论文部分内容阅读
目的建立一种预测在治疗药物监测中需要避免的、只能提供很少信息的“最小信息采样点” 的方法。方法进行了一系列基于一室开放模型的仿真, 并且通过比较不同采样方案中预测参数效果的差异来验证建立的预测最小信息采样点的方法是否合理。我们比较了每种采样方案中各个参数的预测值和真实值( 仿真值) , 通过两者的接近程度来评价该采样方案,预测值和真实值越接近说明通过该方案采样获得的信息量越大, 这个采样方案越合理。结果在预测的最小信息采样点附近的时间点采样, 将会得到准确度和精确度均不佳的参数预测值。此外, 还考查了预测的最小信息采样点和参数之间的关系。计算了一些典型情况下的最小信息采样点, 并通过作图, 了解它们与参数之间的变化关系。在本文研究的例子中, 清除率可预测出一个最小信息采样点而表观分布容积和吸收速率常数各有两个。并且还发现最小信息采样点的预测值随表观分布容积的增大或者清除率和吸收速率常数的减小而增大。结论在实际的治疗药物监测中, 可以根据按照本文所描述的方法预测出的最小信息采样点设计更加合理的采样方案。
Objective To establish a “minimum information sampling point” that can only be provided with little information to be avoided in the monitoring of therapeutic drugs. Methods A series of simulations based on one-compartment open model were carried out and the method of verifying the established minimum information sampling points was verified by comparing the differences of the effects of the prediction parameters in different sampling schemes. We compare the predicted value and the true value (simulated value) of each parameter in each sampling plan, and evaluate the sampling plan according to the closeness between the two. The closer the predicted value and the true value are, the more the information obtained by sampling the program Big, the more reasonable this sampling plan. The result is sampled at a point in time near the predicted minimum information sample point, which will result in a predicted value of the parameter with poor accuracy and accuracy. In addition, the relationship between the predicted minimum information sample points and parameters is also examined. Calculate the minimum information sample points in some typical cases, and through mapping, understand the relationship between them and the parameters. In the case study of this article, the clearance rate predicts a minimum information sample and the apparent distribution volume and absorption rate constants are each two. It is also found that the predicted value of the minimum information sampling point increases with the apparent volume of distribution or the decrease of the clearance rate and absorption rate constant. Conclusion In actual therapeutic drug monitoring, a more reasonable sampling plan can be designed based on the minimum information sample points predicted by the method described in this article.