Open Access Protocol

Acupuncture for constipation in patients

with stroke: protocol of a systematic
review and meta-analysis
Jingbo Zhai,1 Wei Mu,2 Jinhua Si,3 Yan Li,4 Chen Zhao,5 Hongcai Shang,6
Huanan Li,7,8 Guihua Tian6

To cite: Zhai J, Mu W, Si J, et al. Abstract

Acupuncture for constipation in Strengths and limitations of this study
Introduction  Constipation is one of the most common
patients with stroke: protocol complications in patients with stroke. Acupuncture
of a systematic review and ►► This review will provide a comprehensive assess-
has gained increased popularity for the management ment regarding the effect of acupuncture for consti-
meta-analysis. BMJ Open
2018;8:e020400. doi:10.1136/ of constipation. However, there is a lack of supportive pation in patients with stroke.
bmjopen-2017-020400 evidence on the efficacy of acupuncture for poststroke ►► Only randomised controlled trials will be included
constipation. This systematic review aims to collect and which are more likely to provide unbiased informa-
►► Prepublication history and
critically appraise all the available evidence about the tion than other study designs.
additional material for this
efficacy and safety of the acupuncture for constipation in ►► The reliability of the results will largely depend on
paper are available online. To
view these files, please visit poststroke patients. the comprehensiveness and the methodological
the journal online (http://​dx.​doi.​ Methods and analysis  A comprehensive search of quality of the primary studies included in this review.
org/​10.​1136/​bmjopen-​2017-​ Pubmed, Embase, Cochrane Central Register of Controlled
020400). Trials, Web of Science, four Chinese databases (National
Knowledge Infrastructure (CNKI), Chinese Biomedical
Received 2 November 2017 Literatures database (CBM), Wanfang Digital Periodicals Constipation is one of the most common
Revised 23 January 2018
(WANFANG) and Chinese Science and Technology complications in patients with stroke.3 The
Accepted 12 March 2018 evidence from a previous systematic review
Periodicals (VIP) database), one Japanese medical
database (National Institute of Informatics, CiNii) and one shows that the incidence of constipation in
Korean medical database (Oriental Medicine Advanced patients with stroke is 48% (95% CI 33%
Searching Integrated System, OASIS) will be conducted to 63%).4 The incidence of constipation in
to identify randomised controlled trials of acupuncture for patients with haemorrhagic stroke is higher
constipation in poststroke patients. There is no restriction than that in patients with ischaemic type.4
on language or publication status. The primary outcome The incidence in rehabilitation stage is
measure will be frequency of bowel movement. The risk of higher than that in acute stage.4
bias will be assessed using the approach recommended Constipation has a negative impact on the
by Cochrane Handbook for Systematic Reviews of patient’s psychological well-being and phys-
Interventions. We will conduct the meta-analysis to
ical health, restricts social activities, reduces
synthesise the evidence for each outcome, if possible.
quality of life, contributes to a poor outcome
The heterogeneity will be statistically assessed using a χ2
test and I2 statistic. This protocol is developed following
and is associated with high healthcare costs.4–6
the guideline of Preferred Reporting Items for Systematic Stool softeners, prokinetic agents, osmotic
Reviews and Meta-analyses Protocols 2015. and stimulant laxatives and lifestyle or dietary
Ethics and dissemination  The ethical approval is not modification are common treatments for
required because no primary data are collected. The constipation.7 Conventional treatments may
findings will be presented at scientific conferences or a be associated with unwanted side effects, such
peer-reviewed scientific journal. as bloating, dehydration, a high recurrence
PROSPERO registration number CRD42017076880. rate after ceasing drugs.7 8 Most of patients
with chronic constipation are not satisfied
with current treatment options in Europe.9
Many patients with constipation seek help
For numbered affiliations see
Introduction from alternative therapies.9
end of article. Stroke is one of the leading causes of death Acupuncture, as an important part of
and disability worldwide.1 The incidence of complementary and complementary medi-
Correspondence to stroke is subject to large variation globally.2 cine, has gained increased popularity for
Dr Huanan Li;
​lihuanan1984@​126.​com and
The overall incidence in low to middle-in- the management of constipation in Western
Prof. Guihua Tian; come countries exceeded that in high-in- countries.9 There are some possible mech-
​Rosetgh@​163.​com come countries by 20% from 2000 to 2008.2 anisms of acupuncture for constipation.8

