Outcomes and Predictors of Fistula Closure After Laser-Assisted Surgery for Anal Fistula: A Retrospective Cohort Study from Nepal

Introduction: Laser-assisted surgery is an emerging option as a sphincter-sparing procedure for anal fistula, with varied treatment outcomes in the literature. The aim of the study was to evaluate the treatment outcomes and complications of laser-assisted surgery for anal fistula and develop a predictive model for fistula closure. Methods: This was a retrospective cohort study of 86 patients who underwent laser-assisted surgery for anal fistula from 14th April 2024 to 12th April 2025 at Sumeru Hospital, Dhapakhel, Lalitpur, Nepal. Demographic, fistula characteristics and clinical variables were recorded in Microsoft Excel and variables were compared between fistula closure and non-closure groups. Independent predictors were identified using logistic regression model. A predictive model was developed using predictors which were significant in univariate analysis. Performance of the model was evaluated using receiver operating characteristic (ROC) curve analysis and a web-based predictive tool was developed. Results: Among 86 cases, urinary retention was seen in 2(2.3%), bleeding in 2(2.3%), and post-operative wound infection in 14(16.3%) cases. Fistula closure was achieved in 72 (83.7%) patients. On univariate analysis, patients with higher grade fistula (p < 0.001), longer fistula tract (p < 0.001), branched fistula tract (p < 0.001), inflammation at the time of surgery (p < 0.001), and post-operative wound infection (p < 0.001) were associated with failure to close fistula. However, multivariable logistic regression analysis identified branched fistula tract (p=0.033; OR= 0.049; 95% CI: 0.003-0.783) and post-operative wound infection (p = 0.017; OR = 0.044; 95% CI: 0.003-0.575) as independent predictors for failure to close fistula (R2 = 0.772). Area under the ROC curve was 0.92 for the predictive model which showed excellent ability to predict fistula closure. Conclusion: Laser-assisted surgery is associated with low complications with high closure rates. This predictive model is effective to predict fistula closure with high accuracy. However, this model needs external validation with large dataset.