浏览全部资源
扫码关注微信
1.新疆医科大学第一附属医院药学部,乌鲁木齐 830011
2.新疆药物临床研究重点实验室,乌鲁木齐 830011
3.新疆医科大学第一附属医院心理医学中心,乌鲁木齐 830011
4.新疆医科大学药学院,乌鲁木齐 830017
副主任药师,博士。研究方向:临床药学。E-mail:1532527694@qq.com
教授,博士。研究方向:药物研究与开发。E-mail:hjp_yft@163.com
收稿日期:2024-09-06,
修回日期:2025-01-11,
录用日期:2025-02-10,
纸质出版日期:2025-03-30
移动端阅览
谯明,靳路,朱毅等.基于机器学习模型预测抑郁症患者度洛西汀的血药浓度 Δ[J].中国药房,2025,36(06):752-757.
QIAO Ming,JIN Lu,ZHU Yi,et al.Prediction of duloxetine blood concentration in patients with depression based on machine learning[J].ZHONGGUO YAOFANG,2025,36(06):752-757.
谯明,靳路,朱毅等.基于机器学习模型预测抑郁症患者度洛西汀的血药浓度 Δ[J].中国药房,2025,36(06):752-757. DOI: 10.6039/j.issn.1001-0408.2025.06.20.
QIAO Ming,JIN Lu,ZHU Yi,et al.Prediction of duloxetine blood concentration in patients with depression based on machine learning[J].ZHONGGUO YAOFANG,2025,36(06):752-757. DOI: 10.6039/j.issn.1001-0408.2025.06.20.
目的
2
为临床尤其是无治疗药物监测条件的新疆基层医疗机构的抑郁症患者提供度洛西汀用药参考。
方法
2
回顾性收集2022年1月至2023年12月在新疆医科大学第一附属医院服用度洛西汀的281例抑郁症住院患者的病历资料,按7∶3比例划分为训练集(196例)和测试集(85例)。通过随
机森林(RF)模型中的“递归特征消除”程序进行特征选择,采用支持向量机、RF、极端梯度提升(XGBoost)、人工神经网络4种机器学习算法构建度洛西汀血药浓度预测模型,并通过决定系数(
R
2
)、平均绝对误差(MAE)、均方根误差(RMSE)评估比较4种模型的预测性能。通过夏普利加性解释方法对筛选出的最优模型的特征进行解释,确定特征的重要性排序及其对度洛西汀血药浓度预测结果的影响大小与方向。
结果
2
最终选择了29个特征变量,包括年龄、民族、体重指数(BMI)等。XGBoost的
R
2
(0.808)最高,MAE(7.644)、RMSE(10.808)最低。影响度洛西汀血药浓度预测的特征重要性排序为:BMI>年龄>其余20个特征集合(包括肝、肾功能和生化指标)>用药日剂量>合并疾病>联合用药>民族>白细胞计数>血红蛋白>身高。
结论
2
XGBoost模型预测度洛西汀血药浓度的预测性能最佳,BMI和年龄对度洛西汀血药浓度预测的影响较大。
OBJECTIVE
2
To provide medication reference for duloxetine use in clinical settings, particularly for patients with depression in primary medical institutions in Xinjiang that lack therapeutic drug monitoring conditions.
METHODS
2
The medical records of 281 depression inpatients taking duloxetine in the First Affiliated Hospital of Xinjiang Medical University from January 2022 to December 2023 were retrospectively collected. They were divided into training set (196 cases) and test set (85 cases) in the ratio of 7∶3. Feature selection was performed by encapsulating random forests (RF) with recursive feature elimination. Four machine learning algorithms, namely support vector machine, RF, extreme gradient boosting (XGBoost) and artificial neural network, were used to construct duloxetine blood concentration prediction model. The prediction performance of the models was evaluated and compared by coefficient of determination (
R
2
), mean absolute error (MAE) and root mean squared error (RMSE). The feature of the selected optimal model was explained by Shapley additive explanation method, and the importance ranking of the features and the influence on the prediction results of duloxetine blood concentration were determined.
RESULTS
2
A total of 29 characteristic variables were selected, including age, ethnicity, bo
dy mass index(BMI), etc. XGBoost showed the highest
R
2
(0.808), and the lowest MAE (7.644) and RMSE (10.808). The ranking of feature importance for predicting the blood concentration of duloxetine was as follows: BMI>age>other 20 feature sets (including liver and kidney function and biochemical indicators)>daily dosage>comorbidities>combination therapy>ethnicity>white blood cell count>hemoglobin>height.
CONCLUSIONS
2
XGBoost model possesses the best prediction performance of duloxetine blood concentration; BMI and age have a greater impact on the prediction of duloxetine blood concentration.
中华医学会行为医学分会 , 中华医学会行为医学分会认知应对治疗学组 . 抑郁症治疗与管理的专家推荐意见:2022 年[J ] . 中华行为医学与脑科学杂志 , 2023 , 32 ( 3 ): 193 - 202 .
