JACM

THE JOURNAL OF ALTERNATIVE AND COMPLEMENTARY MEDICINE

Volume 00, Number 00, 2018, pp. 1–10
ª Mary Ann Liebert, Inc.
DOI: 10.1089/acm.2018.0228

Effectiveness and Safety of Polydioxanone

Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

Thread-Embedding Acupuncture as an Adjunctive

Therapy for Patients with Chronic Nonspecific Neck Pain:
A Randomized Controlled Trial
Eunseok Kim, KMD, PhD,1,* Yong-Suk Kim, KMD, PhD,2,* Young Il Kim, KMD, PhD,1
Ju-Hyun Jeon, KMD, PhD,1 Ho-Ryong Yoo, KMD, PhD,3 Yang-Chun Park, KMD, PhD,4
and In Chul Jung, KMD, PhD5

Abstract
Objective: To evaluate the effectiveness and safety of treatment with thread-embedding acupuncture (TEA)
using polydioxanone in addition to usual care for patients with chronic nonspecific neck pain (CNP) compared
with treatment with usual care alone.
Methods: A single-center, assessor-blinded, two-armed randomized controlled trial was performed. One
hundred and six outpatients with CNP were randomly allocated into the TEA plus usual care (TU) group or the
usual care (UC) group in a 1:1 ratio. TEA treatments in the neck region were provided once a week for 4 weeks,
and usual care, as needed, was allowed. The primary outcome was the mean Neck Pain and Disability Scale
(NPDS) score. Secondary outcomes included clinical relevance measured by using the clinically important
difference (CID), pressure pain threshold (PPT), Hospital Anxiety and Depression Scale (HADS), EuroQol-5
Dimension (EQ-5D), and patient global impression of change (PGIC). Participants were assessed at baseline
and at weeks 3, 5, and 9. Statistical analyses included analysis of covariance with baseline score as a covariate.
Results: The TU group showed significant improvement in NPDS scores compared with the UC group
(adjusted group difference, week 5: 13.74 [95% confidence interval: 7.57–19.90]; p < 0.0001 and week 9: 17.46
[11.15–23.76]; p < 0.0001). The proportion of patients with a decrease on the NPDS score of ‡11.5 points
(minimal CID) was significantly higher in the TU group at weeks 5 and 9 than in the UC group. At weeks 5 and
9, significant differences were observed on the anxiety/depression subscale of HADS, EQ-5D, and PGIC
between the two groups, but not the PPTs at three sites. Temporary stiffness was observed after TEA treatment,
but no serious adverse events occurred.
Conclusions: The results suggest that polydioxanone TEA is a safe and clinically beneficial adjunctive
treatment for patients with CNP.

Keywords: chronic nonspecific neck pain, polydioxanone, thread-embedding acupuncture, adjunctive therapy, randomized
controlled trial

1
Department of Acupuncture and Moxibustion Medicine, College of Korean Medicine, Daejeon University, Daejeon,
Republic of Korea.
2
Department of Acupuncture and Moxibustion Medicine, Kyung Hee University Korean Medicine Hospital, Seoul,
Republic of Korea.
Departments of 3Neurologic Disorder, 4Internal Medicine, and 5Neuropsychiatry, Dunsan Korean Medicine Hospital,
Daejeon University, Daejeon, Republic of Korea.
*These authors contributed equally to this work.

1
 2 KIM ET AL.

Introduction with usual care alone. In addition, adverse effects of TEA

were monitored for safety assessment.

C hronic nonspecific neck pain (CNP) is a common

musculoskeletal disorder and public health problem.
Most people experience some degree of neck pain in their
Materials and Methods

lifetime and tend to report the recurrence of such pain 1–5 Study design and participants
years after the last occurrence.1,2 Also, CNP is often persis- We performed a two-armed, assessor-blinded randomized
tent, the duration of which is characterized by periods of controlled trial (RCT) in accordance with the Helsinki De-
improvement and aggravation and is less likely to be resolved claration and the International Council for Harmonization-
completely.3 In addition to conventional treatments, com- Good Clinical Practice guidelines. The protocol was approved
plementary and alternative medicine (CAM) therapies, es-
Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

by the institutional review board of the university hospital

pecially acupuncture, are being applied in the treatment of (DJDSKH-16-BM-15) and has been registered with the Clin-
patients with CNP.4,5 Various acupoint stimulation methods, ical Research Information Service (identifier: KCT0002320).
such as needle stimulation, injection, and the embedding of This study was conducted between December 2016 and
specific materials, are selectively used. November 2017 at the Dunsan Korean Medicine Hospital,
Thread-embedding acupuncture (TEA), which inserts med- Daejeon University, Daejeon, Korea. The participants were re-
ical threads at acupoints, is a special type of acupuncture. Ab- cruited through websites, bulletin boards, and brochures. Parti-
sorbable suture materials, such as catgut in China and cipants who voluntarily signed the written consent form were
polydioxanone (PDO) in Korea, are commonly used as em- screened based on eligibility criteria and completed the treat-
bedded medical threads and are considered prolonged stimu- ment expectancy questionnaire.11 At visit 1, eligible participants
lators once they have been administered at the targeted points.6,7 were randomly assigned to the TEA plus usual care (TU) group
PDO TEA has been consistently applied for the treatment of or the usual care (UC) group. TEA treatment was provided once
patients with musculoskeletal pain as adjunctive therapy in a week from week 1 (visit 1; baseline) to week 4, and assess-
general medical practice in Korea. Nevertheless, despite several ments were performed at baseline and at weeks 3, 5, and 9. The
case reports,8–10 the current level of evidence supporting the inclusion and exclusion criteria are presented in Table 1.
contribution of PDO TEA to the treatment of patients with
musculoskeletal pain is insufficient. Furthermore, within our
Randomization and blinding
knowledge, no previous well-designed clinical trial on the effect
of PDO TEA treatment on patients with CNP has been reported. Eligible participants were randomly allocated in a 1:1 ratio
The aim of this study was to evaluate whether 4-week by using the stratified block randomization method. Stratifi-
PDO TEA treatment, in addition to usual care, can improve cation (male vs. female) was based on previous RCTs that
neck pain and disability among patients with CNP compared showed female gender was more responsive to acupuncture

