J Acupunct Meridian Stud 2017;10(4):261e275

Available online at www.sciencedirect.com

Journal of Acupuncture and Meridian Studies

journal homepage: www.jams-kpi.com

RESEARCH ARTICLE

Quality Assessment of Randomized

Controlled Trials of Moxibustion Using
STandards for Reporting Interventions in
Clinical Trials of Moxibustion (STRICTOM)
and Risk of Bias (ROB)
So Yun Kim 1,2, Eun Jung Lee 1,2, Ju Hyun Jeon 1,2, Jung Ho Kim 1,2,
In Chul Jung 1,2,*, Young Il Kim 1,2,*

1
College of Korean Medicine, Daejeon University, Daejeon, Republic of Korea
2
Clinical Trial Center, Dunsan Korean Medical Hospital of Daejeon University,
Daejeon, Republic of Korea
Available online 27 July 2017

Received: Feb 16, 2017 Abstract

Revised: May 31, 2017 Objectives: To assess the quality and completeness of published reports of randomized
Accepted: May 31, 2017 controlled trials (RCTs) of moxibustion.
Method: We searched six databases to retrieve eligible RCTs of moxibustion published from
KEYWORDS 2000 to December 2015. We used the STandards for Reporting Interventions in Clinical Trials
moxibustion; of Moxibustion (STRICTOM) and Risk of Bias (ROB) tool to assess the completeness of reporting
randomized controlled of RCTs of moxibustion and evaluate the reporting quality of included RCTs.
trials; Results: Thirty-four studies of moxibustion were analyzed using STRICTOM and ROB. Of the 34
Risk of Bias; studies, the completeness percentage of STRICTOM varied from 33% to 100% (mean 68%, me-
STandards for Reporting dian 67%). The completeness of STRICTOM items showed a rising tendency along with the pub-
Interventions in lication year. The STRICTOM items of setting and context (14.7%), rationale for the control
Clinical Trials of (17.6%), and response (26.4%) showed incomplete reporting. The number of RCTs that rated
Moxibustion a low risk of bias for allocation concealment (n Z 6), blinding of participants and personnel
(n Z 1), and blinding of outcome assessment (n Z 4) appeared to be small.

* Corresponding authors. 75 176 bun-gil, Daedeok-daero, Seo-gu, Daejeon, Republic of Korea.

E-mail: npjeong@dju.kr (I.C. Jung), omdkim01@dju.kr (Y.I. Kim).
pISSN 2005-2901 eISSN 2093-8152
http://dx.doi.org/10.1016/j.jams.2017.05.012
ª 2017 Medical Association of Pharmacopuncture Institute, Publishing services by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
262
Conclusion: The quality of reporting of RCTs of moxibustion remains incomplete according to
the STRICTOM and ROB tool at present. Researchers should consider the STRICTOM and ROB for
improving not only the completeness of reporting but also the study design. General guidelines
for RCTs of moxibustion are also required.
1. Introduction
Moxibustion is a therapeutic form of traditional East Asian
medicine making use of moxa (mugwort; Artemisia argyi
Folium) or other materials to heat, burn, and/or fumigate
either acupoints or specific areas on the surface of the body
[1]. Traditional East Asian medicine doctors have developed
moxibustion therapy through thousands of years of clinical
experience. It is thought that the therapeutic effect of
moxibustion is a combination of heat, tar, aroma, and
psychological reaction [2]. Based on yin and yang, merid-
ian, and qi theory, moxibustion can boost yang qi and
remove pathogenic coldness or dampness. Recent studies
have suggested that moxibustion can regulate the immune
system, stimulate the anti-inflammatory system, improve
blood circulation, and release chemicals that can alleviate
pain [3]. There are many types of moxibustion including
different forms, materials, and techniques. It is used for
various diseases such as musculoskeletal, genitourinary,
digestive, respiratory, and neoplasmic diseases [4]. Owing
to the wide range of applications, a great deal of research
has been focused on moxibustion. The number of random-
ized controlled trials (RCTs) on moxibustion has steadily
increased [5], but they appear to have some limitations
such as lack of moxibustion detail, unclear randomization,
inappropriate control groups. There have been some
studies assessing the reporting quality of RCTs, but most of
these did not focus on moxibustion. There is a need for
guidelines for RCTs on moxibustion to improve the quality
and completeness of the study.
The aim of this work is to evaluate the quality of RCTs of
moxibustion using the STandards for Reporting In-
terventions in Clinical Trials of Moxibustion (STRICTOM) [6]
and the Risk of Bias (ROB) tool [7]. STRICTOM is developed
based on the Consolidated Standards of Reporting Trials
(CONSORT), a statement for RCTs, and STandards for
Reporting Interventions in Clinical Trials of Acupuncture
(STRICTA), a widely used guideline for RCTs of acupuncture
[8e10]. We anticipate that this study will provide the cur-
rent information about RCTs of moxibustion including study
setting and protocols, and valuable guidelines to report
RCTs in a more evidence-based way.
2. Materials and methods
2.1. Types of studies
Randomized and quasi-randomized (which means the allo-
cation method is not strictly random, such as alternation,
date of birth, or case record number) controlled trials were
eligible for this analysis. There is no language limitation.
Trials that were published from 2000 to December 2015
S.Y. Kim et al.
were included. The studies had to meet the following in-
clusion criteria as well:
1. Individuals of any age and sex who have any kind of
symptoms or disease could be participants. Studies that
recruited healthy people were excluded.
2. Studies designed to use moxibustion as a treatment
method were included, and moxibustion combined with
any other intervention (e.g., acupuncture, herbal med-
icine, Western medicine, physiotherapy) were also
included. Studies adopting self-moxibustion therapy
(participants had to practice moxibustion by themselves)
were excluded because this is thought to be far from
medical practice. Studies using warming needle moxi-
bustion (moxa is lit on the top of an acupuncture needle)
were excluded as well because the effect of this type of
moxibustion is mixed with that of acupuncture.
3. Studies that had at least one control with any type of
intervention (e.g., acupuncture, Western medicine,
physiotherapy) or observational control were included.
Studies were excluded if moxibustion was used for both
the intervention group and the control group.
4. The outcome measure of each study can vary depending
on the target symptom or disease.
2.2. Databases and search strategy
Studies in Korean were searched in NDSL, KoreanTK, and
Oasis. Studies in English were found in PubMed, Wiley On-
line Library (Cochrane Library searching engine), and
CINAHL. The search was performed based on the results
obtained on December 28, 2015. These databases were
searched using the following strategy:
[moxibustion OR moxa OR artemisia OR mugwort] AND
[randomized OR controlled OR random OR control]
These searching terms were slightly modified depending
on the database.
2.3. Data extraction quality assessment
General characteristics of the included studies were
extracted systematically by one author including author
name, publication year, target disorder, sample sizes of the
intervention and control groups, and type of the control.
The extraction items are designed not to repeat the items
of STRICTOM. To see how each study is designed to show the
effectiveness of moxibustion compared with the control,
outcome measure and study results were extracted. If
there are multiple follow-ups and multiple outcomes, we
extracted the last follow-up result.
STRICTOM, modified from STRICTA, was used to analyze
the completeness of RCTs of moxibustion [6]. If there was
an item that consisted of multiple subitems, we considered
Assessment of RCTs of Moxibustion
the item to be satisfied when at least one subitem was
reported appropriately. For example, Item Number 3 of
STRICTOM checked if there was a mention about the num-
ber, frequency, and duration of the treatment sessions. We
checked the item as complete if there was any information
about the treatment sessions, regardless of the number,
frequency, or duration. We used the explanation and ex-
amples of STRICTOM as an assessment reference. Item 1c,
which is an item to check if there was a statement about
the extent of moxibustion regimen variation, was consid-
ered as “yes” if there was no statement about the variation
itself and there was one treatment regimen alone. The item
was rated as “not reported” if there were some variations
for moxibustion treatment by participants or some groups
that were not reported specifically. For Item 2a, which
verified the material used for moxibustion, we saw the item
as satisfied if the study mentioned the moxa material form,
such as moxa floss, moxa cone, moxa stick, even if there
Figure 1 PRISMA flow diagram.
263
was no statement about the size or the manufacturers. For
Item 2c, an item for the number of moxibustion units and/
or moxibustion time per point, we rated it as “yes” if we
could assume the moxibustion unit was one and there was a
statement about the moxibustion time. If there was infor-
mation about the target responses, such as “we left the
moxibustion until there is a burning pain,” we considered it
as “yes” even though there was no exact time. If there was
no statement about either the moxibustion unit or the
time, we rated it as “no.” For Item 2f, an item to check the
participant’s posture and treatment environment, we rated
it as “yes” if there was information about the posture of the
treatment environment, because researchers have paid less
attention to the treatment environment until recently. For
Item 4a, details of other interventions administered to the
moxibustion group, such as acupuncture, cupping, herbs,
exercises, or lifestyle advice, we rated it as “yes” if the
researcher presented the name and type of other
264 S.Y. Kim et al.

