Drug and Alcohol Dependence 142 (2014) 14–23

Contents lists available at ScienceDirect

Drug and Alcohol Dependence

journal homepage: www.elsevier.com/locate/drugalcdep

Review

A meta-analysis of ear-acupuncture, ear-acupressure and

auriculotherapy for cigarette smoking cessation
Yuan Ming Di a,1,2 , Brian H. May a,1,3 , Anthony Lin Zhang a,4 , Iris Wenyu Zhou a,5 ,
Christopher Worsnop c,6 , Charlie C.L. Xue a,b,∗
a
School of Health Sciences, Traditional & Complementary Medicine Research Program, Health Innovations Research Institute, RMIT University, Bundoora,
VIC 3083, Australia
b
Guangdong Provincial Academy of Chinese Medical Sciences, Guangzhou, China
c
Department of Respiratory and Sleep Medicine, Austin Hospital, Heidelberg, VIC 3081, Australia

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

Article history: Background: This systematic review evaluated the effects of ear acupuncture, ear acupressure and auricu-
Received 16 October 2013 lotherapy for cigarette smoking cessation (SC) at end-of-treatment (EoT), three, six and 12 months
Received in revised form 2 July 2014 follow-up.
Accepted 3 July 2014
Methods: Searches of six English and Chinese databases located 25 randomized controlled trials (3735
Available online 11 July 2014
participants). Methodological quality was assessed using Cochrane Risk of Bias. Meta-analyses were
conducted in two pools: 1. SC-specific ear acupuncture/acupressure or auriculotherapy (EAP/R) vs. non-
Keywords:
specific/inactive control; and 2. SC-specific EAP/R vs. other SC-specific treatment. Sensitivity analyses
Smoking
Tobacco were conducted based on the validity of interventions as SC-specific treatments or non-specific/inactive
Cigarette interventions; and the use of biochemical SC confirmation.
Ear acupressure Results: Pool 1: the 12 valid SC-specific EAP/R interventions were superior to inactive EAP/R controls at
Ear acupuncture EoT (RR = 1.77 [1.39, 2.25]), three months follow-up (RR = 1.54 [1.14, 2.08]), and six months follow-up
Systematic review (RR = 2.01, [1.23, 3.28]) but data were insufficient at 12 months. In Pool 2: there was no superiority or
inferiority for EAP/R at EoT or at 3 and 6 month follow-ups compared to SC-specific behavioural therapy
or SC-specific body acupuncture.
Conclusions: Pool 1 data appeared most consistent for studies of ear acupressure (EAPR) vs. non-specific
EAPR controls, with confirmed SC rates at 3 months post-treatment of 20.0% for test groups vs. 7.5% for
controls. In Pool 2 the EAP/R interventions appeared neither inferior nor superior to the behavioural inter-
ventions at 3 and 6 month follow-ups. However, meta-analysis results derived from relatively small-sized
trials with no biochemical validation of SC in Pool 2. Larger, well-controlled studies using biochemical
confirmation of SC are needed.
© 2014 Elsevier Ireland Ltd. All rights reserved.

∗ Corresponding author at: Head of School, School of Health Sciences; Director, Traditional & Complementary Medicine Program, RMIT Health Innovations Research
Institute, RMIT University, PO Box 71, Bundoora, VIC 3083, Australia. Tel.: +61 3 99257360; fax: +61 3 99256539.
E-mail addresses: yuan.di@rmit.edu.au (Y.M. Di), brian.may@rmit.edu.au (B.H. May), tony.zhang@rmit.edu.au (A.L. Zhang), iris.zhou@rmit.edu.au (I.W. Zhou),
christopher.worsnop@austin.org.au (C. Worsnop), charlie.xue@rmit.edu.au (C.C.L. Xue).
1
Contributed equally to this work.
2
Discipline of Chinese Medicine, School of Health Sciences, World Health Organization Collaborating Centre for Traditional Medicine RMIT University, PO Box 71, Bundoora,
VIC 3083, Australia. Tel.: +61 399257511; fax: +61 399257178.
3
Discipline of Chinese Medicine, School of Health Sciences, World Health Organization Collaborating Centre for Traditional Medicine RMIT University, PO Box 71, Bundoora,
VIC 3083, Australia. Tel.: +61 399257166; fax: +61 399257178.
4
Discipline of Chinese Medicine, School of Health Sciences, World Health Organization Collaborating Centre for Traditional Medicine RMIT University, PO Box 71, Bundoora,
VIC 3083, Australia. Tel.: +61 399257758; fax: +61 39925 7178.
5
Discipline of Chinese Medicine, School of Health Sciences, World Health Organization Collaborating Centre for Traditional Medicine RMIT University, PO Box 71, Bundoora,
VIC 3083, Australia. Tel.: +61 399256595; fax: +61 399257178.
6
Senior Respiratory Physician Department of Respiratory & Sleep Medicine, Austin Hospital 145 Studley Road, Heidelberg, VIC 3084, Australia. Tel.: +61 394963688;
fax: +61 394965124.

http://dx.doi.org/10.1016/j.drugalcdep.2014.07.002
0376-8716/© 2014 Elsevier Ireland Ltd. All rights reserved.
 Y.M. Di et al. / Drug and Alcohol Dependence 142 (2014) 14–23 15

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1. Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2. Study inclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3. Data extraction and assessment of risk of bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4. Assessment of EAP/R interventions and controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.5. Data analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.1. Characteristics of included studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2. Adverse events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.3. Ear points used in treatment and control groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.4. Effects of interventions – meta-analysis of smoking cessation events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.4.1. Pool 1: comparison of EAP/R with inactive control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.4.2. Pool 2: comparison of EAP/R with other active therapies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.1. Comparison with other reviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.2. Limitations of this review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.3. Implications for future trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.4. Implications for practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Role of funding source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Authors’ contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

