack of research in patients
experiencing neck pain with radicular symptoms.
Why combine manual therapy and exercise? The use of manipu-
lation and mobilisation alone provides short-term pain relief.
Exercise appears to improve pain and function over the long-term
(Kay et al., 2008). The combination of manual therapy and exercise,
however, seems to produce greater short-term pain reduction than
exercise alone and longer-term changes across multiple outcomes
in comparison to manual therapy alone.
What is the best manipulation or mobilisation to utilize in
combination with exercise? The answer to this question remains
unclear. Evidence from our Cochrane Review suggests that mobi-
lisation and manipulation produce similar changes in pain and
function and that one mobilisation technique may be favored over
another. Additional head to head comparisons of different manual
therapy techniques alone and in combination with exercise are
needed to determine the most effective approach.
What is the best exercise to use in combination with manual
therapy? Since the strength of evidence supporting the combina-
tion of manipulation, mobilisation, and exercise continues to grow,
future investigations should look at which exercise techniques are
optimally combined with manual therapy. A Cochrane review (Kay
et al., 2008) has demonstrated the positive effect of specific cervi-
coscapular resisted exercises, C1/2 self-SNAG exercises, craniocer-
vical endurance exercise and low load endurance exercise, and
upper extremity stretching and strengthening exercises, but the
optimal exercises to combine with manual therapy remain
unknown.
5.1. Implications for practice
 Manipulation or mobilisation and exercise produces
a greater long-term improvement in pain and global
perceived effect when compared to no treatment for chronic
neck pain, subacute/chronic neck pain with cervicogenic
headache, and chronic neck pain with or without radicular
findings.
 Manual therapy and exercise produce greater short-term pain
relief than exercise alone but produces no long-term differ-
ence across multiple outcomes for neck pain of chronic and
mixed duration with or without cervicogenic headache.
 The combination of manual therapy and exercise produces
greater improvements in pain, function, quality of life and
patient satisfaction when compared to manipulation or
mobilisation alone for chronic neck pain.
 Manipulations, mobilisations and exercise are favored over
traditional care for reducing pain at short-term follow-up for
acute WAD, but may be no different at long-term follow-up
for neck pain of chronic or mixed duration.
There was insufficient evidence available to draw any conclu-
sions for neck disorder with radicular findings.
352 J. Miller et al. / Manual Therapy 15 (2010) 334e354
5.2. Implications for research
Meta-analysis of data across trials and sensitivity analysis were
hampered by the wide spectrum of comparisons, treatment char-
acteristics and dosages. Factorial design would help determine the
active treatment agent(s) within a treatment mix. Phase II trials
would help identify the most effective treatment characteristics
and dosages for both exercise and manual therapy. The use of
similar validated outcome measures that are sensitive to change
and use of more homogenous diagnostic subgroups would increase
measurement precision. Vigilance to recommended CONSORT
standards would enhance methodological quality.
Acknowledgements
We thank our volunteers, students, and translators. This is one
review of a series conducted by the Cervical Overview Group:
Bronfort G, Burnie SJ, Cameron ID, Eddy A, Ezzo J, Goldsmith CH,
Graham N, Gross A, Haines T, Haraldsson B, Kay T, Kroeling P,
Morien A, Peloso P, Radylovick Z, Santaguida P, Trinh K, Wang E.
Declarations of interest
Two of our authors are authors in included studies. Although
Gert Bronfort and Jan Hoving were authors, they were not involved
in decisions around the inclusion, quality assessment or data
extraction of their studies.
Sources of support
External sources of support (from 1992 to 2009) have included
 Problem-based Research Award; Sunnybrook and Women's
College Health Sciences Foundation, Canada
 Consortial Center for Chiropractic Research e National Insti-
tutes of Health, Bethesda, MD, USA
 Hamilton Hospital Association, Canada
 University of Saskatchewan, Clinical Teaching and Research
Award, Canada
 Hamilton Health Sciences Corporation, ChedokeeMcMaster
Foundation, Canada
Appendix 1. Data imputation rules
The preliminary assumption made for imputation of missing
values was that data were missing completely at random (Little
and Rubin, 1987). In other words, it was assumed that data were
not missing due to some factors confounded with the treatment
effect.
