Hindawi

Pain Research and Management

Volume 2020, Article ID 3825617, 14 pages
https://doi.org/10.1155/2020/3825617

Review Article
Effectiveness and Safety of Acupuncture for Migraine: An
Overview of Systematic Reviews

Yu-Xi Li,1 Xi-li Xiao,2 Dong-Ling Zhong ,3 Liao-Jun Luo,1 Han Yang,1 Jun Zhou,1
Ming-Xing He,3 Li-Hong Shi,3 Juan Li ,3 Hui Zheng ,1 and Rong-Jiang Jin 3
1
School of Acupuncture-Moxibustion and Tuina/The Third Affiliated Hospital,
Chengdu University of Traditional Chinese Medicine, Sichuan, China
2
Hospital of Chengdu University of Traditional Chinese Medicine, Sichuan, China
3
School of Health Preservation and Rehabilitation, Chengdu University of Traditional Chinese Medicine, Sichuan, China

Correspondence should be addressed to Juan Li; [email protected], Hui Zheng; [email protected], and Rong-Jiang Jin;
[email protected]

Received 5 November 2019; Revised 22 February 2020; Accepted 25 February 2020; Published 23 March 2020

Academic Editor: Federica Galli

Copyright © 2020 Yu-Xi Li et al. This is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background. Migraine is a common neurological disease, which burdens individuals and society all over the world. Acupuncture,
an important method in Traditional Chinese Medicine, is widely used in clinical practice as a treatment for migraine. Several
systematic reviews (SRs) have investigated the effectiveness and safety of acupuncture for migraine. Objective. To summarize and
critically assess the quality of relevant SRs and present an objective and comprehensive evidence on the effectiveness and safety of
acupuncture for migraine. Data Sources. MEDLINE, Embase, Cochrane Library, PROSPERO database, Chinese National
Knowledge Infrastructure (CNKI), Chinese Biological Medicine (CBM), China Science and Technology Journal (SCTJ), and
WanFang database (WF) were searched from inception to December 2019 and grey literatures were manually searched. Selection
Criteria. SRs which meet the criteria were independently selected by 2 reviewers according to a predetermined protocol. Data
Extraction. Characteristics of included SRs were independently extracted by 2 reviewers following a predefined data extraction
form. Review Appraisal. The methodological quality, risk of bias, and reporting quality of included SRs were assessed, respectively,
by a Measurement Tool to Assess Systematic Reviews (AMSTAR) 2, the Risk of Bias in Systematic reviews (ROBIS) tool, and the
Preferred Reporting Item for Systematic Review and Meta-analysis-Acupuncture (PRISMA-A) statement. The quality of out-
comes was evaluated by the Grading of Recommendations Assessment, Development, and Evaluation (GRADE). Results. A total
of 15 SRs were included. All the SRs were published between 2011–2019. Based on AMSTAR 2, 14 out of 15 SRs were rated
critically low quality and 1 was rated low quality. According to ROBIS tool, 9 SRs (60%) were low risk of bias. With the PRISMA-A
checklist, we found 11 out of 15 SRs were found adequately reported over 70%. With the GRADE tool, we found high quality of
evidence indicated that the effective rate of acupuncture was superior to western medicine in treatment of migraine. Besides,
acupuncture reduced more headache days and the times of using painkiller and was more effective in reducing the frequency and
degree of headache than western medicine and sham acupuncture. Limitations. There might be some missing information. The
accuracy of the conclusions may be decreased reduced since we were unable to synthesis all the evidence. Conclusions. Based on
high quality of evidence, we concluded that acupuncture may be an effective and safe therapy for migraine. However, the quality of
SRs in acupuncture for migraine still needs more improvement.

1. Introduction the epidemiological statistics, the prevalence of migraine is

3.3%∼32.6% in female and 0.7%∼16.1% in male [2]. At the
Migraine is a common neurological disease characterized by same time, it costs between 6.5 and 17 billion dollars annually
unilateral, throbbing recurrent headache, often accompanied in the USA which severely burdened individuals and society
by photophobia, phonophobia, or nausea [1]. According to [3, 4]. In the Global Burden of Disease Survey 2010, migraine
2 Pain Research and Management

ranked as the 3rd most prevalent disorder and 7th highest 2. Methods
specific cause of disability worldwide [5, 6]. The standard
treatments for migraine include nonsteroidal anti-inflam- 2.1. Registration. A predetermined written protocol of this
matory drugs, antiepileptic drugs, ergotamines, and triptans. overview was registered in the PROSPERO (International
However, these pharmacotherapies were often accompanied prospective register of systematic overview) database (https://
with undesirable adverse effects [7], such as fatigue, sleep www.crd.york.ac.uk/PROSPERO/), registration number:
disturbance, nausea, and vomiting [8] which lead to poor CRD42017077218. This overview was reported in accordance
compliance of patients. Therefore, more and more patients with the guideline of the pilot version checklist with Preferred
are seeking for effective nonpharmacological alternative Reporting Items for overview of systematic reviews (PRIO-
treatments. harms) [19].
Acupuncture, an important method in Traditional
Chinese Medicine, is widely used in clinical practice as a
treatment for migraine. It is reported that acupuncture 2.2. Ethics. Ethics approval is not required in overview of
was one of the most common complementary therapies in SRs, since it does not involve individual patient data.
worldwide [9]. In German-speaking countries, acu-
puncture has high utilization and is one of the most
primarily used methods to relieve pain [10]. Now, it is 2.3. Inclusion Criteria
increasingly accepted in western countries as an alter-
2.3.1. Types of Reviews. SRs with or without meta-analysis of
native treatment for migraine and other pain conditions
randomized controlled trials (RCTs) were included in which
[11]. The results of clinical studies have demonstrated that
acupuncture was used as treatment for migraine.
acupuncture is an effective and safe therapeutic approach
to treat migraine [12–14]. With the development of evi-
dence-based medicine, numerous systematic reviews 2.3.2. Types of Participants. SRs included RCT recruiting
(SRs) have been conducted to investigate the effectiveness participants diagnosed with migraine according to standard
and safety of acupuncture for migraine. However, the diagnostic criteria (e.g., the International Classification of
results of these SRs often have limitations which may Headache Disorders released by the International Headache
lower the quality of conclusions and mislead the patients, Society or other domestic standards). There was no re-
clinical doctors, and policy makers. striction on the gender, age, race, duration, intensity, con-
Overviews of SRs, a method to evaluate the quality of dition, and source of the patients.
evidence [15], is becoming more and more prevalent in
evidence-based medicine [16]. The overview is a compre-
hensive approach to reassess the quality of SRs by collecting 2.3.3. Types of Interventions. There was no restriction on the
the information of relevant SRs dealing with the same types of acupuncture (e.g., body acupuncture, electro-
disease or health problem [17]. While SR has always been acupuncture, auricular acupuncture, warm-acupuncture,
regarded as one of the most important sources of high and scalp acupuncture).
quality and reliable information in the evidence-based
medicine [18], there are many factors in the evaluation
process which can decrease the quality of SRs, such as 2.3.4. Types of Comparators. SRs included control groups
incomprehensive source of literature, inadequate evaluation which were treated with sham-acupuncture, placebo,
method, and publication bias. Overview of SRs compre- medicine, and other types of nonpharmaceutical therapy or
hensively integrated the evidence of SRs, which contains placed in the waiting list.
more information and can provide more high-quality evi-
dence for clinical work.
This is the first overview which comprehensively 2.3.5. Types of Outcomes. The primary outcome was effective
assessed SRs of acupuncture for migraine with a Mea- rate. Secondary outcomes included intensity, frequency or
surement Tool to Assess Systematic Reviews (AMSTAR) duration of headache, times of using painkiller, quality of
2, the Risk of Bias in Systematic reviews (ROBIS), Pre- life, recurrent rate, and adverse effects of acupuncture in
ferred Reporting Item for Systematic Review and Meta- migraine.
analysis-Acupuncture (PRISMA-A), and the Grades of
Recommendations, Assessment, Development and Eval-
uation (GRADE). The objective of this overview is to 2.4. Exclusion Criteria. The SRs were excluded if one of the
critically assess the quality of relevant SRs and present an following criteria was met: did not use the diagnostic criteria
objective and comprehensive evaluation on effectiveness of migraine mentioned above; SRs with network meta-
and safety of acupuncture for migraine, which can help analysis or indirect comparison; SRs that included retro-
the public and policy-makers understand whether acu- spective studies, prospective studies, cross-sectional clinical
puncture should be recommended as a treatment for studies, and case reports; SRs whose data could not be
migraine. extracted; duplicated publication; review comments.
Pain Research and Management 3