Zhai J, et al. BMJ Open 2018;8:e020400. doi:10.1136/bmjopen-2017-020400 1

Open Access

Acupuncture can promote intestinal canal peristalsis before the stroke diagnosis is not investigated but excluded
through contacting the intestinal wall and regulating trials reporting on patients with a history of constipation
nervous and body fluids.8 Acupuncture can increase before the stroke diagnosis. Stroke is defined as ‘rapidly
rectal internal pressure to restore the defecation sense by developed signs of focal or global disturbance of cerebral
stimulating parasympathetic nerve.8 function, lasting more than 24 hours or leading to death
A systematic review shows that acupuncture is safe for with no apparent cause other than that of vascular origin’,
chronic functional constipation.10 A randomised trial according to WHO criteria.17 We will include patients
supports the use of electroacupuncture for chronic severe with stroke irrespective of any type (ischaemic or haem-
functional constipation.11 However, whether the evidence orrhagic) or phase (acute, subacute or chronic). Acute
is transferable to the stroke population remains unclear. and subacute stroke is defined as less than 6 months since
The recent studies mainly focus on the incidence of onset, and chronic stroke lasts more than 6 months since
constipation after stroke.4 5 The management strategies onset.18
for constipation in patients with stroke remain poorly All of patients should be diagnosed as constipation
investigated.3 according to at least one of the current or past defini-
A 2014 systematic review evaluated the efficacy and tions or guidelines of constipation, such as Rome II/III
safety of acupuncture and moxibustion for poststroke diagnostic criteria or guidelines for clinical research on
constipation.12 This review had evident flaws that threat- Chinese new herbal medicine.12
ened the authenticity of their findings. First, meta-analysis There is no restriction on age, sex or ethnicity of the
found that acupuncture and moxibustion were signifi- enrolled subjects.
cantly more effective than other treatments (OR=2.10,
95% credible interval 1.25 to 3.54, p=0.005) for consti-
Types of interventions
pation in patients with stroke. Acupuncture and moxi-
Experimental interventions
bustion were addressed as a whole, and the efficacy of
acupuncture alone was not systematically investigated, We will include trials using either traditional or contem-
despite that five of the eight included trials compared porary acupuncture. Traditional acupuncture refers to
acupuncture alone with another treatment. Second, the needles inserted in classical meridian points.19 Contempo-
methodological quality of eight included articles was very rary acupuncture refers to needles inserted in non-meridian
poor. Third, the total effective rate, a subjective outcome or trigger points regardless of the source of stimulation (for
measure, was chosen as the primary outcome. Fourth, example, hand, electrical stimulation or fine needle).19 We
the sample size was small. To our knowledge, several new will exclude trials in which treatment without needling,
randomised controlled trials (RCTs) have been published such as acupressure, tap-pricking, point injection and laser
since the meta-analysis was published.13 14 Overall, there acupuncture. No restrictions are imposed on times of treat-
is a lack of supportive evidence on the efficacy and safety ment and length of treatment period.
of acupuncture for constipation in patients with stroke.
Comparator interventions
The aim of this study is to systematically review current
The control interventions could be placebo acupuncture,
available literature to assess the efficacy and safety of the
sham acupuncture, no treatment, another active treat-
acupuncture treatment for post-stroke constipation.
ment or medication.
Placebo acupuncture refers to a needle attached to the
skin surface without penetrating the skin.20
Methods
Sham acupuncture is defined as a needle placed in an
This protocol is developed following the guideline of
area close to but not in acupuncture points or sublim-
Preferred Reporting Items for Systematic Reviews and
inal skin electrostimulation via electrodes attached to the
Meta-analyses Protocols (PRISMA-P) 2015.15
skin.20
Inclusion criteria We consider another active treatment or medication to
Types of studies be pharmacological and non-pharmacological treatment
We will only include RCTs which are more likely to provide or medication, such as laxatives, emollients, lubricants,
unbiased information than other study designs.16 We will lifestyle or dietary modification.
exclude quasirandomised RCTs, such as those allocating We will investigate the comparisons listed below:
by alphabetical order, alternate days of the week or date 1. Acupuncture only compared with no treatment;
of birth. Cross-over trials will be excluded because of 2. Acupuncture only compared with placebo or sham
potential for a carry-over effect. There is no restriction on treatment;
language or publication status. 3. Acupuncture plus another active treatment or med-
ication compared with another active treatment or
Types of participants medication alone;
We will include adults (over 18 years old) suffering from 4. Acupuncture plus another active treatment or medi-
constipation after a first or recurrent stroke. We also cation compared with placebo or sham treatment plus
consider RCTs in which a prior history of constipation another active treatment or medication.