Table 1. Inclusion/Exclusion Criteria

Inclusion criteria Exclusion criteria
Age between 19 and 65 years. Radicular pain in the upper limb with the distribution
Presence of neck paina for more than 3 months. of a specific nerve root.
Pain intensity on a VAS of more than 40 mm Neurological abnormality such as weakness, paresthesia
at screening. (sensory deficits), special test (Spurling’s test, Bakody’s
Diagnosis of nonspecific neck pain by clinicians based sign), muscle atrophy, and pathological reflex.
on medical history, physical examination, and cervical Major spinal pathology such as myelopathy, spondylitis,
spine radiography. neoplasm, and fracture.
Voluntary agreement via written informed consent with History of spinal surgery or scheduled procedures during
a full understanding of the trial. the trial.
Complaints of musculoskeletal pain more severe than neck
pain.
Uncontrolled low-back pain.
History of receiving TEA for treatment of neck pain in the past
6 months.
Hypersensitive reaction to previous acupuncture treatment,
metal allergy, and other exaggerated responses of the skin.
Hemorrhagic illness and/or factors that can affect hemostasis,
such as intake of anticoagulant or antiplatelet.
Pregnancy or having plan for pregnancy during the trial.
Uncontrolled diabetes or cardiovascular disease.
Major psychiatric disability or neurosis.
Alcohol or drug abuse.
Other unsuitability for the study as determined by the
researcher.
a
Neck pain is defined as pain and/or stiffness in the area between the occiput and the third thoracic vertebra and between the medial
borders of the scapulae, which can be aggravated with movement or applied pressure.12,13
TEA, thread-embedding acupuncture; VAS, visual analogue scale.
 THREAD-EMBEDDING ACUPUNCTURE FOR PATIENTS WITH CHRONIC NECK PAIN 3

Table 2. Acupuncture Treatment Details Based on Standards for Reporting Interventions in Clinical
Trials of Acupuncture 2010 Checklist
Item Detail
1. Acupuncture rationale 1a. Polydioxanone TEA
1b. The common acupoints were selected on the basis of meridian theories, the textbook,6
related studies,16–20 and consensus by the acupuncture experts; individualized
acupoints were based on the individual’s symptoms.
1c. Common acupoints plus individualized acupoints.
Two different kinds of TEA needles were selectively applied based on both the
patient’s physique and the general insertion depth for each acupoint.21
Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

2. Details of needling 2a. Eight to 10 needles were inserted on the affected side of neck per session.
2b. Local acupoints on either one or both sides of the neck, depending on the affected site,
were used.
Six common acupoints: GB21, SI14, GB20, BL10, TE16, and LI17.
Two to four individualized acupoints: two paravertebral points (0.5 body-cun lateral to
the spinous process) between the second cervical vertebra and the third thoracic
vertebra16 and a maximum of two Ashi-points (0–2) per side.
2c. The depth ranged from 25 to 38 mm.
Fascia depth by oblique insertion: GB21 (oblique medial insertion), SI14 (oblique
inferior insertion in the direction of the superior part of the medial border of the
scapula), TE16 (oblique superior insertion across the sternocleidomastoid muscle),
Li17 (oblique inferior insertion along the scalene muscle), and other Ashi-points.
Deep layer of muscle by perpendicular insertion: GB20, BL10, paravertebral point, and
other Ashi-points.
2d–2f. Not applicable. Simple insertion technique without manipulation/retention was applied;
a needle with an attached PDO thread was inserted to the depth at which the thread was
fully inserted, after which the needle was immediately removed.
2g. Disposable sterile PDO thread-embedding devices (OV World Co., Seoul, Republic of
Korea) with 29-gauge needles and USP size 6-0 PDO thread (Samyang Biopham Co.,
Seongnam-si, Republic of Korea).
Length: (1) 25-mm needle with 30-mm thread (15 mm · 2; folded in half).
(2) 38-mm needle with 54-mm thread (27 mm · 2; folded in half).
3. Treatment regimen 3a. Four treatment sessions.
3b. Once a week for 4 weeks, about 10 min for each session.
4. Other components 4a. Usual care, based on the needs of the patients or the prescriptions of their attending
of treatment physicians, was allowed (not provided by this study). It included analgesics, physical
therapy, spinal manipulation, massage, etc.
4b. An education brochure regarding neck exercise and daily care was provided
[Link] background 5. The TEA treatments were performed by an individual Korean Medicine Doctor. The
practitioner had received 6 years of university education for Traditional Korean
Medicine and had at least 4 years of clinical experience with TEA.
6. Control or comparator 6a. Setting a control group to allow usual care is appropriate for a pragmatic clinical trial
interventions to evaluate the effectiveness of a certain treatment as an adjunctive therapy.
6b. Participants in the UC group did not receive the TEA treatments, but usual care was
allowed as it was in the experimental group.
TEA, thread-embedding acupuncture; UC, usual care.