2.5. Search Strategy. An electronic literature search was and selection of studies,” “data collection and study
conducted in the MEDLINE, Embase, Cochrane Library, appraisal,” and “synthesis and findings”. Each do-
PROSPERO database, Chinese National Knowledge Infra- main has signaling questions and a judgment of
structure (CNKI), Chinese Biological Medicine (CBM) concerns about risk of bias of the domain, and the
database, Chinese Science and Technology Periodical Da- results are rated as “high risk,” “low risk,” or
tabase (SCTJ), and WanFang database, all from the incep- “unclear risk” [21].
tion to December 2019. Details of search strategy were (3) PRISMA-A statements an extension of PRISMA
presented in Supplementary Table S1. In addition, reference especially for acupuncture, which was published in
lists/bibliographies of included studies, study registries, and 2019 [22]. It consists of a 27-item checklist and a 4-
grey literature, such as dissertations and conference reports, phase flow diagram, aiming to help authors improve
were also searched to avoid missing studies. Besides, the the reporting quality of SRs on acupuncture inter-
experts in the field were also consulted. No language re- ventions. Seven aspects of SRs include title, abstract,
strictions were applied. introduction, methods, results, discussion and
funding. Response options for each item are “yes,”
“no,” and “not applicable”. The completion of each
2.6. Screening. The reviewer (JZ) searched the databases
item was presented as a ratio.
according to the predeveloped standardized search strategy.
All the retrieved literatures were imported into Endnote X8. (4) The quality of primary outcomes of included SRs was
Two reviewers (HY and YXL) independently screened for evaluated by the GRADE system [23]. The assessment
candidates according to the inclusion and exclusion criteria of included SRs was carried out independently by
by reading the title and abstract. Then, the full texts were qualified reviewers (JL and DLZ) who were trained in
downloaded for further screening. At the same time, bib- the GRADE Center in China (Lanzhou). The 5 key
liographic references were also reviewed to identify possible elements of GRADE influenced the quality of evidence
SRs. The disagreements were resolved by discussion. If including study limitations, inconsistency of results,
necessary, the discrepancies were resolved by consulting the indirectness of evidence, imprecision, and reporting
third reviewer (DLZ). bias. The quality of evidences of SRs was rated as
“High,” “Moderate,” “Low,” and “Very Low”. Evidence
based on RCTs began as high quality.
2.7. Data Extraction. A data extracted form was predefined,
including the characteristics of SRs, such as author, title, 3. Results
published year, sample size, intervention, outcome indica-
tors, quality evaluation method, and conclusion. Data was 3.1. Literature Search. We retrieved 457 records according to
independently extracted by two reviewers (HY and LJL) the search strategy. 11 duplicates were excluded by filtration,
using Microsoft Excel. After extraction, the two reviewers 445 papers were screened by titles and abstracts. 65 articles
(LHS and YXL) cross checked to eliminate mis-entry. were considered eligible, and full-text papers were down-
Discrepancies were resolved by team discussion or arbitrated loaded. After being reviewed by two reviewers indepen-
by the third reviewer (DLZ). dently, 50 SRs were excluded and 15 SRs [24–38] were
included for further analyses (Figure 1). The reasons for
exclusion are presented in Supplementary Table S2.
2.8. Assessment of SRs. The assessment of included SRs was
carried out independently by qualified reviewers who were
trained in the Chinese Cochrane Center. Before the evalu- 3.2. Characteristics of SRs. The characteristics of included
ation, each topic of the assessment tools was intensively SRs are presented in Table 1. All the included SRs were
discussed to achieve consensus. After evaluation, two re- published between 2011–2019, 6 of which were published in
viewers cross checked the results. Discrepancies were re- 2016 [21, 31–34, 36]. The number of RCTs in SRs ranged
solved by team discussion or an independent decision form a from 2 to 33. 3 SRs were on prophylactic treatment for
third reviewer. migraine [31, 33, 35], 1 SR on acute migraine [33], 1 SR for
menstrual migraine [36], 1 SR for migraine without aura
(1) AMSTAR 2 [20] was used to assess the methodo- [37], and the others did not clearly stated the type of mi-
logical quality of included SRs. The checklist has 16 graine. 11 SRs specified the diagnostic criteria of HIS (In-
items, including 7 critical items (items 2, 4, 7, 9, 11, ternational Headache Society) or ICHD (International
13, and 15), which are used to critically assess the Classification of Headache Disorders), while 4 SRs
validity of an SR. Each item was evaluated as “yes” (a [25, 27, 28, 32] did not report the diagnostic criteria. All the
positive result), “partial yes” (partial adherence to the 15 SRs performed meta-analysis, 10 out of 15 SRs
standard), and “no” (no information is provided to [24–29, 31, 32, 35, 38] performed subgroup analysis, and
rate an item) according to adherence to the standard. only 4 SRs [30, 33, 34, 36, 38] conducted sensitivity analysis.
(2) The aim of the ROBIS tool is to evaluate the level of The intervention was acupuncture, while comparators were
bias presented in a systematic review. This tool mainly sham acupuncture and medications (ergotamine,
assesses the level of bias across 4 domains of 2 ibuprofen, flunarizine, nimodipine, celecoxib, aspirin,
phases: “study eligibility criteria,” “identification somedon, sodium valproate, metoprolol, and topiramate).
4 Pain Research and Management

Records identified through

database searching

Identification
(n = 457):
Cochrane library (n = 82); Medline
(n = 124); EMBASE (n = 176); CBM Additional records identified
(n = 10); CNKI (n =19); VIP (n = 8); through other sources
Wanfang (n = 38) (n = 0)
Screening

Records after duplicates removed

(n = 445)

Records screened Records excluded

(n = 445) (n = 380)
Eligibility

Full-text articles excluded

with reasons (n = 50):
Full-text articles assessed
not an SR (n = 19); no
for eligibility
separate data of
(n = 65)
acupuncture for migraine
(n = 4); migraine was not
the main research object
(n = 8); acupuncture was
Included

not main intervention

Studies included in (n = 10); commentary
qualitative synthesis (n = 2); updated SR
(n = 15) (n = 3) narrative SR
(n = 2); comparison of
different acupuncure
methods (n =1);
insufficient data (n =1)

Figure 1: Flowchart of the selection process of included SRs.