Table 1 General characteristics of included studies (n Z 34)

Author, year Disorder Intervention, N Control, N Outcomes Results
(analyzed)
Son (2011) Chronic fatigue 25 Sham, 20 (1) CBC (1) NS
[11] (2) Biochemical parameters (2) NS
(3) Urinalysis data (3) NS
Lee (2011) Low back pain of 49 Acu þ Herb þ Physio, (1) NRS (1) p Z 0.003
[12] herniated 46 (2) ODI (2) p Z 0.013
intervertebral disc
Doh (2008) Degenerative 10 Acu þ Hotpack, 10 (1) LSS (1) NS
[13] arthritis of knee (2) VAS (2) p < 0.001
Lee (2011) Chronic cancer 7 Sham, 7 (1) BPI total score (1) p Z 0.016
[14] pain (2) BPI pain intensity score (2) p Z 0.023
(3) BPI pain interference (3) p Z 0.037
score (4) NS
(4) FACT-G (5) NS
(5) Opioid use
Ryu (2013) Chronic back pain 24 NT, 26 (1) VAS (1) p < 0.001
[15] (2) ODI (2) NS
Ren (2013) Senile symptoms 33(31) NT, 30(25) (1) Senile symptom score (1) p < 0.01
[16] (2) Serum melatonin (2) p < 0.05
content
Lu (2015) Irritable bowel 43 electroAcu., 42 (38) (1) VAS (1) NS
[17] syndrome (2) Bristol Stool Form Scale (2) D: p < 0.001; C: NS
Diarrhea (D)
Constipation (C)
Xu (2015) Lower back pain SL 10 (8) Acu, 11 (9) (1) VAS (1) SL: p < 0.05
[18] ML 10 (8) (2) RMQ ML: NS
LL 9 (8) LL: p < 0.05
(Heating time (2) NS
length:
Short/
Moderate/
Long length)
Yi (2014) Irritable bowel 40 (15) Sham, 40 (13) (1) Birmingham IBS scale (1) p < 0.01
[19] syndrome with (2) IBS QOL scale (2) p < 0.01
diarrhea (3) Rectal distention (3) (i) NS
threshold (ii) p < 0.001
(i) First sensation (iii) NS
threshold (iv) p < 0.05
(ii) Defecation urge
threshold
(iii) Pain score (50-mL
balloon)
(iv) Pain score (100-mL
balloon)
(4) fMRI with rectal balloon
distention
Bao (2014) Crohn’s disease 46 (40) Sham þ Acu, 46 (42) (1) CDAI (1) p < 0.001
[20] (2) Total treatment efficacy (2) p < 0.001
(3) QOL (3) p Z 0.017
(4) Lab (4) (i) p Z 0.029
(i) HGB (ii) p Z 0.008
(ii) CRP (iii) NS
(iii) ESR (5) NS
(5) Endoscopic rating (6) p Z 0.029
Assessment of RCTs of Moxibustion 265