therapy.14,15 An independent blinded statistician generated a points. Disposable sterile PDO thread-embedding devices
list of random numbers and prevented it from being exposed. (Fig. 1) were applied at local acupoints on either one or both
Each patient’s allocation was undisclosed to the evaluator, sides of the neck, depending on the affected site. The practi-
data manager, and statistician. However, due to the nature of tioner selected individualized acupoints based on the individ-
the study design, the participants and the practitioner could ual’s symptoms, such as pain distribution and cervical motion,
not be blinded. Therefore, the TEA practitioner was excluded and on palpation with controlled finger pressure. Oblique in-
from assessments, and patients were not allowed to disclose sertion was administered to acupoints on the thorax or the
their assigned group to the assessor. lateral aspect of the neck at the appropriate angle and to the
appropriate depth to avoid pneumothorax or major injury to
surrounding blood vessels. A simple insertion technique
Intervention
without any manipulation was applied so that neither needle
A summary of information detailing the TEA treatment is retention nor induction of the de qi sensation occurred. On the
given in Table 2. This protocol adopted a relatively prag- day of treatment, the application of physical pressure on the
matic method that allowed the selective use of two different treated sites was prohibited to prevent the embedded thread
TEA needles for common acupoints and individualized acu- from protruding.
 4 KIM ET AL.

that evaluate neck movement, neck pain intensity, neck pain on

emotion and cognition, and the level of interference with daily
activities.22 Each item consists of a 10-cm visual analogue
scale (VAS) with numeric anchors of 0, 1, 2, 3, 4, and 5, with a
distance of 20 mm between anchors. Item scores range from 0
(normal) to 5 (most severe) in quarter-point increments, giving
a maximum score of 100 points.23 The NPDS used in this study
utilized a credible and validated Korean translation.24 The
NPDS score was considered to be valid if no more than 15% of
the items were missing,25 and the missing NPDS items were
calculated by using mean imputation.26
Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

Secondary outcomes included the NPDS scores at sec-

ondary end points (weeks 3 and 9), clinical relevance,
pressure pain threshold (PPT), Hospital Anxiety and De-
pression Scale (HADS), EuroQol-Five Dimension 3 level
version (EQ-5D-3L), patient global impression of change
(PGIC), and usual care history. Clinical relevance was
measured by using the clinically important difference
(CID27) based on the ratio of the number of patients with
decreases on the NPDS score of ‡11.5 points (minimal CID)
or with percentages ‡30% (moderate CID) and ‡50%
(substantial improvement) relative to baseline to the total
number of participants, respectively.28,29 The PPT was
measured twice on both sides of the neck, and the mean
value (kg/cm2) was used. The three points at which the PPT
was measured were (1) the insertion point of the levator
scapulae muscle, (2) GB21: the upper trapezius, and (3) the
area extending 1.5 cm outward from the sixth cervical ver-
tebra.30 In HADS, the anxiety subscale (HADS-A) and the
depression subscale (HADS-D) were designated as outcome
variables. The validated Korean version of the EQ-5D-3L,
which consists of the EQ-5D index and the EQ-5D VAS, was
used.31 The EQ-5D index score was calculated based on the
South Korean population-based preference weights for EQ-
5D proposed by Lee et al.32 The history and the frequency of
usual care since the last visit were examined. To ensure that
we obtained an accurate record of the usual care history, we
provided each participant with a sample leaflet including the
medication list and physical therapy list with photos during
assessments. The participants were encouraged to bring
receipts, prescriptions, medical statements, and so on, per-
taining to the usual care. In addition, at each study visit, all
adverse events were monitored and recorded on an adverse
event report form.

FIG. 1. Polydioxanone thread-embedding acupuncture.

Color images available online at [Link]/acm Sample size calculation
Due to the absence of previous TEA studies using the NPDS as
Usual care was defined as routine medical care to alle-
an outcome measure, we estimated the mean difference and the
viate neck pain, such as analgesics, physical therapy, spinal
standard deviation (SD) of the NPDS score between the two
manipulation, and massage. Usual care, based on the needs
groups based on a similar previously published study on the use of
of the patients or the prescriptions of their attending phy-
acupuncture to treat patients with CNP. This study used the
sicians, was allowed for both groups during the study.
NPDS to evaluate the effect of adding 15 sessions of acupuncture
Brochures containing suggestions for neck exercise and
treatment to routine primary care. Considering that TEA treat-
daily care were provided to educate the participants. All
ment tends to require fewer sessions than acupuncture treat-
participants were instructed not to undergo any invasive
ment,8–10,33–35 we assumed that the effect of TEA in our study
treatments applied directly to the neck, for example, injec-
would be comparable to that of the acupuncture in the previous
tions, acupuncture, and wet cupping.
study. For detection of the mean difference in NPDS score of 12.3
points (SD = 20) reported in the previous study with a 5% sig-
Outcomes
nificance level (two-tailed) and with 80% power, a minimum of
The primary outcome was the Neck Pain and Disability 42 participants was needed for each group. Considering a dropout
Scale (NPDS) score at week 5. The NPDS consists of 20 items rate of 20%, we recruited a total of 106 participants for this study.
 THREAD-EMBEDDING ACUPUNCTURE FOR PATIENTS WITH CHRONIC NECK PAIN 5