The outcomes of SRs were effective rate, intensity, and 3.4. Risk of Bias of Included SRs. The ROBIS tool, containing
frequency or duration of headache. For the assessment of 3 phases with 4 domains, was used to assess the risk of bias of
methodological quality, 2 SRs [24, 29] used the Jadad included SRs. Phase 1 assesses the relevance of research
scale, 12 SRs [25, 26, 30–34, 36, 38] used the Cochrane risk question, which is optional and was not performed in our
of bias tool, and the remaining 1 SR [27] did not report study. Table 3 and Figure 2 present the assessment of risk of
any specific tool but described 6 aspects of quality as- bias of each SR. Domain 1 assessed concerns regarding
sessment, including randomization, allocation conceal- specification of study eligibility criteria, and 12 of 15 SRs
ment, blind method, data integrity, selective reporting, (80%) were rated low risk of bias. Domain 2 assessed
and other biases. concerns regarding methods used to identify and select
studies, in which 9 SRs (60%) were in low risk of bias.
Domain 3 assessed concerns regarding methods used to
3.3. Methodological Quality of Included SRs. An overview of collect data and appraise studies, and 11 SRs (73%) were at
methodological quality of included SRs is presented in low risk of bias and 1 [24] unclear risk of bias. Domain 4
Table 2. Among the 15 SRs, 14 were rated critically low assessed concerns regarding the synthesis and findings, and
quality and 1 was rated low quality [37]. Items 2, 3, 7, 10, and 8 SRs (53%) were rated as low risk of bias. The final phase
16 were rated particularly low quality. All SRs used satis- considered the overall risk of bias of SRs, and 9 SRs (60%)
factory techniques to assess the risk of bias. Only 1 SR [37] were low risk of bias.
established a prior study protocol and 2 [28, 36] reported the
funding sources of the included studies. No SR explained the
reasons for selection of study types or provided a complete 3.5. Reporting Quality of Included SRs. Table 4 presents the
list of excluded studies with reasons. And, few SRs assessed overview of PRISMA-A checklist items. 11 out of 15 SRs
publication bias by a funnel plot. were adequately reported over 70%. The section of title,
 Table 1: Characteristics of included SRs.
First author Included No. Assessment of Meta- Subgroup Sensitivity
No. of Diagnostic
and year of study of Intervention Comparator Outcomes methodological analysis analysis analysis Safety
patient criteria
publication design study quality conducted? conducted? conducted?
IHS; Sham Effective rate;
Gao (2011) RCT 12 1744 Acupuncture Jadad scale Yes Yes No NR
ICHD-10 acupuncture headache days
Herbs; sham Cochrane risk of
Zheng (2012) RCT 33 3593 NR Acupuncture Effective rate Yes Yes No NR
acupuncture bias tool
Medications Effective rate;
(flunarizine, headache
ibuprofen, frequency;
Cochrane risk of
Pain Research and Management

Chen (2014) RCT 18 1672 IHS Acupuncture nimodipine, headache intensity Yes Yes Yes Yes
bias tool
naproxen, and duration;
ergotamine accompanying
caffeine); herbs symptoms
Short-term effect;
Yang (2014) RCT 10 893 NR Acupuncture Flunarizine long-term effect; NR Yes No No Yes
headache score
Cochrane risk of
Zhao (2014) RCT 17 1719 NR EA Medications Effective rate; VAS Yes Yes No Yes
bias tool
Medications
Cure rate; effective
Dai (2015) RCT 2 140 IHS Acupuncture (flunarizine, Jadad scale Yes No No NR
rate
nimodipine)
Effective rate;
headache intensity;
headache
Sham frequency; Cochrane risk of
Yang (2015) RCT 10 997 IHS Acupuncture Yes Yes No Yes
acupuncture headache duration; bias tool
medication use;
accompanying
symptoms
Sham
interventions;
Headache
medications
frequency; the Cochrane risk of
Linde (2016) RCT 22 4985 IHS Acupuncture (metoprolol, Yes Yes No NR
proportion of bias tool
flunarizine,
“responders”
valproic acid);
waiting list
Sham Cochrane risk of
Pu J (2016) RCT 5 618 NR Acupuncture VAS Yes No No No
acupuncture bias tool
Medications Effective rate;
(flunarizine, headache
Cochrane risk of
Pu (2016) RCT 7 1285 IHS Acupuncture valproate, frequency; Yes No Yes Yes
bias tool
topiramate, headache intensity;
metoprolol) medication use
5
 6

Table 1: Continued.
First author Included No. Assessment of Meta- Subgroup Sensitivity
No. of Diagnostic
and year of study of Intervention Comparator Outcomes methodological analysis analysis analysis Safety
patient criteria
publication design study quality conducted? conducted? conducted?
Medications
(flunarizine, Long-term effective
Acupuncture; ibuprofen, rate; short-term Cochrane risk of
Song (2016) RCT 18 1470 IHS; ICHD Yes Yes Yes Yes
EA nimodipine, effective rate; bias tool
metoprolol); headache days
placebo
Sham
acupuncture; Effective rate;
medications headache
(flunarizine, frequency; VAS; Cochrane risk of
Xian (2016) RCT 26 3657 IHS; ICHD Acupuncture Yes No Yes
sodium valproate, medication use; bias tool
metoprolol, PDI; MIDAS; PF-
topiramate, SF36; MH-SF36
nimodipine)
Medications
Acupuncture;
(celecoxib, Effective rate; VAS;
auricular Cochrane risk of
Zhao (2016) RCT 18 1268 IHS flunarizine, headache intensity; Yes Yes Yes
acupuncture; bias tool
aspirin, ibuprofen, headache frequency
EA
somedon); herbs
Medications
Frequency of
(ergotamine,
migraine attacks;
Acupuncture; ibuprofen, Cochrane risk of
Xu (2018) RCT 14 1155 ICHD number of migraine Yes No No Yes
EA flunarizine, bias tool
days; VAS; effective
nimodipine);
rate
sham acupuncture
Sham
Frequency of
acupuncture;
migraine attacks;
Acupuncture; medications Cochrane risk of
Lu (2019) RCT 17 2226 ICHD-3 duration of Yes Yes Yes NR
EA (ergotamine, bias tool
migraine; headache
flunarizine,
intensity
nimodipine)
IHS � International Headache Society; ICHD � International Classification of Headache Disorders; NR � no report; AA � auricular acupuncture; EA � electro-acupuncture; acupuncture � classical manual
acupuncture; VAS � Visual Analogue Scale/Score; PDI � Pain Disability Index; MIDAS � the Migraine Disability Assessment; SF-36/12 � Short Form 36/12 Questionnaire.
Pain Research and Management
Pain Research and Management 7

Table 2: Methodological quality of included SRs on acupuncture for migraine.

