Hindawi Publishing Corporation

Evidence-Based Complementary and Alternative Medicine

Volume 2017, Article ID 9519285, 10 pages
http://dx.doi.org/10.1155/2017/9519285

Review Article
Clinical Evidence of Chinese Massage Therapy (Tui Na) for
Cervical Radiculopathy: A Systematic Review and Meta-Analysis

Xu Wei,1 Shangquan Wang,2 Linghui Li,3 and Liguo Zhu3

1
Department of Scientific Research, Wangjing Hospital, China Academy of Chinese Medical Sciences, Huajiadi Street,
Chaoyang District, Beijing 100102, China
2
Department of General Orthopedics, Wangjing Hospital, China Academy of Chinese Medical Sciences, Huajiadi Street,
Chaoyang District, Beijing 100102, China
3
Department of Spine, Wangjing Hospital, China Academy of Chinese Medical Sciences, Huajiadi Street, Chaoyang District,
Beijing 100102, China

Correspondence should be addressed to Liguo Zhu; tcmspine@163.com

Received 28 October 2016; Accepted 26 January 2017; Published 20 February 2017

Academic Editor: Ciara Hughes

Copyright © 2017 Xu Wei et al. This is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Objective. The review is to assess the current evidence of Chinese massage therapy (Tui Na) for cervical radiculopathy. Methods.
Seven databases were searched. Randomised controlled trials incorporating Tui Na alone or Tui Na combined with conventional
treatment were enrolled. The authors in pairs independently assessed the risk of bias and extracted the data. Results. Five studies
involving 448 patients were included. The pooled analysis from the 3 trials indicated that Tui Na alone showed a significant lowering
immediate effects on pain score (SMD = −0.58; 95% CI: −0.96 to −0.21; 𝑍 = 3.08, 𝑃 = 0.002) with moderate heterogeneity
compared to cervical traction. The meta-analysis from 2 trials revealed significant immediate effects of Tui Na plus cervical traction
in improving pain score (MD = −1.73; 95% CI: −2.01 to −1.44; 𝑍 = 11.98, 𝑃 < 0.00001) with no heterogeneity compared to cervical
traction alone. No adverse effect was reported. There was very low quality or low quality evidence to support the results. Conclusions.
Tui Na alone or Tui Na plus cervical traction may be helpful to cervical radiculopathy patients, but supportive evidence seems
generally weak. Future clinical studies with low risk of bias and adequate follow-up design are recommended.

1. Introduction because most treatments had limited success in regard to

decreasing levels of pain and disability, increasing function,
In 1817, Parkinson first observed the clinical feature of cervical
radiculopathy [1]. Despite this long history of awareness, cer- and range of motion [9–16]. Thus, a general consensus
vical radiculopathy caused by degenerative changes remains had not been reached for treating cervical radiculopathy
one of the major contributors for neck pain, commonly worldwide [12]. In response to this situation, a new Chinese
seen condition across many patient populations [2, 3]. Pain clinical consensus of the treatments for cervical radiculopa-
became a common presenting symptom and might be asso- thy was established which will help clinicians to address the
ciated with motor or sensory disorders in areas innervated common problems in clinical practice [17]. Nevertheless, the
by the affected nerve root [4, 5]. In addition, neck pain Chinese clinical consensus was still limited to a narrow band
was more prevalent among women and prevalence peaked which mainly provides instructive suggestions to conven-
in middle age [6]. It has been estimated that patients with tional therapeutic methods, including neck immobilization,
cervical radiculopathy accounts for 60% to 70% of cervical physiotherapy, analgesic medications, and surgery; further,
spondylosis in China [7]. little is known about the effectiveness of complementary
In the absence of myelopathy or obvious muscle weakness and alternative medicine (CAM) interventions for patients
all patients with cervical radicular pain should be treated with cervical radiculopathy. In recent years, however, there
conservatively [8]. At present, the use of conventional modal- has been a growing tendency for improving the associated
ities for cervical radiculopathy remains debatable, primarily symptoms and clinical signs with CAM therapies [18–20].
2 Evidence-Based Complementary and Alternative Medicine