1. Introduction Ear acupuncture (EAP), ear acupressure (EAPR) and other forms
of auriculotherapy have been used for SC and nicotine withdrawal
The effectiveness of acupuncture and related therapies for symptoms (Han, 2006; Li et al., 2009). The National Acupuncture
cigarette smoking cessation (SC) has not been established and the Detoxification Association in the US uses ear acupuncture (EAP) for
results of studies can appear contradictory. A systematic review of drug addictions (McLellan et al., 1993) and EAP is the most common
33 randomized clinical trials (RCTs) of acupuncture, acupressure form of acupuncture for various kinds of substance abuse in the US
and related techniques for SC found an overall positive result when and the UK (D’alberto, 2004; Margolin, 2003).
acupuncture was compared with sham acupuncture at end-of- The meta-analyses in this review focus on interventions for SC
treatment (EoT) but not at follow-up. The review authors reported that use ear acupuncture or acupressure (EAP/R) or other types of
that the results were inconclusive and recommended high quality auriculotherapy. To ensure the clinical relevance of the data, stud-
trials (White et al., 2011). A more recent meta-analysis showed a ies were assessed to determine whether the ‘real’ intervention was
significant effect of acupuncture point (acupoint) stimulation on representative of SC treatments in the clinical literature. To improve
SC rates and cigarette consumption at EoT and at 3- and 6-month the meaningfulness of meta-analyses, the ‘sham’ or ‘placebo’ con-
follow-ups (Cheng et al., 2012) and a meta-analysis of 6 trials found trol intervention was assessed to determine whether it was likely
a benefit for real acupuncture vs. sham acupuncture at longest to have been truly inactive or non-specific for SC.
follow-up (Tahiri et al., 2012). This review aims to evaluate the effects of EAP/R for SC to deter-
These reviews included a variety of acupuncture methods in the mine whether ‘real’ EAP/R is more effective than: (1) ‘sham/placebo’
test arms. Some studies used points on the ear, i.e., ‘ear acupunc- EAP/R, no intervention, wait-list, or other ‘inactive’ control; (2)
ture’ or ‘auriculotherapy’, some used points on other parts of the other SC therapies (medical or behavioural therapies or body
body, i.e., ‘body acupuncture’, while others used a combination of acupuncture).
acupuncture types. In some studies, needles were used to pierce the
skin, i.e., ‘acupuncture’, while others used ‘acupressure’ with beads
attached to the points which were pressed to produce stimulation. 2. Methods
Machines that emit laser light or electrical current were used in
This review based its methods on the Cochrane Handbook for Systematic
some studies. The control arms in some studies employed ‘placebo’ Reviews of Interventions 5.1.0 (Higgins and Green, 2011).
or ‘sham’ interventions designed to mimic the ‘real’ interventions
while other studies compared acupuncture to no treatment or to
interventions that the investigators considered likely to be effec- 2.1. Search strategy
tive for SC. This diversity of interventions and controls presents
Six databases were searched from their inceptions to January 2013: Cochrane
a challenge for reviewers and raises the issue of the meaningful-
Central Register of Controlled Trials (CENTRAL) on The Cochrane Library (2013),
ness of pooled data for clinicians who need to know whether or not PubMed (1996–2013), EMBASE (1998–2013), Scopus (2004–2013), Chinese VIP
the evidence supports acupuncture for SC. For researchers aiming information (www.cqvip.com, 1989–2013) and China National Knowledge Infra-
to design controlled trials, there is the additional issue of how to structure (www.cnki.net, 1988–2013). Search terms were grouped into three blocks:
design plausible ‘sham’ controls in acupuncture trials. This ques- 1 – intervention (including acupuncture, auriculotherapy, acupressure, laser ther-
apy, electrotherapy); 2 – clinical condition (including SC, nicotine withdrawal, quit
tion has received considerable attention (Zhang et al., 2014) but smoking); and 3 – trial design (including clinical trial, random, control, placebo).
it remains unclear which, if any, approach approximates an actual Complete lists of search terms are available from the authors. Reference lists of
placebo. review articles were checked for additional references.
16 Y.M. Di et al. / Drug and Alcohol Dependence 142 (2014) 14–23

Potentially relevant published Potentially relevant published

studies from electronic search of 4 studies from electronic search of 2
English databases Chinese databases
n = 772 n = 298

No. of records after duplicates No. of records after duplicates Other sources
removed - English databases removed- Chinese databases n=1
n= 684 n=174

No. of records after duplicates

removed from English and
Chinese databases
n=859

Studies excluded by screening title

and abstract
n=617

Studies excluded (not related to

Studies screened and categorised acupuncture, not related to
n=242 smoking cessation, not controlled
study)
n= 201

Full text articles assessed for Not randomised, n=1

eligibility Follow up study n=2
n=41 Did not use ear acupuncture n=12
Smoking cessation not an outcome n=1
Studies excluded = 16

Studies included in qualitative

synthesis
n=25

Studies included in quantitative

synthesis (meta-analysis)
n=25

Fig. 1. Flow chart of the search and selection process of RCTs for inclusion in this review.

2.2. Study inclusion 2.4. Assessment of EAP/R interventions and controls

Randomized and quasi-randomized controlled trials of any duration published
in English, French or Chinese were included. Trial participants were smokers aiming
to quit regardless of age or gender. Studies that used a form of EAP/R as intervention
or part of the intervention and used SC as an outcome measure were included. The
process of trial inclusion and exclusion is shown in Fig. 1.
The validity of comparisons between the ‘real’ EAP/R intervention(s), the ‘sham’
or ‘placebo’ intervention(s), or the other ‘active’ SC therapy(s) were assessed for
each study. Textbooks on acupuncture were used as the basis for judgments on
point usage for SC (see Appendix 2).

