Construction and validation of a column chart prediction model for the factors of postoperative recurrence of com-plex anal fistula
10.3969/j.issn.1009-9905.2024.12.008
- VernacularTitle:复杂性肛瘘术后复发因素及列线图预测模型的构建与验证
- Author:
Yang LIU
1
;
Shao-wen WANG
1
;
Hua CHEN
1
;
Xin-chun LI
1
;
Wen-tao HE
1
;
Yan-min CHEN
1
;
Ting TIAN
1
Author Information
1. 中南大学湘雅医学院附属常德医院(常德市第一人民医院)普通外科(湖南 常德 415000)
- Publication Type:Journal Article
- Keywords:
Anal fistula,complexity;
Recurrence;
Influencing factors;
Nomogram prediction model
- From:
Chinese Journal of Current Advances in General Surgery
2024;27(12):962-966
- CountryChina
- Language:Chinese
-
Abstract:
Objective:Toexplore the influencing factors of postoperative recurrence of com-plex anal fistula surgery and construct a column chart model.Methods:A total of 421 patients undergoing complex anal fistula surgery admitted to our hospital from March 2018 to October 2023 were selected as the study objects,and were divided into the modeling group(n=295)and the verification group(n=126)according to a ratio of about 7∶3.In addition,the patients in the modeling group were divided into the non-recurrence group(n=252)and the recurrence group(n=43).The influencing factors of postoperative recurrence of complex anal fistula in the modeling group were analyzed by multi-factor Logistic regression RMS package in R software was applied to construct a column chart prediction model for the risk of postoperative recurrence of complex anal fistula.The predictive performance of the model was evaluated using ROC curves,H-L fit tests,and calibration curves.Results:There was no significant difference between the modeling group and the validation group in terms of comorbidities,anal fistula location,surgical method,and whether to regularly change dressing at the hospital after surgery(P＞0.05).There were obvious differences between the recurrence group and the non recurrence group in terms of the location of anal fistula(x2=6.259),history of anal fistula surgery(x2=6.310),presence of perianal abscess(x2=5.287),and regularly visit the hospital for dressing changes after surgery(x2=3.974)(P＜0.05).Multi-variate logistic regression showed that high anal fistula(OR=8.232),history of anal fistula surgery(OR=11.658),concurrent perianal abscess(OR=11.245),and failure to regularly visit the hospital for dressing changes after surgery(OR=5.997)were independent risk factors for postoperative recur-rence of complex anal fistula(P＜0.05).The area under ROC curve of the modeling group was 0.840(95％CI:0.793～0.880),and the sensitivity and specificity were 81.40％and 86.51％,respectively.In the verification group,the area under ROC curve was 0.858(95％CI:0.785～0.914),and the sen-sitivity and specificity were 87.50％and 81.82％,respectively The two calibration curves showed minor difference between the predicted and actual values,and the H-L fit test showed well con-sistency in the model predictions.Conclusion:Complex anal fistula patients who have a high-position anal fistula,history of anal fistula surgery,concurrent perianal abscess,and failure to regu-larly visit the hospital for dressing changes after surgery have a higher risk of postoperative recur-rence.A risk model constructed based on these factors is helpful for clinical prediction and inter-vention,reducing the risk of postoperative recurrence.