Statistical analyses ANCOVA with the baseline score and the usual care as
All analyses were performed by an independent statistician covariates. The safety set was presented as descriptive sta-
who used SAS, version 9.4 (SAS Institute, Inc., Cary, NC), and tistics and included participants from the TU group who
the level of significance was p < 0.05 (two-tailed). Statistical received at least one TEA treatment.
analyses were implemented using a full analysis set (FAS) and
a per-protocol (PP) set. The FAS was used in main analyses to Results
evaluate the between-group differences, and the PP was used
Recruitment and baseline characteristics
for supplementary analyses. The last-observation-carried-
forward method was used to handle any missing data. Between December 2016 and September 2017, a total of
Baseline assessments for the two groups were performed 151 potential participants were screened for eligibility. Of
Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

using the t-test for continuous variables and either the chi- those, 45 patients were found to be ineligible for the trial,
squared test or Fisher’s exact test for categorical variables. mainly because they did not meet the participation criteria. Of
The differences in outcomes between the two groups were the 106 eligible patients, 90 completed the study while the
analyzed using an analysis of covariance (ANCOVA) with remaining 16 (15%), 8 in each group, dropped out. The
the baseline score as a covariate. A repeated-measures progress of the patients throughout the study is shown in
analysis of variance (ANOVA) was used to assess differ- Figure 2. No significant differences between the groups were
ences between the two groups over time. A post hoc analysis found with regard to demographic characteristics, clinical
compared the NPDS scores between the groups by using features, or outcome variables (Table 3).

FIG. 2. Study flowchart. FAS, full analysis set; PP, per protocol; TEA, thread-embedding acupuncture.
 6 KIM ET AL.

Table 3. Baseline Demographic and Clinical Characteristics

Characteristic TU group (n = 51) UC group (n = 53) pa
Age, mean (SD), years 42.08 (13.10) 43.81 (10.96) 0.4654
Female, n (%) 40 (78.43) 40 (75.47) 0.7203
Body mass index, mean (SD), kg/m2 23.68 (3.23) 23.17 (3.54) 0.4449
College or higher, n (%) 32 (62.75) 36 (67.92) 0.5789
Work status, n (%) 0.2586
Employed/self-employed 30 (58.82) 29 (54.72)
Housewife 5 (9.80) 1 (1.89)
Student 13 (25.49) 17 (32.08)
Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

Others 3 (5.88) 6 (11.32)

Marriage/partner, n (%) 38 (74.51) 42 (79.25) 0.5666
Exercise, mean (SD), min/week 224.90 (181.70) 185.60 (95.87) 0.2562
Current smoking, n (%) 3 (5.88) 4 (7.55) 1.0000
Duration of neck pain, mean (SD), months 71.76 (73.59) 69.17 (62.86) 0.8469
Pain intensity (VAS), mean (SD), mm 64.24 (13.00) 64.40 (15.09) 0.9537
Past experience with neck-pain-specific treatment (multiple response possible), n (%)
Conventional treatment 26 (50.98) 36 (67.92) 0.0783
Korean Medicine treatment 28 (54.90) 31 (58.49) 0.7119
Other 6 (11.76) 8 (15.09) 0.6190
NPDS, mean (SD) 54.44 (15.24) 57.13 (15.56) 0.3747
PPT, mean (SD), kg/cm2
Levator scapulae 4.28 (1.49) 3.81 (1.30) 0.0887
Upper trapezius 3.57 (1.64) 3.15 (1.18) 0.1376
Sixth cervical vertebra 3.54 (1.56) 3.15 (1.26) 0.1719
HADS-A, mean (SD) 6.84 (3.27) 7.42 (2.98) 0.3533
HADS-D, mean (SD) 7.20 (3.42) 7.30 (3.22) 0.8712
EQ-5D index, mean (SD) 0.74 (0.15) 0.73 (0.16) 0.7436
EQ-5D VAS, mean (SD), mm 56.86 (18.88) 57.62 (18.65) 0.8368
a
p Values are from the t-test for continuous variables and from either the chi-squared test or Fisher’s exact test for categorical variables.
EQ-5D, EuroQol-5 Dimension; HADS, Hospital Anxiety and Depression Scale (A, Anxiety subscale; D, Depression subscale); NPDS,
Neck Pain and Disability Scale; PPT, pressure pain threshold; SD, standard deviation; TU, thread-embedding acupuncture plus usual care;
UC, usual care; VAS, visual analogue scale.