2.3. Data extraction and assessment of risk of bias

Data were extracted using a pre-designed form. Y. Di extracted articles in English
and Chinese and I. Zhou checked the data against original articles. AH. Tan and S.
Gianfreda independently extracted data for French articles. Risk of bias was eval-
uated using the Cochrane Collaboration’s procedures (Higgins and Green, 2011).
Judgements were made independently by Y. Di and I. Zhou (English and Chinese)
and by AH Tan and S. Gianfreda (French). Disagreements were resolved by discus-
sion or by a third reviewer (AZ). In two cases, attempts were made to contact authors
to obtain additional information.
2.5. Data analysis
Studies were grouped for pooling based on comparable types of interventions
and controls. SC outcome data were pooled for: end-of-treatment (EoT), 3-, 6- and
12 month follow-ups (FU). Meta-analysis of SC rates was conducted in Review Man-
ager 5.1 using risk ratio (RR). All drop-outs were assumed to have resumed smoking.
Heterogeneity was assessed using I2 . Sensitivity analyses were conducted for com-
parisons judged as using valid interventions and controls and for studies reporting
biochemical confirmation of SC. SC rates are given for intervention and control
 Y.M. Di et al. / Drug and Alcohol Dependence 142 (2014) 14–23
groups and risk difference (RD) is reported as a measure of the percentage difference
between groups.
3. Results
3.1. Characteristics of included studies
All 25 studies tested a form of ear-acupoint stimulation intended
to assist smokers to quit (Supplementary Material 1) and seven
of these studies combined ear and body points (Supplementary
Material 2). Methods of stimulation included needles, pressure,
electricity, laser, and combinations of these. Two main types of
ear acupuncture controls were used: ‘sham’ acupuncture on loca-
tions that were not listed as acupuncture points, and ‘sham’
acupuncture/acupressure using real points considered not specific
for SC. Other controls included wait-list or no intervention, oral
placebo, medical therapy, SC-specific body acupuncture, advice or
behavioural therapy (Supplementary Material 2).
Treatment durations ranged from 1 to 30 days. One study used
a cross-over design (Steiner et al., 1982). The others were paral-
lel designs. Four studies had no dropouts (Han, 2006; Parker and
Mok, 1977; Tian and Krishnan, 2006; Waite and Clough, 1998). Oth-
erwise, dropout rates ranged from 5 to 72%. SC was reported at
time-points ranging from EoT to 5 years after the treatment. 12
studies used a biochemical method to confirm SC (Supplementary
Material 1), seven used exhaled carbon monoxide (CO), four used
cotinine, and one used urine thiocyanate. The remaining studies
used self-report only. Risk of bias assessments are in Supplemen-
tary Material 3 and assessments of the validity of test and control
interventions are in Supplementary Material 2.
3.2. Adverse events
Of the 13 studies that reported AEs, two reported no AEs
were associated with the intervention and eight studies (n = 779)
reported 115 minor AEs (Supplementary Material 1). Two studies
reported types of AEs but did not specify the number of events.
AEs caused by EAP included sore ears, tenderness, sensation around
residual needles, bruising, facial swelling, headache, dizziness, nau-
sea, giddiness, vomiting, euphoria, decreased migraine, and change
in insomnia. EAPR-related AEs included skin allergy to adhesive
tape, itchy or sore ear, and discomfort due to ear beads. One study
reported a participant withdrawal due discomfort.
3.3. Ear points used in treatment and control groups
Interventions used from 1 to 6 ear points. In total, 21 differ-
ent ear points were used in the test groups and 18 different ear
points were used in the controls (Supplementary Material 2). The
most frequently used test points were: Lung (23 studies), Shen-
men (14 studies) and Mouth (9 studies). In the 13 studies that used
non-SC-specific ear points in the controls, the most frequently used
were: Shoulder (6 studies), Eye (3 studies) and Knee (2 studies). Five
studies used non-points as controls but the exact locations were
not specified (Bier et al., 2002; Gillams et al., 1984; Steiner et al.,
1982; Wing et al., 2010; Yeh et al., 2009). Ten studies treated both
ears, nine studies used points unilaterally and seven studies did not
specify.
Treatment sessions ranged from 15 to 30 min. Ear press-needles
were retained from 3 to 4 days up to 3 weeks. Ear beads in EAPR
studies were retained from 3 to 14 days. Studies that used electro-
stimulation used 20 min duration (Fritz et al., 2013; He et al.,
1997; Martin and Waite, 1981; Parker and Mok, 1977; Waite and
Clough, 1998; White et al., 1998; Yeh et al., 2009) but Antoniou
et al. (2005) provided no information on stimulation time (Sup-
plementary Material 2). For laser treatments, Cai et al. (2000) and
17
Kerr and colleagues (2008) both stimulated each point for 1 min.
Only Steiner et al. (1982) mentioned acquiring Deqi (i.e., Tehchi)
sensation in the treatment group. In most studies that used ear
press-needles or EAPR, participants were instructed to stimulate
the points when cravings occurred (Gilbey, 1977; Gillams et al.,
1984; Han, 2006; Li et al., 2009; Tian and Krishnan, 2006; Waite
and Clough, 1998; White et al., 2007; Wing et al., 2010; Zhang et al.,
2013a,b). One study specified the number of stimulations a day (He
et al., 1997).
Although there was considerable diversity in the points used
in the 25 studies, this was reduced in the 12 studies that were
judged to include ‘valid comparisons’. The main ear points were
Lung (n = 12), Shenmen (n = 7) and Mouth (n = 5) and the NS ear
points included Shoulder (n = 5) and Eye (n = 3).
3.4. Effects of interventions – meta-analysis of smoking cessation
events
Studies in Pool 1 all compared a SC-specific EAP/R intervention
with a control that was intended to be inactive for SC. In Pool 2, the
studies all compared SC-specific EAP/R with a non-EAP/R interven-
tion that was intended to be active for SC. In each pool, body points
could be combined with ear points in the test group. Within the
pools, studies that shared similar designs were grouped together.
Results for all 25 studies are reported in Appendix 4 and results for
the 12 validated studies are in Table 1.
3.4.1. Pool 1: comparison of EAP/R with inactive control. At EoT for
all 21 comparisons, there were benefits for the groups EAPR vs. NS-
EAPR, EAP vs. no intervention/wait-list, EAP + Body acupuncture vs.
placebo medication and active vs. inactive machine. The total pool
showed a benefit in favour of the test groups (RR = 1.85, 95% CI
[1.58, 2.16], I2 = 72%) but the heterogeneity was high (Appendix
4). The funnel plot showed no evidence of publication bias. For
the 10 validated comparisons the total pool showed a significant
benefit for the test interventions (RR = 1.77, 95% CI [1.39, 2.25],
I2 = 53%) (Fig. 2). Cessation rates for valid studies were 22.7% (test)
vs. 12.6% (control), RD 10.1% (Table 1). When the 6 comparisons that
used a biochemical confirmation of SC were pooled, the difference
remained significant (RR = 1.56, 95% CI [1.15, 2.11], I2 = 3.9%).
At 6 weeks to 3 months follow-up for all 16 comparisons, there
were benefits for the EAPR vs. NS-EAPR and active vs. inactive
machine groups (Appendix 4). The pooled result showed a ben-
efit (RR = 2.16, 95% CI [1.72, 2.71], I2 = 69%). For the 9 validated
comparisons, benefit remained for EAPR vs. NS-EAPR and bene-
fit became evident for EAP vs. no intervention (Table 1). The valid
pool showed a benefit for the test interventions (RR = 1.54, 95% CI
[1.14, 2.08], I2 = 11%) and the cessation rates were 15.8% (test) vs.
10.1% (control), RD 4.7%. The pool for the 3 studies using biochem-
ical confirmation was not significant (RR = 1.28, 95% CI [0.85, 1.93])
but heterogeneous (I2 = 52%), whereas the EAPR vs. NS-EAPR group
showed a benefit (RR = 2.49, 95% CI [1.13, 5.49], I2 = 0%).
At 6 months follow-up, the only group that showed a benefit was
active vs. inactive machine (Appendix 4, 1.3.8). The pooled result for
11 comparisons was significant but heterogeneous (RR = 2.68, 95%
CI [1.97, 3.64], I2 = 75%). For the five validated comparisons, none
of the groups showed a significant benefit but the pooled result
was significant (RR = 2.01, 95% CI [1.23, 3.28], I2 = 0%) with cessation
rates of 12.4% (test) vs. 6.1% (control), RD 6.3% (Table 1). Biochemical
confirmation was available for one study which was not significant.
At 12 months follow-up, only three comparisons provided
useable data. A benefit was evident for the group EAP + Body
acupuncture vs. no intervention (Supplementary Material 4, 1.4.5)
but there was no benefit for the total pool. For the two validated
comparisons, the same group remained significant and there was a
benefit for the pool (RR = 1.64, 95% CI [1.06, 2.55], I2 = 60%; RD 6.4%)
18 Y.M. Di et al. / Drug and Alcohol Dependence 142 (2014) 14–23

