AIM: To establish a population pharmacokinetics (PPK) model of vancomycin in patients with hematological diseases who developed neutropenia. METHODS: Patients from department of hematology with neutropenia in our hospital were taken into oue study.The patients (n=77) were performed trough and peak serum concentration of vancomycin, and their clinical data and medication information were collected. The Nonlinear mixed effect modeling approach (NONMEM) was used to establish the PPK model of those patients and model assessment and validation was carried out. Goodneess of fit plots and visual predictive check plus Bootstrap approach were used to assess validate our model. RESULTS: The model was a two compartment model, the final formulas were: clearance rate CL=6.84×(BW/70)

AIM: To discuss the application and clinical significance of the ratio of area under the curve over 24 hours to minimum inhibitory concentration (AUC

OBJECTIVE： To evaluate the application of individualization dosage auxiliary system JPKD and SmartDose in individualization administration of vancomycin. METHODS： A retrospective study was conducted among adult inpatients in Hainan Provincial People’s Hospital from Apr. 2018 to Mar. 2019 with intravenous use of vancomycin. SmartDose was used to predict the steady blood trough concentration of vancomycin in the initial dosage regimen， and the absolute weight deviation and relative prediction error between the measured concentration and the predicted concentration were calculated. The effects of body mass index （BMI） and acute kidney injury （AKI） on absolute weight deviation were analyzed by χ2 test or continuously corrected χ2 test. Vancomycin drug delivery scheme was adjusted for patients with ungualified steady blood drug trough concentration. JPKD and SmartDose system were used to predict the blood concentration of vancomycin after adjusting the dosage regimen. The absolute weight deviation and relative prediction error between the measured concentration and the predicted concentration were calculated. The prediction ability of the two systems was evaluated and 3 examples was analyzed. RESULTS： Predicted steady blood trough blood concentration of 85 included patients in SmartDose predicted initial dosage regimen were （11.36±5.96） μg/mL （2.34-29.33   μg/mL）； the measured concentration was （11.44±6.57） μg/mL （3.10-29.50 μg/mL）； absolute weight deviation was 22.95％， and the relative prediction error was 2.72％. Whether BMI was normal or not had significant effects on the absolute weight deviation   （χ2＝4.75， P＝0.029）， and whether AKI occurred or not had no significant effects on the absolute weight deviation （χ2＝0.236， P＝0.627）. JPKD and SmartDose predicted that predicted steady blood trough concentrations of vancomycin in 22 included patients were （11.06±3.58） and （12.15±4.35） μg/mL， and the measured concentration was （12.57±4.50） μg/mL； absolute weight deviations were 18.30％ and 18.68％； relative prediction errors were －8.65％ and －0.44％， respectively. The absolute weight deviations of the predicted values of the two systems were less than 30％. The absolute weight deviations of prediction results were also less than 30％ in 3 patients. CONCLUSIONS： JPKD and SmartDose system have good predictive ability for blood concentration of vancomycin in clinical application， and can be used to optimize the individualized administration of vancomycin.