McKinney et al. (1989) and Provinciali et al. (1996)
The standard deviation was estimated for the outcome pain
intensity using the reported p value boundary to establish the z
value. Since z equals the experimental mean minus the control
mean divided by the variation measure, the standard deviation
could be calculated. It was assumed that the variance in both the
control and experimental group were the same. The SMD was then
calculated using the reported median and estimated SD values.
Vasseljen et al. (1995)
Estimating Standard Deviations from 90% Confidence Intervals
on Medians (Gardner and Altman, 1989): The basic idea is to
estimate the order number for the lower bound (r) and the upper
bound (s) for a 90% Confidence Interval (CI) on the median where 1
less than or equal to r less than or equal to s less than or equal to n,
where n is the sample size.
Once r and s are known, the expected value of the r-th [rounded
down] and s-th (rounded up) order statistic can be used to estimate
the standard deviation from an assumed normal distribution,
where Z subscript 0.95 ¼ 1.645 is the 0.95 percentile from a stan-
dard normal distribution. (Similarly for 95% CI).
Examples used here.
n r s K2
90% CI
37 12 22 0.7713
34 12 23 0.8283
35 13 23 0.7255
39 14 26 0.7846
95% CI
12 3 10 1.5833
From r and s, an estimate of the empirical distribution function
can be obtained and from these an expectation for a normal
distribution function with unknown standard deviation.
Using Minitab (version 9.2) and the inverse cumulative distri-
bution function one can estimate the number of standard devia-
tions associated with that particular r and s for the specific sample
size.
One may then estimate the standard deviation by dividing the
difference in the confidence interval bounds by the constant K2.
Eg; 90% CI is 3e8 for n ¼ 35 the length is 8 e 3 ¼ 5 so the
standard deviation is 5/0.7255 ¼ 6.89
References
Allison GT, Nagy BM, Hall T. A randomized clinical trial of manual therapy for
cervico-brachial pain syndrome e a pilot study. Manual Therapy 2002;7
(2):95e102.
Bonk AD, Ferrari R, Giebel GD, Edelmann M, Huser R. Prospective, randomized,
controlled study of activity versus collar, and the natural history for whiplash
injury, in Germany. Journal of Musculoskeletal Pain 2000;8(1/2):123e32.
Brodin H. Cervical pain and mobilisation. International Journal of Rehabilitation
Research 1984;7(2):190e1.
Brodin H. Cervical pain and mobilisation. Manual Medicine 1985;2:18e22.
Bronfort G, Evan R, Nelson B, Aker PD, Goldsmith CH, Vernon H. A randomized
clinical trial of exercise and spinal manipulation for patients with chronic neck
pain. Spine 2001;26(7):788e99.
D’Sylva J, Miller J, Gross A, Burnie SJ, Goldsmith CH, Graham N, et al. Manual
therapy with or without physical medicine modalities for neck pain:
a systematic review. Manual Therapy, in this issue.
Evans R, Bronfort G, Nelson B, Goldsmith C. Two-year follow-up of a randomized
clinical trial of spinal manipulation and two types of exercise for patients with
chronic neck pain. Spine 2002;27:2383e9.
Giebel GD, Edelmann M, Huser R. Die distorsion der halswirbelsaule: Fruhfunk-
tionalle vs. ruhigstellende behandlung. Zentralbibliotak Chiropractic
1997;122:517e21.
Hoving JL, Koes BW, de Vet HCW, van der Windt DAWM, Assendelft WJJ, van
Mameren H, et al. Manual therapy, physical therapy, or continued care by
a general practitioner for patients with neck pain. A randomized controlled
trial. Annals of Internal Medicine 2002;136:713e22.
Hoving JL, Koes BW, de Vet HCW, van der Windt DAWM, Assendelft WJJ, van
Mameren H, et al. Manual therapy, physical therapy, or continued care by
a general practitioner for patients with neck pain: long-term results from
a pragmatic randomized clinical trial. Clinical Journal of Pain 2006;22(4):370e7.
