结直肠癌保肛术后低位前切综合征风险预测模型的系统评价

A systematic review of the risk prediction model for low anterior resection syndrome after anal preservation for colorectal cancer

目的系统评价结直肠癌保肛术后低位前切综合征风险预测模型。方法系统检索中国知网、万方数据知识服务平台、中国期刊服务平台、中国生物医学文献数据库、Embase、PubMed、Web of Science和The Cochrane Library,检索时限为建库至2023年4月16日。由2名研究者独立进行文献筛选和提取结直肠癌保肛术后低位前切综合征预测模型相关数据,根据预测模型研究偏倚风险评估工具(predictive model research bias risk assessment tool,PROBAST)对纳入文献的偏倚风险和适用性进行评价。结果共纳入13篇文献,涉及17个结直肠癌保肛术后低位前切综合征预测模型,各模型受试者工作特征曲线的曲线下面积均大于0.7(0.707~0.858),其中13个为Logistic回归模型,支持向量机、决策树、随机森林、人工神经网络模型各1个。5篇文献进行了内部验证,4篇文献进行了外部验证。PROBAST工具评估结果显示,纳入的13篇文献均为高偏倚风险;模型适用性方面,2篇为不清楚,1篇为高适用性,10篇为低适用性。模型重复报告的独立预测因子频率高的前3个因子为术前新辅助治疗、低位肿瘤和吻合口漏。结论结直肠癌保肛术后低位前切综合征预测模型具有较好的预测性能,但整体偏倚风险高,建议未来研究对现有模型进行验证和更新,并开发偏倚风险低、临床实用性强的本土化风险预测模型。

Objective To systematically evaluate the prediction model of low anterior resection syndrome after anal preserving surgery for colorectal cancer,and to provide reference for the construction and application of related risk prediction models for medical staff.Method Systematic searches were conducted on CNKI,Wanfang Data Knowledge Service Platform,China Journal Service Platform,China Biomedical Literature Database,Embase,PubMed,Web of Science and The Cochrane Library.The search period was from database establishment to April 16,2023.Two researchers independently screened the literature and extracted data related to the prediction model of low anterior incision syndrome after anus-preserving surgery for colorectal cancer,and evaluated the risk of bias and applicability of the included literature according to the predictive model research bias risk assessment tool(PROBAST).Result A total of 13 articles were included,involving 17 prediction models of low anterior incision syndrome after anus-preserving surgery for colorectal cancer,and the area under the receiver operating characteristic curve of each model was greater than 0.7(0.707-0.858).Among them,13 were logistic regression models,and support vector machine,decision tree,random forest,and artificial neural network model was 1 respectively.5 papers were internally validated and 4 were externally verified.The results of the PROBAST showed that the 13 included studies were at high risk of bias.In terms of model applicability,2 papers were unclear,1 was high applicability,and 10 were low applicability.The top three factors with high frequency of independent predictors reported by the model were preoperative neoadjuvant therapy,low-lying tumors,and anastomotic leak.Conclusion The prediction model of low anterior resection syndrome after anal preserving surgery for colorectal cancer has good predictive performance,but the overall risk of bias is high.It is suggested that future studies should validate and update the existing model,and develop a localized risk prediction model with low risk of bias and strong clinical practicability.