One of the sought-after CAM therapies is massage [21, trials that evaluated the therapeutic effect of Tui Na, including
22]. Massage, dating back to near 2500 BC, is recorded in one or more than two types of Tui Na methods, compared
the oldest existing medical works called Huangdi Nei jing with no treatment, placebo, or conventional therapies were
in China [23]. There are several types of massage, including considered. Combined therapy of Tui Na and other conven-
but not limited to traditional Chinese massage (known as tional interventions compared with other conventional inter-
Tui Na), Shiatsu, Thai massage, Swedish massage, reflexology, ventions in randomised controlled trials was also enrolled.
and myofascial trigger point release [24, 25]. For the patients The interventions containing other CAM treatments (manip-
with neck pain, massage therapy are effective for relieving ulation, mobilization, Chinese herbal medicine, acupuncture,
immediate or short-term pain symptoms, increasing range of Tai Chi, Wuqinxi exercise, qigong, cupping, etc.) in the Tui Na
motion, and improving neck dysfunction [26–29]. Clinicians or comparison group were excluded. Multiple publications
and physiotherapists who have been strictly trained usually reporting the same data were also excluded.
provide Tui Na for the patients with cervical radiculopathy.
The operator often uses the finger, hand acting on the muscle, 2.4. Types of Outcome Measures. The primary outcomes were
or soft tissue of body parts, based mainly on pain location neck and arm pain improvement, intensity of pain evaluated
and muscle tightness. A possible mechanism for the beneficial by at least one of the internationally recognized scales such as
effects of massage seems to be increased blood flow, relief of visual analogue scale (VAS), numerical rating scales (NRS),
muscular spasm, and pain suppression via moderate release and McGill pain questionnaire (MGPQ), or similar tools.
𝛽-endorphin [30]. As such, massage also can relieve anxiety The secondary outcomes analyzed in this review were neck
and depression resulting from cervical radicular neuralgia [31]. disability index (NDI), quality of life (SF-36, SF-12) for assess-
Currently, massage is widely applied to treat cervical ing treatment of cervical radiculopathy recommended by the
degenerative disc disease which included cervical radiculopa- guideline, or adverse events reported by the included studies
thy [32]. In China, Tui Na has been practiced for several [2]. The timing of outcome assessment was defined for four
years in the treatment of cervical radiculopathy. It involved time periods: immediately after treatment (up to one day),
a wide range of skilled and methodical manipulations best short-term follow-up (between one day and three months),
performed by an operator’s finger, hand, elbow, knee, or foot intermediate-term follow-up (between three months and one
applied to muscle or soft tissue at specific parts of the body year), and long-term follow-up (one year and beyond) [25, 26,
[33]. Meanwhile, there were a large number of published 28].
randomised controlled trials testing the effectiveness of Tui
Na in patients with cervical radiculopathy. To date, there 2.5. Information Sources and Search Strategy. We searched
were no known systematic reviews examining the efficacy seven electronic databases from their inception until January
of massage, specifically Chinese massage therapy, for the 31, 2016: PubMed, Cochrane CENTRAL, EMBASE, Chinese
management of cervical radiculopathy. Consequently, this National Knowledge Infrastructure (CNKI, 1979-), Wanfang
systematic review was undertaken to summarize the clinical database (1998-), Chinese Scientific Journals Database (VIP,
evidence of Tui Na in the treatment of cervical radiculopathy. 1989-), and Chinese Biomedical Literature Database (CBM,
1978-). These Chinese databases included medical and physi-
cal therapy, traditional Chinese medicine (TCM) articles, and
2. Methods full manuscripts plus conference proceedings which could
The protocol of systematic review was published in the PROS- be retrieved. The search terms included “cervical radicu-
PERO database which was available on https://www.crd.york lopathy”, “cervical spondylotic radiculopathy”, “nerve-root-
.ac.uk/PROSPERO/display record.asp?ID=CRD42016034004 type cervical spondylosis”, “cervicobrachial pain”, “cervico-
[34]. brachialgia”, “neck and arm pain”, “brachialgia”, “brachial
neuralgia”, “brachial plexus neuropath∗ ”, “neck pain with
2.1. Types of Studies. All completed randomised controlled radiculopathy”, or “neck disorder with radiculopathy” com-
trials comparing the efficacy of Tui Na for the treatment of bined with “massage”, “Chinese massage”, “Tui Na”, “TuiNa”,
cervical radiculopathy were enrolled. The studies reported in “manual therapy”, and “chiropractic”. There was no limitation
abstracts only were also considered, provided that there was on language, publication type, and status. Reference lists of
sufficient information in the abstract, or available from the review articles and included trials were searched.
author. Animal experiments were not inclusive. The strategy for searching PubMed was listed as follows:
(1) ((((((((((cervical radiculopathy [Title/Abstract]) OR
2.2. Types of Participants. The clinical diagnosis for cervical cervical spondylotic radiculopathy [Title/Abstract])
radiculopathy was required to be in accordance with the OR nerve-root-type cervical spondylosis [Title/
criteria of the North American Spine Society in 2011 (NASS Abstract]) OR cervicobrachial pain [Title/Abstract])
2011) [2]. If the criterion was not NASS 2011, the diagnosis OR cervicobrachialgia [Title/Abstract]) OR (neck
should be Chinese recognized criteria or definitions from [Title/Abstract] AND arm pain [Title/Abstract]))
national projected teaching materials. OR brachialgia [Title/Abstract]) OR brachial
neuralgia [Title/Abstract]) OR brachial plexus
2.3. Types of Interventions. In this review, Tui Na was defined neuropath∗ [Title/Abstract]) OR neck pain with
as the finger, hand, elbow, knee, or foot applied to muscle radiculopathy [Title/Abstract]) OR neck disorder
or soft tissue surrounding the neck. Randomised controlled with radiculopathy [Title/Abstract]
Evidence-Based Complementary and Alternative Medicine 3

