betway必威登陆网址 (betway.com )学报››2022,Vol. 43››Issue (1): 56-62.DOI:10.3969/j.issn.2097-0005.2022.01.015

• 公共卫生与管理学 •上一篇下一篇

食管癌发病影响因素分析及风险预测模型的构建

张可昕(), 柴静, 沈兴蓉, 刘荣, 王德斌()

  1. 安徽医科大学卫生管理学院,安徽 合肥 230032
  • 收稿日期:2021-08-18出版日期:2022-01-25发布日期:2022-02-23
  • 通讯作者:王德斌
  • 作者简介:张可昕,硕士研究生,研究方向:社会医学与卫生事业管理,E-mail:zkx1176126994@163.com
  • 基金资助:
    国家自然科学基金(71774002);城市癌症早诊早治项目(K2020034);国家自然科学基金(71503008)

Analysis of influencing factors of esophageal cancer and construction of risk prediction model

Kexin ZHANG(), Jing CHAI, Xingrong SHEN, Rong LIU, Debin WANG()

  1. Anhui Medical University,Hefei 230032,China
  • Received:2021-08-18Online:2022-01-25Published:2022-02-23
  • Contact:Debin WANG

摘要: 目的

探索食管癌((esophageal cancer,EC)发病的影响因素并构建EC风险预测模型。

方法

采用基于社区的分层随机抽样方法纳入研究对象。利用logistic回归分析探讨EC的影响因素,并运用改良哈佛风险指数与年龄-因子加权指数构建EC风险预测模型。

结果

长期居住地是农村、体重过轻、吞咽轻度疼痛或不顺畅、经常食用的不健康食物类型 ≥ 2种、不良饮食喜好 ≥ 2种、每月饮白酒量 > 60两、熬夜、情绪易怒是EC的危险因素(P< 0.05);体重过重、确诊过糖尿病、高血压或高血脂、慢性胃炎,规律饮食、情绪易紧张是EC的保护因素(P< 0.05)。构建的EC风险预测模型受试者工作曲线下面积(area under the ROC curve,AUC)为0.913,95%CI为0.897 ~ 0.930。

结论

EC是多因素共同作用的结果,本研究构建的EC风险预测模型具有较好的预测效果,可为整合资源开展EC筛查和“三早”防治工作提供适宜的理论依据。

关键词:食管癌,影响因素,预测模型

Abstract: Objective

To explore the influencing factors of esophageal cancer and to construct the risk prediction model of esophageal cancer.

Methods

A community-based stratified random sampling method was used to select research objects. Logistic regression analysis was used to explore the influencing factors of esophageal cancer. The modified Harvard cancer risk index and age-specific weighted index were used to construct the risk prediction model.

Results

Long-term residence in rural areas, underweight, mild pain or difficulty in swallowing, more than 2 types of unhealthy foods often eaten, more than 2 bad diet preferences, more than 60 taels of monthly alcohol consumption, stay uping late, emotional irritability are risk factors for esophageal cancer (P< 0.05). Overweight, diagnosed diabetes, hypertension or hyperlipidemia, diagnosed chronic gastritis, regular diet and emotional stress are the protective factors for esophageal cancer (P< 0.05). The area under the ROC curve (AUC) of the constructed esophageal cancer risk prediction model was 0.913, and the 95 %CIwas 0.897~0.930.

Conclusion

Esophageal cancer is the result of multiple factors. The risk prediction model of esophageal cancer constructed in this study has good predictive effect, which can provide a suitable theoretical basis for integrating resources to carry out esophageal cancer screening and “three early” prevention and control work.

Key words:esophageal cancer,influencing factors,prediction model