2.1 Search strategy
We searched electronic database of PubMed, Cochrane Library, Web of Science, Embase. All included studies were limited to randomized controlled trails (RCTs) to May 12th, 2022, which referred to compare the efficacy and safety of PTNS versus shame for FI. At the same time, the references cited in the included literatures were traced back to ensure that all eligible articles were included. The included articles must be English without geographical and ethnic restrictions. The search terms including “tibial nerve stimulation”, “Faecal incontinence tibial”, “tibial nerve”, “faecal incontinence”, “Bowel Incontinence”, “Incontinence, Bowel”, “Fecal Soiling”, and “Soilings, Fecal”. RevMan5.4 software (Cochrane Collaboratio’s Information Management System) was used for the available pooled date. A narrative review was completed if the data cannot be synthesized. Two independent authors (C.L and FP. Y) searched the electronic databases and assessed the eligibility of the studies. Discussed with a third author (DM. W), if any disagreements existed.
2.2 Inclusion and exclusion criteria
Inclusion criteria
1.As for the studies types, which must be randomized controlled trails (RCTs). 2. As for the subjects (≥18 years old), who were suffered from FI symptoms and feeble for the previously conservatives treatments. 3. As for the experimental group, which must received PTNS treatments. 4. As for the control group, the procedures were the same as PTNS without current powered. 5. As for the outcome indicators, which must include at least one primary outcome measures. 6. As for the provided data, which were sufficient so as to calculate the effect size. 7. full text must be available.
exclusion criteria
1.Duplicated studies, review articles, comment, case series and observational studies; 2. Patients had a history of malignancy; 3. The subjects were related with neurological diseases, including diabetic nephropathy, multiple sclerosis and Parkinson disease; 4. Pregnancy, women during child-bearing age should firstly finished pregnancy test; 5. Anatomical limitations which failed to provide placements of the needle electrode; 6. Other diseases misfitted for stimulation including bleeding disorders, cardiac pacemakers, peripheral vascular diseases and lower leg cellulitis; 7. Absence of anorectum for surgical resection or birth defect; 8. Chronic bowel diseases with chronic uncontrolled diarrhoea and systemic infections.
2.3 Quality assessment
All researchers independently evaluated the quality of included RCTs by using Cochrane risk of bias tool which including 7 domains: sequence generation and allocation sequence concealment (selection bias), blinding of participants and personnel, blinding of outcome assessments, selective reporting (reporting bias), incomplete outcome data (attrition bias), and other bias (e.g. fund assistance and baseline comparability). “Low risk”, “high risk” or “unclear risk” were marked in all domains. We used the RevMan 5.4 (Cochrane Collaboration’s Information Management System) to analyze risks and applicability.
2.4 Data extraction
Three authors (C.L, FP.Y and DM.W) independently screened and extracted information and data from the included studies, in order to ensure the integrity and accuracy, if any disagreements existed, a final decision was made with a fourth investigator (KY. P). The extracted information included first author, publication years, country, gender, mean age, sample size, experimental and intervention procedures and duration of follow-up. If necessary, we contacted the authors within the eligible trials by phone, fax, or e-mail to obtain missed valuable information.
2.5 Statistical analysis
We operated Meta-analysis by using RevMan 5.4. For the continuous data, mean and standard deviation (SD) were used to calculate the Mean Difference (MD) and 95% confidence interval (CI). And the Risk Ratio (RR) with 95% CI was applied to analyze the dichotomous data among the two groups. P<0.05 represented a statistical significance in all indicators. I2 statistic test was used to detect the heterogeneity among the included studies, when the results shown more than 50%, we seen it as the high heterogeneity with random-effects model (Inverse-Variance method), if meta-analysis was possible. Conversely, we used the fixed-effects model (Inverse-Variance method) to analysis. Unfortunately, only 4 RCTs[14-17] included in our review, sensitivity analysis and publication bias were not been conducted.