Table 1
Meta-analysis results for the 12 validated studies# in Pools 1 and 2.

Comparisons Time points (EoT, FU) No. comparisons Meta-analysis results (between SC incidence n/N (%) RD % Studies (groups
(participants) groups, risk ratio [95% CI] compared)

Pool 1: EAP/R vs. NS/inactive-control

EAP vs. NS-EAP EoT 2 (257) 1.18 [0.69, 2.04], I2 = 0% T: 23/127 (18.1%) 2.7 Martin and Waite
C: 20/130 (15.4%) (1981) (2 vs. 4) and Wu
et al. (2007)
EAP + electro vs. EoT 1 (46) 13.22 [0.81, 216.26] T: 8/26 (30.8%) 30.8 He et al. (1997)
NS-EAP + electro C: 0/20 (0%)
2
EAPR vs. NS-EAPR EoT 2 (183) 2.55 [1.23, 5.29], I = 0% T: 21/90 (23.3%) 17.7* Li et al. (2009) and
C: 8/93 (8.6%) Zhang et al. (2013a,b)
EAP vs. no EoT 1 (63) 10.66 [2.73, 41.55] T: 22/32 (68.8%) 63.3* Leung (1991) (1 vs. 3)
intervention/wait C: 2/31 (6.5%)
list
EAP + Body Ac vs. EoT 1 (280) 2.14 [1.19, 3.86] T: 30/140 (21.4%) 11.4* Cottraux et al. (1983)
placebo medication C: 14/140 (10.0%) (1 vs. 3)
2
EAP vs. EoT 2 (227) 2.41 [0.94, 6.19], I = 38% T: 12/103 (11.7%) 6.9 Martin and Waite
NonPt-Body Ac C: 6/124 (4.8%) (1981) (1 vs. 5) and
Waite and Clough
(1998)
Active vs. inactive EoT 1 (330) 0.99 [0.62, 1.58] T: 28/160 (17.5%) −0.1 Cai et al. (2000)
machine C: 30/170 (17.6%)

Pool 1: total 1 EoT 10 (1386) 1.77 [1.39, 2.25], I2 = 53% T: 154/678 (22.7%) 10.1*
C: 89/708 (12.6%)
EAP vs. NS-EAP 6 wks-3 months FU 2 (144) 2.00 [0.74, 5.43], I2 = 0% T: 10/72 (13.9%) 7.0 Martin and Waite
C: 5/72 (6.9%) (1981) (2 vs. 4) and
Parker and Mok (1977)
(3 vs. 4)
EAP + electro vs. 6 wks FU 1 (23) 0.36 [0.02, 8.04], I2 = 0% T: 0/11 (0%) −8.3 Parker and Mok (1977)
NS-EAP + electro C: 1/12 (8.3%) (1 vs. 2)
2
EAPR vs. NS-EAPR 3 months FU 2 (183) 2.49 [1.13, 5.49], I = 0% T: 18/90 (20.0%) 12.5* Li et al. (2009) and
C: 7/93 (7.5%) Zhang et al. (2013a,b)
EAP vs. no 3 months FU 1 (63) 7.75 [1.03, 58.38] T: 8/32 (25.0%) 21.8* Leung (1991) (1 vs. 3)
intervention C: 1/31 (3.2%)
EAP + Body Ac vs. 3 months FU 1 (280) 1.57 [0.84, 2.94] T: 22/140 (15.7%) 5.7 Cottraux et al. (1983)
placebo medication C: 14/140 (10.0%) (1 vs. 3)
EAP vs. 3 months FU 1 (149) 1.91 [0.64, 5.75] T: 7/63 (11.1%) 5.3 Martin and Waite
NonPt-Body Ac C: 5/86 (5.8%) (1981) (1 vs. 5)
Active vs. inactive 3 months FU 1 (330) 0.95 [0.58, 1.56] T: 25/160 (15.6%) −0.9 Cai et al. (2000)
machine C: 28/170 (16.5%)

Pool 1: total 2 6 wks-3 months FU 9 (1172) 1.54 [1.14, 2.08], I2 = 11% T: 90/568 (15.8%) 4.7*
C: 61/604 (10.1%)
EAP vs. NS-EAP 6 months FU 1 (126) 2.50 [0.50, 12.41] T: 5/63 (7.9%) 4.7 Martin and Waite
C: 2/63 (3.2%) (1981) (2 vs. 4)
EAP vs. no 6 months FU 1 (63) 3.88 [0.46, 32.77] T: 4/32 (12.5%) 9.2 Leung (1991) (1 vs. 3)
intervention C: 1/31 (3.2%)
EAP + Body Ac vs. 6 months FU 1 (280) 1.62 [0.84, 3.10] T: 21/140 (15.0%) 5.7 Cottraux et al. (1983)
placebo medication C: 13/140 (9.3%) (1 vs. 3)
2
EAP vs. non 6 months FU 2 (227) 2.41 [0.94, 6.19], I = 38% T: 12/103 (11.7%) 6.9 Martin and Waite
Pt-Body Ac C: 6/124 (4.8%) (1981) (1 vs. 5) and
Waite and Clough
(1998)