Behavioral Medicine Branch of the Chinese Medical Asso-ciation , Cognitive Coping Therapy Group of Behavioral Medicine Branch of the Chinese Medical Association . Expert recommendations on treatment and management of major depressive disorder:2022 [J ] . Chin J Behav Med Brain Sci , 2023 , 32 ( 3 ): 193 - 202 .
XU T , GAO N Y , LI Y H , et al . Inhibitory effects of fluoxe-tine and duloxetine on the pharmacokinetics of metoprolol in vivo and in vitro [J ] . Fundam Clin Pharmacol , 2022 , 36 ( 6 ): 1057 - 1065 .
郑恩雨 , 王可 , 冯雪竹 , 等 . 躯体症状障碍相关研究进展 [J ] . 中国药物依赖性杂志 , 2022 , 31 ( 6 ): 407 - 410,416 .
ZHENG E Y , WANG K , FENG X Z , et al . Advances of research on somatic symptom disorders [J ] . Chin J Drug Depend , 2022 , 31 ( 6 ): 407 - 410,416 .
中国药理学会治疗药物监测研究专业委员会 , 中国医师协会精神科医师分会 , 中国药理学会药源性疾病学委员会 , 等 . 中国精神科治疗药物监测临床应用专家共识:2022年版 [J ] . 神经疾病与精神卫生 , 2022 , 22 ( 8 ): 601 - 608 .
Division of Therapeutic Drug Monitoring of Chinese Pharmacological Society , Psychiatry Branch of Chinese Medical Doctor Association , Division of Drug-induced Di-seases of Chinese Pharmacological Society , et al . Expert consensus on clinical application of psychiatric drug monitoring in China:2022 [J ] . Neural Dis Men Heal , 2022 , 22 ( 8 ): 601 - 608 .
MOELLER S B , GBYL K , HJORTHØJ C , et al . Treatment of difficult-to-treat depression:clinical guideline for selected interventions [J ] . Nord J Psychiatry , 2022 , 76 ( 3 ): 177 - 188 .
MACIASZEK J , PAWŁOWSKI T , HADRYŚ T , et al . The impact of the CYP2D6 and CYP1A2 gene polymorphisms on response to duloxetine in patients with major depression [J ] . Int J Mol Sci , 2023 , 24 ( 17 ): 13459 .
RODRIGUES-AMORIM D , OLIVARES J M , SPUCH C , et al . A systematic review of efficacy,safety,and tolerabili-ty of duloxetine [J ] . Front Psychiatry , 2020 , 11 : 554899 .
秦娟娟 , 庄璐 , 姚翀 , 等 . 233例SSRIs和SNRIs类药物药品不良反应/事件报告分析 [J ] . 中国药物应用与监测 , 2020 , 17 ( 4 ): 249 - 252 .
QIN J J , ZHUANG L , YAO C , et al . Analysis on 233 cases of ADR/ADE associated with SSRIs and SNRIs [J ] . Chin J Drug Appl Monit , 2020 , 17 ( 4 ): 249 - 252 .
罗轶凡 , 杨舒 , 初阳 . 模型引导的精准用药在临床血药浓度预测中的应用 [J ] . 药物评价研究 , 2023 , 46 ( 7 ): 1620 - 1628 .
LUO Y F , YANG S , CHU Y . Application of model-informed precision dosing in prediction of serum concentration [J ] . Drug Eval Res , 2023 , 46 ( 7 ): 1620 - 1628 .
中华医学会精神科分会 . CCMD-3中国精神障碍分类与诊断标准 [M ] . 3版 . 济南 : 山东科学技术出版社 , 2001 : 9 - 168 .
Psychiatry Branch of Chinese Medical Society . CCMD-3:Chinese classification and diagnostic criteria of mental dis-orders [M ] . 3rd Edition . Jinan : Shandong Science and Technology Press , 2001 : 9 - 168 .
HIEMKE C , BERGEMANN N , CLEMENT H W , et al . Consensus guidelines for therapeutic drug monitoring in neuropsychopharmacology:update 2017 [J ] . Pharmacopsychiatry , 2018 , 51 ( 1 ): 9 - 62 .
黄志伟 , 王蕊 , 余一旻 , 等 . 度洛西汀在中国健康受试者中的群体药代动力学研究 [J ] . 中国临床药理学杂志 , 2024 , 40 ( 4 ): 598 - 602 .
HUANG Z W , WANG R , YU Y M , et al . Population pharmacokinetics of duloxetine in Chinese healthy subjects [J ] . Chin J Clin Pharmacol , 2024 , 40 ( 4 ): 598 - 602 .
梁海月 , 杨青军 , 王姣 . 盐酸度洛西汀肠溶胶囊在中国健康受试者体内药代动力学与安全性研究 [J ] . 天津药学 , 2024 , 36 ( 2 ): 1 - 4,75 .