Hoving JL, de Vet HCW, Koes BW, van Mameren H, Deville WJLM, van der
Windt DAWM, et al. In: Hoving JL, editor. Manual therapy, physical therapy, or
continued care by the general practitioner for patients with neck pain: long--
term results from a pragmatic randomized trial. 1st ed. Wageningen: Pons &
Looijen BV; 2001. p. 59e73.
Jull G, Trott P, Potter H, Zito G, Niere K, Shirley D, et al. A randomized controlled trial
of exercise and manipulative therapy for cervicogenic headache. Spine 2001;27
(17):1835e43.
 J. Miller et al. / Manual Therapy 15 (2010) 334e354
Jull G, Trott P, Potter H, Zito G, Niere K, Shirley D, et al. A randomized controlled trial
of exercise and manipulative therapy for cervicogenic headache. Spine 2002;27
(17):1845e943.
Jull G, Sterling M, Kenardy J, Beller E. Does the presence of sensory hypersensitivity
influence outcomes of physical rehabilitation for chronic whiplash? e A
preliminary RCT. Pain 2007;129:28e34.
Karlberg M, Magnusson M, Eva-Maj M, Melander A, Moritz U. Postural and symp-
tomatic improvement after physiotherapy in patients with dizziness of sus-
pected cervical origin. Archives of Physical Medicine and Rehabilitation
1996;77:874e82.
Korthals-de Bos IB, Hoving JL, van Tulder MW, Rutten-van Molken MP, Ader HJ, de
Vet HC, et al. Cost effectiveness of physiotherapy, manual therapy, and general
practitioner care for neck pain: economic evaluation alongside a randomized
controlled trial. British Medical Journal 2003;326(7395):911.
Korthals-de Bos IB, Hoving JL, van Tulder MW, Rutten-van Molken MP, Ader HJ, de
Vet HC, et al. Manual therapy is more cost-effective than physical therapy and
GP care for patients with neck pain. 1st ed. Wageningen: Pons & Looijen BV;
2001. pp. 75e89.
McKinney LA, Dornan JO, Ryan M. The role of physiotherapy in the management of
acute neck sprains following road-traffic accidents. Archives of Emergency
Medicine 1989;6(1):27e33.
McKinney MB. Behandlung der HWS-Distorsionen bei sog. Schleuderverletzungen.
Orthopade 1994;23:287e90.
Mealy K, Brennan H, Fenelon GC. Early mobilisation of acute whiplash injuries.
British Medical Journal 1986;92:656e7.
Palmgren PJ, Sandstrom PJ, Lundqvist FJ, Heikkila H. Improvement after chiropractic
care in cervicocephalic kinesthetic sensibility and subjective pain intensity in
patients with nontraumatic chronic neck pain. Journal of Manipulative and
Physiological Therapeutics 2006;29:100e6.
Persson L, Karlberg M, Magnusson M. Effects of different treatments on postural
performance in patients with cervical root compression. A randomized
prospective study assessing the importance of the neck in postural control.
Journal of Vestibular Research 1996;6(6):439e53.
Persson LCG, Carlsson CA, Carlsson JY. Long-lasting cervical radicular pain managed
with surgery, physiotherapy, or a cervical collar. A prospective, randomized
study. Spine 1997;22(7):751e8.
Persson LCG, Lilja A. Pain, coping, emotional state and physical function in patients
with chronic radicular neck pain. A comparison between patients treated with
surgery, physiotherapy or neck collar e a blinded, prospective randomized
study. Disability and Rehabilitation 2001;23(8):325e35.
Persson LCG, Moritz U, Brandt L, Carlsson CA. Cervical radiculopathy: pain, muscle
weakness and sensory loss in patients with cervical radiculopathy treated with
surgery, physiotherapy or cervical collar. A prospective, controlled study.
European Spine Journal 1994;6(4):256e66.
Provinciali L, Baroni M, Illuminati L, Ceravolo MG. Multimodal treatment to prevent
the late whiplash syndrome. Scandinavian Journal of Rehabilitation Medicine
1996;28:105e11.