Pool 1: total 3 6 months FU 5 (696) 2.01 [1.23, 3.28], I2 = 0% T: 42/338 (12.4%) 6.3*
C: 22/358 (6.1%)
EAP + Body Ac vs. 12 months FU 1 (280) 2.56 [1.23, 5.33] T: 23/140 (16.4%) 10.0* Cottraux et al. (1983)
no intervention C: 9/140 (6.4%) (1 vs. 4)
EAP + Body Ac vs. 12 c FU 1 (280) 1.21 [0.69, 2.12] T: 23/140 (16.4%) 2.8 Cottraux et al. (1983)
placebo medication C: 19/140 (13.6%) (1 vs. 3)

Pool 1: total 4 12 months FU 2 (560) 1.64 [1.06, 2.55], I2 = 60% T: 46/280 (16.4%) 6.4*
C: 28/280 (10.0%)

Pool 2: EAP/R vs. non-EAP/R SC therapy

EAP vs. behavioural EoT (valid) 3 (396) 0.95 [0.72, 1.26], I2 = 43% T: 61/200 (30.5%) −1.6 Cottraux et al. (1983)
therapy C: 63/196 (32.1%) (1 vs. 2), Gillams et al.
(1984) (1 vs. 3) and
Leung (1991) (1 vs. 2)
EAP vs. Body Ac EoT (valid) 1 (50) 0.78 [0.34, 1.76] T: 7/25 (30.2%) −5.8 Lamontagne et al.
C: 9/25 (36.0%) (1980) (1 vs. 2)

Pool 2: total 1 EoT (valid) 4 (446) 0.93 [0.71, 1.21], I2 = 22% T: 68/225 (30.2%) −5.1
C: 78/221 (35.3%)
EAP vs. behavioural 3 months FU 3 (396) 2
0.98 [0.64, 1.50], I = 0% T: 35/200 (17.5%) −0.4 Cottraux et al. (1983)
therapy C: 35/196 (17.9%) (1 vs. 2), Gillams et al.
(1984) (1 vs. 3), and
Leung (1991) (1 vs. 2)
 Y.M. Di et al. / Drug and Alcohol Dependence 142 (2014) 14–23 19

Table 1 (Continued)

Comparisons Time points (EoT, FU) No. comparisons Meta-analysis results (between SC incidence n/N (%) RD % Studies (groups
(participants) groups, risk ratio [95% CI] compared)

EAP vs. Body Ac 3 months FU 1 (50) 1.00 [0.33, 3.03] T: 5/25 (20%) 0 Lamontagne et al.
C: 5/25 (20%) (1980) (1 vs. 2)

Pool 2: total 2 3 months FU 4 (446) 0.98 [0.66, 1.46], I2 = 0% T: 40/225 (17.8%) −0.3
C: 40/221 (18.1%)
EAP vs. behavioural 6 months FU 3 (396) 1.09 [0.67, 1.77], I2 = 43% T: 30/200 (15%) 1.2 Cottraux et al. (1983)
therapy C: 27/196 (13.8%) (1 vs. 2), Gillams et al.
(1984) (1 vs. 3), and
Leung (1991) (1 vs. 2)
EAP vs. Body Ac 6 months FU 1 (50) 0.50 [0.10, 2.49] T: 2/25 (8.0%) −8.0 Lamontagne et al.
C: 4/25 (16.0%) (1980) (1 vs. 2)

Pool 2: total 3 6 months FU 4 (446) 1.01 [0.64, 1.61], I2 = 31% T: 32/225 (14.2%) 0.2
C: 31/221 (14.0%)
EAP vs. behavioural 12 months FU 1 (278) 2.27 [1.12, 4.58] T: 23/140 (16.4%) 9.2* Cottraux et al. (1983)
therapy C: 10/138 (7.2%) (1 vs. 2)

Pool 2: total 4 12 months FU 1 (278) 2.27 [1.12, 4.58] T: 23/140 (16.4%) 9.2*
C: 10/138 (7.2%)
#
Validated studies were judged as in employing valid smoking cessation treatment and controls, see Section 3.3.
Ac, acupuncture; Acpr, acupressure; EAP, ear acupuncture; EAPR, ear acupressure; Electro, electro-acupuncture; EoT, end of treatment; FU, follow up; NS, non-specific for
smoking cessation.
*
Significant difference.