Table 1 (continued )
Author, year Disorder Intervention, N Control, N Outcomes Results
(analyzed)
(6) Intestinal histology
score
Kim (2014) OA of knee 102 Usual care, 110 (1) K-WOMAC (1) p < 0.01
[21] (2) NRS (2) p < 0.01
(3) Physical performance (3) (i) p Z 0.0006
test (ii) NS
(i) Stand test (iii) NS
(ii) Balance test (4) NS
(iii) Walk test (5) Two domains of
(4) BDI eight p < 0.01,
(5) SF-36 others NS
Zhao (2014) OA of knee 55 (49) Sham, 55 (50) (1) WOMAC (1) (i) p Z 0.002
[22] (i) Pain (ii) NS
(ii) Function
Chen (2013) Chronic persistent 144 (138) Medication, 144 (1) ACT (1) p Z 0.00003
[23] asthma (139) (2) Lung function (2) (i) p Z 0.042
(i) FEV1 (ii) p Z 0.0012
(ii) PEF (iii) p Z 0.047
(iii) Attack frequency
Kim (2013) Chronic fatigue 25 Placebo, 20 (1) NRS by questionnaire (1) p < 0.05
[24] (2) VAS (2) p < 0.01
(3) (i) Total ROS (3) (i) NS
(ii) MDA (ii) p Z 0.021
(iii) Total antioxidant (iii) NS
cap (iv) p Z 0.001
(iv) Total GSH content (v) NS
(v) GSH peroxidase (vi) p Z 0.012
(vi) GSH reductase (vii) NS
(vii) catalase activity (viii) NS
(viii) superoxide
dismutase
Park (2011) Functional 13 (11) Sham, 13 (1) Defecation frequency (1) NS
[25] constipation (2) BSS (2) NS
(3) CAS (3) NS
Zhang Hypomenorrhea 76 Acu only, 76 Therapeutic effect p < 0.05
(2009) [26]
Ju (2009) Insomnia 40 Estazolam, 35 Curative effect NS
[27]
Zhang Xerophthalmia 30 Artificial tear, 31 (1) Breakup time of tear (1) p < 0.05
(2007) [28] film (2) p < 0.01
(2) Secretary amount of
tears
Gao (2007) Chronic atrophic 30 Acu only, 30 (1) Symptom score (1) p < 0.05 (compar-
[29] gastritis Drug, 28 (2) Pathologic effect ing with drug)
(3) Serum gastrin (2) NA
(3) p < 0.05 (compar-
ing with drug)
Jia (2006) Ankylosing Drug þ moxa, 30 Drug only, 30 (1) Therapeutic effect (1) NS comparing
[30] spondylosis Drug þ Acu, 30 (2) Symptoms and signs drug þ acu,
(3) ESR and CRP p < 0.05
comparing drug
only
(2) NA
(3) NS
Zeng (2006) Cervical vertigo 40 Acu only, 38 Therapeutic effect p < 0.05
[31]
(continued on next page)
266 S.Y. Kim et al.

Table 1 (continued )
Author, year Disorder Intervention, N Control, N Outcomes Results
(analyzed)
Liang (2006) Bell’s palsy Basic Basic treatment, 128 (1) Therapeutic effect Basic
[32] treatment þ acu- (2) HouseeBrackmann treatment þ acu-
moxa, 155 grade moxa
acu-mox, 156 (3) FDIP, FDIS
(1) p < 0.05
(2) p < 0.01
(3) p < 0.01
Ding (2005) Multiple aortitis 40 Drug, 40 (1) Therapeutic effect (1) p < 0.01
[33] (2) Clinical symptoms (2) p < 0.05
(3) Therapeutic effect on (3) p < 0.05
signs
Edson Semen 9 (8) Placebo (indifferent Sperm analysis: (1) Significant (not in
(2003) [34] abnormalities points), 10 p value)
(1) % of normal form sperm (2) NS
(2) Volume, concentration,
progressive motility,
number of round cell
Kang (2000) Poststroke voiding 10 NA (Herb, Acu, (1) Removal time (1) NS
[35] dysfunction Physio), 10 (2) Residual urine (2) NS
Wu (2000) Ulcerative colitis 30 SASP, 16 (1) Therapeutic effect (1) p < 0.01
[36] (2) Histologic change (2) NA
Choi (2003) Hemiplegic upper NA (20) Herb, Acu, Physio, (1) Fugl-Myer test (1) p Z 0.038
[37] extremity after NA (20) (2) Motricity index test (2) p Z 0.002
stroke (3) Modified Barthel index (3) NS
Kim (2005) Blood pressure 30 Bed rest, 31 (1) Systolic blood pressure, (1) NS/p < 0.01/
[38] elevation in 30/60/90/120 min p < 0.01/p < 0.01
hypertensive (2) Diastolic blood pressure, (2) NS/NS/NS/
patients 30/60/90/120 min p < 0.05
Lee (2008) Hemiplegia on NA (21) H-med, Acu, Physio, Functional Independence p Z 0.001
[39] stroke NA (21) Measure
Shin (2009) Premenstrual 10(8) Hand Acu, 10 (7) (1) Menstrual symptom (1) p < 0.018
[40] syndrome Obs, 10 (7) severity (comparing Obs)
(2) Change in body temper- (2) p < 0.05 clavicle,
ature (clavicle, chest, chest, abdomen,
lower abdomen, shoul- shoulder
der, back, lumbar) (comparing Obs)/
p < 0.05 abdomen
(comparing Hand
Acu)
Shi (2012) Aging symptoms 31 NT, elderly control, (1) Therapeutic effect (1) p < 0.01(elderly
[41] 30 (2) Bcl-2 expression control)
NT, young control, (3) PKC expression (2) p < 0.01 (both
30 control)
(3) p < 0.01 (both
control)
Bian (2013) Postoperative 50 Routine nursing care, (1) Automatic micturition (1) p < 0.001
[42] dysuria 50 (2) Moxibustion correlation (2) p Z 0.003
with automatic (3) p Z 0.001
micturition
(3) Rate of urethral
catheterization
Cai (2014) Subhealth fatigue 60 (58) Herbal pill (Buzhong (1) Fatigue assessment in- (1) p < 0.05
[43] status Yiqi pill), 60(57) strument (FS-14) (2) In some items (4
(2) SF-36 items) p < 0.05,
others NS
Assessment of RCTs of Moxibustion 267

