Predictors of breast cancer screening utilization among female at high risk of developing breast cancer: application of a Lasso Logistic model
-
摘要: 目的 探讨Lasso Logistic回归模型在乳腺癌高风险人群筛查利用相关因素研究中的应用。方法 基于健康风险评估模型筛选乳腺癌高风险人群,利用Lasso Logistic回归模型进行变量选择,通过交叉验证选择模型中的最优调和参数λ,再建立传统Logistic回归模型分析筛查利用情况的影响因素。结果 经健康风险评估模型筛选后,共纳入771名乳腺癌高风险人群,乳腺癌筛查利用率为72.1%。交叉验证选择的最优λ为0.044,经Lasso Logistic回归模型进行变量筛选后纳入的自变量为年龄、文化程度、既往乳腺疾病史和乳房自检行为,赤池信息准则(akaike information criterion,AIC)和贝叶斯信息准则(bayesian information criterion,BIC)分别为762.44和785.68,均低于传统Logistic回归模型(762.73,804.55)。结论 Lasso Logistic回归模型可用于乳腺癌高风险人群筛查利用情况相关因素研究。年龄、文化程度、既往乳腺疾病史和乳房自检行为影响乳腺癌高风险人群的筛查利用情况。
-
关键词:
- Lasso Logistic回归模型 /
- 乳腺癌筛查 /
- 相关因素
Abstract: Objective To evaluate the application of a Lasso Logistic model for potential factors related with breast cancer screening among female at high risk of developing breast cancer. Methods A health risk appraisal model was used to measure women's objective risk of developing breast cancer. Lasso Logistic model was used to analyze the predictors of breast cancer screening utilization. Cross validation method was used to choose λ for Lasso Logistic model. In addition, the akaike information criterion (AIC) and bayesian information criterion (BIC) were chosen to evaluate the model fitting of Lasso Logistic model, compared with full Logistic model. Results Of the 771 women at high risk of developing breast cancer, 72.1% attended screening. The cross validation resulted in λ=0.044 for Lasso Logistic model. Age, education level, personal history of breast disease and breast self-examination were associated with breast cancer screening utilization in Lasso Logistic model and full Logistic model. AIC of Lasso Logistic model and full Logistic model were 762.44 and 762.73, respectively. And BIC were 785.68 and 804.55, respectively. Conclusions Lasso Logistic model was a sound fitting model for predictors of breast cancer screening utilization among women at high risk of developing breast cancer. Age, education level, personal history of breast disease and breast self-examination were predictors of breast cancer screening utilization.-
Key words:
- Lasso Logistic model /
- Breast cancer screening /
- Relative fators
期刊类型引用(18)
1. 闫慈,古丽努尔·阿卜杜热合曼,张旭,孙刚. 基于LASSO变量选择联合贝叶斯网络构建乳腺癌患者5年预后风险模型的建立与预测. 重庆医学. 2024(03): 405-410+417 . 百度学术
2. 邵文倩,钱金平,徐伟民. 临床护士乳腺癌筛查保护动机现状及影响因素分析. 护士进修杂志. 2024(02): 135-140 . 百度学术
3. 戎成振,卢家忠,杨静静. 基于Nomgram模型对老年阵发性房颤病人合并高尿酸血症临床预测模型的初步探索. 蚌埠医学院学报. 2024(03): 352-357 . 百度学术
4. 张茹,吴高春,张星光,张子英,闫涛,田梓璇,张楠. 内蒙古自治区兴安盟35~64岁农牧区妇女乳腺癌高危人群检出情况及其影响因素分析. 中国公共卫生. 2024(05): 540-544 . 百度学术
5. 蔡紫庭,王梦婷,闫慧姣,王苏蒙,张钹,张乐,白志荣,乔友林,王辰. 鄂尔多斯市妇幼保健院医务人员乳腺癌认知现状及其影响因素分析. 中国公共卫生. 2024(06): 742-749 . 百度学术
6. 蒲星月,马原,钟志刚. 2006—2020年中国女性乳腺癌死亡趋势分析——基于年龄-时期-出生队列模型. 卫生经济研究. 2023(02): 28-33 . 百度学术
7. 陈寿莉,王国蓉,李巧巧,张甜,王映印,刘瑶,Tsu-Yin Wu. 四川省女性乳腺癌筛查参与行为现状及影响因素分析. 现代预防医学. 2023(02): 272-276+291 . 百度学术
8. 王莹,马霞,宫舒萍,蔡雷琦,亓爱玲,张先慧. 2011—2021年山东省济南市女性乳腺癌死亡趋势分析. 中国肿瘤. 2023(10): 766-772 . 百度学术
9. 林倩,陈爱华,张婷婷. DCE-MRI纹理分析对乳腺癌分子分型的诊断价值. 磁共振成像. 2023(12): 40-48 . 百度学术
10. 张俐丽,冯国琴. 个体化预测非小细胞肺癌患者化疗期间肺部感染风险Nomogram模型的建立与验证. 中国感染控制杂志. 2022(02): 171-179 . 百度学术
11. 李兰君,刘赪,赵联文. 基于惩罚极大Lq似然的Logistic回归系数的估计及变量选择. 甘肃科学学报. 2022(03): 21-25 . 百度学术
12. 黄冬玲,刘奕仕,黄雪韵,梁玉玲,杜婉燕. 2015—2017年东莞市石碣镇乳腺癌筛查率及影响因素分析. 实用预防医学. 2021(01): 77-80 . 百度学术
13. 张娜,刘朝兴. 石家庄市乳腺癌的筛查结果分析. 当代医学. 2021(19): 102-105 . 百度学术
14. 刘妍琛,张晓曙,崔旭东,金娜,赵祥凯,赵昕,郑洪淼,李娟生,申希平,孟蕾,任晓卫. 基于Group LASSO Logistic回归分析模型分析流行性乙型脑炎早期临床症状与预后的关联. 中华疾病控制杂志. 2021(08): 891-897+934 . 本站查看
15. 黄娅,李运明,雷丽. Lasso Logistic回归模型识别脂肪肝风险因素效果研究. 甘肃科学学报. 2021(04): 45-51 . 百度学术
16. 何明艳,朱碧琪,钟媛,王雷,杨柳,廖先珍,让蔚清. 2005-2013年中国女性乳腺癌发病及死亡趋势分析. 中华疾病控制杂志. 2019(01): 10-14 . 本站查看
17. 康文博,赵静雅,吕雪峰,陈勇,韩雪琳,田曙光,陈芳艳,苏雪婷,王洪源,韩黎. Lasso-logistic模型在医院下呼吸道感染预测中的应用. 中国感染控制杂志. 2019(07): 619-624 . 百度学术
18. 朱珠,刘晴,庹吉妤,张敏,谭晓东,余秋丽,李沁梅. 柴湖地区食管癌及癌前病变风险评估模型建立. 现代预防医学. 2019(15): 2701-2704+2712 . 百度学术
其他类型引用(26)
-

计量
- 文章访问数: 1021
- HTML全文浏览量: 141
- PDF下载量: 118
- 被引次数: 44