2 Zhai J, et al. BMJ Open 2018;8:e020400. doi:10.1136/bmjopen-2017-020400

 Open Access

Types of outcome measures Data extraction and management

Primary outcome Two review authors (YL and CZ) will independently extract
The primary outcome measure will be frequency of bowel data from the included studies. Calibration exercises
movement.21 22 Bowel movement frequency is defined as will be conducted to ensure consistency across reviewers
the mean number of spontaneous bowel movements per before starting the review. The following information will
week.23 be extracted using a predetermined data form: general
information (title, authors, country of study, funding,
Secondary outcome year of publication, registry number); details of study
Secondary outcome measures include proportion of (aim, design, inclusion and exclusion criteria, method
patients with stool consistency, proportion of patients of randomisation and allocation); study population (age,
using rescue medication such as laxatives or rectal evac- sex, sample size, number for analysis, type of stroke,
uants, quality of life (QoL) and mean transit time. phase of stroke); intervention characteristics (type, dura-
Stool consistency will be defined by trialists or measured tion, dose, follow-up time points, compliance); outcome
based on scales such as Bristol Stool Form Scale.24 (primary and secondary outcomes, time points, method
QoL will be measured by generic or condition-specific of outcome assessments, blinding of outcome assessment,
scales, such as Short Form 36 Health Surveys.25 adverse effects).
Transit time is defined as the time from the first percep- It is possible that patients in different phases of stroke
tion of wanting to defecate to the end of defecation.26 were enrolled in one trial. Whenever necessary the authors
We will sum up the number of adverse events (AEs) and of the original trial will be contacted for additional infor-
calculate the proportion of AEs.23 mation and clarification of the data. Any disagreement
will be resolved by consensus or consultation with a third
Search strategy review author.
Electronic searches
The published literature will be identified by searching Assessment of risk of bias in included studies
Pubmed, Embase, Cochrane Central Register of Two authors (HL and HS) will independently assess
Controlled Trials and Web of Science. Four Chinese the risk of bias using the approach recommended by
databases (CNKI, CBM, WANFANG and VIP database), Cochrane Handbook for Systematic Reviews of Inter-
one Japanese medical database (CiNii) and one Korean ventions.31 The following risk of bias domains will be
medical database (OASIS) will also be systematically assessed: sequence generation (selection bias); allocation
searched to identify any relevant study. concealment (selection bias); blinding of participants
The search strategy is developed by a medical librarian and personnel (performance bias); blinding of outcome
(JS) according to key terms from previous literature assessment (detection bias); incomplete outcome data
reviews.27 28 The detailed search strategy is attached (attrition bias); selective outcome reporting (reporting
(see online supplementary appendix 1). The terms will bias) and other bias.
be modified as necessary for other databases. We will not We will attempt to describe what is reported to have
apply any language or date restrictions. happened in each study for each domain of risk of bias.
Thus, we will be able to provide the rationale for the judge-
Searching other resources ment of whether this domain is at low, high or unclear
We will check the reference lists of identified relevant risk of bias. Where necessary, we will contact authors of
RCTs and reviews for additional studies. We will contact included studies for missing information or clarification.
experts in the field of stroke and constipation to identify If all domains are at low risk of bias, the overall risk of
any additional trials. bias of individual studies will be categorised as low risk
The WHO International Clinical Trials Registry Plat- of bias. Otherwise, overall risk of bias will be categorised
form (ICTRP), ​ClinicalTrials.​gov will also be checked to as high risk of bias.32 The ‘risk of bias’ summary will be
identify planned, ongoing or unpublished trials. Google presented graphically.
Scholar will be searched to identify any grey literature.
Measures of treatment effect
Data collection and analysis We will use the risk ratio with 95% CI to express the esti-
Selection of studies mate of the effect for dichotomous outcome.
Two review authors (WM and JS) will independently For continuous outcome, we will express the estimate
assess abstracts and titles of studies identified by literature of the effect as mean difference with 95% CI. When the
search. Duplicates will be omitted using EndNote soft- same outcome is measured in a variety of ways, the stan-
ware (V.X7.0).29 Relevant studies will be selected against dardised mean difference with 95% CI will be used to
the predefined inclusion criteria. If necessary, reviewers express the size of the intervention effect.
will examine full-text reports to identify eligible studies.
EndNote software will also be used to manage records. We Dealing with missing data
will illustrate the selection process in a PRISMA diagram.30 We will attempt to contact study authors for missing data
Any disagreement will be resolved by consensus. or clarification, where feasible. The following strategies