Interventions measures ANOVA, the NPDS scores did not satisfy the as-
Patients in the TU group received an average of 16.50 (SD sumption of sphericity in the Mauchly test and showed a
3.80) TEA needlings per session. During the study, the usual significant group · time interaction effect ( p < 0.0001), with
care that patients received was neck-pain-specific prescrip- the between-group difference becoming more pronounced
tions, physical therapy, spinal manipulation, massage, and with increasing follow-up time (Fig. 3). The supplementary
medicated patches. Compared with the participants in the TU analysis of the NPDS with the PP set supports the statistically
group, those in the UC group received more treatment in all significant superiority of the TU group at each evaluation
aspects; moreover, a significant difference in the use of point. In a post hoc analysis, with adjustments for the baseline
medical patches was found between the groups (0.00% vs. score and the usual care treatment, significant differences
11.3%; p = 0.0270). The proportion of patients who did not between the groups were noted at every visit.
receive any usual care treatment during the study was sig- No significant differences between the two groups were
nificantly higher in the TU group than in the UC group (86.3% found for the PPT at the three sites at which it was measured.
vs. 66.0%; p = 0.0158). In both HADS-A and HADS-D, the TU group showed sig-
nificant improvement compared with the UC group at weeks
3, 5, and 9. Significant group differences favoring the TU
Outcomes
group were found in the EQ-5D index and the EQ-5D VAS
The patients in the TU group reported significantly lower scores at weeks 5 and 9 (Table 4). Significant group · time
NPDS scores at weeks 3, 5, and 9 (Table 4). The observed interaction effects were found for the following outcome:
mean NPDS scores showed a 5-week reduction from baseline HADS-A ( p = 0.0047), HADS-D ( p < 0.0001), EQ-5D index
of 15.44 points for the TU group compared with 2.18 points ( p = 0.0001), and EQ-5D VAS ( p = 0.0390).
for the UC group. The adjusted mean difference in the NPDS The proportions of patients with decreased NPDS more
scores between the two groups at week 5 was 13.74 (95% than minimal CID/moderate CID/substantial improvement
confidence interval [CI]: 7.57–19.90), and the TU group was were significantly higher in the TU group than in the UC
confirmed to have been significantly improved compared group at weeks 5 and 9 (Fig. 4). In PGIC, a statistically
with the UC group ( p < 0.0001). At week 9, patients in the significant larger number of patients in the TU group than in
TU group reported significantly lower NPDS scores than the UC group reported their impressions of change as
those in the UC group did, the adjusted mean difference being ‘‘minimally improved’’ or ‘‘much improved’’ or ‘‘very
17.46 (95% CI: 11.15–23.76; p < 0.0001). In the repeated- much improved’’ at weeks 5 and 9.
 Table 4. Observed Outcomes and Adjusted Group Differences
Adjusted mean
Variable Weeka TU group (n = 51) UC group (n = 53) differenceb (95% CI) pb
NPDS Baseline 54.44 (15.24) 57.13 (15.55)
3 49.04 (18.47) 56.17 (16.31) 5.08 (0.03–10.13) 0.0487
5 39.00 (21.58) 54.95 (18.73) 13.74 (7.57–19.90) <0.0001
9 35.52 (20.67) 54.95 (18.60) 17.46 (11.15–23.76) <0.0001
PPT
Levator scapulae Baseline 4.28 (1.49) 3.81 (1.29)
3 4.09 (1.56) 3.70 (1.40) -0.02 (-0.40 to 0.37) 0.9338
Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

5 4.09 (1.63) 3.46 (1.49) -0.24 (-0.65 to 0.17) 0.2516

9 4.05 (1.72) 3.47 (1.54) -0.20 (-0.67 to 0.26) 0.3890
Upper trapezius Baseline 3.57 (1.64) 3.15 (1.18)
3 3.44 (1.62) 3.13 (1.34) 0.00 (-0.41 to 0.42) 0.9940
5 3.51 (1.67) 2.91 (1.46) -0.28 (-0.73 to 0.17) 0.2180
9 3.41 (1.87) 2.79 (1.18) -0.31 (-0.77 to 0.16) 0.1916
Sixth cervical vertebra Baseline 3.54 (1.56) 3.15 (1.26)
3 3.33 (1.31) 3.12 (1.35) 0.06 (-0.30 to 0.42) 0.7415
5 3.43 (1.44) 2.92 (1.50) -0.24 (-0.66 to 0.18) 0.2535
9 3.40 (1.49) 2.94 (1.40) -0.20 (-0.62 to 0.22) 0.3478
HADS-A Baseline 6.84 (3.27) 7.42 (2.98)
3 6.39 (3.14) 7.72 (3.69) 0.82 (0.02–1.62) 0.0442
5 5.86 (3.02) 7.47 (3.99) 1.15 (0.17–2.12) 0.0219
9 5.25 (2.92) 7.47 (3.78) 1.81 (0.81–2.80) 0.0005
HADS-D Baseline 7.20 (3.42) 7.30 (3.22)
3 6.86 (3.31) 7.92 (3.63) 0.98 (0.09–1.87) 0.0320
5 6.10 (3.04) 7.58 (3.64) 1.42 (0.42–2.42) 0.0059
9 5.29 (2.92) 7.77 (3.57) 2.41 (1.44–3.38) <0.0001
EQ-5D index Baseline 0.74 (0.15) 0.73 (0.16)
3 0.76 (0.15) 0.73 (0.16) -0.03 (-0.08 to 0.02) 0.3049
5 0.81 (0.15) 0.71 (0.16) -0.09 (-0.14 to -0.04) 0.0005
9 0.83 (0.15) 0.71 (0.16) -0.12 (-0.17 to -0.07) <0.0001
EQ-5D VAS Baseline 56.86 (18.88) 57.62 (18.65)
3 57.94 (18.68) 56.08 (18.61) -2.17 (-8.83 to 4.49) 0.5190
5 65.86 (17.72) 57.40 (19.85) -8.82 (-15.34 to -2.30) 0.0085
9 66.10 (18.66) 57.62 (18.72) -8.74 (-15.59 to -1.88) 0.0130
Observed outcomes are presented as mean (SD).
a
Treatment period was from baseline to week 4.
b
Adjusted mean difference between the groups, and p-value were calculated from analysis of covariance with the baseline score as a
covariate.
CI, confidence interval; EQ-5D, EuroQol-5 Dimension; HADS, Hospital Anxiety and Depression Scale (A, Anxiety subscale; D,
Depression subscale); NPDS, Neck Pain and Disability Scale; PPT, pressure pain threshold; SD, standard deviation; TU, thread-embedding
acupuncture plus usual care; UC, usual care; VAS, visual analogue scale.