LIANG H Y , YANG Q J , WANG J . The pharmacokinetics and safety of enteric-coated duloxetine hydrochloride capsules in healthy Chinese volunteers [J ] . Tianjin Pharm , 2024 , 36 ( 2 ): 1 - 4,75 .
刘斐烨 , 凌静 , 丁可 . 基于样本筛选优化的支持向量机预测万古霉素血药浓度 [J ] . 中国医院药学杂志 , 2022 , 42 ( 10 ): 1015 - 1019 .
LIU F Y , LING J , DING K . Prediction of vancomycin serum concentration based on SVM with sample selection optimization [J ] . Chin J Hosp Pharm , 2022 , 42 ( 10 ): 1015 - 1019 .
XUE B , LI D W , LU C Y , et al . Use of machine learning to develop and evaluate models using preoperative and intraoperative data to identify risks of postoperative complications [J ] . JAMA Netw Open , 2021 , 4 ( 3 ): e212240 .
杨建宁 , 洪豆豆 , 李杨 , 等 . 甘肃省不同地区糖尿病肾脏疾病的机器学习预测模型的研究 [J ] . 中国糖尿病杂志 , 2025 , 33 ( 1 ): 8 - 15 .
YANG J N , HONG D D , LI Y , et al . Machine learning prediction model of diabetic kidney disease in different regions of Gansu Province [J ] . Chin J Diabetes , 2025 , 33 ( 1 ): 8 - 15 .
CHANG L Y , HAO X , YU J , et al . Developing a machine learning model for predicting venlafaxine active moiety concentration:a retrospective study using real-world evidence [J ] . Int J Clin Pharm , 2024 , 46 ( 4 ): 899 - 909 .
沈黎 , 李智平 . 儿童他克莫司个体化给药模型的研究进展 [J ] . 中国药房 , 2025 , 36 ( 1 ): 124 - 128 .
SHEN L , LI Z P . Advances in individualized dosing mo-dels of tacrolimus for children [J ] . China Pharm , 2025 , 36 ( 1 ): 124 - 128 .
ASRIAN G , SURI A , RAJAPAKSE C . Machine learning-based mortality prediction in hip fracture patients using biomarkers [J ] . J Orthop Res , 2024 , 42 ( 2 ): 395 - 403 .
肖桃 , 倪晓佳 , 黄善情 , 等 . 基于治疗药物监测的氟伏沙明血药浓度结果及其影响因素分析 [J ] . 中国临床药理学杂志 , 2021 , 37 ( 22 ): 3064 - 3067 .
XIAO T , NI X J , HUANG S Q , et al . Research on the therapeutic drug monitoring and influence factors of fluvoxamine [J ] . Chin J Clin Pharmacol , 2021 , 37 ( 22 ): 3064 - 3067 .
梁海 , 夏茹楠 , 吴炜 , 等 . 影响文拉法辛血药浓度相关因素的研究进展 [J ] . 中国医院药学杂志 , 2023 , 43 ( 15 ): 1758 - 1762 .
LIANG H , XIA R N , WU W , et al . Research progress of the influencing factors for blood concentration of vanlafaxine [J ] . Chin J Hosp Pharm , 2023 , 43 ( 15 ): 1758 - 1762 .
钟璐莲 , 陈政 , 郭媛媛 , 等 . 艾司西酞普兰血药浓度/剂量比的影响因素研究 [J ] . 中国医院用药评价与分析 , 2022 , 22 ( 6 ): 668 - 671 .
ZHONG L L , CHEN Z , GUO Y Y , et al . Influencing factors of serum concentration/dose ratio of escitalopram [J ] . Eval Anal Drug Use Hosp China , 2022 , 22 ( 6 ): 668 - 671 .
MOLDEN E , SPIGSET O . Interactions between metoprolol and antidepressants [J ] . Tidsskr Nor Laegeforen , 2011 , 131 ( 18 ): 1777 - 1779 .
QIN X , XIE C , HAKENJOS J M , et al . The roles of CYP1A2 and CYP2D in pharmacokinetic profiles of serotonin and norepinephrine reuptake inhibitor duloxetine and its metabolites in mice [J ] . Eur J Pharm Sci , 2023 , 181 : 106358 .
赵婷 , 李红健 , 翁振群 , 等 . 基于人工神经网络的新疆维吾尔族癫痫患儿奥卡西平血清浓度预测研究 [J ] . 中国药学杂志 , 2020 , 55 ( 16 ): 1376 - 1380 .
ZHAO T , LI H J , WEN Z Q , et al . Prediction of serum concentration of oxcarbazepine in Uygur children with epilepsy in Xinjiang based on artificial neural network [J ] . Chin Pharm J , 2020 , 55 ( 16 ): 1376 - 1380 .
0
浏览量
1
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构