Table 1 (continued )
Author, year Disorder Intervention, N Control, N Outcomes Results
(analyzed)
Gao (2015) Androgenic 42 Western medicine, 42 (1) Serum testosterone (T), (1) All p < 0.05
[44] alopecia estradiol (E2), T/E2 (2) p < 0.001
(2) Curative effect
ACT Z asthma control test; Acu Z acupuncture; Bcl-2 Z B-cell lymphoma gene-2; BDI Z Beck Depression Inventory; BPI Z Brief pain inventory;
BSS Z Bristol Stool Form Scale; CAS Z Constipation Assessment Scale; CBC Z complete blood count; CDAI Z Crohn’s disease Activity Index;
CRP Z C-reactive protein; electroAcu Z electroacupuncture; ESR Z erythrocyte sedimentation rate; FACT-G Z Functional Assessment of
Cancer Therapy-General score; FDIP Z facial disability index somatic function score; FDIS Z facial disability index social function score;
FEV1 Z forced expiratory volume in 1 second; fMRI Z functional magnetic resonance imaging; GSH Z glutathione; Herb Z herbal medicine;
HGB Z hemoglobin; IBS Z irritable bowel syndrome; K-WOMAC Z Korean Western Ontario and McMaster Universities Questionnaire;
LSS Z Lysholm Scoring Scale; LL Z long length; MDA Z malondialdehyde; ML Z moderate length; NA Z not available; NRS Z numeric rating
scale; NS Z not significant; NT Z no additional treatment; OA Z osteoarthritis; Obs Z observation; ODI Z Oswestry Disability Index; PEF Z peak
expiratory flow; Physio Z physiotherapy; PKC Z protein kinase C; Placebo Z placebo moxibustion; QOL Z quality of life; RMQ Z Roland Morris
Questionnaire; ROS Z reactive oxygen species; SASP Z salicylate fapyridine; SF-36 Z 36-Item Short Form Health Survey; Sham Z sham mox-
ibustion; SL Z short length; Temp Z temperature; VAS Z Visual Analog Scale; WOMAC Z Western Ontario and McMaster Universities
Questionnaire.

interventions, such as “we allowed the participant to take
herbal medicine.”
The Cochrane ROB tool was used to assess the methodo-
logical quality of the studies. The assessed items were as
follows: (1) random sequence generation for selection bias,
(2) allocation concealment for selection bias, (3) blinding of
participants and personnel for performance bias, (4) blinding
of outcome assessment for detection bias, (5) incomplete
outcome data for attribution bias, and (6) selective outcome
reporting for reporting bias. The assessment for each item
was “Low,” “Unclear,” and “High.” “Low” indicated low risk
of bias, “Unclear” was marked when the study mentioned
the item but not details, and “High” was for a high risk of bias
when there was no mention about the item.
This quality assessment was performed by two authors
(K.S.Y. and J.I.C.) independently. Any disagreement be-
tween authors was resolved with discussion.
3. Results
3.1. General characteristics
There were 442 potentially relevant studies in total. Of
these, 35 RCTs [11e44] that fulfilled the inclusion criteria
were retrieved for further analysis (Fig. 1). The general
characteristics of the included RCTs are presented in
Table 1. A total of 2,981 participants were involved in the
studies, of which 1,555 were in the intervention groups
and 1,426 were in the control groups (these are not the
exact numbers because two studies [37,39] did not specify
the number of participants enrolled). The data from 1,863
participants were analyzed, where 1,498 were in the
intervention groups and 1,365 in the control groups. The
number of participants ranged from 7 to 311 in the inter-
vention groups (mean  standard deviation,
45.73  54.16) and 7 to 144 in the control groups
(41.94  32.53). The median sample sizes in the inter-
vention and control groups were 30.5 and 36.5, respec-
tively. There were 75 intervention groups in 35 studies in
which one of these studies consisted of three intervention
groups [18], and another study had two intervention
groups [32]. There are 38 control groups in which four
studies designed two control groups [29,30,40,41]. In
total, there are five three-armed studies [29,30,32,40,41]
and one four-armed study [18].
Based on the 10th revision of the International Statistical
Classification of Diseases and Related Health Problems, we
categorized the disorder of participants. The most
frequent disorders were the diseases of the musculoskel-
etal system and connective tissue (Chapter XIII, n Z 8); the
diseases of the digestive system (Chapter XI, n Z 6); and
the symptoms, signs, and abnormal clinical and laboratory
findings, not elsewhere classified (Chapter XVIII, n Z 6).
The diseases of the circulatory system (Chapter IX, n Z 5),
the diseases of the genitourinary system (Chapter XIV,
n Z 3), neoplasms (Chapter II, n Z 1), the diseases of the
eye and adnexa (Chapter VII, n Z 1), the diseases of the
respiratory system (Chapter X, n Z 1), and the diseases of
the skin and subcutaneous tissue (Chapter XII, n Z 1) fol-
lowed. As an individual condition, osteoarthritis of knee
(n Z 3) and fatigue state (n Z 3) were most frequently
involved. Conditions associated with stroke (n Z 3) were
also frequently involved, but different disorders were
included, such as voiding dysfunction (n Z 1) and hemi-
plegia (n Z 2).
We classified the 38 control groups into three types: (1)
traditional East Asian medical treatment (n Z 20), such as
acupuncture (n Z 5), electroacupuncture (n Z 1), herbal
medicine (n Z 1), sham moxibustion (n Z 7), and combined
traditional therapies such as acupuncture, herbal medicine,
or physiotherapy (n Z 6). (2) Usual care (n Z 12) such as
Western medication (n Z 8), artificial tear (n Z 1), and
combined routine care such as conventional medication,
therapy, exercise (n Z 3). (3) No treatment (n Z 6) such as
bed rest or observation group.
The study outcomes were represented mostly by p
value except one study [24]. Thirteen studies (30.23%)
[12,16,23,26,28,31e33,39e42,44] reported that moxibus-
tion was effective for all the outcome measures designed
for the study. The remaining studies (n Z 30, 69.76%) re-
ported that either some of or all of their outcome
268 S.Y. Kim et al.

measures yielded results of no significance for the effec-
tiveness of moxibustion when applied to specific
conditions.
3.2. Details and reporting completeness of
STRICTOM items
The STRICTOM checklist consists of 15 items, and the items
are shown in Table 2. Of 34 studies, four used direct mox-
ibustion, where ignited moxibustion is placed above the
skin directly [12,14,37,38]. Three studies did not mention
about which type of moxibustion is used such as direct,
indirect, heat sensitive [29,33,34]. The remaining 27
studies used indirect moxibustion, where there is some
material, mostly ginger, salt, garlic, etc., between the
moxa and the skin. Concerning the material used for
moxibustion, moxa cone is the most widely used material in
16 studies [11e15,17,19e21,24,30,31,35e38], followed by
moxa stick in nine studies [16,18,23,27e29,32,43,44]. Two
studies used moxa pillar [22,25], two studies mentioned the
raw material of moxibustion, not the form [34,39], and two
did not present any information. One study used moxa roll
[42]. The most frequently used acupuncture point for
moxibustion was CV4 (n Z 9), followed by ST36 (n Z 7) and
CV6 (n Z 6). Twelve studies used only one acupoint, seven
studies used two points and three points, respectively, and
one study used the greatest number, 17 points. Seventeen
studies stated the number of moxibustion unit per point. It
varied from one to seven, three studies used one unit, two
studies used two units, eight studies used three units, and
three studies used five units. One study used from one to
seven moxibustion units per point. Of the 17 studies that
reported the number of moxibustion unit, six studies