(2) (((((massage [Title/Abstract]) OR Chinese massage 2.10. Quality of Evidence. The overall quality of evidence was
[Title/Abstract]) OR Tui Na [Title/Abstract]) OR assessed for each important outcome using the grading of
TuiNa [Title/Abstract]) OR manual therapy [Title/ recommendations assessment, development, and evaluation
Abstract]) OR chiropractic [Title/Abstract] (GRADE) approach [41–45]. Levels of quality of evidence
(3) (1) AND (2) recommended by the GRADE Working Group were defined
as high, moderate, low, and very low [46, 47].
2.6. Study Selection. Two authors independently searched
and assessed all the retrieved studies in an unblinded 3. Results
standardized manner. During title and abstract screening,
relevant studies were saved. Then the studies with accessible 3.1. Search Strategy. The electronic search found a total of
full text were retrieved and further assessed according to 6935 titles and abstracts. One additional clinical study was
the eligibility criteria. The PRISMA flow diagram was used identified from a conference proceeding. The full text of 76
to record the selection processes. Any disagreement was articles was retrieved and assessed the studies for inclusion in
resolved through discussion or, if required, we consulted a the review. As a result of limitation in the inclusion criteria we
third review author. excluded 71 studies: nonrandomised controlled trials (𝑛 = 2),
inappropriate intervention group (𝑛 = 34) such as Tui Na
2.7. Data Extraction and Management. Data collection was combined with manipulation/mobilization, incorrect control
independently conducted by two authors. For eligible studies, group (𝑛 = 15) including CAM interventions (Chinese herbal
two review authors extracted the data using the agreed form. medicine, acupuncture, moxibustion, cupping, plaster for
We extracted the following information: (1) the first author external use, iontophoresis of TCM, and functional exercise),
or the first two authors’ names and year of publication; (2) unacknowledged outcomes (𝑛 = 20) covering self-compile
basic characteristics of the studies—sample size, diagnosis assessment scale, and clinical effect evaluation criteria which
criteria, intervention characteristics (duration and number were not reported in the international guideline or expert
of treatment sessions), and outcome measures; (3) basic consensus. A total of 5 reports involving 448 patients with
characteristics of the patients—population characteristics relevant outcome data were eligible for inclusion. All included
(age, sex), baseline, and before and after treatment; (4) the trials were conducted in China and published in the Chinese
detailed description of Tui Na techniques. When information journals from 2011 to 2014 [35–39]. See Figure 1 for summary
regarding any of the above was unclear, we contacted authors of search results.
of the original studies.
3.2. Patient Characteristics. One study used the diagnosis
2.8. Risk of Bias. Two authors assessed the methodological criterion of State Administration of Traditional Chinese
quality using the criteria outlined in the Cochrane Handbook Medicine in 1994 (SATCM-1994) [37]; another used the
for Systematic Reviews of Interventions [40]. And two authors criterion from National Projected Teaching Materials of
compared the results and discussed difference according China in 2007 (NPTMC-2007) [38], while the other studies
to the Cochrane criteria until agreement was reached. The applied the criterion issued by Chinese Medical Association
domains included random sequence generation, allocation in 1993 (CMA-1993) [35, 36, 39]. However, the three criteria
concealment, blinding of participants and personnel, blind- used to diagnose cervical radiculopathy were almost the same
ing of outcome assessment, incomplete outcome data, selec- for the main symptoms (neck and arm pain, paresthesias,
tive reporting, and other bias. For other sources of bias, two numbness, and sensory changes) and signs (Spurling test,
aspects were identified: (1) a trial stopped early owing to Jackson test, and Eaton test) and imaging examination
some data-dependent process; (2) the baseline was extreme (cervical spine X-ray photograph, computed tomography,
imbalance. We reported the judgment for each criterion as or magnetic resonance imaging) compared with the NASS
“yes (low risk of bias),” “no (high risk of bias),” or “unclear criteria. The three diagnosis criteria were depicted for each
(information is insufficient to evaluate)”. study in the characteristics of included studies as shown in
Table 1. Participants were generally adult patients over 45
2.9. Data Synthesis. We carried out data analysis using years of age, explicitly excluding patients with myelopathy
the Review Manager 5.2.0 software. For all outcomes we or obvious muscle weakness. The prevalence of women
conducted the analysis based an intention-to-treat principle. participants was higher than that of men. No significant
For the continuous data, mean difference (MD) was used to difference on baseline was identified in all the studies. The
assess the difference in the same way between the groups. basic characteristics of included participants were described
Standardized mean difference (SMD) was chosen if clinical in Table 2. Of the studies which had outcomes included in the
outcome was the same but measured using different methods meta-analysis (𝑛 = 228/220), the main intervention strategies
in the different trials. 𝐼2 ≥ 50% was identified moderate or were categorized as Tui Na (𝑛 = 138), Tui Na plus cervical
significant heterogeneity. And the 95% confidence intervals computer traction (𝑛 = 90), and cervical computer traction
(CI) were calculated in the meta-analysis. The random-effect (𝑛 = 220). Cervical computer traction meant that the subjects
model was used to calculate the treatment effect across received mechanical intermittent cervical traction guided
trials when substantial heterogeneity existed. Where there are by computer programme. In three studies the intervention
high levels of heterogeneity we would advise caution in the was classified as “Tui Na” as the arm received only cervical
interpretation of results. computer traction [35, 37, 39]. In two studies, the primary
4 Evidence-Based Complementary and Alternative Medicine

Table 1: Basic characteristics of the included trials.

Sample Diagnostic Treatment
Reference Intervention Control Outcome measures
size criteria duration
Zhang et al. Cervical computer traction (20
120 CMA-1993 Tui Na (once every other day) 20 days Pain score (NRS)
2011 [35] minutes, once every other day)
Jiang 2013 Cervical computer traction (30
60 CMA-1993 Tui Na (once a day) + C 2 weeks Pain score (VAS)
[36] minutes, once a day)
Mi and Bi Cervical computer traction (20–30
60 SATCM-1994 Tui Na (once a day) 2 weeks Pain score (VAS)
2013 [37] minutes, once a day)
Huang 2013 NPTMC- Cervical computer traction (15
120 Tui Na (once a day) + C 28 days Pain score (VAS)
[38] 2007 minutes, once a day)
Cervical computer traction (20
Liu 2014 [39] 88 CMA-1993 Tui Na (once every other day) 2 weeks Pain score (VAS)
minutes, once a day)
CMA-1993 = diagnosis criterion issued by Chinese Medical Association in 1993; SATCM-1994 = diagnosis criterion of State Administration of Traditional
Chinese Medicine in 1994; NPTMC-2007 = diagnosis criterion from National Projected Teaching Materials of China in 2007; C = control group; NRS =
numerical rating scale; VAS = visual analogue scale.
Identification

Records identified through Additional records through

electronic search other sources
(n = 6935) (n = 1)

Records after duplicates removed

(n = 2512)
Screening

Records screened Records excluded

(n = 4424) (n = 4348)

Full-text articles excluded

Eligibility

with the following reasons

(n = 71)
Full-text articles Nonrandomised
assessed for eligibility controlled trials (n = 2)
(n = 76) Inappropriate
Intervention group (n = 34)
Incorrect control group
(n = 15)
Unacknowledged outcome
(n = 20)
Included

Studies included in the

systematic review
(n = 5)

Figure 1: Flow diagram of the literature searching and study selection.

