Bailey J M
Department of Anesthesiology, Emory University School of Medicine, Atlanta, Georgia 30322.
Anesthesiology. 1993 Jan;78(1):116-23. doi: 10.1097/00000542-199301000-00017.
There is increasing interest among anesthesiologists in the use of continuous infusion of intravenous drugs. The therapeutic effect of most drugs is a function of the concentration at the site of drug effect, which in turn is determined by the plasma concentration. Constant plasma concentrations can be maintained by computer-controlled infusion pumps. However, such equipment is not yet widely available and will be expensive.
A technique is presented to enable the anesthesiologist to maintain approximately a desired plasma concentration after an arbitrary bolus dose by using a series of infusions with rates decreasing in a stepwise fashion. The algorithm is based on approximating the exact infusion needed to maintain the target plasma concentration by producing this concentration at discrete, specific times. Equations are derived for calculating the sequential rates of the infusion scheme. The equations assume linear pharmacokinetics, and the starting point for derivation of the equations is the assumption that the plasma concentration is given by the convolution of the drug infusion and the unit dose-response function.
The accuracy of the technique was assessed by simulating the infusion of fentanyl and midazolam. By using an infusion scheme of three steps, the error was no greater than 38% for fentanyl and no greater than 10% for midazolam.
Other than the assumption of linear kinetics, the algorithm is independent of pharmacokinetic models. Implementation does not require computer-based numerical analysis.
麻醉医生对静脉药物持续输注的应用兴趣日益浓厚。大多数药物的治疗效果取决于药物作用部位的浓度，而该浓度又由血浆浓度决定。通过计算机控制的输注泵可维持恒定的血浆浓度。然而，此类设备尚未广泛普及且价格昂贵。
本文介绍了一种技术，通过一系列速率逐步降低的输注，使麻醉医生在给予任意推注剂量后能维持大致所需的血浆浓度。该算法基于在离散的特定时间产生目标血浆浓度来近似维持该浓度所需的精确输注量。推导了用于计算输注方案顺序速率的方程。这些方程假定为线性药代动力学，方程推导的出发点是假设血浆浓度由药物输注与单位剂量反应函数的卷积给出。
通过模拟芬太尼和咪达唑仑的输注评估了该技术的准确性。采用三步输注方案时，芬太尼的误差不超过38%，咪达唑仑的误差不超过10%。
除了线性动力学假设外，该算法独立于药代动力学模型。实施不需要基于计算机的数值分析。