but this was based on only one study (Table 1) and there was no
biochemical confirmation.
3.4.2. Pool 2: comparison of EAP/R with other active therapies. In
White et al. (2007), two EAPR groups were used – one used 1 ear pel-
let and the other used 2 pellets. These two groups were analyzed as
one (Supplementary Material 4, 2.1.1). None of the included studies
used biochemical confirmation of SC (see Supplementary Material
1).
At EoT, there was a benefit for the group EAP/R vs. behavioural
therapy/advice (Supplementary Material 4, 2.1.1). There was no dif-
ference between groups for the total pool of eight comparisons but
heterogeneity was 70%. For the four valid comparisons there were
no differences between EAP and behavioural therapy or EAP and
body acupuncture (Table 1). The pool was not significant (Fig. 3).
At three and six months follow-up the total and valid pools were
the same. At three months, there was no difference between EAP
and behavioural therapy or body acupuncture. The total pool was
not significant (Table 1). Cessation rates were 17.5% for EAP vs.
17.9% for the behavioural therapies. At six months follow-up, there
were no significant differences for the groups or the pool.
For the three studies that reported data for 12 months follow-up,
there was no difference between EAP and medical intervention but
the EAP appeared superior to the behavioural interventions (Sup-
plementary Material 4). The pool was significant but heterogeneous
(RR = 2.16, 95% CI [1.45, 3.21], I2 = 86%). However, only Cottraux
et al. (1983) was judged as valid (Table 1). This study showed a
benefit in favour of EAP (RR = 2.27, 95% CI [1.12, 4.58]) with an SC
rate of 16.4% for EAP and 7.2% for relaxation therapy (RD 9.2%).
4. Discussion
Overall, the results suggest that EAP/R provides a benefit in
assisting smoking cessation with absolute SC rates of around
22–30% at EoT falling to 15–18% at 3 months and 12–14% at 6
months (Table 1). However, the overall magnitude of the EAP/R
specific effect was difficult to calculate due to the variety of inter-
ventions and controls. In Pool 1, at 3 months follow-up the EAPR
group showed a 12.5% biochemically confirmed benefit over inac-
tive control but this was based on only two studies and there
were no data for EAPR at 6 months. In Pool 2, EAP/R was neither
superior nor inferior to forms of behavioural therapy at 6 months
follow-up with SC rates of 13.8–15%. This suggests EAP/R could
be a clinical alternative to established SC interventions. However,
this result was based on data without biochemical confirmation,
the behavioural therapies were diverse and none of the trials pro-
vided direct comparisons with nicotine replacement or other SC
pharmacotherapy.
The removal from Pool 1 of studies that did not pass the validity
assessment did not affect the direction of the pooled result at EoT
but the RD was reduced from 13.4% to 10.1% for the valid pool (Sup-
plementary Material 4). When biochemical validation was added
as a criterion, the RD decreased to 8.5% but the pool remained sig-
nificant in favour of the EAP/R groups. Similarly, heterogeneity was
reduced from 72% to 53% and finally 3.9%. This suggests that the pro-
cedures used to narrow the focus of the meta-analyses produced a
more meaningful pool.
Within Pool 1 at EoT, four of the nine groups showed superior-
ity for the test groups (Supplementary Material 4). EAPR was more
effective than controls that used pressure on non-specific points
(NS-EAPR) but when ear acupuncture (EAP) was compared directly
with ‘sham’ ear acupuncture (NS-EAP), there was no difference
between groups. Cessation rates were similar in the test groups
for EAP (24.1%) and EAPR (23.3%) but there were large differences
between the control groups. The SC rate of 21.5% in the NS-EAP
appears very high for a ‘sham’ whereas 8.6% for NS-EAPR seems
more realistic and is not dissimilar the rates for the placebo medi-
cation (10.0%) and no intervention/wait list (12.5%) groups. When
only the validated comparisons were considered, the SC rates for
the EAPR vs. NS-EAPR group did not change while those for EAP vs.
NS-EAP declined but there remained no difference between sham
and control (Fig. 2). A possible explanation is that NS-EAP produces
a greater non-specific effect due to needling the ear than does the
non-invasive NS-EAPR and this accounts for the apparently greater
efficacy of EAPR compared to EAP.
At the 3 and 6 month follow-ups, EAP/R remained superior to
the controls for the total pools and the validated studies but at 12
months there remained too few studies in the pools for the results
to be meaningful (Supplementary Material 4). The group EAPR vs.
NS-EAPR showed significant effects at 3 months in the total and
validated pools and following biochemical confirmation but there
were no data for longer follow-ups (Table 1). In contrast, EAP vs.
NS-EAP showed no significant differences at any time-point in the
total and validated pools although the SC rates remained higher in
20 Y.M. Di et al. / Drug and Alcohol Dependence 142 (2014) 14–23

Fig. 2. EAP/R vs. non-specific/inactive control. Outcome smoking cessation – end of treatment.