Table 2 STandards for Reporting Interventions in Clinical Trials of Moxibustion (STRICTOM) items and completeness of
reporting or included studies (n Z 34)
Item Item No. Detail Number of reported
randomized controlled
trials [n (%)]
Moxibustion rationale 1a Type of moxibustion (direct moxibustion, indirect 31 (91.17)
moxibustion, heat-sensitive moxibustion, moxa burner
moxibustion, natural moxibustion)
1b Reasoning for treatment provided, based on historical 17 (50.00)
context, literature sources, and/or consensus method, with
references where appropriate
1c Extent to which treatment was varied 34 (100.00)
Details of moxibustion 2a Materials used for moxibustion (moxa floss, moxa cone, 29 (85.29)
moxa stick, herbal patches, and their sizes and
manufacturers)
2b Names of acupoints (or location if no standard name) for 34 (100.00)
moxibustion (unilateral/bilateral)
2c Number of moxibustion units and/or moxibustion time per 32 (94.11)
point (mean and range where relevant)
2d Procedure and technique for moxibustion (direct/indirect, 29 (85.29)
warming/sparrow-pecking technique, warming needle,
moxa box, heat-sensitive moxibustion)
2e Responses sought (warm feeling, skin reddening, burning 9 (26.47)
pain, heat-sensitization phenomenon)
2f Patient posture and treatment environment 13 (38.23)
Treatment regimen 3 Number, frequency, and duration of treatment sessions 34 (100.00)
Other components of 4a Details of other interventions administered to the 31 (91.17)
treatment moxibustion group (acupuncture, cupping, herbs, exercises,
lifestyle advice)
4b Setting and context of treatment protocol, and information 5 (14.70)
and explanation to patients
Treatment provider 5 Description of treatment provider (qualification or 11 (32.35)
background professional affiliation, years in moxibustion practice and
other relevant experience for professional, or any special
training in advance for layman)
Control and comparator 6a Rationale for the control or comparator in the context of 6 (17.64)
interventions the research question, with sources that justify the choice
6b Precise description of the control or comparator. If another 34 (100.00)
form of moxibustion or moxibustion-like control is used,
provide details as for Items 1e3 above
Precaution measures 7 Precise description of the precaution measures, if any 0
Assessment of RCTs of Moxibustion
Figure 2 Percentage of randomized controlled trials with complete reporting of the STandards for Reporting Interventions in
Clinical Trials of Moxibustion (STRICTOM) items.
mentioned the moxibustion time per point. A total of 18
studies reported how long the moxibustion time was per
point; 30 minutes was the most widely used moxibustion
time (n Z 7). The time varied from 5 minutes to 60 minutes.
Two studies did not use moxibustion time, but considered
moxibustion length, like “burning moxibustion until it left
0.5 cm or 30 mm” [15,16]. One study mentioned both
moxibustion unit number and moxibustion length limit for
burning [15]. Four studies used a response to moxibustion
time, like “burning moxibustion until pain appears”
[13,14,24,39]. Three studies represented both moxibustion
unit number and response to burning time length
[13,14,39]. Moxibustion treatment was done mostly one
time/d (n Z 13), or three times/wk (n Z 10). The duration
of treatment sessions varied from 7 days to 3 months,
mostly 4 weeks (n Z 11). The co-intervention administered
to most members of the moxibustion group was acupunc-
ture (n Z 12), followed by Western medication (n Z 4),
physiotherapy (n Z 4), usual care (n Z 2), etc. Eleven
studies described the treatment provider as licensed,
certified or trained therapist, or Korean/Oriental medical
doctor.
The number and percentage of reported RCTs for each
item are represented in Table 2. The least reported item
(Item 4bdSetting and context of treatment protocol, and
information and explanation to patients) was reported by
269
only five studies. Four items (Item 1cdExtent to which
treatment was varied; Item 2bdNames of acupoints for
moxibustion; Item 3dNumber, frequency and duration of
treatment sessions; and Item 6bdPrecise description of the
control or comparator) were reported in all the 34 studies.
The overall distribution of the completeness for STRICTOM
items is shown in Fig. 2. Five items were fulfilled between
five and 13 studies (from 15% to 38%), six items between 17
and 32 studies (from 50% to 94%), and four items were
fulfilled by all the 34 studies (100%), as mentioned earlier.
The mean of the number of reported studies was 23.26
(68%), and the median was 29 (85%).
As for the completeness per study, 13 studies fulfilled
between five and nine items (from 33% to 60% complete-
ness), 10 studies between 10 and 11 (from 67% to 73%
completeness), and 11 studies between 12 and 15 (from 80%
to 100% completeness). The mean of the number of satis-
fied items was 10.26 (68%), and the median was 10 (67%).
The percentage is shown in Fig. 3.
There was a significant association between the
completeness of reporting for STRICTOM items and the year
when the study published (r Z 0.613, p < 0.001). The
distribution of STRICTOM satisfactory percentage per study
by publication year is shown in Fig. 4. The more recent the
publication year is, the higher the STRICTOM percentage
appears to be.
270 S.Y. Kim et al.

Figure 3 Percentage of STandards for Reporting Interventions in Clinical Trials of Moxibustion (STRICTOM) items with complete
reporting per randomized controlled trials.

Figure 4 Correlation between the publication year and STandards for Reporting Interventions in Clinical Trials of Moxibustion
(STRICTOM) completeness percentage of included studies (n Z 34).

