[Population pharmacokinetics of vancomycin from severe in patients with lower respiratory tract infection]

Objective: To develop a population pharmacokinetic (PPK) model of vancomycin in Chinese inpatients with severe lower respiratory tract infection. Methods: We gathered serum concentrations of vancomycin from inpatients who received vancomycin during Nov 2011 to Nov 2012.Vancomycin serum concentrations was measured by high performance liquid chromatography. Vancomycin PPK analysis was performed using nonlinear mixed effects model (NONMEM) program. Results: We gathered the data of 70 inpatients with lower respiratory tract infection at respiratory ward or respiratory intensive care unit(RICU) between Nov 2011 to Nov 2012 [58 males, 12 females; 78(23-91) years old; the mean of APACHⅡ score was 16.4±4.8]. A total of 267 concentrations of vancomycin were gathered from 70 patients. A 2-compartment model fit the concentration data best. The final vancomycin PPK model was: CL(i)=1.67×e(η1), V(1i)=33.04×[1-0.199×(60/Scr)] ×e(η2), Q(i)=7.08×2.053(AI)×e(η3), V(2i)=19.29×e(η4).(CL: vancomycin clearance; V(1): distribution volume of the central compartment; Q: intercompartment clearance; V(2): distribution volume of the intercompartment). The population mean values were 1.67 L/h, 33.04 L; 7.08 L/h and 19.29 L respectively. Serum creatinine was the covariate to affect vancomycin apparent distribution volume of the central compartment. Intercompartment clearance was 2.053 times larger in patients with albumin infusion than that in patients without. Conclusions: We found that a 2-compartment model fit the concentration data best in Chinese patients with severe lower respiratory tract infection. Serum creatinine and albumin infusion were the most significant covariates to affect vancomycin PK.