aim of the study was to compare the effect of Tui Na plus and only pain score outcomes were included [35–39]. All
cervical computer traction with cervical computer traction the treatment duration of the included studies was beyond 2
[36, 38]. The treatment duration of traction ranged from 15 weeks. In our review, we summarized the massage techniques
to 30 minutes once every other day or once a day in the trials. as shown in Table 3.
In clinic application, the traction load was mainly determined
by different individuals’ weight and pain levels. These studies 3.3. Risk of Bias. Table 4 showed the summary of method-
were reported as separate comparisons with cervical traction ological quality, respectively. In the included studies the
 Table 2: Basic characteristics of the included subjects.
Pain scores
Reference T/C (M/F) Age (yrs) Baseline difference
BT AT
T/C: 60/60 Total population: 18–35 T: 8.67 ± 5.02 T: 3.11 ± 3.08
Zhang et al. 2011 [35] NSD
M/F: 48/72 M/F: NR C: 8.87 ± 5.27 C: 5.02 ± 3.51
Evidence-Based Complementary and Alternative Medicine

T: 30 (18/12) T: 51 ± 6.7 T: 7.2 ± 2.6 T: 3.1 ± 2.4

Jiang 2013 [36] NSD
C: 30 (16/14) C: 48 ± 7.1 C: 7.5 ± 2.3 C: 5.3 ± 2.5
T: 30 (14/16) T: 47.5 T: 4.53 ± 2.16 T: 0.62 ± 0.57
Mi and Bi 2013 [37] NSD
C: 30 (13/17) C: 45.6 C: 4.56 ± 1.80 C: 0.81 ± 1.15
T: 60 (31/29) T: 45.62 ± 11.64 T: 6.35 ± 1.11 T: 2.35 ± 0.78
Huang 2013 [38] NSD
C: 60 (26/34) C: 45.34 ± 12.06 C: 6.43 ± 1.05 C: 4.05 ± 0.84
T: 48 (20/28) T: 51.23 ± 5.20 T: 5.93 ± 1.08 T: 1.87 ± 1.03
Liu 2014 [39] NSD
C: 40 (16/24) C: 51.54 ± 5.15 C: 5.98 ± 1.01 C: 2.97 ± 1.34
T = treatment group, C = control group, NR = no reported, M = male, F = female, yrs = years, NSD = no statistical difference, BT = before treatment, and AT = after treatment.
5
 6

Table 3: Treatment with massage techniques in intervention group.

Reference Treatment with massage techniques
(1) Press and knead the soft tissue around the neck and shoulder using the thumb (the patient is in prostrate position)
(2) knead the acupoints: Fengchi (GB20), Fengfu (GV16), and Ashi points using a moderate force (the patient is in supine position)
Zhang et al. 2011 [35]
(3) Stretch and rotate the neck (the patient is in supine position)
(4) Press and knead the neck and shoulder, followed by grasping the shoulder (the patient is in sitting position)
(1) knead and palm-rub the neck and shoulder muscle (the patient is in sitting position all the time)
(2) Press and knead the acupoints: Fengchi (GB20), Tianzhu (BL10), Tianzong (SI11), Dazhui (GV14), Jianjing (GB21), Fengmen (BL12),
Jiang 2013 [36] Houxi (SI3), and Hegu (LI4)
(3) Stretch and rotate the neck
(4) knead the neck and shoulder muscle again
(1) knead and roll cervical spinous process and paraspinous tissue
(2) Grasp and knead the acupoints and meridians in the patient’s affected upper extremities: Jianjing (GB21), pericardium meridian, triple
Mi and Bi 2013 [37] energizer meridian, large intestine meridian, and small intestine meridian
(3) Roll the acupoints: Tianzong (SI11) and Zhongfu (LU1) using metacarpophalangeal joints
(4) Press the acupoints: Futu (LI18), Tianding (LI17), Quchi (LI11), Hegu (LI4), Yangxi (LI5), and Yanggu (SI5)
(1) Grasp and knead the soft tissue around the neck and shoulder and press and knead the acupoints: Tianzong (SI11), Jianjing (GB21), and
Dazhui (GV14); pull and stretch the upper extremities (the patient is in supine position all the time)
Huang 2013 [38] (2) Hold the distal end of the patient’s affected upper extremities and shake in small amplitude
(3) Rap the soft tissue around the neck and shoulder
(4) Pat the back using the hollow palm
(1) Grasp and knead the acupoints: Fengfu (GV16), Fengchi (GB20), Tianding (LI17), Tianzhu (BL10), Jianjing (GB21), Quepen (ST12),
Quchi (LI11), Xiaohai (SI8), Shousanli (LI10), Waiguan (TE5), Neiguan (PC6), Shenmen (HT7), and Hegu (LI4) (the patient is in sitting
position all the time)
Liu 2014 [39] (2) Roll, grasp, and knead the muscle around the neck and shoulder; especially press and knead the scalene muscle and pressure point
(3) Stretch and rotate the neck
(4) Grasp and knead the acupoints: Jianjing (GB21); hold the distal end of the patient’s affected upper extremities and shake in small
amplitude and pat the back and the upper extremities
Evidence-Based Complementary and Alternative Medicine
Evidence-Based Complementary and Alternative Medicine 7

Table 4: Methodological quality of the included trials based on the Cochrane Handbook.

Reference A B C D E F G
Zhang et al. 2011 [35] ? ? − ? + ? +
Jiang 2013 [36] ? ? − ? + ? +
Mi and Bi 2013 [37] ? ? − ? + ? +
Huang 2013 [38] + ? − ? + ? +
Liu 2014 [39] ? ? − ? + ? +
A = random sequence generation; B = allocation concealment; C = blinding of participants and personnel; D = blinding of outcome assessment; E = incomplete
outcome data; F = selective reporting; G = other bias; “+”, low risk of bias; “−”, high risk of bias; “?”, unclear risk of bias.