Zhai J, et al. BMJ Open 2018;8:e020400. doi:10.1136/bmjopen-2017-020400 3

Open Access

will be used to evaluate the potential influence of missing The intervention effect will be analysed using the χ2
data.32 test, with p<0.05, indicating statistically significant differ-
1. Worst-case scenario analysis: all participants with miss- ences between subgroups.
ing data counted as failures.
2. Extreme worst-case analysis: participants with missing Sensitivity analysis
data in experimental group counted as failures and We will perform sensitivity analyses to evaluate the robust-
participants with missing data in control group count- ness of the pooled results excluding trials with high risk of
ed as successes. bias, trials in which a prior history of constipation before
3. Extreme best-case analysis: participants with missing the stroke diagnosis is not investigated and the option of
data in experimental group counted as successes and using missing data (worst-case scenario analysis, extreme
participants with missing data in control group count- worst-case analysis or extreme best-case analysis).32 34
ed as failures.
Summary of findings table
Assessment of heterogeneity We will prepare ‘summary of findings’ tables including
We will assess the statistical heterogeneity using a χ2 a grade of the overall quality of the body of evidence for
test.33 In addition, we will quantify heterogeneity using each outcome using Grading of Recommendations Assess-
the I2 statistic value which ranges from 0% to 100%.34 ment, Development and Evaluation profiler (GRADEpro)
P<0.1 of χ2 test or I2>50% indicates statistically significant Guideline Development Tool.36
heterogeneity.33 34 Potential clinical heterogeneity will be Two review authors will independently assess the quality
assessed by prespecified subgroup analyses. of the body of evidence according to five GRADE criteria:
study limitations, imprecision, inconsistency, indirectness
Assessment of reporting biases and publication bias. It will fall into one of four possible
When a meta-analysis includes 10 or more RCTs, we will ratings (high, moderate, low and very low). Any discrep-
assess asymmetry using funnel plots visually.32 In addition, ancy will be resolved by consensus or consultation with a
we will test asymmetry using the Harbord modified test third review author.
for dichotomous outcomes and Egger test for continuous
outcomes.32 Amendments
We will provide the date of any amendment, a description
Data synthesis of the change and the rationale in the event of protocol
We will combine more than one trial to estimate pooled amendments.
intervention effect using the meta-analysis when studies
examine the same intervention and outcomes with
comparable methods in similar populations.31
Ethics and dissemination
We will pool the continuous data using the inverse vari-
Ethical approval is not required because no primary data
ance method and dichotomous data using the Mantel-
are collected.
Haenszel method.31
This review will provide a comprehensive assessment
We will use the fixed-effect model to combine data
regarding the effect of acupuncture for constipation in
when statistical heterogeneity is low. However, when
patients with stroke. The results will be fundamental for
P<0.1 or I2>50%, the random-effect model will be used to
reliable recommendations in the management of post-
provide a more conservative estimate of effect.35
stroke constipation.
All analyses will be conducted with Review Manager
We will present findings from this systematic review
V.5.3 software. If a meta-analysis is not possible, we will
at scientific conferences and publish the findings in a
provide a narrative summary of the results from indi-
peer-reviewed scientific journal according to the PRISMA
vidual studies.
guidelines.
Subgroup analysis and investigation of heterogeneity Author affiliations
We will perform the following subgroup analyses to 1
Research institute of Traditional Chinese Medicine, Tianjin University of Traditional
investigate heterogeneity when sufficient data are avail- Chinese Medicine, Tianjin, China
2
able. We will conduct subgroup analyses based on age, Department of Clinical Pharmacology, The Second Affiliated Hospital of Tianjin
University of Traditional Chinese Medicine, Tianjin, China
sex, type of stroke (haemorrhagic and ischaemic stroke), 3
Library, Tianjin University of Traditional Chinese Medicine, Tianjin, China
different definitions of constipation, phase of stroke 4
School of Nursing, Tianjin University of Traditional Chinese Medicine, Tianjin, China
(acute, subacute or chronic), type of acupuncture 5
Hong Kong Chinese Medicine Clinical Study Centre, Hong Kong Baptist University,
(manual acupuncture, electroacupuncture, etc) and Hong Kong, China
6
type of control group (placebo, sham acupuncture, no Key Laboratory of Chinese Internal Medicine of Ministry of Education and Beijing,
Dongzhimen Hospital, Beijing University of Chinese Medicine, Beijing, China
treatment or another active treatment or medication). A 7
First Teaching Hospital of Tianjin University of Traditional Chinese Medicine, Tianjin,
subgroup analysis based on population with different diet China
habits is under plan considering that diet habits may play 8
Laboratory for Biological Effects of Tuina, State Administration of Traditional
an important role in constipation development. Chinese Medicine, Tianjin, China