Skargren EI, Carlsson PG, Oberg BE. One-year follow-up comparison of the cost and
effectiveness of chiropractic and physiotherapy as primary management for
back pain. Spine 1998;23(17):1875e84.
Skargren EI, Oberg BE, Carlsson PG, Gade M. Cost and effectiveness analysis of
chiropractic and physiotherapy treatment for low back and neck pain, Six-
month follow-up. Spine 1997;22(18):2167e71.
Skargren EI, Oberg BE. Predictive factors for 1-year outcome of low-back and neck
pain in patients treated in primary care: comparison between the treatment
strategies chiropractic and physiotherapy. Pain 1998;77:201e7.
Vasseljen O, Johansen BM, Westgaard RH. The effect of pain reduction on perceived
tension and EMG-recoded trapezius muscle activity in workers with shoulder
and neck pain. Scandinavian Journal of Rehabilitation Medicine 1995;27:243e52.
Walker MJ, Boyles RE, Young BA, Strunce JB, Garber MB, Whitman JM, et al. The
effectiveness of manual physical therapy and exercise for mechanical neck pain.
Spine 2008;33(22):2371e8.
Ylinen J, Takala EP, Nykanen M, Hakkinen A, Malkia E, Pohjolainen T, et al. Active
neck muscle training in the treatment of chronic neck pain in women. JAMA
2003;289(19):2509e16.
Additional references
Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, Elbourne D, et al, Consort
Group (Consolidated Standards of Reporting Trials). The revised CONSORT
statement for reporting randomized trials: explanation and elaboration. Annals
of Internal Medicine 2001;134(8):663e94.
Assendelft WJ, Bouter LM, Knipschild PG. Complications of spinal manipulation:
a comprehensive review of the literature. Journal of Family Practice 1996;42
(5):475e80.
Beattie P. Measurement of health outcomes in the clinical setting: applications to
physiotherapy. Physiotherapy Theory and Practice 2001;17:173e85.
Bogduk N. Whiplash: why pay for what does not work? Journal of Musculoskeletal
Pain 2001;8(1/2):29e53.
Bombardier C. Outcome assessment in the evaluation of treatment of spinal disorders:
summary and general recommendations. Spine 2000;25(24):3100e3.
Borghouts JAJ, Koes BW, Bouter LM. The clinical course and prognostic factors of
non-specific neck pain: a systematic review. Pain 1998;77:1e13.
353
Borghouts JAJ, Koes BW, Bouter LM. Cost-of-illness in neck pain in the Netherlands
in 1996. Pain 1999;80:629e36.
Bronfort G. Efficacy of spinal manipulation and mobilisation for low back and neck
pain: a systematic review and best evidence synthesis. Amsterdam: Thesis
Publishers; 1997.
Bronfort G, Haas M, Evans RL, Boutar LM. Efficacy of manipulation and mobilisation
for low back pain and neck pain: a systematic review and best evidence
synthesis. The Spine Journal 2004;4(3):335e56.
Carlesso L. Adverse events associated with cervical manipulation or mobilization
for neck pain. Manual Therapy, in this issue.
Childs JD, Cleland JA, Elliott JM, Flynn TW, Teyhen DS, Wainner RS, et al. Neck pain:
a clinical practice guideline linked to the international classification of function,
disability, and health from the orthopaedic section of the American Physical
Therapy Association. Journal of Orthopaedic and Sports Physical Therapy
2008;38(9):A1e34.
Cicchetti DV. Assessing inter-rater reliability for rating scales: resolving some basic
issues. British Journal of Psychiatry 1976;129:452e6.
Cleland JA, Childs JD, Fritz JM, Whitman JM, Eberhart SL. Development of a clinical
prediction rule for guiding treatment of a subgroup of patients with neck pain:
use of thoracic spine manipulation, exercise, and patient education. Physical
Therapy 2007;87(1):9e23.
Cohen J. Statistical power analysis for the behavioural sciences. 2nd ed. Hillsdale,
NJ: Lawrence Erlbaum Associates; 1988.