Author (year) I1 I2∗ I3 I4∗ I5 I6 I7∗ I8 I9∗ I10 I11∗ I12 I13∗ I14 I15∗ I16 Ranking of quality
Gao (2011) Y N N PY N N N Y Y N N Y Y N Y N Critically low
Zheng (2012) Y N N PY Y Y N N Y N N Y Y N N N Critically low
Chen (2014) Y N N PY Y Y N PY Y N Y Y Y N Y N Critically low
Yang (2014) Y N N PY Y N N PY Y N N N N N N N Critically low
Zhao (2014) Y N N PY Y Y N PY Y Y Y N N N N N Critically low
Dai (2015) N N N PY N N N N Y N N N N Y N N Critically low
Yang (2015) Y N N PY N Y N PY Y N Y Y Y Y N N Critically low
Linde (2016) Y N N PY Y Y N PY Y N Y Y Y N N N Critically low
PuJ (2016) Y N N PY Y Y N N Y N Y N N Y N N Critically low
Pu (2016) Y N N PY Y Y N Y Y N Y Y N Y N N Critically low
Song (2016) Y N N N Y Y N PY Y N Y Y N Y Y N Critically low
Xian (2016) Y N N PY Y Y N Y Y N Y N N Y N N Critically low
Zhao (2016) Y N N PY Y Y N Y Y Y Y Y Y N Y N Critically low
Xu (2018) Y Y N PY Y Y N Y Y N Y N Y Y Y Y Low
Lu (2019) Y N N PY Y Y N Y Y N Y N Y Y Y N Critically low
∗
The key items of the AMSTAR 2; I: item; Y: yes; N: no; PY: partial yes. Item 1: did the research questions and inclusion criteria for the review include the
components of PICO? Item 2: did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the
review and did the report justify any significant deviations from the protocol? Item 3: did the review authors explain their selection of the study designs for
inclusion in the review? Item 4: did the review authors use a comprehensive literature search strategy? Item 5: did the review authors perform study selection
in duplicate? Item 6: did the review authors perform data extraction in duplicate? Item 7: did the review authors provide a list of excluded studies and justify
the exclusions? Item 8: did the review authors describe the included studies in adequate detail? Item 9: did the review authors use a satisfactory technique for
assessing the risk of bias (RoB) in individual studies that were included in the review? Item 10: did the review authors report on the sources of funding for the
studies included in the review? Item 11: if meta-analysis was performed did the review authors use appropriate methods for statistical combination of results?
Item 12: if meta-analysis was performed, did the review authors assess the potential impact of RoB in individual studies on the results of the meta-analysis or
other evidence synthesis? Item 13: did the review authors account for RoB in individual studies when interpreting/discussing the results of the review? Item
14: did the review authors provide a satisfactory explanation for, and discussion of, any heterogeneity observed in the results of the review? Item 15: if they
performed quantitative synthesis, did the review authors carry out an adequate investigation of publication bias (small study bias) and discuss its likely impact
on the results of the review? Item 16: did the review authors report any potential sources of conflicts of interest, including any funding they received for
conducting the review?

abstract, and introduction were all well reported (100%). [31, 35]. One SR [25] reported more effective rate of acu-
Though in Section 2, topic of protocol and registration, search puncture than Chinese herbal medicine in treating migraine
strategy, study selection, data items, risk of bias, and additional (RR = 1.29, 95% CI = (1.14, 1.45), P < 0.00001).
analyses were reported inadequately, three topics (study se-
lection, risk of bias across studies, and additional analysis) in
Section 3 were reported under 70%. Of all the items, protocol 3.7. Evidence Quality of Included SRs. We evaluated the
and registration (13.33%), search strategy (33.33%), risk of bias quality of primary outcomes extracted from included studies.
in individual studies (33.33%), and risk of bias across studies Table 6 shows the level of evidence quality of studies reported
(33.33%) accounted for the main reporting limitations. effective rate. The high level of evidence quality indicated that
Overall, 4 SRs [26, 36–38] reached over 85% compliance. the effective rate of acupuncture was superior than western
medicine, both in short-term and long-term. The inconsis-
tency and imprecision were the main reasons for down-
grading. Significant heterogeneity downgraded inconsistency
3.6. Effectiveness of Acupuncture for Migraine. We summa- and imprecision was downgraded because the total sample
rized the outcomes from the included SRs and presented size did not meet the optimal information size.
them in Table 5. The evidence suggested that the effective
rate of acupuncture was superior to western medicine (risk
ratio (RR) = 1.17, 95% confidence interval (CI) = (1.12, 1.22), 3.8. Safety of Acupuncture for Migraine. Of all the 15 SRs, 8
P � 0.71) [26]. Acupuncture had better long-term effective SRs [26, 30, 31, 33–37] mentioned the adverse events of acu-
rate for migraine (RR = 4.17, 95% CI (2.80, 6. 20), puncture in the treatment of migraine. 2 SRs [30, 36] did not
P < 0.00001) [34, 35] and reduced more headache days further analyze the safety evaluation due to the small number of
(standardized mean difference (SMD) = −0.13, 95% studies. 6 SRs [26, 31, 33–35, 37] concluded that acupuncture
CI = (−0.25, −0.02), P � 0.02) and the times of using treatment had fewer adverse events than medication, which
painkiller (SMD = −0.73, 95% CI = (−2.14, 0.69), P � 0.31) indicated that acupuncture was a safe therapy for migraine.
than western medicine and sham acupuncture, both in
short-term and long-term follow-up [33, 35]. Besides, 4. Discussion
acupuncture was more effective in reducing frequency
(SMD = −2.18, 95% CI = (−2.61, −1.75), P < 0.00001) and 4.1. Summary of Main Findings. This is the first overview of
degree of headache (SMD = −1.93, 95% CI = (−2.53, −1.36), SRs that investigate the effectiveness and safety of acu-
P � 0.005) than western medicine and sham acupuncture puncture for migraine. We rigorously appraised the
8 Pain Research and Management

Table 3: Tabular presentation of risk of bias of included SRs.

Phase 2 Phase 3
Review 1. Study eligibility 2. Identification and selection 3. Data collection and 4. Synthesis and Risk of bias in the
criteria of studies study appraisal findings review
Gao (2011) ☺ ☹ ? ☺ ☹
Zheng (2012) ☺ ☺ ☺ ☹ ☹
Chen (2014) ☺ ☺ ☺ ☺ ☺
Yang (2014) ☺ ☹ ☹ ☹ ☹
Zhao (2014) ☹ ☹ ☺ ☺ ☹
Dai (2015) ☺ ☹ ☹ ☹ ☹
Yang (2015) ☺ ☺ ☺ ☺ ☺
Linde (2016) ☺ ☺ ☺ ☺ ☺
PuJ (2016) ☹ ☹ ☺ ☹ ☹
Pu (2016) ☺ ☺ ☺ ☺ ☺
Song (2016) ☺ ☺ ☺ ☺ ☺
Xian (2016) ☺ ☺ ☺ ☹ ☺
Zhao (2016) ☺ ☺ ☺ ☺ ☺
Xu (2018) ☺ ☺ ☺ ☹ ☺
Lu (2019) ☺ ☺ ☺ ☹ ☺
☺ � low risk; ☹ � high risk; ? � unclear risk.

