基于LASSO回归对非小细胞肺癌患者术后恶心呕吐危险因素分析
Analysis of risk factors for postoperative nausea and vomiting in non-small cell lung cancer patients based on LASSO regression
李会娟 1赵迪 1王文华 1景向娜 1杨淑君2
作者信息
- 1. 郑州大学第一附属医院麻醉与围手术期及疼痛医学部,郑州 450052
- 2. 郑州大学第一附属医院肿瘤科,郑州 450052
- 折叠
摘要
目的 基于最小绝对收缩和选择算子(LASSO)回归构建非小细胞肺癌患者PONV预测列线图并验证其预测效果.方法 收集2021年4月至2022年2月郑州大学第一附属医院胸外科接受肺叶切除术成年非小细胞肺癌患者临床数据.选取符合纳入标准的738例患者,根据随机数表,患者按7∶3的比例分为训练队列(517例)和验证队列(221例).在对训练队列进行LASSO回归分析后,筛选PONV的独立危险因素,建立列线图模型.使用受试者工作特征曲线下面积(AUC)、校准曲线、决策曲线分析(DCA)来评估该模型的预测效率、准确性和临床实用性.结果 通过LASSO回归筛选非小细胞肺癌患者PONV 3个独立的危险因素:性别、体重及单核细胞计数,并且构建PONV风险预测列线图,该模型训练集和内部验证集的ROC曲线下面积分别为0.687[95％可信区间(CI):0.641～0.714],0.683(95％CI:0.656～0.725);校准图显示预测观测值与实际观测值之间具有良好一致性;DCA的结果显示阈值概率8％～25％时,该模型可以最大临床获益.结论 本研究构建并验证非小细胞肺癌患者PONV列线图的预测效能和临床实用性.
Abstract
Objective To construct a predictive nomogram for postoperative nausea and vomiting(PONV)in non-small cell lung cancer(NSCLC)patients based on least absolute shrinkage and selection operator(LASSO)regression and validate its predictive effect.Methods Retrospective clinical data of a-dult NSCLC patients who underwent pulmonary lobectomy surgery at the First Affiliated Hospital of Zheng-zhou University from April 2021 to February 2022 were collected.A total of 738 eligible patients were se-lected and divided into a training cohort(n=517)and a validation cohort(n=221)in a 7∶3 ratio using a random number table method.After performing LASSO regression analysis on the training cohort,inde-pendent risk factors for PONV were identified,and a nomogram model was established.The predictive effi-ciency,accuracy,and clinical utility of the model were evaluated using the area under the receiver operat-ing characteristic curve(AUC),calibration curve,and decision curve analysis(DCA).Results LASSO regression identified three independent risk factors for PONV in NSCLC patients:gender,weight,and monocyte count.A nomogram predicting PONV risk was constructed,with AUCs of 0.687[95％confi-dence interval(CI):0.641-0.714]for the training set and 0.683(95％CI:0.656-0.725)for the inter-nal validation set.The calibration plot showed good consistency between predicted and observed values.DCA results indicated maximum clinical benefit when the threshold probability ranged from 8％to 25％.Conclusion This study constructed and validated the predictive efficacy and clinical utility of a PONV no-mogram for NSCLC patients.The predictive model provides a reliable basis for the prevention and treatment of PONV in NSCLC patients.
关键词
非小细胞肺癌/术后恶心呕吐Key words
Non-small cell lung cancer/Postoperative nausea and vomiting引用本文复制引用
出版年
2024
NETL
NSTL
万方数据