应用Bayesian反馈法预测非稳态下苯妥英钠浓度

Application of Bayesian feedback method in the prediction of non-steadyphenytoin concentrations

采用Ｂｅａｌ提出的计算非稳态下苯妥英钠药物浓度的简化公式，应用Ｂａｙｅｓｉａｎ反馈公式结合先验的群体参数编制了预测非稳态下苯妥英钠浓度的程序。对６例中毒病人取一点浓度进行反馈预测，预测相关性达０．９６，预测最大误差在１０％以内。对临床上复杂用药背景下病人苯妥英钠浓度的成功预测，显示了Ｂａｙｅｓｉａｎ反馈在实际的治疗药物监测（ＴＤＭ）中广泛的应用前景。

A Bayesian approach was developed to predict the non-steady phenytoin. TheBayesian program was developed from the simplified model of professor Beal. The non-steadyphenytoin concentration was calculated by using the Bayesian combined with the population phar-macokinetic parameters of phenytoin. The relationship between the observed and predicted con-centration reached 96%in 6 patients with one point of concentration. The predicted error was lessthan 10%. The results show that the method can be successfully used in the predictoin of pheny-toin under the complex conditoin of phenytoin administration.