Types of studies

All clinical RCTs of acupuncture for discomfort in patients with gastroscopy will be included in the review, while randomised crossover studies and quasi-RCTs will be excluded. Because of the particularity of acupuncture manipulation, it is difficult to blind the acupuncturists. Therefore, blinding will not be part of the inclusion criteria, although it will be evaluated as an item in the risk of bias assessment.

Types of participants

Participants who underwent gastroscopy will be included, with no age limitation. Excluded participants will be those who have had a colonoscopy, and those who have suffered from chronic pharyngolaryngitis, severe digestive system diseases, persistent hiccups, severe nausea and retching, proven tumours in the upper digestive tract or uncontrolled cardiopulmonary disease.

Types of interventions

Any method of acupuncture usage will be included, such as manual acupuncture, electroacupuncture, fire needling, warm needling, pyonex, scalp acupuncture, auricular acupuncture and intradermal needling, without limitations on the treatment length and frequency. The control groups with no intervention, sham acupuncture, placebo control and drug therapy (such as lubricant use, pharyngeal anaesthesia and sedation) will be included. Acupuncture combined with another therapy will also be included if the combined therapy is the same in both groups. The combined interventions include drugs, but exclude complementary and alternative therapy such as relaxation and music therapy. To assess the efficacy of acupuncture therapy, we intend to compare acupuncture treatment with either no intervention or sham acupuncture. To assess the effectiveness of acupuncture therapies, we plan to compare them with conventional positive intervention consisting of lubricant use, pharyngeal anaesthesia and/or sedation.

Types of outcome assessments

The primary outcome measurement is the proportion of patients with discomfort as assessed via completed questionnaire or visual analogue scale (VAS). If the patient's condition permits, the proportion of throat discomfort, nausea, salivation, retching, emesis or hiccupping will be assessed. The secondary outcomes consist of the proportion of patients satisfied with the whole process and the proportion of patients who would opt for the same procedure again if required. If the included trials permit, the doses of anaesthesia or sedatives, the operation time of gastroscopy examination and the revival time of patients after gastroscopy will also be evaluated. The incidence of adverse events will be assessed as a safety outcome.

Electronic searches

We will electronically search the following databases from their founding date to 30 April 2014, regardless of the publication status: OVID MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials (CENTRAL), Chinese Biomedical Literature Database (CBM), Chinese Medical Current Content (CMCC), Chinese Scientific Journal Database (VIP database), Wan-Fang Database and China National Knowledge Infrastructure (CNKI). The search strategy has been decided on after a discussion among all reviewers according to the guidance of the Cochrane handbook.21 The key words include “gastroscopy”, “gastroscope”, “endoscopy” or “endoscope” combined with “acupuncture”, “manual acupuncture”, “electroacupuncture”, “fire needling”, “warm needling”, “pyonex”, “scalp acupuncture”, “auricular acupuncture” or “intradermal needling”. The search strategy for OVID MEDLINE is shown in table 1. The search words used in Chinese databases have the same meaning as those used in the English databases.

Other sources

Potentially eligible studies will also be found through searching:

• The reference list of previously published reviews related to endoscopy and acupuncture;

• Conference abstracts that may have ongoing or unpublished trials in relation to gastroscopy and acupuncture. The data could be collected by contacting the author if applicable.

Selection of studies

We plan to conduct this systematic review between 28 February 2014 and 30 November 2014. Prior to the selection of the studies, all reviewers will undergo training to ensure a basic understanding of the background and purpose of the review. Reviewers will then reach a consensus on screening and later procedures. After electronic searching, the records will be moved to a database set up by EndNote software (V.X6). Records found through other sources will also be moved to the same database. Two reviewers (WW and TZ) will then independently determine the included studies using the following steps: first, remove duplicates (the same study published in different languages, or studies sharing the same title and abstract both in a journal and conference proceedings, or different articles reporting the same trial in different aspects); second, exclude studies in which participants accepted acupuncture treatment during colonoscopy, or studies in which participants received acupoint injections or other excluded interventions according to the exclusion criteria described above; third, remove studies that were not designed as randomised controlled trials, and trials in which no data can be extracted. The details of the study selection procedure are shown in a PRISMA flow chart (figure 1). During this procedure, any disagreement between the reviewers will be discussed and judged by a third author (ZL).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flow diagram of the study selection process.