Study or subgroup Experimental Control Weight Std. Mean Difference Std. Mean Difference
Mean SD Total Mean SD Total IV, Random, 95% CI IV, Random, 95% CI
Liu 2014 1.87 1.03 48 2.97 1.34 40 32.8% −0.92 [−1.37, −0.48]
Mi and Bi 2013 0.62 0.57 30 0.81 1.15 30 28.6% −0.21 [−0.71, 0.30]
Zhang et al. 2011 3.11 3.08 60 5.02 3.51 60 38.5% −0.57 [−0.94, −0.21]

Total (95% CI) 138 130 100.0% −0.58 [−0.96, −0.21]

2 2 2
Heterogeneity: 𝜏 = 0.06; 𝜒 = 4.38, df = 2 (P = 0.11); I = 54% −2 −1 0 1 2
Test for overall effect: Z = 3.08 (P = 0.002) Favours [experimental] Favours [control]

Figure 2: Forest plot of the comparison of Tui Na versus cervical computer traction for pain score.

Experimental Control Mean Difference Mean Difference

Study or subgroup Weight
Mean SD Total Mean SD Total IV, Fixed, 95% CI IV, Fixed, 95% CI
Huang 2013 2.35 0.78 60 4.05 0.84 60 94.8% −1.70 [−1.99, −1.41]
Jiang 2013 3.1 2.4 30 5.3 2.5 30 5.2% −2.20 [−3.44, −0.96]

Total (95% CI) 90 90 100.0% −1.73 [−2.01, −1.44]

2 2
Heterogeneity: 𝜒 = 0.59, df = 1 (P = 0.44); I = 0% −2 −1 0 1 2
Test for overall effect: Z = 11.98 ( P < 0.00001) Favours [experimental] Favours [control]

Figure 3: Forest plot of the comparison of Tui Na plus cervical computer traction versus cervical computer traction for pain score.

method used to generate the randomisation sequence was lowering immediate effects on pain score (SMD = −0.58;
not described or was not clear. Only one study reported 95% CI: −0.96 to −0.21; 𝑍 = 3.08, 𝑃 = 0.002) in cervical
allocation by completely random number table, implying radiculopathy patients (Figure 2). A random-effects model
random sequence generation [38]. We did not identify the was used for statistical analysis according to the moderate
methods of concealing the study group allocation in the heterogeneity (𝐼2 = 54%). SMD was chosen because pain
studies. All the studies included in the review were not score was measured by different tools in the different trials;
placebo-controlled. In the physiotherapy study, blinding one was NRS [35] and the other two studies used VAS
might not have been convincing to patient and clinical staff. [37, 39]. For the pain measures, however, the difference of
The blinding of outcome assessment was not reported in the change scores between the intervention and control groups
trials. The amount of incomplete outcome data was generally should reach the 2 points that was generally accepted as
low, with attrition levels below 5%. But all the trials did not clinically meaningful [24]. So the result of meta-analysis did
report the follow-up data on the some outcome. Although the not support the clinical significance of Tui Na therapy.
studies provided some data on pain assessment, information
on other outcomes was sparse, such as quality of life. None of 3.5. Tui Na Plus Cervical Computer Traction versus Cervical
the studies stopped early. Baseline imbalance was not found Computer Traction for Pain Score. Two studies compared
in the demographic characteristics or the outcomes between Tui Na plus cervical computer traction to cervical computer
the study groups. traction for the outcome of pain score [36, 38]. The meta-
analysis from the 2 independent trials revealed significant
3.4. Tui Na versus Cervical Computer Traction for Pain Score. immediate effects of the combination therapy in improving
Three studies compared Tui Na to cervical computer traction pain score (MD = −1.73; 95% CI: −2.01 to −1.44; 𝑍 =
for the outcome of pain score [35, 37, 39]. Results from the 11.98, 𝑃 < 0.00001) (Figure 3) for cervical radiculopathy
pooled analysis indicated that Tui Na showed a significant patients. A fixed-effects model was used to analyze the data
8 Evidence-Based Complementary and Alternative Medicine