FIG. 3. Change over time in the Neck Pain

and Disability Scale. Data are unadjusted
means – SEM. The treatment period was
from baseline to week 4. {The p-value in-
dicates a significant group · time interaction
effect as determined from analyses using a
repeated-measures analysis of variance.
SEM, standard error of the mean; TU,
thread-embedding acupuncture plus usual
care; UC, usual care.

7
 8 KIM ET AL.
Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

FIG. 4. Clinical relevance. Ratios of the number of patients with decreases on the NPDS score of ‡11.5 points (minimal
CID) or with percentages ‡30% (moderate CID) and ‡50% (substantial improvement) relative to baseline to the total number
of participants, respectively, are presented. *Significant difference ( p < 0.05, chi-squared test). CID, clinically important
difference; NPDS, Neck Pain and Disability Scale; TU, thread-embedding acupuncture plus usual care; UC, usual care.
Safety data
Among 191 sessions of TEA treatment, 17 adverse
events, which were evaluated as being mild, were reported
by 12 patients (22.6%) in the TU group. Stiffness in the neck
Table 5. Summary of Intervention-Related region was the most common discomfort, whereas others
Adverse Events were bruising, irritation caused by the thread, skin flares,
and pruritus (Table 5). Other adverse events not related to
TU group
the intervention were coughing, other musculoskeletal pain,
Variable (n = 51)
dyspepsia, headache, insomnia, and so on. None of the re-
Patients who reported adverse events 12 (22.6) ported adverse events were considered to be serious.
associated with the TEA treatment, n (%)
Total sessions of TEA treatment, n 191
Discussion
Adverse eventa
Stiffness, n (mean duration in days) 10 (8.2) Adding 4 weeks of TEA treatment to usual care led to
Bruise, n (mean duration in days) 2 (8.0) significant improvements in NPDS scores compared with
Irritation caused by the thread, n (mean 2 (5.0) usual care alone. The trends of the NPDS score in the two
duration in days) groups over time were significantly different, with the
Skin flare, n (mean duration in days) 1 (4.0)
Pruritus, n (mean duration in days) 1 (4.0) between-group difference becoming more pronounced with
A thread sticking out from the skin, 1 (1.0) increasing follow-up time. Significant improvements in
n (mean duration in days)b psychological distress and quality of life were observed in
the TU group compared with the UC group, but no signifi-
a
Number of cases reported as adverse events (multiple answers cant differences were found in the PPT at the three points.
possible). Some discomfort after TEA treatment was reported, but, in
b
Occurred after the patient had received massage with strong
pressure the day before the third treatment. all cases, it was mild and temporary.
TEA, thread-embedding acupuncture; TU, thread-embedding When the stubborn progress of CNP is considered, the
acupuncture plus usual care. observation that the between-group difference in NPDS
 THREAD-EMBEDDING ACUPUNCTURE FOR PATIENTS WITH CHRONIC NECK PAIN 9

scores at week 9 was greater than it was at week 5 is treatment. Thus, more studies investigating the physiologic
noteworthy. Within the TU group, all variables, except reaction caused by the embedded PDO thread, the optimal
PPTs, showed greater improvements at week 9 than at week modality of treatment, and the efficacy of TEA by using a
5. Concluding that this was due to the effect of the PDO placebo control (e.g., inserting needles without PDO thread)
threads in the acupoints would be premature at this time; are needed. Despite such methodological weaknesses, the
however, a notable observation was that the therapeutic effect study design was relatively pragmatic so that it reflected the
increased over time after termination of treatment rather than daily medical practice in Korea, where conventional medi-
simply being maintained. Moreover, even though the propor- cine and Korean Medicine, part of CAM, coexist.
tion of patients who received usual care was significantly
higher in the UC group than in the TU group, the outcomes Conclusions
observed in the TU group were consistently superior to those TEA in addition to usual care showed clinically relevant
Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