Côté P, Cassidy D, Corroll L. The Saskatchewan health and back pain survey. The
prevalence of neck pain and related disability in Saskatchewan adults. Spine
1998;23(15):1689e98.
Côté P, Kristman V, Vidmar M, Van Eerd D, Hogg-Johnson S, Beaton D, et al. The
prevalence and incidence of work absenteeism involving neck pain: a cohort of
Ontario lost-time claimants. Spine 2008;33(4S):S192e8.
Dixon WJ, Massey Jr FJ. Introduction to statistical analysis. 3rd ed. Toronto:
McGraw-Hill Book Co.; 1969.
Dupont WD, Plummer WD. Power and sample size calculations: a review and
computer program. Controlled Clinical Trials 1990;11:116e8.
Farrar JT, Young Jr JP, LaMoureaux L, Worth JL, Poole RM. Clinical importance of
changes in chronic pain intensity measured on an 11-point numerical rating
scale. Pain 2001;94:149e58.
Felson DT, Anderson JJ, Boers M, Bombardier C, Furst D, Goldsmith C, et al. American
college of rheumatology: preliminary definition of improvement in rheumatoid
arthritis. Arthritis and Rheumatism 1995;38(6):727e35.
Finch E, Brooks D, Stratford P, Mayo N. Physical rehabilitation outcome measures:
a guide to enhanced clinical decision making. 2nd ed. Lippincott Williams and
Wilkins; 2002.
Furlan AD, Pennick V, van Tulder MW, Shekelle P, Bombardier C, Bouter L. Updated
method guidelines for systematic reviews in the cochrane back review group.
Cochrane Back Review Group Editorial Board; 2007.
Gardner MJ, Altman DG. Statistics with confidence: confidence intervals and
statistical guidelines. British Medical Journal 1989;72(4):119e21.
Gemmell H, Miller P. Comparative effectiveness of manipulation, mobilisation and
the Activator instrument in the treatment of non-specific neck pain: a system-
atic review. Chiropractic and Osteopathy 2006;14(7).
Goldsmith CH, Boers M, Bombardier C, Tugwell P. Criteria for clinically important
changes in outcomes. Development, scoring and evaluation of rheumatoid
arthritis patients and trial profiles. Journal of Rheumatology 1993;20(3):
561e5.
Gross AR, Aker PD, Quartly C. Manual therapy in the treatment of neck pain.
Rheumatic Diseases Clinics of North America 1996;22(3):579e98.
Gross A, Goldsmith C, Kay T. Cervical overview group. conservative management of
mechanical neck disorders: a series of systematic reviews. In: Association of
chiropractic colleges and research agenda conference (ACC-RAC) 2003.
Davenport, IA: Consortial Center for Chiropractic Research; 2003. p. 29.
Gross AR, Kay TM, Kennedy C, Gasner D, Hurley L, Yardley K, et al. Clinical practice
guideline on the use of manipulation or mobilisation in the treatment of adults
with mechanical neck disorder. Manual Therapy 2002;7(4):193e205.
Gross AR, Hoving JL, Haines TA, Goldsmith CH, Kay T, Aker P, et al. Cervical overview
group. Manipulation and mobilisation for mechanical neck disorders. Cochrane
Database of Systematic Reviews 2004;(1). CD004249.
Gross AR, Goldsmith C, Hoving JL, Haines T, Peloso P, Aker P, et al. Cervical Overview
Group. Conservative management of mechanical neck disorders: a systematic
review. Journal of Rheumatology 2007;34(5):1083e102.
Gross A, Miller J, D'Sylva J, Burnie SJ, Goldsmith CH, Graham N, et al. Manipulation or
mobilisation for neck pain: a cochrane review. Manual Therapy 2010;15(4):315e33.
earlier article in this issue.
Guzman J, Hurwitz EL, Carroll LJ, Haldeman S, Cote P, Carragee EJ, et al. A new
conceptual model of neck pain: linking onset, course, and care: the bone and
joint decade 2000e2010 task force on neck pain and its associated disorders.
Spine 2008;33(4S):S14e23.