Risk of bias in the review

4. Synthesis and findings

3. Data collection and study

appraisal

2. Identification and selection of

studies

1. Study eligibility criteria

0 20 40 60 80 100
(%)

High Darker colours

Low indicate overall ROB
Unclear rating; lighter colours
concern judgments
Figure 2: Graphical presentation of risk of bias of included SRs.

published SRs with AMSTAR 2, ROBIS, PRISMA-A, and identification and selection of studies. The results indicated
GRADE. Based on AMSTAR 2, 14 out of 15 SRs were rated that the reviewers of SR should pay attention to whether the
critically low quality and 1 was rated low quality. By using search includes an appropriate range of databases or elec-
the ROBIS tool, 9 SRs were rated low risk bias. With tronic sources for published reports. Instead of database
PRISMA-A checklist, we found 11 out of 15 SRs were found searching, the additional methods should also be used to
adequately reported over 70%. The results of GRADE identify relevant reports, including conference reports and
suggested that acupuncture was is an effective and safe clinical trial registration platforms. In domain 4, the risk of
method for migraine. bias in synthesis of findings was high. Even though the data
was synthesized in all the SRs, we were not able to determine
whether data synthesis and analysis methods have been
4.2. Implications for Further Study. This overview presents followed in advance, which may ignore the results of some
several challenges for producers of SRs that should be studies. The robustness of the findings should be assessed
considered. By using the ROBIS tool, we found that the risk through funnel plot or sensitivity analyses, and the biases in
of bias in domain 2 and domain 4 of phase 2 were relatively primary studies should be minimized or addressed in the
high. In domain 2, we focused on the risk of bias in synthesis.
 Table 4: Compliance of included SRs with PRISMA-A checklist.
Gao Zheng Chen Yang Zhao Dai Yang Linde PuJ Pu Song Xian Zhao Xu Lu Compliance
Section/topic
(2011) (2012) (2014) (2014) (2014) (2015) (2015) (2016) (2016) (2016) (2016) (2016) (2016) (2018) (2019) (%)
1 Title Title Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y 100.00
Structured
2 Abstract Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y 100.00
summary
3 Introduction Rationale Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y 100.00
4 Objectives Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y 100.00
Protocol and
5 Methods N N N N N N N Y N N N N N Y N 13.33
registration
Eligibility
Pain Research and Management

6 Y N Y Y Y N Y Y Y Y Y Y Y Y Y 86.67
criteria
Information
7 Y Y Y N Y Y Y N N Y Y Y Y Y Y 80.00
sources
8 Search N N Y N N N Y Y N N N N Y N Y 33.33
9 Study selection Y N Y Y N N N Y N Y N Y N Y Y 53.33
Data collection
10 N Y Y Y Y N Y Y Y Y Y N Y Y Y 80.00
process
11 Data items N N Y Y N N Y Y N Y N Y Y Y Y 60
Risk of bias in
12 individual Y Y Y Y Y N Y Y Y Y Y Y Y Y Y 93.33
studies
Summary
13 Y Y Y Y N Y Y Y Y Y Y Y Y Y Y 93.33
measures
Synthesis of
14 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y 100.00
results
Risk of bias
15 Y N Y N N N N N N N Y N Y Y N 33.33
across studies
Additional
16 Y Y Y N N N Y Y N Y Y Y Y Y Y 73.33
analyses
17 Results Study selection Y N Y Y N N N Y N Y Y N Y N N 53.33
Study
18 Y N Y Y Y N Y Y Y Y Y Y Y Y Y 86.67
characteristics
Risk of bias
19 Y Y Y Y Y N Y Y Y Y Y Y Y Y Y 93.33
within studies
Results of
20 individual Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y 100.00
studies
Synthesis of
21 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y 100.00
results
Risk of bias
22 Y N Y N N N N N N N Y N Y Y N 33.33
across studies
Additional
23 Y N Y N N N Y Y N Y N N Y N Y 46.67
analysis
9
 10

Table 4: Continued.
Gao Zheng Chen Yang Zhao Dai Yang Linde PuJ Pu Song Xian Zhao Xu Lu Compliance
Section/topic
(2011) (2012) (2014) (2014) (2014) (2015) (2015) (2016) (2016) (2016) (2016) (2016) (2016) (2018) (2019) (%)
Summary of
24 Discussion Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y 100.00
evidence
25 Limitations N Y Y Y Y N Y Y N Y Y Y Y Y Y 86.67
26 Conclusions Y Y Y Y Y N Y Y Y Y Y Y Y Y Y 93.33
27 Funding Funding N N N Y N N N Y Y N Y N Y Y Y 46.67
77.78% 59.26% 92.59% 74.07% 59.26% 37.03% 77.78% 88.89% 59.26% 81.48% 81.48% 70.37% 92.59% 88.89% 88.89%
Pain Research and Management
Pain Research and Management 11

Table 5: Summary of evidence.

