MONITORING DRUG THERAPY THROUGH SMARTCHART

在今天的医学中，许多老年患者肾功能减退 或肝功能，并服用多种药物。 药物不良 相互作用和毒性是常见的。 SmartChart（R）是一个允许护士和医生 有效地将真实的临床数据从 床边 对于第一种情况，纸质图表是通过 而不是用笔或听写。 在第一阶段，我们 将首先修改SmartChart， 患者参考中心列表，该列表将自动扩展 随着新药的使用和进入。 然后我们将开发一种药物 监测程序（SmartDose）使用完整、准确和 及时的临床信息，以估计适当的药物剂量， 监测潜在的药物相互作用。 每次治疗 更改或输入，程序将计算出适当的剂量 根据一些公式，确定是否有 与其他药物的相互作用，并显示消息，如果药物 处方可能会导致临床问题。 该数据库将 通过简化斯坦福大学药物相互作用数据库创建， 增加了普遍接受的药物剂量和动力学公式， 药品名称 消息的准确性和实用性将使用 来自透析患者人群的临床信息。 在第二阶段，我们将测试神经网络模拟程序， 医生和护士最终能否“训练”计算机 在他们自己的专科单位和医院监测药物使用情况。 我们还将测试药物剂量计算的贝叶斯方法， 预测.

In today's medicine, many elderly patients have diminished kidney or liver function and take numerous medications. Adverse drug interactions and toxicities are common. SmartChart(R) is a program which allows nurses and physicians to efficiently enter real time clinical data into PC's from the bedside. For the first the paper chart is generated through a microcomputer rather than by pen or dictation. During Phase I we will first modify SmartChart so that therapy entries of all patients refer to a central list which will automatically expand as new drugs are used and entered. We will then develop a drug monitoring program (SmartDose) to use the complete, accurate and timely clinical information to estimate proper drug dosages and monitor for potential drug interactions. Each time a therapy is changed or entered, a program will calculate the proper dosage according to a number of formulas, determine whether there are interactions with other drugs, and display messages if the drug prescription could cause clinical problems. The database will be created by simplifying the Stanford drug interaction database and adding generally accepted drug dosing and kinetic formulas for each drug title. Accuracy and utility of messages will be tested using clinical information from a dialysis patient population. In Phase II, we will test neural network simulation programs to see whether physicians and nurses could eventually "train" a computer to monitor drug use in their own specialty units and hospitals. We will also test Bayesian methods of drug dose calculation and prediction.