3.3. Risk of bias
Risk of bias assessed for each study is presented in Table 3.
One study fulfilled all criteria [22]. In general, studies did
not report which random sequence tool they used, and only
reported that they “randomly divided the participants.”
Fifteen of the 34 studies reported random sequence gen-
eration [11,12,14,16e22,24,25,37,42,44]. Many studies did
not mention allocation concealment. Only six studies
[14,17,19e22] reported allocation concealment by
Assessment of RCTs of Moxibustion 271

Table 3 Risk of bias graph of included studies (n Z 34)

Random Allocation Blinding of participants Blinding of Incomplete Selective
sequence concealment and personnel outcome outcome data outcome
generation assessment reporting
Son 2011 [11]

Lee 2011 [12]

Doh 2008 [13]

Lee 2011 [14]

Ryu 2013 [15]

Ren 2013[16]

Lu 2015 [17]

Xu 2015 [18]

Yi 2014 [19]

Bao 2014 [20]

Kim 2014 [21]

Zhao 2014 [22]

Chen 2013 [23]

Kim 2013 [24]

Park 2011 [25]

Zhang 2009 [26]

Ju 2009 [27]

(continued on next page)

272 S.Y. Kim et al.

Table 3 (continued )
Random Allocation Blinding of participants Blinding of Incomplete Selective
sequence concealment and personnel outcome outcome data outcome
generation assessment reporting
Zhang 2007 [28]

Gao 2007 [29]

Jia 2006 [30]

Zeng 2006 [31]

Liang 2006 [32]

Ding 2005 [33]

Edson 2003 [34]

Kang 2000 [35]

Wu 2000 [36]

Choi 2003 [37]

Kim 2005 [38]

Lee 2008 [39]

Shin 2009 [40]

Shi 2012 [41]

Bian 2013 [42]

Cai 2014 [43]

Gao 2015 [44]