Author Outcomes (total patient number in the intervention group/total patient number in the control group or total participants in
(year) both groups, number of studies)
Acupuncture vs sham-acupuncture
Gao (2011) Effective rate (OR � 1.28, 95% C (1.02, 1.61), P � 0.03) (650/603, 8)
Zheng (2012) Effective rate (RR � 1.87, 95% CI (1.17, 2.98), P � 0.009) (91/54, 3)
Effective rate (RR � 1.19, 95% CI (1.13, 1.25), P � 0.06), (596/438, 13); headache times (SMD � 0.75, 95% CI (0.42, 1.08),
Chen (2014) P � 0.001), (362/288, 7); headache degree (SMD � 0.47, 95% CI (−0.17, 1.10), P < 0.00001) (330/258, 6); headache duration
(SMD � 0.62, 95% CI (0.46, 0.78), P � 0.008), (362/288, 7)
Short-term effective rate (RR � 1.27, 95% CI (1.11, 1.45), P < 0.0004), (414/409, 9); long-term effective rate (RR � 1.76, 95%
Yang (2014)
CI (1.05, 2.94), P � 0.03), (117/115, 4)
Zhao (2014) Effective rate (RR � 1.18, 95% CI (1.09, 1.27), P � 0.007) (619/410, 11)
Dai (2015) Effective rate (OR � 4.85, 95% CI (1.69, 13.94), P � 0.003), (65/51, 2)
Not effective rate (RR � 0.24, 95% CI (0.15, 0.38), P � 0.61), (19/93, 4); recurrence rate (RR � 0.47, 95% CI (0.28, 0.81),
Yang (2015)
P � 93), (14/53, 2)
Headache frequency after treatment (SMD � −0.18, 95% CI (−0.28, −0.08), I2 � 47%), (952/694, 12); headache frequency
Linde (2016)
after follow-up (SMD � −0.18, 95% CI (−0.28, −0.08), I2 � 47%), (896/638, 10)
VAS score 2 h after acupuncture (MD � −0.38, 95% CI (−0.83, 0.07), P � 0.10), (350/349, 4); reduced VAS score 2 h after
acupuncture (MD � 0.36, 95% CI (0.08, 0.65), P � 0.01), (290/289, 3); VAS score 4h after acupuncture (MD � −0.42, 95% CI
PuJ (2016)
(−0 .96, 0.12), P � 0.12), (350/349, 4); reduced VAS score 4h after acupuncture (MD � 0.49, 95% CI (0.14, 0.84), P � 0.007),
(290/289, 3)
Effective rate at 1–2 months follow-up (RR � 1.06, 95% CI (0.92, 1.24), P � 0.42), (508/462, 5); effective rate at 3–4 months
follow-up (RR � 1.06, 95% CI (0.91, 1.22), P � 0.48), (525/476, 6); effective rate at 5–6 months follow-up (RR � 1.11, 95% CI
Xian (2016)
(0.96, 1.29), P � 0.17), (515/470, 5); effective rate of more than 6 months follow-up (RR � 2.03, 95% CI (1.10, 3.74),
P � 0.02), (24/11, 2)
Headache frequency (MD � 1.05, 95% CI (1.75, 0.34); P < 0.01), (120/120, 3); VAS score (MD � 1.19, 95% CI (1.75, 0.63);
Xu (2018)
P < 0.01), (84/84, 3)
Headache frequency (SMD � −0.97, 95% CI (−1.60,−0.34), P � 0.002), (95/69, 3); headache duration (SMD � −0.73, 95% CI
Lu (2019)
(−1.25,−0.21),P � 0.006) (86/82, 3); headache intensity (SMD � −0.67, 95% CI (−1.15, −0.19),P � 0.006), (553/490, 6)
Acupuncture vs western medicine
Zheng (2012) Effective rate (RR � 1.24, 95% CI (1.16, 1.34), P < 0.00001), (1602/925, 28)
Headache frequency after treatment (SMD � −0.25, 95% CI (−0.39, −0.10)), (431/308, 3); headache frequency after follow-up
Linde (2016)
(SMD � −0.13, 95% CI (−0.28, −0.01)), (436/308, 3)
Effective rate after 3–4 months(RR � 1.24, 95% CI (1.04, 1.47), P � 0.02), (449/323, 4); effective rate after 5–6 months
(RR � 1.18, CI (0.97, 1.43), P � 0.11), (344/220, 2); headache days after 3–4 months (SMD � -0.30, 95% CI (−0.45,−0.16),
P < 0.0001), (439/316, 4); headache days after 5–6 months (MD � −0.66, 95% CI (−1.18,−0.13), P � 0.01), (344/220, 2);
headache times after 3–4 months (MD � −0.32, 95% CI (−0.59,-0.04), P � 0.03), (171/145, 3); headache times after 3–4
Pu (2016)
months (MD � −0.47, 95% CI (−1.22,−0.28), P � 0.22), (131/106, 2); headache degree after 3–4 months (SMD � −0.11, 95%
CI (−0.56, 0.33), P � 0.01), (495/370, 4); headache degree after 5–6 months (SMD � −0.31, 95% CI (−0.47, −0.15),
P � 0.0001), (385/261, 3); Times of using painkiller after 3–4 months(MD � −0.64, 95% CI (−1.93, 0.65), P � 0.33), (207/181,
4); times of using painkiller after 5–6 months(SMD � −0.22, 95% CI (−0.44, 0.00), P � 0.06), (174/147, 3)
Short-term effective rate (RR � 2.76, 95% CI (2.03, 3.77), P < 0.00001), (616/602, 15); long-term effective rate(RR � 4.17, 95%
Song (2016)
CI (2.80, 6.20), P < 0.00001), (331/311, 7); headache times (RR � −0.79, 95% CI (−1.39, −0.20), P � 0.009), (92/72, 2)
Effective rate at 0–1 months follow-up (RR � 1.66, 95% CI (1.16, 2.37), P � 0.005), (180/160, 4); effective rate at 1–2 months
Xian (2016) follow-up (RR � 1.25, 95% CI (1.01, 1.55), P � 0.04), (162/76, 2); effective rate at 3–4 months follow-up (RR � 1.55, 95% CI
(1.09, 2.20), P � 0.01), (239/125, 5); effective rate at 5–6 months follow-up (RR � 1.30, 95% CI (0.77, 2.19), P � 0.32), (169/87, 2)
Zhao (2016) Effective rate (RR � 1.18, 95% CI (1.09, 1.27), P � 0.007), (649/497, 11);
Headache frequency (MD � 1.50; 95% CI (2.32, 0.68); P < 0.01), (110/110, 2); VAS score (MD � 0.97, 95% CI (0.63, 1.31);
Xu (2018)
P < 0.01), (198/163, 3); effective rate (RR � 1.30; 95% CI (1.16, 1.45); P < 0.01), (178/178, 6)
Headache frequency (SMD � −1.29, 95% CI (−1.85,−0.73), P < 0.00001), (512/486, 8); headache duration (SMD � −0.88,
Lu (2019)
95% CI (−1.32, −0.45), P < 0.0001) (445/427, 7)
Acupuncture vs Chinese herbal medicine
Zheng (2012) Effective rate (RR � 1.29, 95% CI (1.14, 1.45), P < 0.00001), (111/81, 3)
CI, confidence interval; OR, odds ratio; RR, relative risk; MD, mean difference; WMD, weighted mean difference; SMD, standardized mean difference, HR,
hazard ratio; VAS, visual analogue scale.

The PRISMA-A statement provided the basis for the offer a protocol or registration number of SR. An advance
author to improve the reporting quality of the SRs with registration helps promote transparency, minimize potential
acupuncture as intervention. According to the results of the bias in the conducting and reporting review, reduce du-
PRISMA score, the lowest report rate (13.33%) was in the plication of effort between groups, and keep SRs updated.
protocol and registration section. Only 2 SRs managed to [39] A free and open database, the International Prospective
12 Pain Research and Management

Table 6: Evidence quality of included studies.