with no heterogeneity (𝐼2 = 0%). Similarly, the result only 4.2. Comparison with the Literature. Tui Na used alone
suggested a trend in favour of Tui Na therapy rather than or combined with analgesics drugs or cervical computer
clinical significance. traction has been widely used as adjunctive treatment for
cervical radiculopathy in China. And until now two sys-
3.6. Adverse Effects of Tui Na. Adverse effects were poorly tematic reviews reported the effectiveness or safety of Tui
reported in the studies. As shown in Table 1, no trial paid Na or manipulation for cervical radiculopathy [49, 50].
attention to the side effects from treatment or adverse events However, a systematic review has not been evaluated based
for the participants. on the PRISMA statement [49]. The other systematic review
confirmed the potential benefit and safety of massage or
3.7. Publication Bias. For the same intervention and outcome, manipulation, but only enrolling cervical computer traction
funnel plot analysis could not be conducted due to the as a control group [50]. Even more important, the interven-
small number of included studies (less than 10) in the meta- tions which also include high-velocity and low-amplitude
analysis. manipulation do not belong to massage in the two systematic
reviews. Therefore, neither of the systematic reviews has
3.8. Quality of Evidence. According to the GRADE approach, commented on the role of Chinese massage therapy in the
each comparison for the same outcome was assessed. When treatment of cervical radiculopathy.
Tui Na is compared to cervical computer traction, very low
quality evidence (3 trials, 268 participants) was identified to 4.3. Strengths and Limitations. As we all know, this is the
support the effect of Tui Na alone for improving the neck and first systematic review which analyzes the effect of Tui Na
arm pain. We downgraded the quality rating by three levels, for cervical radiculopathy. According to the results of current
including the limitations in study design and execution, literatures, we have come to the tentative conclusion about
inconsistency of results, and imprecision. When Tui Na plus the effect of Tui Na alone or Tui Na combined with cervical
cervical computer traction is compared to cervical computer computer traction.
traction, low quality evidence (2 trials, 180 participants) Nevertheless, our study had several limitations. First
was identified to confirm the effect of Tui Na plus cervical of all, the included studies used a relative small number
computer traction for relieving the neck and arm pain. The of patients, varying between 60 and 120 patients. None
quality of evidence for this outcome was low because of the of the studies performed a power calculation. Further,
previous studies with high risk of bias and imprecise result. multiple diagnosis criteria might introduce some bias into
the study. In the second place, there was a lack of well-
4. Discussion designed randomised controlled trials. For example, concrete
random method and allocation concealment were unclear.
4.1. Summary of Main Results. Massage is one of the fre- The majority of the studies were rated “unclear” or “high”
quently applied nonoperative interventions to relieve the risk of bias with regard to blinding. Generally, results need
symptoms of cervical radiculopathy [48]. However, evidence to be downgraded as to level of effectiveness as no blind
to confirm the efficacy of massage including Tui Na for cer- placebo-controlled trials were included in the meta-analysis.
vical radiculopathy is scarce. This systematic review included And no dose-response study design was found in all the
5 studies that investigated the effect of Tui Na alone or Tui literatures. In the third place, our review clearly described
Na plus cervical computer traction on cervical radiculopathy massage techniques as stand-alone treatments without cer-
patients. All the primary studies were from China. The vical manipulation or mobilization techniques, but there is
treatment duration was within 1 month in the studies. All the lack of consistent terminology for massage therapy, as
the randomised studies reported a reduction in pain scores reported by others [51, 52]. Fourthly, the included scales were
between the groups. mainly used to evaluate the variation of pain, whereas other
These results suggested that Tui Na or the combination scales in relation to cervical spine function or quality of life
therapy alleviated the pain symptom and that the focus were not enrolled. Last but not the end, the previous articles
should lie on the immediate effects on pain score. According were limited related to adverse events in the studies. However,
to the results of meta-analysis, the highest reduction in pain the known or expected adverse effects of Tui Na might include
dizziness, nausea, increasing pain in the clinical practice and
scores was associated with the combination therapy. But
palpitation, and headache reported in the literature [53].
the difference between the intervention and control groups
did not reach what is generally considered the minimally
clinically important difference. This review included no
5. Conclusion
studies to investigate the short-, intermediate-, and long-term In conclusion, there was weak endorsement of Tui Na
effect of Tui Na. From the Tui Na’s safety point of view, we did alone or Tui Na plus cervical computer traction for cervical
not determine the adverse event or adverse drug reaction in radiculopathy patients. And the safety of Tui Na could not
the course of treatment so far. However, there was no high yet be judged. Based on the available randomised trials, there
quality of evidence to support the effect of Tui Na, whether is a lack of strong evidence for Tui Na alone or Tui Na plus
Tui Na alone or Tui Na plus cervical computer traction in the cervical computer traction according to GRADE approach.
treatment of cervical radiculopathy on the basis of GRADE As a result, there is a need for well-designed randomised
approach. controlled studies with sufficient power so as to confirm the
Evidence-Based Complementary and Alternative Medicine 9