observed in the UC group, regardless of whether the effect of benefits in treating patients with CNP. Although some patients
usual care was statistically adjusted for or not. As for NPDS experienced temporary post-treatment discomfort, TEA can be
scores, improvements in the TU group, as well as the group considered as a valuable option for treating such patients.
differences before and after the treatment, were greater than
minimal CID. Therefore, TEA treatment in addition to usual Acknowledgments
care is not only statistically significant but also clinically This study was supported by a grant from the Traditional
meaningful in the treatment of patients with CNP. Korean Medicine R&D Project, Ministry of Health and
In the TEA group, the NPDS scores were lower and the Welfare, Republic of Korea (HI15C0006). We especially
HADS-A/HADS-D scores were also significantly reduced thank Sung Hoon Lee for critical advice on the use of TEA
compared with the UC group. This result supports the finding in the treatment of patients with neck pain.
of a previous cross-sectional study that neck pain is signifi-
cantly correlated with depression and anxiety.36 However, the Authors’ Contribution
PPTs at the three points were not significantly different be- E.K. and Y.-S.K. are responsible for the conception, de-
tween the groups after TEA treatment. This may be due to the sign, and drafting of the article. Y.I.K. and J.-H.J. are re-
fact that the mean PPT from both sides was applied without sponsible for the interpretation of the data and critical
considering the affected side. Furthermore, the overall fre- revision of the article. H.-R.Y. is responsible for interpre-
quencies of usual care in both groups were relatively low tation of the data and revision of the article. Y.-C.P. and
compared with those in other pragmatic RCTs.37–40 This may I.C.J. are responsible for the conception, interpretation of
be due to the local patients who participated in this study the data, and final approval of the article to be published.
preferring fewer conventional treatments. Moreover, the
participants in the TU group may have received less usual care Author Disclosure Statement
than those in the UC group because their neck pain gradually
improved over the duration of the study. No competing financial interests exist.
While no study of the add-on effects of TEA for treating
patients with CNP has been published, two RCTs reported References
the clinical effects of acupuncture as an adjunctive therapy 1. Hogg-Johnson S, van der Velde G, Carroll LJ, et al. The
for patients with neck pain and disability.15,41 In those burden and determinants of neck pain in the general pop-
studies, acupuncture treatment was given at acupoints at the ulation: Results of the Bone and Joint Decade 2000–2010
discretion of the practitioners 10–15 times over 3 months, Task Force on Neck Pain and Its Associated Disorders.
whereas TEA treatment in this study was provided 4 times Spine 2008;33(4 Suppl):S39–S51.
in a partially individualized manner. Witt et al. stated that 2. Carroll LJ, Hogg-Johnson S, van der Velde G, et al. Course
improvements at the end of acupuncture treatment lasted up and prognostic factors for neck pain in the general popu-
to 3 months; they also noted that acupuncture potentially lation: Results of the Bone and Joint Decade 2000–2010
had a long-term effect. Despite the fewer treatments in our Task Force on Neck Pain and Its Associated Disorders.
study group, the NPDS scores and changes were similar to Spine 2008;33(4 Suppl):S75–S82.
those in the randomized patients of Witt’s study. However, 3. Cote P, Cassidy JD, Carroll LJ, Kristman V. The annual
because of the relatively short follow-up period of this incidence and course of neck pain in the general population:
study, no conclusion that TEA treatment has a long-term A population-based cohort study. Pain 2004;112:267–273.
effect or is superior to acupuncture treatment can be drawn. 4. Thomas K, Coleman P. Use of complementary or alterna-
tive medicine in a general population in Great Britain.
This study has some limitations. Neither the practitioner
Results from the National Omnibus survey. J Public Health
nor the subjects were blinded to the intervention, and no
(Oxf) 2004;26:152–157.
participants in the UC group were provided with TEA 5. Plastaras CT, Schran S, Kim N, et al. Complementary and
treatments. Therefore, the results of the study could be bi- alternative treatment for neck pain: Chiropractic, acu-
ased in favor of the experimental group. Because this study puncture, TENS, massage, yoga, Tai Chi, and Feldenkrais.
did not deal with a sham TEA device, we could not deter- Phys Med Rehabil Clin N Am 2011;22:521–537, ix.
mine to what extents the dry needling and the embedding 6. Kim YJ, Kim HN, Shin MS, Choi BT. Thread embedding
with PDO thread individually contributed to the effect of the acupuncture inhibits ultraviolet B irradiation-induced skin
TEA treatment. In addition, the 5-week follow-up after TEA photoaging in hairless mice. Evid Based Complement Al-
treatment was insufficient to assess the long-lasting effects ternat Med 2015;2015:539172.
of TEA. We also have some concerns about the temporary 7. Committee KAMSTC. Acupuncture Medicine. Seoul, Re-
discomfort experienced by some participants after the TEA public of Korea: Hanmi Medical Publishing Co., 2016:181.
 10 KIM ET AL.

8. Jang H, Woo C, Ahn H, Kwon H. A clinical four case studies 27. Jaeschke R, Singer J, Guyatt GH. Measurement of health
on chronic low back pain treated by needle embeeding therapy. status. Ascertaining the minimal clinically important dif-
J Korea CHUNA Manual Med Spine Nerves 2014;9:45–55. ference. Control Clin Trials 1989;10:407–415.
9. Lee SM, Ji YS, Jeon JH, et al. Effect of needle-embedding 28. Jorritsma W, Dijkstra PU, de Vries GE, et al. Detecting
& acupuncture therapy on shoulder pain in Behcet disease relevant changes and responsiveness of Neck Pain and
patient: A case report. J Acupunct Res 2013;30:219–224. Disability Scale and Neck Disability Index. Eur Spine J
10. Lee JH, Yang TJ, Lee DG, et al. The effect of needle- 2012;21:2550–2557.
embedding therapy on osteoarthritis of the knee combined 29. Dworkin RH, Turk DC, Wyrwich KW, et al. Interpreting
with Korean Medical Treatment: Report of five cases. J the clinical importance of treatment outcomes in chronic
Acupunct Res 2014;31:195–204. pain clinical trials: IMMPACT recommendations. J Pain
11. Kim YJ, Lee IS, Kim HS, et al. Validation of the Korean 2008;9:105–121.
Downloaded by Wegner Health Science Information Center/ University of South Dakota multisite from [Link] at 12/20/18. For personal use only.

