医学生烟草依赖的Nomogram风险预测模型构建  

Construction of a risk predictive nomogram model for nicotine dependence among medical students

作　　者：崔梦晶[1] 费高强 孙倩男 钱依宁[1] 蒙沿羱 顾嘉昌 张徐军[1] CUI Meng-jing;FEI Gao-qiang;SUN Qian-nan;QIAN Yi-ning;MENG Yan-yuan;GU Jia-chang;ZHANG Xu-jun(School of Public Health,Southeast University,Nanjing,Jiangsu 210009,China)

机构地区：[1]东南大学公共卫生学院,江苏南京210009

出　　处：《中国预防医学杂志》2021年第2期125-130,共6页Chinese Preventive Medicine

基　　金：江苏省卫生和计划生育委员会医学科研基金资助项目(H201412)。

摘　　要：目的探讨医学生烟草依赖的影响因素,并构建Nomogram风险预测模型。方法 2017年5—6月,研究采用分层随机整群抽样的方法,对江苏省内4所高校医学生进行问卷调查收集有关烟草依赖行为(连续或累积吸烟≥6个月)的142名医学生作为研究组;按照1∶2比例,选择与研究组性别、年级和专业匹配的无烟草依赖行为的284名医学生作为对照组。采用logistic回归分析筛选其独立危险因素,构建Nomogram风险预测模型,并对模型进行验证。结果 142名(3.67%)医学生有烟草依赖行为,其中轻度烟草依赖112人(78.87%),中度烟草依赖26人(18.31%),重度烟草依赖4人(2.82%)。Logistic回归分析显示:少数民族、家庭月收入6 000~10 000元、父亲吸烟、大多数朋友吸烟、父母不反对吸烟、认为吸烟是适应社会的必须、感到焦虑/抑郁、性格外向为医学生烟草依赖的危险因素(OR=3.24, 2.82, 1.86, 13.93, 2.98/5.19, 2.26, 5.79/5.06/14.79, 3.94)。基于上述8项烟草依赖的独立危险因素,构建Nomogram风险预测模型,模型C-index(即AUC)为0.83。结论医学生烟草依赖的主要危险因素为少数民族、家庭月收入高、父亲吸烟、大多数朋友吸烟、父母不反对吸烟、认为吸烟是适应社会的必须、感到焦虑/抑郁、性格外向等。Nomogram风险预测模型具有良好的校准度、区分度和风险预测能力,且有简便、直观和实用等优点,可用于预测个体烟草依赖风险,应用价值较高。Objective To explore the risk factors of nicotine dependence among medical students, and to construct a risk predictive nomogram model. Methods A stratified random cluster sampling method was used to recruit participants from 4 medical universities in Jiangsu province from May 2017 to June 2017, and a questionnaire survey was conducted to gather study associated information.A total of 142 medical students with nicotine dependence behaviors(continuous or cumulative smoking for 6 months and longer) were enrolled in case group, while 284 non-nicotine dependent and gender-grade-major matched students were recruited into control group.Logistic regression analysis was used to screen its independent risk factors, a risk predictive nomogram model was constructed and then verified. Results Among 142(3.67%) medical students in case group, 112(78.87%), 26(18.31%) and 4(2.82%) had mild, moderate and severe nicotine dependence.Multivariate logistic regression analysis showed that ethnic minorities, family monthly income of 6 000-10 000 RMB, father smoking, friends smoking, parents not against smoking, a thought of smoking a social must, feeling anxious/depressed, and outgoing personality were the major risk factors for nicotine dependence in medical students(OR=3.24, 2.82, 1.86, 13.93, 2.98/5.19, 2.26, 5.79/5.06/14.79, 3.94).Based on these risk factor, a risk predictive nomogram model was constructed and it′s C-index(AUC) was 0.83.Conclusions A variety of risk factors involve in the nicotine dependence of medical students.The risk predictive nomogram model constructed based on risk factors has strong calibration and discrimination ability,which can be simply used to predict the nicotine independence risk of an individual.

关 键 词：烟草依赖 医学生 吸烟 烟草依赖评估量表 病例对照 Nomogram风险预测模型 

分 类 号：R181[医药卫生—流行病学]