effectiveness of Tui Na to improve the clinical manifestation, [6] S. Hogg-Johnson, G. Van Der Velde, L. J. Carroll et al.,
including the pain symptom, cervical function, and quality of “The burden and determinants of neck pain in the general
life. population: results of the bone and joint decade 2000-2010 task
force on neck pain and its associated disorders,” Spine, vol. 33,
no. 4, pp. S39–S51, 2008.
Abbreviations [7] L. Zhu, X. Wei, and S. Wang, “Does cervical spine manipulation
reduce pain in people with degenerative cervical radiculopathy?
CAM: Complementary and alternative medicine
a systematic review of the evidence, and a meta-analysis,”
NASS: North American Spine Society Clinical Rehabilitation, vol. 30, no. 2, pp. 145–155, 2016.
VAS: Visual analogue scale
[8] B. I. Woods and A. S. Hilibrand, “Cervical radiculopathy:
NRS: Numerical rating scales epidemiology, etiology, diagnosis, and treatment,” Journal of
MGPQ: McGill pain questionnaire Spinal Disorders and Techniques, vol. 28, no. 5, pp. E251–E259,
NDI: Neck disability index 2015.
CNKI: Chinese National Knowledge [9] I. Nikolaidis, I. P. Fouyas, P. A. Sandercock, and P. F. Statham,
Infrastructure “Surgery for cervical radiculopathy or myelopathy,” Cochrane
VIP: Chinese Scientific Journals Database Database of Systematic Reviews, no. 1, Article ID CD001466,
CBM: Chinese Biomedical Literature Database 2010.
MD: Mean difference [10] A. D. Furlan, F. Yazdi, A. Tsertsvadze et al., “Complemen-
SMD: Standardized mean difference tary and alternative therapies for back pain II,” Evidence
CI: Confidence intervals Report/Technology Assessment, no. 194, pp. 1–764, 2010.
GRADE: Grading of recommendations assessment, [11] E. Salt, C. Wright, S. Kelly, and A. Dean, “A systematic
development, and evaluation literature review on the effectiveness of non-invasive therapy for
SATCM: State Administration of Traditional cervicobrachial pain,” Manual Therapy, vol. 16, no. 1, pp. 53–65,
Chinese Medicine 2011.
NPTMC: National Projected Teaching Materials of [12] R. Boyles, P. Toy, J. Mellon Jr., M. Hayes, and B. Hammer,
China “Effectiveness of manual physical therapy in the treatment of
CMA: Chinese Medical Association. cervical radiculopathy: a systematic review,” Journal of Manual
and Manipulative Therapy, vol. 19, no. 3, pp. 135–142, 2011.
[13] R. J. Rodine and H. Vernon, “Cervical radiculopathy: a system-
Competing Interests atic review on treatment by spinal manipulation and measure-
ment with the Neck Disability Index,” Journal of the Canadian
The authors have no conflict of interests to disclose. Chiropractic Association, vol. 56, no. 1, pp. 18–28, 2012.
[14] E. J. Thoomes, W. Scholten-Peeters, B. Koes, D. Falla, and
A. P. Verhagen, “The effectiveness of conservative treatment
Acknowledgments for patients with cervical radiculopathy: a systematic review,”
The study was financially supported by the National Science Clinical Journal of Pain, vol. 29, no. 12, pp. 1073–1086, 2013.
and Technology Program of China (no. 2006BAI04A09 and [15] L. Manchikanti, F. J. E. Falco, S. Diwan, J. A. Hirsch, and H.
no. 2014BAI08B06), the Science and Technology Program of S. Smith, “Cervical radicular pain: the role of interlaminar and
transforaminal epidural injections,” Current Pain and Headache
China Academy of Chinese Medical Sciences (no. YS1304),
Reports, vol. 18, no. 1, article 389, 2014.
and the Key Program of China Academy of Chinese Medical
[16] S. J. Snodgrass, J. A. Cleland, R. Haskins, and D. A. Rivett, “The
Sciences (no. ZZ0808018). The funding agency was not clinical utility of cervical range of motion in diagnosis, prog-
involved in data collection, data analysis, data interpretation, nosis, and evaluating the effects of manipulation: a systematic
or manuscript development. review,” Physiotherapy, vol. 100, no. 4, pp. 290–304, 2014.
[17] L. Zang, N. Fan, Y. Hai et al., “Using the modified Delphi
References method to establish a new Chinese clinical consensus of the
treatments for cervical radiculopathy,” European Spine Journal,
[1] K. Radhakrishnan, W. J. Litchy, W. M. O’Fallon, and L. T. Kur- vol. 24, no. 6, pp. 1116–1126, 2015.
land, “Epidemiology of cervical radiculopathy: a population- [18] X. Wei, S. Wang, J. Li et al., “Complementary and alternative
based study from Rochester, Minnesota, 1976 through 1990,” medicine for the management of cervical radiculopathy: an
Brain, vol. 117, no. 2, pp. 325–335, 1994. overview of systematic reviews,” Evidence-Based Complemen-
[2] G. Evans, “Identifying and Treating the Causes of Neck Pain,” tary and Alternative Medicine, vol. 2015, Article ID 793649, 10
Medical Clinics of North America, vol. 98, no. 3, pp. 645–661, pages, 2015.
2014. [19] S.-M. Gao, H. Guo, X.-F. Zhao, L.-L. Huang, L.-L. Zhu, and
Z.-P. Wang, “Clinical study on curing the cervical spondylotic
[3] “Cervical radiculopathy: a common cause of neck pain,” Amer-
radiculopathy with the methods of Blade needle closed loosing
ican Family Physician, vol. 93, no. 9, 2016.
and reduction with traction at the same time,” Zhongguo gu
[4] D. L. Corey and D. Comeau, “Cervical radiculopathy,” Medical shang, vol. 25, no. 1, pp. 14–17, 2012.
Clinics of North America, vol. 98, no. 4, pp. 791–799, 2014. [20] X. Cui, K. Trinh, and Y.-J. Wang, “Chinese herbal medicine for
[5] J. M. Rhee, T. Yoon, and K. D. Riew, “Cervical radiculopathy,” chronic neck pain due to cervical degenerative disc disease,”
Journal of the American Academy of Orthopaedic Surgeons, vol. Cochrane Database of Systematic Reviews (Online), no. 1, Article
15, no. 8, pp. 486–494, 2007. ID CD006556, 2010.
10 Evidence-Based Complementary and Alternative Medicine

