摘要
目的探索已发表的成人万古霉素群体药代动力学模型在不同人群中的预测性能。方法共收集来自342个患者的937个血药浓度数据,模型预测仅在测浓度的时间点进行,用预测偏倚(ME)、精密度(MAE)和拟合优度图来评估模型的预测性能。结果在年龄<65岁亚组(ME=-0. 76,MAE=2. 02),体重≥65 kg组(ME=-0. 65,MAE=1. 85)和肌酸酐清除率(CLCR)≥80m L·min^(-1)·1. 73 m^(-2)组(ME=-0. 59,MAE=1. 64) 3个组中,模型的预测性能较好。而在年龄≥65岁组(ME=-0. 87,MAE=2. 80)和CLCR <80m L·min^(-1)·1. 73 m^(-2)组(ME=-1. 10,MAE=3. 05) 2个组中,预测较差。尚还不能证明模型在体重<65 kg组(ME=-0. 86,MAE=2. 38)中预测较好。结论该模型在年龄<65岁,体重≥65 kg和CLCR≥80 m L·min^(-1)·1. 73 m^(-2)的人群中预测性能好。
Objective To explore the predictive performance of a previously reported vancomycin population pharmacokinetic model in different Chinese adult patients.Methods A new dataset including 937 serum concentration points from 342 patients was used,model prediction was performed only at the time points of concentration measurement.The predictive bias( ME),accuracy( MAE) and goodness of fit plot were used to assess the predictive performance of the model.Results The predictive performance of the model in three subgroups,including the age < 65 age subgroup( ME =-0.76,MAE = 2.02),body weight≥ 65 kg subgroup( ME =-0.65,MAE = 1.85) and creatinine clearance( CLCR)≥ 80 mL·min^-1·1.73 m^-2 subgroup( ME =-0.59,MAE = 1.64) was better.However,the predictive performance of the model in other two subgroups,the age ≥ 65 age subgroup( ME =-0.87,MAE = 2.80) and CLCR < 80mL·min^-1·1.73m^-2 subgroup( ME =-1.10,MAE =3.05) was poor.In this study,it hasn’t proven that the predicted performance of the model was good in body weight < 65 kg subgroup( ME =-0.86,MAE = 2.38).Conclusion The predictive performance of the model could be accepted in age < 65 age,body weight ≥ 65 kg and CLCR≥ 80 mL·min^-1·1.73 m^-2 patient subgroups.
作者
经力
刘滔滔
郭晴
庞惠媚
韦文兴
覃小玲
陈铭
张韧
JING Li;LIU Tao-tao;GUO Qing;PANG Hui-mei;WEI Wen-xing;QIN Xiao-ling;CHEN Ming;ZHANG Ren(Department of Pharmacy, The FirstAffiliated Hospital of Guangxi MedicalUniversity, Nanning 530021, China)
出处
《中国临床药理学杂志》
CAS
CSCD
北大核心
2019年第6期577-580,共4页
The Chinese Journal of Clinical Pharmacology
基金
国家自然科学基金资助项目(81460569)
