BPG is committed to discovery and dissemination of knowledge
Meta-Analysis
©Author(s) (or their employer(s)) 2026.
World J Gastrointest Surg. Feb 27, 2026; 18(2): 113979
Published online Feb 27, 2026. doi: 10.4240/wjgs.v18.i2.113979
Figure 1
Figure 1 The selection process of studies for the meta-analysis. CNKI: China National Knowledge Infrastructure; RCT: Randomized controlled trial.
Figure 2
Figure 2 Forest plot. A: Cure rate. Forest plot depicting pooled relative risk (RR) of cure following ligation of the intersphincteric fistula tract (LIFT) vs conventional surgery, with 95% confidence intervals (CIs). Each square represents an individual study estimate, and the diamond indicates the overall pooled effect; B: Recurrence rate. Forest plot depicting the pooled RR for recurrence between LIFT and conventional surgery, with 95%CIs. A lower RR indicates reduced recurrence in the LIFT group; C: Complication rate. Forest plot comparing the complication rates between LIFT and conventional surgery. The pooled RR and 95%CIs demonstrate the safety profile of LIFT; D: Postoperative pain. Forest plot of pooled mean differences in the postoperative pain scores (Visual Analog Scale) between LIFT and conventional surgery, with 95%CIs. Negative values favor LIFT. LIFT: Ligation of the intersphincteric fistula tract; CI: Confidence intervals; RR: Relative risk.
Figure 3
Figure 3 Sensitivity analysis for complication rate. Leave-one-out analysis was conducted by sequentially omitting each of the included studies. The pooled relative risk and 95% confidence intervals remained stable, indicating that no single trial significantly influenced the overall estimate. CI: Confidence interval.
Figure 4
Figure 4 Funnel plots for publication bias assessment. Funnel plots evaluating the potential publication bias across the following eight outcomes: Healing time, postoperative pain, operative time, overall clinical effectiveness, anal function, recurrence rate, complication rate, and cure rate. Symmetrical distributions suggest no significant risk of publication bias.


Write to the Help Desk