首页 | 本学科首页   官方微博 | 高级检索  
     

食管癌高危人群列线图预测模型的建立与应用研究
引用本文:刘兵,孙中明,文进博,缪丹丹,张芹,胡伟,唐勇,潘恩春. 食管癌高危人群列线图预测模型的建立与应用研究[J]. 现代预防医学, 2022, 0(14): 2527-2534. DOI: 10.20043/j.cnki.MPM.202112274
作者姓名:刘兵  孙中明  文进博  缪丹丹  张芹  胡伟  唐勇  潘恩春
作者单位:淮安市疾病预防控制中心,江苏 淮安 223001
摘    要:目的 分析重度异型增生及以上病变(severe dysplasia and above, SDA)的影响因素,建立列线图预测模型,为进一步开展食管癌筛查和防治工作提供依据。方法 选取2009—2017年度淮河癌症早诊早治项目中3个项目点进行上消化道内镜检查(碘染色)并经病理确诊的人群,采用logistic回归分析找出SDA发病的影响因素,根据影响因素建立列线图预测模型。结果 通过单因素和多因素logistic回归分析,总人群风险预测模型显示年龄、人均年收入、直系亲属人数、女性、饮用自来水和经常食用腌晒食品共6个因素与SDA发病密切相关。≤60岁人群中,年龄、人均年收入、饮用自来水共3个因素与SDA发病密切相关。>60岁人群中,年龄、直系亲属人数、BMI、饮用自来水共4个因素与SDA发病密切相关。根据以上影响因素构建列线图预测模型,三种预测模型的AUC分别为0.712、0.714、0.626,均具有较高的预测价值(P<0.05)。结论 SDA列线图预测模型具有较好的准确度、灵敏度及特异度,可以提高食管癌检出率,从而进一步提高癌症早诊早治工作的社会经济效益,促进筛查工作的可持续性发展。

关 键 词:食管癌  早诊早治  回归模型  检出率  列线图

Establishment and application of nomogram prediction model in high risk population of esophageal cancer
LIU Bing,SUN Zhong-ming,WEN Jin-bo,MIAO Dan-dan,ZHANG Qin,HU Wei,TANG Yong,PAN En-chun. Establishment and application of nomogram prediction model in high risk population of esophageal cancer[J]. Modern Preventive Medicine, 2022, 0(14): 2527-2534. DOI: 10.20043/j.cnki.MPM.202112274
Authors:LIU Bing  SUN Zhong-ming  WEN Jin-bo  MIAO Dan-dan  ZHANG Qin  HU Wei  TANG Yong  PAN En-chun
Affiliation:Huai’an Center for Disease Control and Prevention, Huai’an, Jiangsu 223001, China
Abstract:Objective To analyze the influencing factors of severe dysplasia and above lesions (SDA) and establish the nomogram prediction model of SDA, so as to provide a basis for further screening and prevention of esophageal cancer. Methods A logistic regression analysis was used to investigate the influencing factors of the incidence of esophageal cancer and to establish a risk prediction model for esophageal cancer among the patients who underwent upper gastrointestinal endoscopy (iodine staining) and were confirmed by pathology in three projects of early diagnosis and treatment of cancer in Huaihe River from 2009 to 2017. The nomogram prediction model was established according to the influencing factors. Results By single factor and multivariate logistic regression analysis, the total population risk prediction model showed that age, per capita annual income, the number of direct relatives, women, drinking tap water, and frequent consumption of salted food were closely related to the incidence of esophageal cancer. Among people under 60 years old, age, per capita annual income, and drinking tap water were closely related to the incidence of esophageal cancer. Among people over 60 years old, age, number of immediate relatives, BMI and drinking tap water were closely related to the incidence of esophageal cancer. The nomogram prediction model was constructed based on the above influencing factors. The AUC of the three nomogram prediction models were 0.712, 0.714, and 0.626, respectively, which had high predictive value(P<0.05). Conclusion The nomogram prediction model of esophageal cancer has good accuracy, sensitivity, and specificity. It can improve the detection rate of esophageal cancer, further improve the social and economic benefits of early diagnosis and treatment of cancer, and promote the sustainable development of screening.
Keywords:Esophageal cancer  Early diagnosis and treatment  Regression model  Detection rate  Nomogram
点击此处可从《现代预防医学》浏览原始摘要信息
点击此处可从《现代预防医学》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号