安徽省结直肠癌发病影响因素分析及风险评估模型的构建

Analysis of influence factors and risk assessment model of colorectal cancer in Anhui Province

目的探讨结直肠癌发病的影响因素,构建一个稳定且可靠的结直肠癌风险评估模型。方法采用Logistic回归模型分析结直肠癌的影响因素,结合哈佛风险指数模型和年龄因子加权指数构建结直肠癌的风险评估模型,根据受试者工作曲线(ROC)下面积(AUC)大小选择最优的预测模型。结果确诊过肠息肉、大便带血等是结直肠癌的危险因素,确诊过胃炎、经常紧张等是结直肠癌的保护因素,结直肠癌的风险评估模型ROC曲线的AUC=0.789(95%CI为0.764~0.813),预测效果较好。结论结直肠癌的影响因素众多,预测能力较好的风险评估模型有利于帮助基层医生快速确定高危人群,为实施个性化的干预措施提供可能。

Objective:To explore the influence factors of colorectal cancer,build a stable and reliable risk assessment model of colorectal cancer,and to provide rapid screening colorectal cancer high-risk groups.Methods:Logistic regression model was used to analyze the influence factors of colorectal cancer,with Harvard risk index model and the age factor weighted index,build risk assessment model of colorectal cancer and choose the best the optimal prediction model according to the ROC curve area size.Results:Diagnosis of intestinal polyps and bloody stool are risk factors of colorectal neoplasms,and confirmed gastritis and nervous feeling always are protective factors of colorectal cancer.AUC of ROC curve of Colorectal cancer risk assessment model was 0.789(95%CI 0.764 to 0.813)and the prediction effect was excellent.Conclusion:The influence factors of colorectal cancer is numerous,and a better risk assessment model is conducive to help grassroots doctors quickly identify high-risk groups to make the implementation of personalized intervention possible.