Author Interventions vs
Outcomes (number of Risk of Publication Quality of
Inconsistency Indirection Imprecision
(date) comparisons studies) bias bias evidence
Effective rate at the
0 0 0 −1① 0 Moderate
Gao Acupuncture vs sham end of treatment (8)
(2011) acupuncture Effective rate at the
0 −1② 0 −1① 0 Low
end of follow-up (4)
Acupuncture vs
Effective rate (8) 0 −1② 0 0 0 Low
western medicine
Acupuncture vs
Zheng
Chinese medicine Effective rate (3) 0 0 0 −1① −1③ Low
(2012)
therapy
Acupuncture vs sham
Effective rate (3) 0 0 0 −1① −1③ Low
acupuncture
Chen Acupuncture vs
Effective rate (13) 0 0 0 0 0 High
(2014) western medicine
Short-term effective
0 −1② 0 0 0 Low
Yang Acupuncture vs rate (9)
(2014) western medicine Long-term effective
0 −1② 0 —1① 0 Very low
rate (4)
Zhao Acupuncture vs
Effective rate (11) 0 0 0 0 0 High
(2014) Western medicine
Dai Acupuncture vs
Effective rate (2) 0 −1② 0 −1① −1③ Very low
(2014) western medicine
Effective rate after 3–4
0 −1② 0 −1① −1③ Very low
Acupuncture vs months follow-up (4)
Pu (2016)
western medicine Effective rate after 5–6
0 0 0 −1① −1③ Low
months follow-up (2)
Short-term effective
0 0 0 0 0 High
rate (15)
Song Acupuncture vs Long-term effective
0 0 0 0 0 High
(2016) western medicine rate (7)
Long-term headache
0 −1② 0 0 −1③ Low
times (2)
Effective rate at 1–2
0 0 0 −1① 0 Moderate
months follow-up (5)
Effective rate at 3–4
0 0 0 −1① 0 Moderate
months follow-up (6)
Acupuncture vs sham
Effective rate at 5–6
acupuncture 0 −1② 0 −1① 0 Low
months follow-up (5)
Effective rate of more
than 6 months follow- 0 −1② 0 0 −1③ Low
Xian
up (2)
(2016)
Effective rate at 1–2
0 −1② 0 0 −1③ Low
months follow-up (4)
Effective rate at 3–4
0 −1② 0 0 −1③ Very low
Acupuncture vs months follow-up (2)
western medicine Effective rate at 5–6
0 −1② 0 −1① −1③ Very low
months follow-up (5)
Effective rate of 0–1
0 −1② 0 −1① −1③ Very low
months follow-up (2)
Zhao Acupuncture vs
Effective rate (8) 0 −1② 0 0 0 Moderate
(2016) western medicine
Xu Acupuncture vs
Effective rate (6) 0 0 0 −1① 0 Moderate
(2018) western medicine
①
The optimal information size was not enough. ②I2 value of the combined results was large, and/or confidence intervals overlapped difference. ③Suspicion of
publishing bias.

Register of Systematic Reviews (PROSPERO, http://www. tool when conducting SRs/meta-analyses. In accordance
crd.york.ac.uk/prospero), has been advocated and recom- with the requirements of the PRISMA-A statement, writing
mended for reviewers to avoid bringing bias in SRs. In order a SR/meta-analysis helps to get better reporting quality.
to achieve a better quality of evidence, the researchers need In the assessment of evidence quality with GRADE tool,
to strictly control the risk of bias with reference to the ROBIS we found that the biggest reason for downgrading was
Pain Research and Management 13

inconsistencies among studies, owing to the high I2 value and References

statistically significant heterogeneity of effect estimates. The
GRADE guideline suggests, when it comes to inconsistency, [1] Headache Classification Subcommittee of the International
SR authors should generate and test a small number of a priori Headache, “The international classification of headache dis-
orders: 2nd edition,” Cephalalgia, vol. 24, no. 9, pp. 9–160,
hypotheses related to patients, interventions, outcomes, and
2004.
methodology to explore the sources of heterogeneity [40]. [2] B. K. Rasmussen, “Epidemiology of headache,” Cephalalgia,
vol. 15, no. 1, pp. 45–68, 1995.
4.3. Strength and Limitations. There are some strength in our [3] W. F. Stewart, C. Bruce, A. Manack, D. C. Buse, S. F. Varon,
study; firstly, this overview is the first to systematically and R. B. Lipton, “A case study for calculating employer costs
evaluated the methodological quality and reporting quality for lost productive time in episodic migraine and chronic
of SRs in acupuncture for migraine. Secondly, we combined migraine,” Journal of Occupational and Environmental
the latest high-quality evidence of SRs to provide a more Medicine, vol. 53, no. 10, pp. 1161–1171, 2011.
[4] P. Tfelt-Hansen, J. Pascual, N. Ramadan et al., “Guidelines for
convinced evidence for clinical work. Thirdly, we started this
controlled trials of drugs in migraine: third edition. A guide
overview with a predesigned protocol, which helped reduce
for investigators,” Cephalalgia, vol. 32, no. 1, pp. 6–38, 2012.
the risk of bias. [5] J. F. Murphy, “The global burden of disease,” Irish Medical
In addition to the strengths, there are several limitations Journal, vol. 106, no. 106, p. 4, 2013.
to be noted. Firstly, there might be some missing information [6] Headache Classification Committee of the International
since we only gathered studies in English and Chinese. Headache, “The international classification of headache dis-
Secondly, we were unable to synthesize all the evidence, which orders, 3rd edition (beta version),” Cephalalgia, vol. 33, no. 9,
may decrease the accuracy of the conclusions. pp. 629–808, 2013.
[7] B. Gilmore and M. Michael, “Treatment of acute migraine
5. Conclusions headache,” American Family Physician, vol. 83, no. 83,
pp. 271–280, 2011.
Based on high quality of evidence, we conclude that acu- [8] P. C. Tfelt-Hansen, “Evidence-based guideline update:
puncture is more effective and safer than medication or pharmacologic treatment for episodic migraine prevention in
sham acupuncture in the treatment of migraine. However, adults: report of the Quality Standards subcommittee of the
the methodological quality, risk of bias, and reporting American Academy of Neurology and the American Head-
quality of SRs in acupuncture for migraine still needs im- ache Society,” Neurology, vol. 80, no. 9, pp. 869-870, 2013.
provement in future. [9] G. Bodeker, G. Burford, C. Grundy, C.-K. Ong, and K. Shein,
WHO Global Atlas of Traditional, Complementary and Al-
Conflicts of Interest ternative Medicine,WHO, Geneva, Switzerland, 2005.
[10] M. Frass, R. P. Strassl, H. Friehs, M. Müllner, M. Kundi, and
The authors declare that they have no conflicts of interest. A. D. Kaye, “Use and acceptance of complementary and al-
ternative medicine among the general population and medical
Authors’ Contributions personnel: a systematic review,” The Ochsner Journal, vol. 12,
no. 12, pp. 45–56, 2012.
Yu-Xi Li, Xi-li Xiao, and Dong-Ling Zhong contributed [11] L. Zhao, F. W. Zhang, Y. Li et al., “Adverse events associated
equally to this work. JL, HZ, and RJJ were involved con- with acupuncture: three multicentre randomized controlled
ceptualization. YXL, XLX, and DLZ wrote the original draft. trials of 1968 cases in China,” Trials, vol. 12, p. 87, 2011.
JL, MXH, and LHS wrote, reviewed, and edited the article. [12] J. Alecrim-Andrade, J. Maciel-Júnior, X. Cladellas, H. Correa-
Filho, and H. Machado, “Acupuncture in migraine prophy-
HY, LJL, and JZ were involved in the methodology. DLZ and
laxis,” Cephalalgia, vol. 26, no. 5, pp. 520–529, 2006.
YXL performed the formal analysis. MXH and LHS were
[13] L. Zhao, J. Chen, Y. Li et al., “The long-term effect of acu-
responsible for the data curation. All authors read and
puncture for migraine prophylaxis,” JAMA Internal Medicine,
approved the publication of this overview. vol. 177, no. 4, pp. 508–515, 2017.
[14] H.-C. Diener, K. Kronfeld, G. Boewing et al., “Efficacy of
Acknowledgments acupuncture for the prophylaxis of migraine: a multicentre
randomised controlled clinical trial,” The Lancet Neurology,
We would like to thank Professor Yao-Long Chen and Mr. vol. 5, no. 4, pp. 310–316, 2006.
NanYang from GRADE center in China (Lanzhou) for their [15] J. Pt Higgins and S. Green, Cochrane Handbook for Systematic
assistance with the GRADE assessment. This work was Reviews of Interventions, John Wiley & Sons, Hoboken, NJ,
supported by the National Natural Science Foundation of USA, 2009.
China (grant nos. 81873356 and 81704137) and the Sichuan [16] L. Hartling, A. Chisholm, D. Thomson, and D. M. Dryden, “A
Science and Technology Program (grant nos. 2019YFS0019 descriptive analysis of overviews of reviews published between
and 2019YJ0331). 2000 and 2011,” PLoS One, vol. 7, no. 11, Article ID e49667,
2012.
Supplementary Materials [17] L. A. Becker and A. D. Oxman, “Overviews of reviews,” in
Cochrane Handbook for Systematic Reviews of Interventions,
S1 Appendix: search strategies. S2 Appendix 2: excluded J. P. Higgins and S. Green, Eds., pp. 607–631, Wiley Online
systematic reviews. (Supplementary Materials) Library Hoboken, NJ, USA, 2008.
14 Pain Research and Management

