浏览全部资源
扫码关注微信
1.西安国际医学中心医院药学部,西安 710100
2.长安大学信息工程学院,西安 710064
3.长安大学经济与管理学院,西安 710064
主管药师,硕士。研究方向:精准药学服务。E-mail:cxi9@foxmail.com
主任药师,硕士。研究方向:精准药学服务与药事管理。E-mail:aliceyuanhailing@163.com
收稿日期:2025-01-22,
修回日期:2025-05-06,
录用日期:2025-05-07,
纸质出版日期:2025-06-15
移动端阅览
陈曦,袁申奥,袁海玲,等.丙戊酸血药浓度预测的小样本多分类机器学习模型对比 [J].中国药房,2025,36(11):1399-1404.
CHEN Xi,YUAN Shen’ao,YUAN Hailing,et al.Comparison of small-sample multi-class machine learning models for plasma concentration prediction of valproic acid[J].ZHONGGUO YAOFANG,2025,36(11):1399-1404.
陈曦,袁申奥,袁海玲,等.丙戊酸血药浓度预测的小样本多分类机器学习模型对比 [J].中国药房,2025,36(11):1399-1404. DOI: 10.6039/j.issn.1001-0408.2025.11.20.
CHEN Xi,YUAN Shen’ao,YUAN Hailing,et al.Comparison of small-sample multi-class machine learning models for plasma concentration prediction of valproic acid[J].ZHONGGUO YAOFANG,2025,36(11):1399-1404. DOI: 10.6039/j.issn.1001-0408.2025.11.20.
目的
2
构建用于预测丙戊酸(VPA)血药浓度的三分类(不足、正常、超限)和二分类(不足、正常)模型,并比较这2种模型的性能,为临床制定用药方案提供参考。
方法
2
收集2022年11月-2024年9月在西安国际医学中心医院接受VPA治疗并进行血药浓度检测的480名患者的临床数据(共695份数据)。分别针对三分类和二分类模型的目标变量构建预测模型,利用XGBoost特征重要性评分进行特征排名和选取,采用12种机器学习算法进行训练和验证,并通过准确率、F1分数及受试者工作特征曲线下面积(AUC)3个指标对模型的性能进行评价。
结果
2
在三分类模型中,合并肾病和合并电解质紊乱的XGBoost特征重要性评分排名较高;然而在二分类模型中,这些特征的重要性排名显著降低,提示其与VPA血药浓度超限之间存在紧密的关联。在三分类模型中,随机森林法表现最佳,但其测试集F1分数仅达到0.704 0,AUC仅为0.519 3;而在二分类模型中,CatBoost方法表现最佳,其测试集F1分数为0.785 7,AUC达到了0.819 5。
结论
2
本研究构建的三分类模型具有预测VPA血药浓度超限的能力,但预测及模型泛化能力较差;构建的二分类模型仅能对血药浓度不足和正常情况进行分类预测,但模型预测性能较强。
OBJECTIVE
2
To construct three-class (insufficient, normal, excessive) and two-class (insufficient, normal) models for predicting plasma concentration of valproic acid (VPA), and compare the performance of these two models, with the aim of providing a reference for formulating clinical medication strategies.
METHODS
2
The clinical data of 480 patients who received VPA treatment and underwent blood concentration test at the Xi’an International Medical Center Hospital were collected from November 2022 to September 2024 (a total of 695 sets of data). In this study, predictive models were constructed for target variables of three-class and two-class models. Feature ranking and selection were carried out using XGBoost scores. Twelve different machine learning algorithms were used for training and validation, and the performance of the models was evaluated using three indexes: accuracy, F1 score, and the area under the working characteristic curve of the subject (AUC).
RESULTS
2
XGBoost feature importance scores revealed that in the three-class model, the importance ranking of kidney disease and electrolyte disorders was higher. However, in the two-class model, the importance ranking of these features significantly decreased, suggesting a close association with the excessive blood concentration of VPA. In the three-class model, Random Forest method performed best, with F1 score of 0.704 0 and AUC of 0.519 3 on the test set; while in the two-class model, CatBoost method performed optimally, with F1 score of 0.785 7 and AUC of 0.819 5 on the test set.
CONCLUSIONS
2
The constructed three-class model has the ability to predict excessive VPA blood concentration, but its prediction and model generalization abilities are poor; the constructed two-class model can only perform classification prediction for insufficient and normal blood concentration cases, but its model performance is stronger.
MARSON A , BURNSIDE G , APPLETON R , et al . The SANAD Ⅱ study of the effectiveness and cost-effectiveness of valproate versus levetiracetam for newly diagnosed generalised and unclassifiable epilepsy:an open-label,non-inferiority,multicentre,phase 4,randomised controlled trial [J ] . Lancet , 2021 , 397 ( 10282 ): 1375 - 1386 .
