基于Nomogram对男性痛风患病风险预测模型的构建及验证  

Construction and validation of a Nomogram-based prediction model for the risk of gout in men

作　　者：陈昱名 蒋坪飞 刘璐 何爽 夏依代·图尔荪 李珍珍 苗蕾[1] Chen Yuming;Jiang Pingfei;Liu Lu;He Shuang;Xiayidai·Tuersun;Li Zhenzhen;Miao Lei(College of Public Health,Xinjiang Medical University,Urumqi 830000,China;The First Affiliated Hospital of Xinjiang Medical University,Urumqi 830000,China)

机构地区：[1]新疆医科大学公共卫生学院,乌鲁木齐830000 [2]新疆医科大学第一附属医院,乌鲁木齐830000

出　　处：《中华内分泌代谢杂志》2023年第4期310-314,共5页Chinese Journal of Endocrinology and Metabolism

基　　金：国家自然科学基金项目(81460153)。

摘　　要：目的探讨痛风的危险因素及建立预测痛风发病风险的列线图模型。方法选取2018年至2020年新疆医科大学附属中医医院、新疆维吾尔自治区人民医院、新疆医科大学第一附属医院就诊的1032例汉族男性为研究对象,采用简单随机抽样方法按7∶3比例分为训练集(722例)和验证集(310例)。收集受试者的一般资料和生化指标。收集的资料用于评估痛风患病的风险。使用RStudio软件,用LASSO回归分析来筛选最佳预测因子,引入从LASSO回归筛选出的预测因子,构建预测痛风风险列线图模型,采用受试者工作特征(ROC)曲线,Hosmer-Lemeshow检验评估列线图模型的区分度和校准度。最后采用rmda程序包进行决策曲线分析(DCA),以在验证数据中评估模型的临床实用性。结果年龄、尿酸、体重指数、总胆固醇、腰臀比是痛风的危险因素(P<0.05)。基于上述5个独立危险因素建立的列线图预测模型具有较好的区分度(AUC值:训练集验证为0.923,验证集验证为0.922)和准确度(Hosmer-Lemeshow检验:验证集验证P>0.05);决策曲线分析显示预测模型曲线在大于15%的阈值概率区间具有临床实用价值。结论联合年龄、尿酸、体重指数、总胆固醇、腰臀比指标建立的列线图模型可以较为准确地预测痛风发病风险。Objective To investigate the risk factors of gout and establish a columnar graph model to predict the risk of gout development.Methods A total of 1032 Han Chinese men attending the Affiliated Hospital of Traditional Chinese Medicine of Xinjiang Medical University,People′s Hospital of Xinjiang Uygur Autonomous Region,and the First Affiliated Hospital of Xinjiang Medical University from 2018 to 2020 were selected as study subjects and divided into training set(722 cases)and validation set(310 cases)by simple random sampling method in the ratio of 7∶3.General information and biochemical indices of the subjects were collected.The collected information was used to assess the risk of gout prevalence.LASSO regression analysis of R Studio software was used to screen the best predictors,and was introduced to construct a column line graph model for predicting gout risk using receiver operating characteristic(ROC)curves,and the Hosmer-Lemeshow test was used to assess the discrimination and calibration of the column line graph model.Finally,decision curve analysis(DCA)was performed using the rmda program package to assess the clinical utility of the model in validation data.Results Age,uric acid,body mass index,total cholesterol,and waist-to-hip ratio were risk factors for gout(P<0.05).The column line graph prediction model based on the above five independent risk factors had good discrimination(AUC value:0.923 for training set validation and 0.922 for validation set validation)and accuracy(Hosmer-Lemeshow test:P>0.05 for validation set validation);decision curve analysis showed that the prediction model curve had clinical practical value.Conclusion The nomogram model established by combining age,uric acid,body mass index,total cholesterol,and waist-to-hip ratio indicators can predict the risk of gout more accurately.

关 键 词：痛风 LASSO回归 列线图 风险模型 

分 类 号：R589.7[医药卫生—内分泌]