[18] J. Zhang, H. C. Shang, and B. L. Zhang, “How to assess the western medicine for migraine,” Liaoning Journal of Tradi-
quality of systematic review and meta-analysis,” Journal of tional Chinese Medicine, vol. 43, no. 3, pp. 821–826, 2016.
Chinese Integrative Medicine, vol. 6, no. 4, pp. 337–340, 2008. [35] J. Xian, A Systematic Review and Analysis of Selected Acupoint
[19] K. I. Bougioukas, A. Liakos, A. Tsapas, E. Ntzani, and of Acupuncture for Migraine Prophylaxis, Shandong Uni-
A.-B. Haidich, “Preferred reporting items for overviews of versity of Traditional Chinese Medicine, Jinan, Shandong,
systematic reviews including harms checklist: a pilot tool to be China, 2013.
used for balanced reporting of benefits and harms,” Journal of [36] Y. Y. Zhao, Data Mining and Systematic Review of Modern
Clinical Epidemiology, vol. 93, pp. 9–24, 2018. Acupuncture Treatment for Menstrual Migraine, Chengdu
[20] B. J. Shea, B. C. Reeves, G. Wells et al., “Amstar 2: a critical University of Traditional Chinese medicine, Chengdu,
appraisal tool for systematic reviews that include randomised Sichuan, China, 2016.
or non-randomised studies of healthcare interventions, or [37] J. Xu, F.-q. Zhang, J. Pei, and J. Ji, “Acupuncture for migraine
both,” BMJ, vol. 358, p. j4008, 2017. without aura: a systematic review and meta-analysis,” Journal
[21] P. Whiting, J. Savović, J. P. T. Higgins et al., “ROBIS: a new of Integrative Medicine, vol. 16, no. 5, pp. 312–321, 2018.
tool to assess risk of bias in systematic reviews was developed,” [38] L. Tingting, Y. Xinmin, Z. Yanyan et al., “Efficacy of acu-
Journal of Clinical Epidemiology, vol. 69, pp. 225–234, 2016. puncture for improving migraine symptoms and cerebral
[22] X. Wang, Y. Chen, Y. Liu et al., “Reporting items for sys- blood flow velocity: a meta-analysis,” Chinese Journal of
tematic reviews and meta-analyses of acupuncture: the Evidence-Based Medicine, vol. 19, no. 06, pp. 665–672, 2019.
[39] Editors PLM, “Best practice in systematic reviews: the im-
PRISMA for acupuncture checklist,” BMC Complementary
portance of protocols and registration,” PLoS Medicine, vol. 8,
and Alternative Medicine, vol. 19, no. 1, p. 208, 2019.
no. 2, Article ID e1001009, 2011.
[23] D. Atkins, D. Best, P. A Briss et al., “Grading quality of ev-
[40] G. H. Guyatt, A. D. Oxman, R. Kunz et al., “Grade guidelines:
idence and strength of recommendations,” BMJ, vol. 328,
7. Rating the quality of evidence-inconsistency,” Journal of
no. 328, p. 1490, 2004.
Clinical Epidemiology, vol. 64, no. 12, pp. 1294–1302, 2011.
[24] X. M. Gao, B. S. Wang, Y. Y. Song, Q. I. Hong, and Z. X. Rong,
“Therapeutics. Acupuncture for migraine:a systematic review
and meta-analysis of randomized controlled trials,” Chinese
Journal of Clinical Pharmacology, vol. 16, no. 5, pp. 530–537,
2011.
[25] S. M. Zheng and H. Cui, “Acupuncture for migraine:A meta
analysis,” Chinese Journal of Information on Traditional
Chinese Medicine, vol. 19, no. 6, pp. 20–23, 2012.
[26] W. S. Chen, Systematic Reviews for Short-Term Effects of
Acupuncture in Treating Migraine, Chengdu University of
Traditional Chinese Medicine, Chengdu, Sichuan, China,
2014.
[27] J. Yang, Y. Shen, and S. Wang, “Systematic review on efficacy
for migraine treatment by acupuncture and flunarizine,”
World Science and Technology/Modernization of Traditional
Chinese Medicine and Materia Medica, vol. 16, no. 07,
pp. 1608–1613, 2014.
[28] D. Zhao, System Evaluation and Regularity of Selected Acu-
point about Electro-Acupuncture Treatment on Migraine,
Shandong University of Traditional Chinese Medicine, Jinan,
Shandong, China, 2014.
[29] X. C. Dai and C. Lin, “A Meta-analysis of treating migraine by
acupuncture Clinical,” Journal of Chinese Medicine, vol. 1,
no. 35, pp. 1–3, 2015.
[30] Y. Yang, Q. H. Que, X. D. Ye, and G. H. Zheng, “Verum versus
sham manual acupuncture for migraine: a systematic review
of randomised controlled trials. Acupuncture in Medicine,”
Journal of the British Medical Acupuncture Society, vol. 34,
no. 2, p. 76, 2015.
[31] K. Linde, G. Allais, B Brinkhaus et al., “Acupuncture for the
prevention of episodic migraine,” The Cochrane Database of
Systematic Reviews, no. 6, p. CD001218, 2016.
[32] S. X. Pu, T. Ge, D. Y. Wang, J. J. Chen, and J. Li, “Analgesic
effect of acupuncture during migraine acute attack period:a
Meta analysis,” Chongqing Medicine, vol. 45, no. 10,
pp. 1353–1356, 2016.
[33] S. X. Pu, Effectiveness and Safety of Acupuncture versus Drug
Treatment for Migraine Prophylaxis: A Meta-Analysis,
Chongqing Medical University, Chongqing, China, 2016.
[34] Q. Song, S. Zhao, L. I. Li, Y. Shen, and S. Wang, “Meta-
analysis on prevention comparison of acupuncture with