MARSON A G , BURNSIDE G , APPLETON R , et al . Lamotrigine versus levetiracetam or zonisamide for focal epilepsy and valproate versus levetiracetam for generalised and unclassified epilepsy:two SANAD Ⅱ non-inferiority RCTs [J ] . Health Technol Assess , 2021 , 25 ( 75 ): 1 - 134 .
陈策 , 郑丽丹 , 谢作良 , 等 . 双相情感障碍住院患者的临床特征及用药情况的调查研究 [J ] . 中国全科医学 , 2020 , 23 ( 2 ): 245 - 250 .
HIEMKE C , BERGEMANN N , CLEMENT H W , 等 . 神经精神药理学治疗药物监测共识指南: 2017版 [J ] . 实用药物与临床 , 2022 , 25 ( 2 ): 97 - 118 .
LI Z , GAO W , LIU G , et al . Interaction between valproic acid and carbapenems:decreased plasma concentration of valproic acid and liver injury [J ] . Ann Palliat Med , 2021 , 10 ( 5 ): 5417 - 5424 .
钱金 , 陈顺 , 谢新芳 , 等 . 建立丙戊酸钠在中国癫痫患者中的群体药代动力学模型 [J ] . 第二军医大学学报 , 2017 , 38 ( 11 ): 1449 - 1453 .
METHANEETHORN J . A systematic review of population pharmacokinetics of valproic acid [J ] . Br J Clin Pharmacol , 2018 , 84 ( 5 ): 816 - 834 .
张仲斌 , 吴晔 , 季双敏 , 等 . 不同年龄组中国癫痫患儿丙戊酸群体药代动力学模型的建立 [J ] . 中华实用儿科临床杂志 , 2014 , 29 ( 9 ): 698 - 703 .
OGUNGBENRO K , AARONS L , CRESim & Epi-CRESim Project Groups . A physiologically based pharmacokinetic model for valproic acid in adults and children [J ] . Eur J Pharm Sci , 2014 , 63 : 45 - 52 .
MA H Y , HUANG S H , LI F X , et al . Development and validation of an automatic machine learning model to predict abnormal increase of transaminase in valproic acid-treated epilepsy [J ] . Arch Toxicol , 2024 , 98 ( 9 ): 3049 - 3061 .
MA P , SHANG S L , HUANG Y F , et al . Joint use of population pharmacokinetics and machine learning for prediction of valproic acid plasma concentration in elderly epileptic patients [J ] . Eur J Pharm Sci , 2024 , 201 : 106876 .
DAMNJANOVIĆ I , TSYPLAKOVA N , STEFANOVIĆ N , et al . Joint use of population pharmacokinetics and machine learning for optimizing antiepileptic treatment in pediatric population [J ] . Ther Adv Drug Saf , 2023 , 14 : 20420986231181337 .
HSU C W , LAI E C , CHEN Y B , et al . Valproic acid monitoring:serum prediction using a machine learning framework from multicenter real-world data [J ] . J Affect Disord , 2024 , 347 : 85 - 91 .
杜欣 , 李欣燚 , 谢晓慧 , 等 . 碳青霉烯类药物对丙戊酸血药浓度影响及两药联用后癫痫发作影响因素:文献Meta分析 [J ] . 中国医院药学杂志 , 2021 , 41 ( 4 ): 395 - 399 .
林小燕 , 林燕玲 , 庄俊鹏 . 癫痫患者丙戊酸钠血药浓度的影响因素 [J ] . 临床合理用药 , 2024 , 17 ( 36 ): 59 - 62 .
张文滨 , 林燕 , 蔡优生 , 等 . 丙戊酸钠血药浓度影响因素及不良反应分析 [J ] . 临床合理用药 , 2024 , 17 ( 10 ): 33 - 36 .
董露露 , 邹素兰 , 凌静 . 肝功能异常患者丙戊酸钠血药浓度的影响因素分析 [J ] . 淮海医药 , 2022 , 40 ( 4 ): 346 - 349 .
王利媛 , 段自皞 , 李晶 , 等 . 基于游离丙戊酸血药浓度开展丙戊酸钠剂量调整的药学实践1例 [J ] . 中国医院药学杂志 , 2023 , 43 ( 1 ): 117 - 120 .
ACHARYA S , BUSSEL J B . Hematologic toxicity of sodium valproate [J ] . J Pediatr Hematol Oncol , 2000 , 22 ( 1 ): 62 - 65 .
NASREDDINE W , ATWEH S F , BEYDOUN A A , et al . Predicting the occurrence of thrombocytopenia from free valproate levels:a prospective study [J ] . Seizure , 2022 , 94 : 33 - 38 .
王之舟 , 程红勤 , 白向荣 , 等 . 丙戊酸钠致高乳酸血症伴代谢性酸中毒 [J ] . 药物不良反应杂志 , 2020 , 22 ( 6 ): 379 - 380 .
0
浏览量
12
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构