Hogg-Johnson S, van der Velde G, Carroll LJ, Holm LW, Cassidy JD, Guzman J, et al.
Bone and joint decade 2000e2010 task force on neck pain and its associated
disorders. the burdon and determinants of neck pain in the general population:
results of the bone and joint decade 2000e2010 task force on neck pain and its
associated disorders. Spine 2008;33(4S):S39e51.
Hoving JL, Gross AR, Gasner D, Kay T, Kennedy C, Hondras MA, et al. A critical
appraisal of review articles on the effectiveness of conservative treatment of
neck pain. Spine 2001;26(2):196e205.
354 J. Miller et al. / Manual Therapy 15 (2010) 334e354
Hurwitz EL, Carragee JEJ, van der Velde G, Carroll LJ, Nordin M, Guzman J, et al.
Treatment of neck pain: noninvasive interventions: results of the bone and joint
decade 2000e2010 task force on neck pain and its associated disorders. Spine
2008;33(4S):S123e52.
Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al.
Assessing the quality of reports of randomized clinical trials: is blinding
necessary? Controlled Clinical Trials 1996;17:1e12.
Jovey RD. General principles of pain management. In: Jovey RD, editor. Managing
pain. 1st ed. Toronto: Healthcare and Financial Publishing, Rogers Media; 2002.
p. 15e20.
Kay TM, Gross A, Goldsmith CH, Hoving JL, Brønfort G. Exercises for mechanical
neck disorders. Cochrane Database of Systematic Reviews 2009;(Issue 4). Art.
No.: CD004250.
Kay T, Rutherford S, McCall R, Voth S, Gross AR, Santaguida L, et al. Exercises for
mechanical neck disorders: a cochrane review update. Orthopaedic Sympo-
sium; 2008.
Kendal MG, Stuart A. The advanced theory of statistics. 2nd ed., vol. 1. New York:
Hafner Publishing Co.; 1963.
Kjellman GV, Skargren EI, Oberg BE. A critical analysis of randomised clinical trials
on neck pain and treatment efficacy. A review of the literature. Scandinavian
Journal of Rehabilitation Medicine 1999;31:139e52.
Lee CE, Simmonds MJ, Novy DM, Jones S. Self-reports and clinician-measured
physical function among patients with low back pain: a comparison. Archives of
Physical Medicine and Rehabilitation 2001;82:227e31.
Linton SJ, Hellsing AL, Hallden K. A population-based study of spinal pain among
35e45 year old individuals. Prevalence, sick leave and health care use. Spine
1998;23(13):1457e63.
Little RJA, Rubin DB. Statistical analysis with missing data. Toronto: J Wiley & Sons
Inc.; 1987.
MacDermid JC, Walton DM, Avery S, Blanchard A, Etruw E, McAlpine C, et al.
Measurement properties of the neck disability index: a systematic review.
Journal of Orthopaedic and Sports Physical Therapy 2009;39(5):400e17.
MacPherson H, White A, Cummings M, Jobst KA, Rose K, Niemtzow RC, STRICTA
Group. Standards for reporting interventions in controlled trials of Acupunc-
ture: the STRICTA recommendations. Journal of Alternative and Complementary
Medicine 2002;8(1):85e9.
Magee DJ, Oborn-Barret E, Turner S, Fenning N. A systematic overview of the
current research evidence on the selected treatment interventions on soft
tissue neck injury following trauma. Physiotherapy Canada 2000;52(2):111e30.
Makela M, Heliovaara M, Sievers K, Impivaara O, Knekt P, Aromaa A. Prevalence,
determinants and consequences of chronic neck pain in Finland. American
Journal of Epidemiology 1991;134:1356e67.
Olesen J. Classification and diagnostic criteria for headache disorders, cranial
neuralgias and facial pain. Cephalgia 1988;8(7):61e2.
Olesen J, Gobel H. ICD-10 Guide for Headaches. Guide to the classification, diagnosis
and assessment of headaches in accordance with the tenth revision of the
International classification of diseases and related health problems and its
application to neurology. Cephalalgia 1997;17(Suppl. 19):29e30.