Z indicates low risk of bias; indicates unclear risk of bias; indicates high risk of bias.
Assessment of RCTs of Moxibustion
envelope or letter. Blinding of participants and personnel
was reported only by one study [22], which used sham
moxibustion treatment. Eight studies used the sham moxi-
bustion method [11,14,19,20,22,24,25,34], but most of
them designed blinding of participants, not the clinicians
involved. Four studies reported blinding of outcome
assessment. There was no study that reported incomplete
outcome data. Only one study had a risk of reporting bias,
because the study reported one of its outcome measures
for limited participants, and its criteria were not fully
described [36].
4. Discussion
The reporting quality of RCT is becoming more important,
especially for studies on traditional East Asian medicine, as
its methodology has not yet been fully established. A few
existing studies addressed the issue of reporting quality of
RCT of moxibustion. Xiong et al [45] conducted a system-
atic evaluation on acupuncture and moxibustion studies in
China. Wang et al [46] collected RCTs of traditional Chinese
medicine and assessed the reporting quality, by analyzing
the abstracts of Chinese medical journals. Previous re-
views, including the aforementioned two studies, sug-
gested that RCTs of traditional East Asian medicine need to
provide more detailed information on research
methodology.
We reviewed 34 studies of moxibustion and found that
some studies reported moxibustion as being effective.
Considering that reporting on the effectiveness of the
moxibustion treatment can be affected by the kind and
number of outcome measures the researcher choose, it is
difficult to make a simple comparison of the effectiveness
among studies. The report “Not significant” also does not
mean moxibustion is not effective for the specific disease
or symptom, but rather the study could not find evidence
supporting the effect on the condition. The study design is
another factor that can affect the results. Of the 34 studies
included in this work, there were studies that had a less-
refined study design than others. Further studies that take
these variables into account should be undertaken.
The STRICTOM checklist suggests various details that
future research of moxibustion should consider. The least
reported items, setting and context (Item 4b), rationale for
the control (Item 6a), and response (Item 2e), should be
considered when reporting and designing a study. For the
item setting and context, researchers should report infor-
mation about the protocol setting, such as which expla-
nation will be given to the participants. It is important to
report the setting and context for other researchers who
try the same or a similar study design. The rationale for the
control, as well as the rationale for the moxibustion
treatment, should be included in the report, such as the
historical context, literature sources, other related papers,
and/or consensus method. The response sought through
moxibustion is not emphasized as much as the response
through acupuncture, so-called de qi [47]. De qi is a
distinctive sensation experienced by a patient or an
acupuncturist during acupuncture treatment. The response
related to acupuncture was named de qi, but the feeling or
response by moxibustion has not been studied sufficiently,
273
and which response affects the action of moxibustion is
more uncertain. Sensations such as heating, burning,
warming feeling, or other phenomena such as redness and
peak temperature can be target responses, and these
should be reported. As moxibustion is thought to have
nonspecific effects such as psychological reactions by the
treatment environment, posture during treatment, rela-
tionship with treatment provider [2], the researcher should
report information about the patient posture and treat-
ment environment. For the same reason, the description of
the provider is important not only for the quality of
treatment but also for other researchers who design
studies of moxibustion. For the items of other interventions
administered to the moxibustion group, there can be many
types of co-intervention with moxibustion such as
acupuncture, cupping, herbal medicine, exercise, or life-
style advice. In this study, we checked whether there is
mention about the other interventions administered, but
strictly speaking the details should be followed. For
example, “we allowed the participant to take herbal
medicine” is not sufficient, as there is no information about
which herbal medicine it is, time of administration, dose,
etc. Many other items of STRICTOM lacked detail, although
the completeness of reporting improved for recently pub-
lished studies.
Concerning the risk of bias, blinding is difficult in trials
using moxibustion because patients will distinguish the
sham moxibustion from the real one. As blinding is difficult
or nearly impossible, assessing ROB may be less meaningful
for studies of moxibustion. Eight studies designed sham
moxibustion method [11,14,19,20,22,24,25,34] using an
insulator to isolate heat [11,19,20,22,24,25], a tempera-
ture control method where the moxibustion was removed
before the heat reached the skin [14] or nontherapeutic
indifferent points [34]. More investigation of sham moxi-
bustion methods is required for future studies, because
estimating the effectiveness of moxibustion is inseparable
from appropriate sham moxibustion.
The first limitation of this study is the possibility of
undetected studies. We tried to retrieve all available RCTs
of moxibustion, but with no limit on patient’s disease or
condition, the searching strategy may be less refined.
Although the six search engines we used offered relatively
broad search features, there may be some missing studies.
Second, even if the study satisfied a reporting item of
STRICTOM or ROB partially because the item has multiple
subitems, we considered the item as satisfied. Strictly
speaking, the full description of the content for a report-
ing item should be required. Our assessment in some items
may be overrated by this point of view. Third, it has only
been a short time since STRICTOM introduced, in 2013. It is
similar to STRICTA but it demands a certain description
only for moxibustion. Some time may be needed for re-
searchers to fully grasp and make use of this reporting
tool. For the RCTs analyzed in this study, there may be not
enough information and experience about details of
STRICTOM items.
This work is the first close investigation to assess the
completeness of reporting RCTs of moxibustion based on
STRICTOM and ROB. Our evaluation indicates that recent
RCTs of moxibustion need to report key components of RCT
and details about the study. Our study results can also
274
provide researchers and clinicians with more information
about the recent studies of moxibustion and the current
study trends. We reviewed 34 RCTs of moxibustion and
recognized the need for proper guidelines for RCT of mox-
ibustion. It would be easier to report studies of moxibustion
if there were guidelines available on the stage of study
design. Therefore, based on this work, we will try to
develop guidelines for RCT of moxibustion for designing
protocols, and reporting the results. We look forward to
more rigorous trials of moxibustion with refined study
design and a high degree of reliability.
Disclosure statement
The authors have no conflicts of interest to declare.
Acknowledgments
This study was supported by a grant of the Traditional
Korean Medicine R&D Project, Ministry of Health & Welfare,
Republic of Korea (HI15C0006).
References
[1] Chang XR, Hong J, Yi SX. Illustrated Chinese Moxibustion
Technique and Methods. London, UK: Singing Dragon; 2012:
16e21.
[2] Yamashita H, Ichiman Y, Tanno Y. Changes in peripheral
lymphocyte subpopulations after direct moxibustion. Am J
Chin Med. 2011;29:227e235.
[3] Zhang JF, Wu YC. Modern progress of mechanism of moxi-
bustion therapy. J Acupunct Tuina Sci. 2006;4:257e260.
[4] Kim SY, Chae Y, Lee SM, Lee H, Park HJ. The effectiveness of
moxibustion: an overview during 10 years. Evid Based Com-
plement Alternat Med. 2011, e306515.
[5] Park HJ, Son CG. Overview for moxibustion-related researches
worldwide. J Acupunct Meridian Stud. 2008;25:167e174.