4 Zhai J, et al. BMJ Open 2018;8:e020400. doi:10.1136/bmjopen-2017-020400

 Open Access

Contributors  JZ, HL and GT conceived the study; these three authors provided 14. Haifei L, Fengjun S, Bihong Y. Therapeutic observation of
general guidance to the drafting of the protocol. JZ and WM drafted the protocol. acupuncture at He-Sea and Front-Mu points for post-stroke
JS designed the search strategy. JZ, YL, CZ and HS drafted the manuscript. JZ, WM, constipation. Shanghai J Acu-mox 2016;35:160–1.
15. Shamseer L, Moher D, Clarke M, et al. Preferred reporting items for
JS, YL, CZ, HL, GT and HS reviewed and revised the manuscript. All authors have
systematic review and meta-analysis protocols (PRISMA-P) 2015:
read and approved the final version of the manuscript. elaboration and explanation. BMJ 2015;349:g7647.
Funding  This study is supported by the National Natural Science Foundation of 16. Gluud LL, Morgan MY. Endoscopic therapy and beta-blockers for
China (grant numbers 81373762, 81603495, 81703936) and Beijing Nova Program secondary prevention in adults with cirrhosis and oesophageal
(grant number xx2014B049). varices. Cochrane Database Syst Rev 2017;6:CD012694.
17. Hatano S. Experience from a multicentre stroke register: a
Competing interests  None declared. preliminary report. Bull World Health Organ 1976;54:541–53.
18. Momosaki R, Yamada N, Ota E, et al. Repetitive peripheral magnetic
Patient consent  Not required. stimulation for activities of daily living and functional ability in people
Provenance and peer review  Not commissioned; externally peer reviewed. after stroke. Cochrane Database Syst Rev 2017;6:CD011968.
19. Cheong YC, Dix S, Hung Yu Ng E, et al. Acupuncture and
Open Access This is an Open Access article distributed in accordance with the assisted reproductive technology. Cochrane Database Syst Rev
terms of the Creative Commons Attribution (CC BY 4.0) license, which permits 2013;7:CD006920.
others to distribute, remix, adapt and build upon this work, for commercial use, 20. Cheuk DKL, Wong V, Chen WX. Acupuncture for autism spectrum
provided the original work is properly cited. See: http://​creativecommons.​org/​ disorders. Cochrane Database Syst Rev 2011;9:CD007849.
licenses/​by/​4.​0/ 21. Woodward S, Norton C, Chiarelli P. Biofeedback for treatment of
chronic idiopathic constipation in adults. Cochrane Database Syst
© Article author(s) (or their employer(s) unless otherwise stated in the text of the Rev 2014;3:CD008486.
article) 2018. All rights reserved. No commercial use is permitted unless otherwise 22. Turawa EB, Musekiwa A, Rohwer AC. Interventions for preventing
expressly granted. postpartum constipation. Cochrane Database Syst Rev
2015;9:CD011625.
23. Chen M, Zheng H, Li J, et al. Non-pharmacological treatments for
adult patients with functional constipation: a systematic review
protocol. BMJ Open 2014;4:e004982.
24. Miller LE, Ibarra A, Ouwehand AC, et al. Normative values for stool
References frequency and form using Rome III diagnostic criteria for functional
1. Strong K, Mathers C, Bonita R. Preventing stroke: saving lives constipation in adults: systematic review with meta-analysis. Ann
around the world. Lancet Neurol 2007;6:182–7. Gastroenterol 2017;30:161–7.
2. Feigin VL, Lawes CM, Bennett DA, et al. Worldwide stroke incidence 25. Coggrave M, Norton C, Cody JD. Management of faecal
and early case fatality reported in 56 population-based studies: a incontinence and constipation in adults with central neurological
systematic review. Lancet Neurol 2009;8:355–69. diseases. Cochrane Database Syst Rev 2014;1:CD002115.