the EAP groups. Active vs. inactive machine showed a benefit for
real ear stimulation for time-points from EoT to 6 months FU in the
total pool (Supplementary Material 4) but following the validation
assessment, only one study remained and there was no difference
(Table 1).
In Pool 2 there were no differences between the valid EAP inter-
ventions and the other valid treatments at EoT, 3 months or 6
months (Table 1). A difference appeared at 12 months follow-up
but it was based on one study (Cottraux et al., 1983). SC rates were
high for EAP at EoT (29.9%) but had declined at 3 months (17.8%)
with a further decline at 6 months (14.2%). For the other therapies
the SC rates were slightly higher at EoT (32.6%) and at 3 months
(18.1%) and very similar at 6 months (14.0%).
Two of the studies that compared EAP with behavioural ther-
apy had contradictory results. Cottraux et al. (1983) showed that
in the short term, Schultz’s relaxation procedure had a greater
 Y.M. Di et al. / Drug and Alcohol Dependence 142 (2014) 14–23
Fig. 3. EAP/R vs. non-EAP/R smoking cessation therapies. Outcome smoking cessation – end of treatment.
effect than EAP plus body acupuncture but in the long term, the
acupuncture produced significantly greater abstinence than the
behavioural therapy. On the other hand, Leung (1991) reported that
post-treatment, EAP had a greater effect on SC than a behavioural
programme which involved participants learning techniques for
avoiding smoking, managing withdrawal, maintaining abstinence
and obtaining social support, but at longer terms (1, 3 and 6 months)
the behavioural therapy was better (see Supplementary Material 2
for more detail on the therapies).
4.1. Comparison with other reviews
White et al. (2011) included 33 studies, 22 of which used EAP/R.
Cheng et al. (2012) included 20 studies, of which 17 were of EAP/R.
Tahiri et al. (2012) included 6 studies. The present review encom-
passes all the EAP/R studies in the previous reviews plus three
additional studies. It also conducted sensitivity analyses based on
assessments of the validity of the comparisons and the presence of
a biochemical confirmation of SC.
The results for Pool 1 at EoT are consistent with those reported
in White et al. (2011) and Cheng et al. (2012). Whereas White et al.
(2011) reported no difference between acupuncture and controls
at any of the follow-ups, Cheng et al. (2012) found a benefit at 3
months and 6 months follow-up and Tahiri et al. (2012) reported
a benefit for EAP/R at longest follow-up (6 or 12 months). These
differences in results are due mainly to the approaches to pooling.
As Supplementary Material 4 shows, there is considerable diver-
sity in the types of comparisons used and the sub-groups within
a pool do not necessarily show effects in the same direction as
the total pool. By dividing the studies into separate groups we
increased the specificity of the meta-analyses to illustrate these
effects. The addition of the validity assessment and the sensitivity
analyses reduced heterogeneity and increased the meaningfulness
of the results. Therefore, the results can be interpreted with greater
confidence than could a combined result for all the studies. Nev-
ertheless, studies tend to be small and there are methodological
reporting issues, so the pooled results should be interpreted with
caution.
4.2. Limitations of this review
From a methodological perspective, this review did not follow
all the procedures specified by the Cochrane collaboration. Searches
of trial registries, ‘grey’ literature and hand searches were not per-
formed, data extraction from English and Chinese articles was not
conducted independently by two of the authors, instead, data were
21
extracted by one author and checked against the original arti-
cles independently by a second author. No additional data were
obtained from authors. The criteria used in judging the validity
of controls were stringent, so it could be argued that some of
the studies excluded (e.g., Kerr et al., 2008; Labadie et al., 1983;
Tian and Krishnan, 2006; Lamontagne et al., 1980) used reason-
able approaches to generating controls but the published reports
lacked the details we required. Another approach would have been
to include these as an intermediate step in the sensitivity analysis.
Although the focus of the review is on EAP/R, studies that com-
bined EAP/R with body acupuncture were included. Consequently
the effects in the total pools are not attributable to EAP/R alone.
Nevertheless, studies were grouped so that the effects EAP/R alone
can be isolated (Table 1; Supplementary Material 4).
4.3. Implications for future trials
In designing a control EAP intervention, the following issues
should be considered: 1 – non-specific EAP/R controls should only
use points that are not used for SC in the literature; 2 – the con-
cepts ‘placebo’, and ‘sham’ should not be used to refer to the use
of non-SC-specific points as controls. Books do not list points that
are ‘inactive for SC’, only points for which there is some empirical
evidence of usefulness. So, just because points are not listed for SC,
it does not mean that they are ‘inactive’; 3 – the meaning of the
term ‘point’ should not be misinterpreted as referring to a single
location, like a geometrical point. In the case of EAP, some charts
represent ear points as dots, whereas others specify zones (Jin and
Jin, 2008; Oleson, 2003) – accuracy in location is important but
variation within the confines of clinical practice should be consid-
ered and point locations selected accordingly; and 4 – the type and
degree of stimulation provided to points should be considered care-
fully. It does not necessarily follow that needling is stronger than
pressure, that two points provide a higher dose than one point, or
that ear plus body acupuncture has an additive effect.
Future trials should provide reasons and references regarding
the selection of the SC-specific EAP/R interventions as well as
details on the procedures used that are sufficient to enable replica-
tion of the EAP intervention. Also the actual location of the control
points should be clearly specified to enable the reader to judge the
meaningfulness of the control.
Based on the pooled results at EoT, the addition of electro-
stimulation appears not to have raised the cessation rate, nor did
combining ear plus body points. When used alone, EAPR produced a
similar effect to that of EAP in the SC-specific groups. In contrast, the
EAP control interventions produced relatively large effects when
22 Y.M. Di et al. / Drug and Alcohol Dependence 142 (2014) 14–23
compared to those of the EAPR controls. It cannot be determined
whether this was an effect of the points used in the EAP controls or it
was due the relatively stronger stimulation produced when needles
are used and this led to an elevated non-specific effect. However,
it does suggest that EAPR controls using non-SC-specific points
produce effects close to the expected 10% ‘placebo effect’. There-
fore, future studies could consider using EAPR vs. NS-EAPR since
this design employs a single type of intervention, is replicable, can
be blinded and appears not to produce a large non-specific effect.
Based on two studies (Li et al., 2009; Zhang et al., 2013a,b), the effect
size (risk ratio) estimate for SC was 2.49 (95% CI 1.13–5.49; 7.5% for
the control group) at the 3-month follow-up. Using GPower 3.0.5,
the required sample size would be 138 participants per group for
80% statistical power or 183 per group for 90% statistical power
with a significance level of 5% (2-tailed).
4.4. Implications for practice
Although the mechanism of action of EAP/R interventions was
not elucidated by any of the studies, in terms of benefit relative to
controls, differences were evident at a number of time-points and
for a number of different comparisons suggesting that the effect
was more than simply a non-specific effect of undertaking an inter-
vention. Of the points used, ear Lung and ear Shenmen were typical
of the valid studies, so the best available evidence suggests that