Peeters GGM, Verhagen AP, deBie RA, Oostendorp RAB. The efficacy of conservative
treatment in patients with whiplash injury. Spine 2001;26(4):E64e73.
Rajala U, Keinanen-Kiukanniemi S, Uusimaki A, Kivela S-L. Muscoloskeletal pains
and depression in a middle-aged Finnish population. Pain 1995;61:451e7.
Rubinstein SM, Pool JJM, van Tulder MW, Riphagen II , de Vet HCW. A systematic
review of the diagnostic accuracy of provocative tests of the neck for diagnosing
cervical radiculopathy. European Spine Journal 2007;16:307e19.
Sackett DL, Straus SE, Richardson WS, Rosenberg W, Haynes RB. Evidence-based
medicine: how to practice and teach EBM. 2nd ed. Edinburgh: Churchill Liv-
ingstone; 2000.
Schumacher HR, Klippel JH, Koopman WJ, editors. Primer on the rheumatic
diseases. 10th ed. Atlanta: Arthritis Foundation; 1993.
Sjaastad O, Fredriksen TA, Pfaffenrath V. Cervicogenic headache: diagnostic criteria.
Headache 1990;30:725e6.
Spitzer WO, Leblanc FE, Dupuis M. Scientific approach to the assessment and
management of activity related spinal disorders. Spine 1987;7:S1e59.
Spitzer WO, Skovron ML, Salmi LR, Cassidy JD, Duranceau J, Suissa S, et al. Scientific
monograph of the Quebec task force on whiplash-associated disorders: rede-
fining “whiplash” and its management. Spine 1995;20:S1e73.
Stratford PW, Riddle DL, Binkley JM, Spadoni G, Westaway MD, Padfield B. Using the
neck disability index to make decisions concerning individual patients. Phys-
iotherapy Canada; 1999:107e19. Spring.
Takala J, Sievers K, Klaukka T. Rheumatic symptoms in the middle-aged population in
southwestern Finland. Scandinavian Journal of Rheumatology 1982;47:15e29.
The Cochrane Collaboration. Cochrane handbook for systematic reviews of inter-
ventions. Version 5.0.1 edition. Cochrane Library; 2008.
van Tulder MW, Assendelft WJJ, Koes BW, Bouter LM, the Editorial Board of the
Cochrane Collaboration Back Review Group. Method guidelines for systematic
reviews in the Cochrane Collaboration Back Review Group for spinal disorders.
Spine 1997;22(20):2323e30.
van Tulder M, Malmivaara A, Esmail R, Koes B. Exercise therapy for low back pain. A
systematic review within the framework of the Cochrane Collaboration Back
Review Group. Spine 2000;25(21):2784e96.
van Tulder M, Furlan A, Bombardier C, Bouter L. Editorial board of the cochrane
collaboration back review group. Updated method guidelines for systematic
reviews in the cochrane collaboration back review group. Spine 2003;28
(12):1290e9.
Verhagen AP, Scholten-Peeters GG, vanWijngaarden S, de Bie RA, Bierma-Zeinstra SM.
Conservative treatment for whiplash. Cochrane Database of Systematic Reviews;
2007;. doi:10.1002/14651858.CD003338.pub3. Issue 2. Art. No.: CD003338.
Vernon HT, Humphreys BK, Hagino CA. A systematic review of conservative treat-
ments for acute neck pain not due to whiplash. Journal of Manipulative and
Physiological Therapeutics 2005;28:443e8.
Vernon H, Humphreys BK, Hagino C. The outcome of control groups in clinical trials
of conservative treatments for chronic mechanical neck pain: a systematic
review. BMC Musculoskeletal Disorders 2006;7(58):1e10.
Vernon H, Humphreys K, Hagino C. Chronic mechanical neck disorders in adults
treated by manual therapy: a systematic review of change scores in randomized
clinical trials. Journal of Manipulative and Physiological Therapeutics
2007;30:215e27.
Westerling D, Jonsson BG. Pain from the neck-shoulder region and sick leave.
Scandinavian Journal of Social Medicine; 1980:8131e6.