[21] L. J. Kong, H. S. Zhan, Y. W. Cheng, W. A. Yuan, B. Chen, and M. [38] H. Huang, “Chiropractic therapy combined with traction treat-
Fang, “Massage therapy for neck and shoulder pain: a systematic ment for cervical radiculopathy,” Hu Nan Zhong Yi Za Zhi, vol.
review and meta-analysis,” Evidence-Based Complementary and 29, no. 5, pp. 68–69, 2013.
Alternative Medicine, vol. 2013, Article ID 613279, 10 pages, 2013. [39] Z. G. Liu, “Scalene muscle-focused massage for the treatment
[22] Y. H. Cheng and G. C. Huang, “Efficacy of massage therapy on of cervical radiculopathy,” Zhongguo Zhong Yi Gu Shang Ke Za
pain and dysfunction in patients with neck pain: a systematic Zhi, vol. 22, no. 3, pp. 65–66, 2014.
review and meta-analysis,” Evidence-Based Complementary and [40] J. P. T. Higgins and S. Green, “Cochrane handbookfor systematic
Alternative Medicine, vol. 2014, Article ID 204360, 13 pages, reviews of interventions, version 5.1.0[updated March 2011],”
2014. The Cochrane Collaboration, 2009, http://www.equator-network
[23] G. C. Goats, “Massage—the scientific basis of an ancient art: .org/reporting-guidelines/cochrane-handbook-for-systematic-
Part 1. The techniques,” British Journal of Sports Medicine, vol. reviews-of-interventions-version-5-1-0/.
28, no. 3, pp. 149–152, 1994. [41] G. H. Guyatt, A. D. Oxman, G. E. Vist et al., “GRADE: an
[24] K. C. Patel, A. Gross, N. Graham et al., “Massage for mechanical emerging consensus on rating quality of evidence and strength
neck disorders,” Cochrane Database of Systematic Reviews, vol. of recommendations,” BMJ, vol. 336, no. 7650, pp. 924–926,
9, Article ID CD004871, 2012. 2008.
[25] S. Kumar, K. Beaton, and T. Hughes, “The effectiveness of [42] A. D. Furlan, V. Pennick, C. Bombardier, and M. Van Tulder,
massage therapy for the treatment of nonspecific low back pain: “2009 Updated method guidelines for systematic reviews in the
a systematic review of systematic reviews,” International Journal cochrane back review group,” Spine, vol. 34, no. 18, pp. 1929–
of General Medicine, vol. 6, pp. 733–741, 2013. 1941, 2009.
[26] L. Brosseau, G. A. Wells, P. Tugwell et al., “Ottawa panel [43] A. Gross, J. Miller, J. D’Sylva et al., “Manipulation or mobilisa-
evidence-based clinical practice guidelines on therapeutic mas- tion for neck pain,” Cochrane database of systematic reviews, no.
sage for neck pain,” Journal of Bodywork and Movement Thera- 1, Article ID CD004249, 2010.
pies, vol. 16, no. 3, pp. 300–325, 2012. [44] D. Atkins, D. Best, P. A. Briss et al., “Grading quality of evidence
[27] M. Topolska, S. Chrzan, R. Sapuła, M. Kowerski, M. Soboń, and and strength of recommendations,” BMJ, vol. 328, no. 7454,
K. Marczewski, “Evaluation of the effectiveness of therapeutic Article ID 1490, 2004.
massage in patients with neck pain,” Ortopedia Traumatologia [45] G. H. Guyatt, A. D. Oxman, S. Sultan et al., “GRADE guidelines:
Rehabilitacja, vol. 14, no. 2, pp. 115–123, 2012. 9. Rating up the quality of evidence,” Journal of Clinical
[28] K. J. Sherman, A. J. Cook, R. D. Wellman et al., “Five-week Epidemiology, vol. 64, no. 12, pp. 1311–1316, 2011.
outcomes from a dosing trial of therapeutic massage for chronic [46] H. Balshem, M. Helfand, H. J. Schünemann et al., “GRADE
neck pain,” Annals of Family Medicine, vol. 12, no. 2, pp. 112–120, guidelines: 3. Rating the quality of evidence,” Journal of Clinical
2014. Epidemiology, vol. 64, no. 4, pp. 401–406, 2011.
[29] A. J. Cook, R. D. Wellman, D. C. Cherkin, J. R. Kahn, and [47] The GRADE Working Group, “Introduction: criteria for apply-
K. J. Sherman, “Randomized clinical trial assessing whether ing or using GRADE,” The GRADE System, 2010, http://www
additional massage treatments for chronic neck pain improve .gradeworkinggroup.org/.
12- and 26-week outcomes,” Spine Journal, vol. 15, no. 10, pp. [48] M. A. Childress and B. A. Becker, “Nonoperative management
2206–2215, 2015. of cervical radiculopathy,” American Family Physician, vol. 93,
[30] G. C. Goats, “Massage—the scientific basis of an ancient art: no. 9, pp. 746–754, 2016.
part 2. Physiological and therapeutic effects,” British Journal of [49] H. M. Song, X. Z. Yang, S. Q. Chen, S. Z. Wang, and T. X. Wu,
Sports Medicine, vol. 28, no. 3, pp. 153–156, 1994. “Effect evaluating system of manipulative therapy for cervical
[31] C. A. Moyer, J. Rounds, and J. W. Hannum, “A meta-analysis of spondylotic radiculopathy,” Xian Dai Zhong Xi Yi Jie He Za Zhi,
massage therapy research,” Psychological Bulletin, vol. 130, no. 1, vol. 17, no. 34, pp. 5251–5253, 2008.
pp. 3–18, 2004. [50] J. Yang, R. C. Zhang, and X. J. Wang, “Meta-analysis on nerve-
[32] R.-M. Avery, “Massage therapy for cervical degenerative disc root-type cervical spondylosis treatment by manipulation or
disease: alleviating a pain in the neck?” International Journal of massage and cervical traction,” Huan Qiu Zhong Yi Yao, vol. 6,
Therapeutic Massage & Bodywork, vol. 5, no. 3, pp. 41–46, 2012. no. 9, pp. 641–648, 2013.
[33] M. Yang, Y. Feng, H. Pei et al., “Effectiveness of Chinese massage [51] D. C. Bervoets, P. A. J. Luijsterburg, J. J. N. Alessie, M. J. Buijs,
therapy (Tui Na) for chronic low back pain: Study Protocol For and A. P. Verhagen, “Massage therapy has short-term benefits
A Randomized Controlled Trial,” Trials, vol. 15, article 418, 2014. for people with common musculoskeletal disorders compared
[34] PLoS Medicine Editors, “Best practice in systematic reviews: the to no treatment: a systematic review,” Journal of Physiotherapy,
importance of protocols and registration,” PLoS Medicine, vol. 8, vol. 61, no. 3, pp. 106–116, 2015.
no. 2, Article ID e1001009, 2011. [52] L. Fang and M. Fang, “Research progress on the standardization
[35] J. Zhang, Q. Lin, and J. Yuan, “Therapeutic efficacy observation of Chinese Tuina therapy: a short review,” Chinese Journal of
on tuina therapy for cervical spondylotic radiculopathy in ado- Integrative Medicine, vol. 19, no. 1, pp. 68–72, 2013.
lescence: a randomized controlled trial,” Journal of Acupuncture [53] X. Yang, H. Zhao, and J. Wang, “Chinese massage (Tuina) for
and Tuina Science, vol. 9, pp. 249–252, 2011. the treatment of essential hypertension: a systematic review and
[36] H. L. Jiang, “60 cases Clinical Observation of massage combined meta-analysis,” Complementary Therapies in Medicine, vol. 22,
with cervical traction for cervical radiculopathy,” Hang Kong no. 3, pp. 541–548, 2014.
Hang Tian Yi Xue Za Zhi, vol. 24, no. 3, pp. 358–359, 2013.
[37] Z. X. Mi and J. W. Bi, “Clinical observation of massage for
cervical spondylotic radiculopathy,” Xi Bu Zhong Yi Yao, vol. 26,
no. 7, pp. 101–102, 2013.