Study registration
This protocol has been registered on PROSPERO (CRD42020170495), and it is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISRMA) Protocol statement guidelines [27-28].
Study eligibility criteria
Types of studies
Without limitations on language and publication status, this study will include randomized controlled trials (RCTs) that explore the effectiveness and safety of NMES for the treatment of patients with IC.
Types of participants
All patients who were diagnosed as IC will be included, regardless gender, race, age, economic status, duration and severity of IC.
Types of interventions
Experimental interventions
All patients in the experimental group were treated with NMES only. Any treatments combined with NMES will be excluded.
Control interventions
All patients in the control group were treated with any interventions, such as oral medication, moxibustion, and physical therapy. However, any management combined with NMES will be excluded.
Type of outcome measurements
Primary outcomes are pain intensity (as measured by Visual Analog Scale or other pain scales), and improvement of overall symptoms (as assessed by patient-reported global response assessment or other tools).
Secondary outcomes are urinary frequency episodes, quality of life (as checked by 36-Item Short Form Survey or other questionnaires), and adverse events.
Search strategy and data management
Search strategy
The electronic databases will be comprehensively retrieved from their commencements to the March 1, 2020 to identify all related potential studies: MEDLINE, EMBASE, Cochrane Library, Web of Science, CINAHL, China National Knowledge Infrastructure, Chinese Biomedical Literature Database, Chinese Scientific Journal Database, and WANFANG Database. There are no limitations to publication language and status. A search strategy for MEDLINE has been established (table 1). Identical search strategies will be applied to all other electronic databases.
We will also search relevant conference abstracts, clinical trial registries for ongoing trials, and reference lists of all related reviews.
Study selection
Two authors will independently import all citations into EndNote X9 to eliminate duplicated ones. Titles/abstracts of all potential records will be screened to remove any irrelevant studies. If necessary, we will obtain and read full-text of remaining literatures according to the eligibility criteria. All excluded studies will be noted and summarized with reasons. Any differences will be worked out with the help of another author and a consensus will be made. The results of study selection will be summarized in a PRISRMA flow diagram.
Data extraction and management
Two authors will independently extract data by a standardized data form developed specifically for this study. Any confusion will be cleared up with the help of another author and a final conclusion will be made. The information includes study general information (e.g. title, first author, year of publication), participant characteristics (age, gender), diagnostic criteria, inclusion and exclusion criteria, study design, sample size, interventions, outcomes, results, findings, adverse events, and funding information.
Dealing with missing data
Any insufficient or missing data will be requested from primary authors. If they can not provide those data, we will perform data analysis based on the available data, and we will discuss its affects on the study findings.
Study quality assessment
The Cochrane Collaboration’s tool will be utilized to assess study quality of included trials by two independent authors. Each study will be identified at seven aspects, and each one is graded as 3 levels: low, unclear, and high risk of bias. Divergences will be arbitrated with the help of another author.
Statistical analysis
Data synthesis
RevMan 5.3 software will be utilized for statistical analysis. Risk ratio with 95% confidence intervals (CIs) will be used to measure the treatment effect for dichotomous outcome data. Mean difference or standardized mean difference and 95% CIs will be suggested to measure the treatment effect for continuous outcome data. Statistical heterogeneity will be examined by I² test. I² ≤ 50% suggests homogeneity, and we will use a fixed-effect model. If possible, we will conduct a meta-analysis when sufficient trials are included. I² >50% indicates considerable heterogeneity, and we will place a random-effect model. In addition, we will perform subgroup analysis to check sources of obvious heterogeneity. If there is still remarkable heterogeneity after subgroup analysis, a meta-analysis will not be carried out. However, we will report outcome results using a narrative synthesis.
Subgroup analysis
If studies are adequate, we will perform a subgroup analysis based on the different types of interventions, controls, and outcome measurements.
Sensitivity analysis
Whenever it is appropriate, we will undertake a sensitivity analysis to test stability of study findings by removing low quality studies, or small sample studies.
Reporting bias
We will adopt a funnel plot and Egger’s regression test to investigate reporting bias if 10 or more trials are included [29-30].
Grading the quality of evidence
Two authors will independently appraise quality of evidence for each outcome by Grading of Recommendations Assessment, Development and Evaluation, which grades quality of evidence as 4 levels: high, moderate, low, and very low [31]. Any uncertainty will be solved with the help of an independent arbitrator.
Dissemination
This study will be disseminated through a peer-reviewed journal or a conference meeting.