From the *Department of Anesthesiology, University Medical Center Groningen, University of Groningen, The Netherlands; †Departmento de Anestesiología, Escuela de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile; and ‡Department of Anesthesia, Ghent University, Gent, Belgium.
Anesth Analg. 2014 Jun;118(6):1221-37. doi: 10.1213/ANE.0000000000000165.
Pharmacokinetic (PK) models are used to predict drug concentrations for infusion regimens for intraoperative displays and to calculate infusion rates in target-controlled infusion systems. For propofol, the PK models available in the literature were mostly developed from particular patient groups or anesthetic techniques, and there is uncertainty of the accuracy of the models under differing patient and clinical conditions. Our goal was to determine a PK model with robust predictive performance for a wide range of patient groups and clinical conditions.
We aggregated and analyzed 21 previously published propofol datasets containing data from young children, children, adults, elderly, and obese individuals. A 3-compartmental allometric model was estimated with NONMEM software using weight, age, sex, and patient status as covariates. A predictive performance metric focused on intraoperative conditions was devised and used along with the Akaike information criteria to guide model development.
The dataset contains 10,927 drug concentration observations from 660 individuals (age range 0.25-88 years; weight range 5.2-160 kg). The final model uses weight, age, sex, and patient versus healthy volunteer as covariates. Parameter estimates for a 35-year, 70-kg male patient were: 9.77, 29.0, 134 L, 1.53, 1.42, and 0.608 L/min for V1, V2, V3, CL, Q2, and Q3, respectively. Predictive performance is better than or similar to that of specialized models, even for the subpopulations on which those models were derived.
We have developed a single propofol PK model that performed well for a wide range of patient groups and clinical conditions. Further prospective evaluation of the model is needed.
药代动力学（PK）模型用于预测术中显示的输注方案的药物浓度，并计算靶控输注系统中的输注率。对于丙泊酚，文献中可用的 PK 模型大多是从特定的患者群体或麻醉技术开发的，并且在不同的患者和临床条件下，模型的准确性存在不确定性。我们的目标是确定一种具有广泛患者群体和临床条件下强大预测性能的 PK 模型。
我们汇总和分析了 21 个先前发表的丙泊酚数据集，其中包含来自幼儿、儿童、成人、老年人和肥胖个体的数据。使用 NONMEM 软件，使用体重、年龄、性别和患者状态作为协变量，对 3 室房室模型进行了估计。设计了一种专注于术中条件的预测性能指标，并与 Akaike 信息准则一起用于指导模型开发。
该数据集包含来自 660 名个体（年龄范围 0.25-88 岁；体重范围 5.2-160 公斤）的 10927 个药物浓度观察值。最终模型使用体重、年龄、性别和患者与健康志愿者作为协变量。对于 35 岁、70 公斤的男性患者，参数估计值分别为：9.77、29.0、134 L、1.53、1.42 和 0.608 L/min 用于 V1、V2、V3、CL、Q2 和 Q3。预测性能优于或与专门模型相似，即使是在这些模型推导的亚群中也是如此。
我们已经开发了一种用于广泛患者群体和临床条件的单一丙泊酚 PK 模型。需要进一步前瞻性评估该模型。