3. Lim SF, Childs C. A systematic review of the effectiveness of bowel 26. Zheng Q, Zheng H, Lu L, et al. Acupuncture for functional
management strategies for constipation in adults with stroke. Int J constipation: protocol of an individual patient data meta-analysis.
Nurs Stud 2013;50:1004–10. BMJ Open 2015;5:e007137.
4. Li J, Yuan M, Liu Y, et al. Incidence of constipation in stroke patients: 27. Zhang SH, Liu M, Asplund K, et al. Acupuncture for acute stroke.
a systematic review and meta-analysis. Medicine 2017;96:e7225. Cochrane Database Syst Rev 2005;2:CD003317.
5. Su Y, Zhang X, Zeng J, et al. New-onset constipation at acute stage 28. Aboumarzouk OM, Agarwal T, Antakia R, et al. Cisapride for intestinal
after first stroke: incidence, risk factors, and impact on the stroke constipation. Cochrane Database Syst Rev 2011;1:CD007780.
outcome. Stroke 2009;40:1304–9. 29. Song GM, Liu XL, Bian W, et al. Systematic review with
6. Shariff U, Aboumarzouk OM, Nelson RL. Antibiotics for the network meta-analysis: comparative efficacy of different enteral
management of constipation predominant irritable bowel immunonutrition formulas in patients underwent gastrectomy.
syndrome and chronic constipation. Cochrane Database Syst Rev Oncotarget 2017;8:23376–88.
2009;1:CD007518. 30. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for
7. Zhou SL, Zhang XL, Wang JH. Comparison of electroacupuncture reporting systematic reviews and meta-analyses of studies that
and medical treatment for functional constipation: a systematic evaluate health care interventions: explanation and elaboration. PLoS
review and meta-analysis. Acupunct Med 2017;35:324–31. Med 2009;6:e1000100.
8. Zhang C, Guo L, Guo X, et al. Short and long-term efficacy of 31. In: Higgins JPT, Green S, eds. Cochrane handbook for systematic
combining Fuzhengliqi mixture with acupuncture in treatment of reviews of interventions version 5.1.0. 2011: The Cochrane
functional constipation. J Tradit Chin Med 2013;33:51–9. Collaboration. http://www.​handbook.​cochrane.​org (updated Mar
9. Tack J, Mueller-Lissner S, Dubois D, et al. Only 27% of European 2011).
patients with chronic constipation are satisfied with current treatment 32. Aamann L, Dam G, Rinnov A, et al. Physical exercise for people with
options. Gut 2009;58(Suppl II):A181. cirrhosis. Cochrane Database Syst Rev 2017;6:CD012678.
10. Zhang T, Chon TY, Liu B, et al. Efficacy of acupuncture for chronic 33. Anthonappa RP, Ashley PF, Bonetti DL, et al. Non-pharmacological
constipation: a systematic review. Am J Chin Med 2013;41:717–42. interventions for managing dental anxiety in children. Cochrane
11. Liu Z, Yan S, Wu J, et al. Acupuncture for chronic severe functional Database Syst Rev 2017;6:CD012676.
constipation: a randomized trial. Ann Intern Med 2016;165:761–9. 34. Boonstra A, van Dulmen-den Broeder E, Rovers MM, et al. Severe
12. Yang JP, Liu JY, Gu HY, et al. [Randomized controlled trials of fatigue in childhood cancer survivors. Cochrane Database Syst Rev
acupuncture and moxibustion for post-stroke constipation: a meta 2017;6:CD012681.
analysis]. Zhongguo Zhen Jiu 2014;34:833–6. 35. Schilder AG, Chong LY, Ftouh S, et al. Bilateral versus unilateral
13. Wang C, Liu M, Wen Q, et al. [Clinical observation of preventing hearing aids for bilateral hearing impairment in adults. Cochrane
and treating constipation by electroacupuncture intervention in Database Syst Rev 2017;12:CD012665.
ischemic stroke patients at acute stage]. Zhongguo Zhen Jiu 36. GRADEpro GDT. GRADEpro Guideline Development Tool, 2015.
2015;35:430–4. https://​gradepro.​org.

Zhai J, et al. BMJ Open 2018;8:e020400. doi:10.1136/bmjopen-2017-020400 5