节点文献
最优子集回归模型在预测肝移植受者他克莫司血药浓度中的应用
Application of Best Subset Multiple Regression Model in Prediction of Tacrolimus Blood Concentration in Liver Transplantation Recipients
【摘要】 目的:建立肝移植受者他克莫司血药浓度简易估算方法。方法:收集37例肝移植受者口服他克莫司的176份稳态全血浓度数据,采用最优子集回归法建立他克莫司稳态血药浓度简易估算公式。结果:以浓度测定前4日他克莫司累积剂量预测他克莫司血药浓度的准确性及精密度较好,平均预测误差(0.04±2.5)ng/ml,平均绝对误差(2.00±1.45)ng/ml,80.8%的血药浓度数据绝对预测误差≤3.0ng/ml。结论:本方法预测他克莫司血药浓度准确性和精密度较好,简便迅捷。
【Abstract】 OBJECTIVE:To establish a simple method for predicting tacrolimus blood concentration in liver transplantation recipients. METHODS:The 176 data samples of steady blood concentration of tacrolimus were collected from 37 liver transplantation recipients. Best subset multiple linear regression method(MLR)was used to establish the simple formula for predicting tacrolimus concentration. RESULTS:The blood concentration of tacrolimus could be predicted by accumulative dose of tacrolimus in 4 days before determined accurately and precisely.Mean prediction error and mean absolute prediction error were(0.04±2.5)ng/ml and (2.00±1.45)ng/ml,respectively. The absolute prediction error of 80.8% of concentration data was less than 3.0 ng/ml. CONCLUSIONS:The method is accurate and precisive,simple and convenient to predict tacrolimus blood concentration.
【Key words】 Tacrolimus; Liver transplantation; Best subset multiple regression;
- 【文献出处】 中国药房 ,China Pharmacy , 编辑部邮箱 ,2013年18期
- 【分类号】R969.1
- 【被引频次】1
- 【下载频次】143