these two points should be employed, a minimalist intervention
using non-invasive EAPR is not inferior or superior to EAP with skin
penetration, and the addition of electrical stimulation appears not
to improve results.
5. Conclusions
When compared to the valid inactive controls, the valid SC-
specific EAP/R interventions were superior at EoT (RD 10.1%), at
three months follow-up (RD 4.7%) and six months follow-up (RD
6.3%) but there was insufficient data at 12 months. Compared to
the SC-specific behavioural interventions, there was no superi-
ority or inferiority for EAP/R at EoT (RD −1.6%), 3 months (RD
−0.4%) or 6 months (RD 1.2%) but there was no comparison with
nicotine replacement or other SC pharmacotherapy. Of the vari-
ous EAP/R interventions and controls, data appear most consistent
for comparisons between EAPR and non-specific EAPR, with SC
rates at 3 months of 20.0% vs. 7.5% but this is based on a rela-
tively small sample. Future trials on EAP/R for SC should follow
rigorous methodology and reporting, use sufficiently large sample
sizes with follow-ups of at least 6 months and employ biochemical
confirmation of SC.
Role of funding source
Funding for this study was provided by Guangdong Provin-
cial Academy of Chinese Medical Sciences (GPACMS), China. The
GPACMS had no further role in study design; in the collection, anal-
ysis and interpretation of data; in the writing of the report; or in
the decision to submit the paper for publication.
Authors’ contributions
AZ and CX conceived of the study. YD and IZ participated in risk
of bias analysis. YD and BM participated in data analysis. YD, BM, AZ,
IZ, CW and CX wrote, reviewed and approved the final manuscript.
Conflict of interest
All authors declare that they have no conflict of interest.
Acknowledgement
We thank Miss A.H. Tan and Mr S. Gianfreda for their time and
language skills for data extraction and analysis of French articles.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.drugalcdep.
2014.07.002.
References
Antoniou, D., Bouros, D., Pavlakou, G., Chaimala, D., Papageorgiou, C., Beltegris, I.,
Dariotaki, F., Chondrou, E., Michalopoulou, P., Gregoratou, T., Gerogianni, A.,
2005. Aural electrical stimulation for smoking cessation – a double blind, placebo
controlled study. Eur. Respir. J. 388.
Bier, I.D., Wilson, J., Studt, P., Shakleton, M., 2002. Auricular acupuncture, education,
and smoking cessation: a randomized, sham-controlled trial. Am. J. Public Health
92, 1642–1647.
Cai, Y., Changxin, Z., Ung, W.S., Lei, Z., Kean, L.S., 2000. Laser acupuncture for ado-
lescent smokers – a randomized double-blind controlled trial. Am. J. Chin. Med.
28, 443–449.
Cheng, H.M., Chung, Y.C., Chen, H.H., Chang, Y.H., Yeh, M.L., 2012. Systematic review
and meta-analysis of the effects of acupoint stimulation on smoking cessation.
Am. J. Chin. Med. 40, 429–442.
Cottraux, J.A., Harf, R., Boissel, J.P., Schbath, J., Bouvard, M., Gillet, J., 1983. Smoking
cessation with behaviour therapy of acupuncture – a controlled study. Behav.
Res. Ther. 21, 417–424.
D’alberto, A., 2004. Auricular acupuncture in the treatment of cocaine/crack abuse: a
review of the efficacy, the use of the National Acupuncture Detoxification Asso-
ciation protocol, and the selection of sham points. J. Altern. Complement. Med.
10, 985–1000.
Fritz, D.J., Carney, R.M., Steinmeyer, B., Ditson, G., Hill, N., Zee-Cheng, J., 2013.
The efficacy of auriculotherapy for smoking cessation: a randomized, placebo-
controlled trial. J. Am. Board Fam. Med. 26, 61–70.
Gilbey, V., 1977. Auricular acupunture for smoking withdrawal. Am. J. Acupunct.,
239–247.
Gillams, J., Lewith, G.T., Machin, D., 1984. Acupuncture and group therapy in stopping
smoking. Practitioner 228, 341–344.
Han, Y., 2006. Combining of acupuncture and auricular points applying for treatment
of 42 cases of tobacco withdrawal syndrome. J. Clin. Acup-Mox (Chin.) 22, 16.
He, D., Berg, J.E., Høstmark, A.T., 1997. Effects of acupuncture on smoking cessation
or reduction for motivated smokers. Prev. Med., 208–214.
Higgins, J., Green, S., 2011. Cochrane Handbook for Systematic Reviews of Interven-
tions Version 5.10. In: The Cochrane Collaboration. Wiley, New York.
Jin, S.Y., Jin, W.C., 2008. A Practical Handbook of Auricular Acupuncture. People’s
Medical Publishing House, China.
Kerr, C.M., Lowe, P.B., Spielholz, N.I., 2008. Low level laser for the stimulation of
acupoints for smoking cessation: a double blind, placebo controlled randomised
trial and semi structured interviews. J. Chin. Med. 86, 46–51.
Labadie, J.C., Dones, J.P., Gachie, J.P., Freour, P., Perchoc, S., Huynh Van Thao, J.P., 1983.
[Desintoxication tabagique: acupuncture et traitement medical]. Gaz. Med. Fr.,
2741–2747.
Lamontagne, Y., Annable, L., Gagnon, M.A., 1980. Acupuncture for smokers: lack of
long-term therapeutic effect in a controlled study. CMAJ 122, 787–790.
Leung, J.P., 1991. Smoking cessation by auricular acupuncture and behavioral ther-
apy. Psychologia 34, 177–187.
Li, Y., Shen, T., Cao, L.Y., Zhang, W., Wu, Q.H., 2009. Clinical research of the seed of
cowherb auricular pressing in treating 70 cases of quitting smoking. J. Tradit.
Chin. Med. 29, 505–506.
Margolin, A., 2003. Acupuncture for substance abuse. Curr. Psychiatry Rep. 5,
333–339.
Martin, G.P., Waite, P.M., 1981. The efficacy of acupuncture as an aid to stopping
smoking. N. Z. Med. J. 93, 421–423.
McLellan, A.T., Grossman, D.S., Blaine, J.D., Haverkos, H.W., 1993. Acupuncture treat-
ment for drug abuse: a technical review. J. Subst. Abuse Treat. 10, 569–576.
Oleson, T., 2003. Auriculotherapy Manual-Chinese and Western Systems of Ear
Acupuncture, Third Edition. Churchhill Livingstone.
Parker, L., Mok, M.S., 1977. The use of acupuncture for smoking withdrawal. Am. J.
Chin. Med. 5, 363–366.
Steiner, R.P., Hay, D.L., Davis, A.W., 1982. Acupuncture therapy for the treatment of
tobacco smoking addiction. Am. J. Chin. Med. 10, 107–121.
Tahiri, M., Mottillo, S., Joseph, L., Pilote, L., Eisenberg, M.J., 2012. Alternative smoking
cessation aids: a meta-analysis of randomized controlled trials. Am. J. Med. 125,
576–584.
Tian, X., Krishnan, S., 2006. Efficacy of auricular acupressure as an adjuvant therapy
in substance abuse treatment: a pilot study. Altern. Ther. Health Med. 12, 66–69.
Waite, N.R., Clough, J.B., 1998. A single-blind, placebo-controlled trial of a sim-
ple acupuncture treatment in the cessation of smoking. Br. J. Gen. Pract. 48,
1487–1490.
 Y.M. Di et al. / Drug and Alcohol Dependence 142 (2014) 14–23
White, A.R., Resch, K.L., Ernst, E., 1998. Randomized trial of acupuncture for nicotine
withdrawal symptoms. Arch. Intern. Med. 158, 2251–2255.
White, A.R., Moody, R.C., Campbell, J.L., 2007. Acupressure for smoking cessation –
a pilot study. BMC Complement. Altern. Med. 7, 8.
White, A.R., Rampes, H., Liu, J.P., Stead, L.F., Campbell, J., 2011. Acupuncture and
related interventions for smoking cessation. Cochrane Database Syst. Rev. 1,
CD000009.
Wing, Y.K., Lee, A., Wong, E.L.Y., Leung, P.C., Zhang, L., Pang, E.S.Y., 2010. Auricular
acupressure for smoking cessation: a pilot randomized controlled trial. Med.
Acupunct. 22, 265–271.
Wu, T.P., Chen, F.P., Liu, J.Y., Lin, M.H., Hwang, S.J., 2007. A randomized controlled
clinical trial of auricular acupuncture in smoking cessation. J. Chin. Med. Assoc.
70, 331–338.
23
Yeh, M.L., Chang, C.Y., Chu, N.F., Chen, H.H., 2009. A six-week acupoint stimulation
intervention for quitting smoking. Am. J. Chin. Med. 37, 829–836.
Zhang, A.L., Di, Y.M., Worsnop, C., May, B.H., Xue, C.C.L., 2013a. Ear-acupressure for
smoking cessation: study protocol for an RCT. Forsch. Komplementarmed. 20,
290–294.
Zhang, A.L., Di, Y.M., Worsnop, C., May, B.H., Da Costa, C., Xue, C.C., 2013b. Ear-
acupressure for smoking cessation: a randomised controlled trial. Evid. Based
Complement. Alternat. Med., http://dx.doi.org/10.1155/2013/637073.
Zhang, C.S., Yang, A.W., Zhang, A.L., May, B.H., Xue, C.C., 2014. Sham control meth-
ods used in ear-acupuncture/ear-acupressure randomized controlled trials: a
systematic review. J. Altern. Complement. Med. 20, 147–161.