version of the Acupuncture Expectancy Scale. Acupunct 30. Irnich D, Behrens N, Molzen H, et al. Randomised trial of
Med 2014;32:51–55. acupuncture compared with conventional massage and
12. Kuorinka I, Jonsson B, Kilbom A, et al. Standardised ‘‘sham’’ laser acupuncture for treatment of chronic neck
Nordic questionnaires for the analysis of musculoskeletal pain. BMJ 2001;322:1574–1578.
symptoms. Appl Ergon 1987;18:233–237. 31. Kim M-H, Cho Y-S, Uhm W-S, et al. Cross-cultural adaptation
13. Lee S, Nam D, Leem J, et al. Efficacy and safety of and validation of the Korean version of the EQ-5D in patients
Myofascial-meridian Release Acupuncture (MMRA) for with rheumatic diseases. Qual Life Res 2005;14:1401–1406.
chronic neck pain: A study protocol for randomized, pa- 32. Lee YK, Nam HS, Chuang LH, et al. South Korean time
tient- and assessor-blinded, sham controlled trial. BMC trade-off values for EQ-5D health states: Modeling with
Complement Altern Med 2016;16:45. observed values for 101 health states. Value Health 2009;
14. White P, Lewith G, Prescott P, Conway J. Acupuncture 12:1187–1193.
versus placebo for the treatment of chronic mechanical 33. Guo AS, Li AH, Chen X, et al. Clinical effect of obser-
neck pain: A randomized, controlled trial. Ann Intern Med vation of acupoint catgut embedding on fibromyalgia syn-
2004;141:911–919. drome [In Chinese]. Chin J Basic Med 2014;20:380–382.
15. Witt CM, Jena S, Brinkhaus B, et al. Acupuncture for pa- 34. Xiao Q. Clinical efficacy of acupuncture and embedding
tients with chronic neck pain. Pain 2006;125:98–106. therapy for hemiplegic shoulder pain [In Chinese]. Chinese
16. Blossfeldt P. Acupuncture for chronic neck pain—A cohort Med Modern Dist Educ China 2006;4:31–33.
study in an NHS pain clinic. Acupunct Med 2004;22:146–151. 35. Yao J, Chen X, Yao J, Xie H. Efficacy of acupuncture
17. Tobbackx Y, Meeus M, Wauters L, et al. Does acupuncture combined with embedding therapy in treatment of shoulder-
activate endogenous analgesia in chronic whiplash- hand syndrome [In Chinese]. China Pract Med 2008;3:67–68.
associated disorders? A randomized crossover trial. Eur J 36. Blozik E, Laptinskaya D, Herrmann-Lingen C, et al. De-
Pain 2013;17:279–289. pression and anxiety as major determinants of neck pain: A
18. Seo BK, Lee JH, Kim PK, et al. Bee venom acupuncture, cross-sectional study in general practice. BMC Muscu-
NSAIDs or combined treatment for chronic neck pain: loskelet Disord 2009;10:13.
Study protocol for a randomized, assessor-blind trial. Trials 37. Thomas KJ, MacPherson H, Thorpe L, et al. Randomised
2014;15:132. controlled trial of a short course of traditional acupuncture
19. Que Q, Ye X, Su Q, et al. Effectiveness of acupuncture in- compared with usual care for persistent non-specific low
tervention for neck pain caused by cervical spondylosis: Study back pain. BMJ 2006;333:623.
protocol for a randomized controlled trial. Trials 2013;14:186. 38. Scharf HP, Mansmann U, Streitberger K, et al. Acu-
20. Cho JH, Nam DH, Kim KT, Lee JH. Acupuncture with puncture and knee osteoarthritis: A three-armed random-
non-steroidal anti-inflammatory drugs (NSAIDs) versus ized trial. Ann Intern Med 2006;145:12–20.
acupuncture or NSAIDs alone for the treatment of chronic 39. MacPherson H, Tilbrook H, Richmond S, et al. Alexander
neck pain: An assessor-blinded randomised controlled pilot technique lessons or acupuncture sessions for persons with
study. Acupunct Med 2014;32:17–23. chronic neck pain: A randomized trial. Ann Intern Med
21. Chou PC, Huang YC, Hsueh CJ, et al. Retrospective study 2015;163:653–662.
using MRI to measure depths of acupuncture points in neck 40. Cherkin DC, Sherman KJ, Avins AL, et al. A randomized
and shoulder region. BMJ Open 2015;5:e007819. trial comparing acupuncture, simulated acupuncture, and
22. Wheeler AH, Goolkasian P, Baird AC, Darden BV, II. usual care for chronic low back pain. Arch Intern Med
Development of the Neck Pain and Disability Scale. Item 2009;169:858–866.
analysis, face, and criterion-related validity. Spine 1999;24: 41. Salter GC, Roman M, Bland MJ, MacPherson H. Acu-
1290–1294. puncture for chronic neck pain: A pilot for a randomised
23. Goolkasian P, Wheeler AH, Gretz SS. The neck pain and controlled trial. BMC Musculoskelet Disord 2006;7:99.
disability scale: Test-retest reliability and construct valid-
ity. Clin J Pain 2002;18:245–250. Address correspondence to:
24. Lee JS, Suh KT, Kim JI, et al. Validation of the korean In Chul Jung, PhD
version of the neck pain and disability scale. Asian Spine J Department of Neuropsychiatry
2013;7:178–183. Dunsan Korean Medicine Hospital
25. Scherer M, Blozik E, Himmel W, et al. Psychometric Daejeon University
properties of a German version of the neck pain and dis- 75 Daedeok-daero 176 beon-gil, Seo-gu
ability scale. Eur Spine J 2008;17:922–929.
Daejeon 35235
26. Blozik E, Himmel W, Kochen MM, et al. Sensitivity to
Republic of Korea
change of the Neck Pain and Disability Scale. Eur Spine J
2011;20:882–889. E-mail: npjeong@[Link]