3.1 Design
A systematic review and meta-analysis will be carried out in this study, which will only include RCTs and will exclude equivalence trials and clinical inferiority trials.
3.2 Registration
We have registered the protocol on the International Prospective Register of Systematic Reviews (PROSPERO) (registration number: CRD42020114900). This study will be conducted by strictly following the requirements of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.[24] The summary of this study will follow the PRISMA for Protocol (PRISMA-P) guidelines.[25]
3.3 Inclusion criteria
This study strictly follows the following criteria for inclusion in the literature.
3.3.1 Types of studies
This study will be evaluated and screened according to the research review guidelines, including participants, interventions, comparisons, and outcomes.
Randomized controlled trials (RCTs) that have used acupuncture as a monotherapy or combined with other treatments to treat intractable hiccups will be included in the study. In case any RCT applied an incorrect random approach, then it will not be included. Furthermore, non-RCTs, animal experiments, human cell or tissue experiments, and repeated publication studies will be excluded.
3.3.2 Types of participants
Adult patients aged >18 years who meet the criteria for persistent and intractable hiccup diagnosis will be included, regardless of gender, nationality, and education level (refer to the Diagnostic Standard for the Clinical Diagnosis and Treatment of Digestive System Diseases[26] for intractable hiccups; the hiccups last for more than 1 month. They could have restarted after 30–60 min). Patients who do not meet the diagnostic criteria or have received other interventions will be excluded.
3.3.3 Types of interventions
Experimental group: This group will use acupuncture at acupoints dominated by Cuanzhu (BL2) to treat intractable hiccups based on the control group. The acupuncture method, treatment time, and treatment frequency are not limited.
Control group: This group will include a variety of different conventional treatments, such as no treatment, placebo, sham acupuncture, and drugs.
The routine treatment in the RCTs need not be necessarily similar, but the intervention method must be consistent.
Several comparisons will be analyzed as follows:
(1) Acupuncture at acupoints dominated by Cuanzhu (BL2) compared with no treatment
(2) Acupuncture at acupoints dominated by Cuanzhu (BL2) compared with placebo
(3) Acupuncture at acupoints dominated by Cuanzhu (BL2) compared with sham acupuncture
(4) Acupuncture at acupoints dominated by Cuanzhu (BL2) compared with drugs
......
3.3.4 Types of outcome measures
(1) Primary outcome measures
①The time when the hiccup symptoms disappear after interventions
②The frequency or some other change in hiccup symptoms after interventions
③The adverse events produced by the intervention
(2) Secondary outcome measures
①The number of diverse methods of interventions that are needed to be taken to stop the hiccups
②Secondary adverse events that do not need withdrawal
3.4 Information sources
Eight electronic databases, including four Chinese databases (CNKI, SinoMed, Wanfang Database, and Chinese Scientific Journal Database (VIP)), four English databases (Web of Science, Medline, Embase, and Cochrane Library) will be searched from their date of establishment to September 2020. In addition, gray literature such as conference papers and bibliographic references will be included. The search was independently completed by two researchers in September 2020. Both Chinese and English RCTs will be included.
3.5 Search strategies
A comprehensive search strategy will be developed for PubMed or MEDLINE (Figure 2)
3.6 Data collection and management
3.6.1 Selection of studies
The evaluators are trained and prescreened to ensure the accuracy and standardization of the literature screening process. The literature screening process requires at least two reviewers (ZY, and WZJ) to conduct independent discussions and discuss the decision with a third commentator (QWX) whenever there is a disagreement. The literature screening process will be reported in detail in the systematic review plan and in the full text. The following steps are included (Figure 3):
(1) Using NoteExpress to classify and organize the initial inspection documents and eliminate duplicate documents
(2) Reading the title and abstract of each study and excluding unrelated studies that do not significantly meet the inclusion criteria
(3) Analyzing and removing repeated publications
(4) Trying to contact the original author to supplement relevant information for studies with incomplete information.
3.6.2 Data extraction
Two reviewers (ZY, and WZJ) will use the predesigned standardized information extraction form for extracting research data. The extracted data will include basic information (such as the title, published , journal, publication time, author, ID number, and author contact information, ), research methods (such as the inclusion or exclusion criteria, randomization methods, blinding, sample size, intervention group, and control group), intervention cycle, frequency of intervention, participant characteristics (such as age, gender, and grouping, ), and other confounding factors (such as the funding source, author’s key conclusions, author’s evaluation of confounding factors, and references to other studies, etc.).
3.7 Assessment of risk of bias and reporting of study quality
We intend to use the Risk of Bias Assessment tool recommended by the Cochrane Handbook 5.1 to evaluate the methodological quality of the included RCTs.
The following seven aspects will be included:
(1) Random sequence generation (selection bias)
(2) Allocation concealment (selection bias)
(3) Blinding of participants and personnel (performance bias)
(4) Blinding of outcome assessment (detection bias)
(5) Incomplete outcome data (attrition bias)
(6) Selective reporting (reporting bias)
(7) Other bias
For each of the included studies, an evaluation of “yes” (low bias), “no” (high bias), and “unclear” (indeterminate lack of relevant information or bias) will be done according to the abovementioned seven terms. The decision will be discussed with the third commentator whenever there is a disagreement. The “Risks of bias graph” and “Risk of bias summary” tools in the RevMan software will be used to represent the percentage of each criterion for all the judged results (“Yes,” “No,” and “Unclear”).
3.8 Comprehensive statistical analysis of data
Data analysis and meta-analysis will be performed using the RevMan 5.3 software through the Cochrane collaboration. We will determine the effect size on the basis of different data. The second classification results will be analyzed at a 95% confidence interval for relative risk (RR) and 95% confidence interval analysis for continuous variables using MD (Standard Deviation).
3.8.1 Handling missing data
If there are incomplete or missing data, we will try to contact the author by e-mail or phone to obtain the most complete data. In case complete data are not available, that particular RCT will be excluded.
3.8.2 Heterogeneity analysis
A chi-square test will be used to estimate the presence of statistical heterogeneity, and the I2 test will be used to estimate heterogeneity on R-3.3.2 software. If there is no heterogeneity, a fixed effects model (I2<50%, P>0.05) will be used. In case heterogeneity occurs, a random effects model will be used and a subgroup analysis or sensitivity analysis will be conducted.
3.8.3 Subgroup analysis
If the heterogeneity is obvious, a subgroup analysis will be performed based on different types of acupuncture (such as electroacupuncture, and fire needle, ) or other interventions using R-3.3.2 software.
3.8.4 Sensitivity analysis
Sensitivity analysis will be performed to determine data reliability based on missing data, sample size, and heterogeneity using R-3.3.2 software.
3.8.5 Publication bias
A funnel plot will be used to determine if there is a publication bias when there are more than 10 studies included. In case of a publication bias, the funnel plot will appear as an asymmetric distribution.
3.8.6 Grading of evidence quality
The quality evaluation in this study will be completed using the Grading of Recommendations, Assessment, Development, and Evaluation established by the World Health Organization. The quality results will be divided into the following four aspects: “high,” “medium,” “low,” and “very low.”