Data extraction and management

Before data extraction, all of the reviewers will jointly discuss and create a unified data extraction form. The following factors will be contained in the data extraction form:

1. General information including reference ID, author, time of publication, journal and location of performance.

2. Characteristics of participants.

3. Randomisation.

4. Method of allocation concealment.

5. Blinding.

6. Interventions, including the name of the intervention, length of treatment and type of control.

7. Outcome measures including primary and secondary outcomes.

8. Results.

9. Adverse events.

10. Conflicts of interest.

11. Other information such as the type of gastroscope.

The applicability of the form created will be tested by extracting information from three or more studies. After testing, WW and TZ will independently extract the data from the included studies and fill in the form. The final results of extraction will be checked, with any disagreement between the reviewers discussed and judged by a third author (ZL). ZL will also check the final data to make sure there are no errors.

Assessment of risk of bias in included studies

Two reviewers (WW and TZ) will evaluate the methodological quality of the included trials independently, with the use of the Cochrane Collaboration's tool for risk of bias assessment.21 The tool consists of six domains of a trial, such as sequence generation, allocation concealment and blinding. The assessments will be categorised into three levels of bias: low risk, high risk and unclear risk.

Measures of treatment effect

For dichotomous outcomes, data will be analysed using a risk ratio (RR) with 95% CIs. For continuous outcomes, a standard mean difference (SMD) with 95% CI will be used.

Unit of analysis issues

Because the gastroscopic procedure is a transient process and the acupuncture is often applied before and/or during the gastroscopy, the outcomes will usually be measured once the intervention is complete. Therefore, we will mainly focus on the instant effect of acupuncture therapy.

Dealing with missing data

If data required for the data extraction form are missing, we will try to contact the first or corresponding authors of the studies by telephone or email to obtain the information. If missing data are unobtainable, complete case analysis for continuous outcomes and dichotomous outcomes will be completed using the methods of Ebrahim et al22 and Akl et al,23 with a sensitivity analysis then conducted.

Assessment of heterogeneity

We will perform the Higgins I² test for heterogeneity prior to the meta-analysis to find out if inconsistencies exist within the included trials. We have set a cut-off point of 50% for the I² statistics, as we consider that there will be significant heterogeneity among trials when an I² value exceeds 50%. If this is the case, meta-analysis will not be suggested.

Assessment of reporting biases

If 10 or more trials are included in a meta-analysis, we will generate funnel plots to assess the reporting biases. The plots will be assessed visually and by using Egger's test.

Data synthesis

Data synthesis will be performed using RevMan V.5.2 software from the Cochrane Collaboration. For dichotomous data, if no heterogeneity can be found, we will combine the RR of each study and compute 95% CI with the fixed effect model. If significant heterogeneity is detected, the random effect model will be used. For continuous data, we will combine the SMD of each study and calculate the 95% CI according to the outcome measurement.

Subgroup analysis

Subgroup analysis will be conducted according to different interventions, controls and outcome measures. If the included trials permit, we will also perform subgroup analysis based on the participants’ ages and different diseases.

Sensitivity analysis

We will carry out a sensitivity analysis to remove the impact of lower quality studies if heterogeneity remains after subgroup analysis and input data verification. The meta-analysis will be repeated and studies of lower quality will be excluded. The result will be compared and discussed according to the pooled effect size.

Ethics and dissemination

This systematic review does not require formal ethical approval because the data we will use are not involved in individual data and privacy. The results of this review will provide a general overview and evidence of the effectiveness and safety of acupuncture therapy for discomfort in patients during gastroscopy. The findings will be disseminated through peer-reviewed publications or conference presentations.