万古霉素个体化给药辅助决策系统在重症患者中的应用

Application of vancomycin individualized dosage auxiliary system in ICU patients

目的:评估个体化给药辅助决策系统"pharmVAN个体化给药方案"(A系统)与"万古霉素剂量推荐和血药浓度预测系统"(B系统)对重症患者万古霉素个体化给药的应用情况。方法:以2019年9月至12月在东部战区总医院重症医学科住院并静脉使用万古霉素的患者为研究对象进行回顾性研究。测定入组患者的稳态血药浓度,并使用上述两款决策系统预测万古霉素初始给药方案的稳态血药浓度,计算实测值与预测值之间的绝对权重偏差和相对预测误差,利用单因素和多因素Cox回归分析探讨影响绝对权重偏差的相关因素。结果:A系统和B系统分别入组23例、15例样本,A系统入组患者的实测浓度为(20.36±13.65)μg·mL-1(1.39~58.66μg·mL-1),预测稳态药物浓度是(11.81±9.14)μg·mL-1(1.1~36.04μg·mL-1),绝对权重偏差为(42.83±23.21)%,相对预测误差为(-39.04±29.37)%;B系统入组患者的实测浓度为(19.67±9.88)μg·mL-1(5.78~41.01μg·mL-1),预测稳态谷浓度是(15.72±8.65)μg·mL-1(4.53~33.83μg·mL-1),绝对权重偏差为(37.29±29.95)%,相对预测误差为(-11.75±47.32)%。单因素分析发现,A系统绝对权重偏差的影响因素包括采样天数、给药次数和药物浓度,B系统绝对权重偏差的影响因素包括年龄和肌酐清除率。再进行多因素Cox回归分析,结果未显示统计学意义,上述因素均不是影响预测性的独立危险因素。结论:A系统和B系统两款个体化给药辅助决策系统不适用于重症患者万古霉素稳态血药浓度的预测。

OBJECTIVE To evaluate the application of individualization dosage auxiliary system‘pharmVAN individualized drug administration scheme’(system A)and‘vancomycin dose recommendation and drug concentration prediction system’(system B)to vancomycin individualized drug administration in ICU patients.Methods:A retrospective study was conducted in patients hospitalized in the Department of Intensive Care,Eastern War Zone General Hospital and intravenously administered vancomycin from September 2019 to December 2019.The steady-state plasma concentrations of vancomycin in patients were measured,and the steady-state plasma concentrations of vancomycin in the initial dosing regimen were predicted using the above two decision systems.The absolute weight deviation and relative prediction error between the measured and predicted values were calculated,and the related factors affecting the absolute weight deviation were explored using univariate and multivariate Cox regression analysis.Results:23 and 15 samples were enrolled in system A and system B,respectively.The measured concentration of patients enrolled in system A was(20.36±13.65)μg·mL-1(1.39-58.66μg·mL-1);the predicted steady drug concentration was(11.81±9.14)μg·mL-1(1.1-36.04μg·mL-1);the absolute weight deviation was(42.83±23.21)%,and the relative prediction error was(-39.04±29.37)%.The measured concentration of patients enrolled in system B was(19.67±9.88)μg·mL-1(5.78-41.01μg·mL-1),the predicted steady-state valley concentration was(15.72±8.65)μg·mL-1(4.53-33.83μg·mL-1),the absolute weight deviation was(37.29±29.95)%,and the relative prediction error was(-11.75±47.32)%.Univariate analysis revealed that the influencing factors of absolute weight deviation of system A included sampling days,number of doses,and drug concentration,and the influencing factors of absolute weight deviation of system B included age and creatinine clearance.Multivariate Cox regression analysis was then performed,and the results did not show statistical significance,and none of the above factors were independent risk factors affecting predictability.Conclusion:Two individualized dosage auxiliary systems,system A and system B,are not suitable for predicting steady-state plasma concentrations of vancomycin in critically ill patients.