[6] Cheng CW, Fu SF, Zhou QH, Wu TX, Shang HC, Tang XD.
Extending the CONSORT statement to moxibustion. J Integr
Med. 2013;11:54e63.
[7] Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D,
Oxman AD, et al. The Cochrane Collaboration’s tool for
assessing risk of bias in randomised trials. BMJ. 2011;343,
d5928.
[8] Schulz KF, Altman DG, Moher D. CONSORT 2010 statement:
updated guidelines for reporting parallel group randomized
trials. Ann Int Med. 2010;152:726e732.
[9] MacPherson H, Altman DG, Hammerschlag R, Youping L,
Taixiang W, White A, et al. Revised STandards for Reporting
Interventions in Clinical Trials of Acupuncture (STRICTA):
extending the CONSORT statement. PLoS Med. 2010;7,
e1000261.
[10] Lee H, Cha SJ, Park HJ, Seo JC, Park JB, Lee HJ. Revised
STandards for Reporting Interventions in Clinical Trials of
Acupuncture (STRICTA): extending the CONSORT statement.
Korean J Acupunct. 2010;27:1e23.
[11] Son CG. Safety of 4-week indirect-moxibustion therapy at CV4
and CV8. J Acupunct Meridian Stud. 2011;4:262e265.
[12] Lee JY, Park SH, Han SY, Park JY, Lee HJ. A case-control study
of the effect of cotreatment with Sinseon moxibustion on low
back pain of HIVD patients. J Korean Acupunct Moxibustion
Soc. 2011;28:77e83.
S.Y. Kim et al.
[13] Doh MH, Kim TY. Effects of the moxibustion therapy on the
pain decrease and joint recovery with degenerative knee
arthritis. J Soc Prev Korean Med. 2009;13:81e92.
[14] Lee JS. The efficacy and safety of moxibustion with regard to
cancer pain in advanced cancer patients: a randomized,
single-blinded, and placebo-controlled pilot study. Seoul,
South Korea: Department of Oriental Medicine, Graduate
School Kyung Hee University; 2011 [master’s thesis]. Available
at: http://dcollection.khu.ac.kr/jsp/common/DcLoOrgPer.
jsp?sItemIdZ000000083198.
[15] Ryu HS, Park KS. The effects of moxibustion on chronic back
pain and activities of daily living in aged. J Korean Acad Soc
Rehabil Nurs. 2015;18:38e45.
[16] Ren YH, Zhong L. Influence of moxibustion at shénquè (神阙
CV 8) on senile symptoms and melatonin. World J Acupunct
Moxibustion. 2013;23:6e10.
[17] Lu ZZ, Yin XJ, Teng WJ, Chen YH, Sun J, Zhao JM, et al.
Comparative effect of electroacupuncture and moxibustion
on the expression of substance P and vasoactive intestinal
peptide in patients with irritable bowel syndrome. J Tradit
Chin Med. 2015;35:402e410.
[18] Xu JF, Lin RZ, Wu YL, Wang YX, Liu J, Zhang YL, et al. Effect of
stimulating acupoint Guanyuan (CV 4) on lower back pain by
burning moxa heat for different time lengths: a randomized
controlled clinical trial. J Tradit Chin Med. 2015;35:36e40.
[19] Zhu Y, Wu ZY, Ma XP, Liu HR, Bao CH, Yang L, et al. Brain
regions involved in moxibustion-induced analgesia in irritable
bowel syndrome with diarrhea: a functional magnetic reso-
nance imaging study. BMC Complement Altern Med. 2014;14:
500.
[20] Bao CH, Zhao JM, Liu HR, Lu Y, Zhu YF, Shi Y, et al. Ran-
domized controlled trial: moxibustion and acupuncture for
the treatment of Crohn’s disease. World J Gastroenterol.
2014;20:11000e11011.
[21] Kim TH, Kim KH, Kang JW, Lee MH, Kang KW, Kim JE, et al.
Moxibustion treatment for knee osteoarthritis: a multi-centre,
non-blinded, randomised controlled trial on the effectiveness
and safety of the moxibustion treatment versus usual care in
knee osteoarthritis patients. PLoS One. 2014;9:1e8.
[22] Zhao L, Cheng K, Wang LZ, Wu F, Deng HP, Tan M, et al.
Effectiveness of moxibustion treatment as adjunctive therapy
in osteoarthritis of the knee: a randomized, double-blinded,
placebo-controlled clinical trial. Arthritis Res Ther. 2014;16,
R133.
[23] Chen RX, Chen MR, Xiong J, Chi ZH, Zhang B, Tian N, et al.
Curative effect of heat-sensitive moxibustion on chronic
persistent asthma: a multicenter randomized controlled trial.
J Tradit Chin Med. 2013;33:584e591.
[24] Kim HG, Yoo SR, Park HJ, Son CG. Indirect moxibustion (CV4
and CV8) ameliorates chronic fatigue: a randomized, double-
blind, controlled study. J Altern Complement Med. 2013;19:
134e140.
[25] Park JE, Sul JU, Kang KW, Shin BC, Hong KE, Choi SM. The
effectiveness of moxibustion for the treatment of functional
constipation: a randomized, sham-controlled, patient blinded,
pilot clinical trial. BMC Complement Altern Med. 2011;11:124.
[26] Zhang HJ. 76 Cases of hypomenorrhea treated by acupuncture
to regulate the menstrual cycle. J Tradit Chin Med. 2009;29:
177e178.
[27] Ju YL, Chi X, Liu JX. Forty cases of insomnia treated by sus-
pended moxibustion at Baihui (GV 20). J Tradit Chin Med.
2009;29:95e96.
[28] Zhang YC, Yang W. Effects of acupuncture and moxibustion on
tear-film of the patients with xerophthalmia. J Tradit Chin
Med. 2007;27:258e260.
[29] Gao XY, Yuan J, Li HJ, Shan R. Clinical research on acupunc-
ture and moxibustion treatment of chronic atrophic gastritis.
J Tradit Chin Med. 2007;27:87e91.
Assessment of RCTs of Moxibustion
[30] Jia J, Wang QY, Zhang TH, Li J. Treatment of ankylosing
spondylitis with medicated moxibustion plus salicylazosulfa-
pyridine and methotrexateda report of 30 cases. J Tradit
Chin Med. 2006;26:26e28.
[31] Zeng XX. Ginger moxibustion for treatment of cervical verti-
goda report of 40 cases. J Tradit Chin Med. 2006;26:17e18.
[32] Liang FR, Li Y, Yu SG, Li CD, Hu LX, Zhou D, et al. A multi-
central randomized control study on clinical acupuncture
treatment of Bell’s palsy. J Tradit Chin Med. 2006;26:3e7.
[33] Ding XR, Gao QF, Li P. Acupuncture treatment for multiple
aortitisda clinical report of 40 cases. J Tradit Chin Med.
2005;25:260e263.
[34] Edson G, Agnaldo CP, Ysao Y, Miguel S. Effects of acupuncture
and moxa treatment in patients with semen abnormalities.
Asian J Androl. 2003;5:345e348.
[35] Kang KS, Jeong EJ, Moo SK, Ko CN, Joh KH, Kim YS, et al. Clinical
study on the effects of moxibustion for post-stroke voiding
dysfunction. J Korean Oriental Med. 2000;21:236e241.
[36] Wu HG, Zhou LB, Shi DR, Liu SM, Liu HR, Zhang BM, et al.
Morphological study on colonic pathology in ulcerative colitis
treated by moxibustion. World J Gastroenterol. 2000;6:861e865.
[37] Choi YS, Kim TK, Jung WS, Moon SK. Effects of moxibustion on
the hemiplegic upper extremity after stroke. Korean J Orient
Int Med. 2003;24:283e289.
[38] Kim BS, Jang IS, Yeo JJ, Lee TH, Son DH, Se ES, et al. Effect of
Choksamni (足三里, ST 36) moxibustion on blood pressure
elevation in hypertensive patients: a randomized controlled
trial. J Korean Oriental Med. 2005;26:66e73.
[39] Lee SH, Kim JK, Son YH, Jeong HY, Kim JH, Kwon JN, et al. A
clinical study of moxibustion therapy’s effect on functional
recovery in hemiplegia on stroke. Korean J Orient Int Med.
2008;29:278e284.
275
[40] Shin KR, Ha JY, Park HJ, Heitkemper M. The effect of hand
acupuncture therapy and hand moxibustion therapy on pre-
menstrual syndrome among Korean women. West J Nurs Res.
2009;31:171e186.
[41] ShiY,CuiYH,WuHG,ZhangW,ZhaoC,LiuHR,etal.Effectsofmild-
warming moxibustion on Bcl-2 and PKC expressions of peripheral
blood in elderly people. J Tradit Chin Med. 2012;32:45e51.
[42] Bian XM, Ling L, Lin WB, Liang HH, Zhang Y, Wang LC. Moxi-
bustion therapy at CV4 prevents postoperative dysuria after
procedure for prolapse and hemorrhoids. Evid Based Com-
plement Alternat Med. 2013:756095.
[43] Cai RL, Hu L, Li ZH, Li M, Cheng HL, Wu ZJ. Effects of warming
moxibustion on scores of fatigue and life quality in patients
with subhealth fatigue status. World J Acupunct Moxibustion.
2014;24:10e14.
[44] Gao JY, Liu HJ, Shi J, Yang Y, Ge CQ, Qie ZW, et al. Clinical
efficacy of catgut embedment combined with moxibustion
and bloodletting in the treatment of androgenic alopecia.
World J Acupunct Moxibustion. 2015;25:1e6.
[45] Xiong J, Du YH, Li B, Shi L, Xu YY, Liu Q, et al. Assessment of
methodology and report quality of systematic evaluation and
meta-analysis of acupuncture-moxibustion in China. Zhongguo
Zhen Jiu. 2009;29:763e768.
[46] Wang L, Li Y, Li J, Zhang M, Xu L, Yuan W, et al. Quality of
reporting of trial abstracts needs to be improved: using the
CONSORT for abstracts to assess the four leading Chinese med-
ical journals of traditional Chinese medicine. Trials. 2010;11:75.
[47] Park JE, Ryu YH, Liu Y, Jung HJ, Kim AR, Jung SY, et al. A
literature review of de qi in clinical studies. Acupunct Med.
2013;31:132e142.