KLINICKÁ FARMAKOLOGIE A FARMACIE / Klin Farmakol Farm 2022;36(3):85-92 / www.klinickafarmakologie.cz 88 The best outcomes in terms of %PE, RMSE, Bland-Altman bias as well as Pearson’s R were obtained with the DOS2 model. WIN1 produced the lowest %PE and Bland-Altman bias (Table 3, Fig. 2) among the WIN models, but the correlation (Pearson’s R) between the predicted and measured vancomycin values was less tight (Fig. 3). RMSE was the same while %PE and Bland-Altman bias were almost the same in the WIN3 model, with a slightly better correlation than the WIN1 model. Comparison of pharmacokinetics parameters The DOS2 and WIN1–3 models use the same population-based pharmacokinetics parameters (Table 2). The DOS1 model uses lower V1 and both rate constants, whereas the one-compartment WIN4 model uses higher V1 and renal fraction. When compared to population-based data, all two-compartment models used higher V1, rate constant k12, and fr, with the exception of fr in WIN2, while k21 was slightly lower. WIN4 used higher V1 and lower fr. WIN1–3 models used slightly higher V1 than DOS1, 0.22–0.24 vs. 0.20 L/kg LBMc, P <0.0001. k12 was higher in WIN1 and WIN3 −1.27 and 1.25, respectively vs. 0.99, P <0.0001. DOS2 used higher V1 and k12, whereas its k21 was similar to DOS1 (Table 2). Comparison to DOS1 model All three two-compartment WIN models (WIN1-WIN3) produced similar %PE and RMSE when compared with the DOS1 model (Table 4). %PE and Bland-Altman bias were the lowest in the WIN2 model, whereas RMSE was the lowest and Pearson’s R was the highest in the WIN3 model. %PE, Bland-Altman bias, and Pearson’s R were higher, but RMSE was similar in the DOS2 model, compared to the WIN models. The one-compartment WIN4 model differed most from the DOS2 model. Bland-Altman plots for all model comparisons are shown in Fig. 4. The Pearson’s R between vancomycin concentrations predictedby theWINandDOSmodels varied from0.631 to0.941, P<0.0001 (Fig. 5). The Pearson’s R between %PE produced by DOS1 and those produced by other models varied from 0.830 to 0.930 (Fig. 6), P <0.0001. Fig. 1. Vancomycin concentration measured and predicted by models Tab. 2. Pharmacokinetic parameters of vancomycin. k12, k21 – rate constants, kelm – elimination rate constant non-renal, kelr – elimination rate constant renal, V1 – volume of distribution related to lean body mass, fr – renal fraction. Individualized data presented as mean ± standard deviation. ‡ P <0.0001, || P <0.01, § P <0.05 compared to population data; † P <0.0001, * P <0.05 compared to DOS1 V1 k12 k21 kelr fr Clm kelm population DOS1 0.17 0.92 0.46 0.0037 DOS2 0.21 1.12 0.48 0.75 0.21 WIN1 0.21 1.12 0.48 0.00327 0.0143 WIN2 0.21 0.75 0.21 WIN3 0.21 1.12 0.48 0.75 0.21 WIN4 0.39 0.79 fitted DOS1 0.20 ± 0.03‡ 0.99 ± 0.08‡ 0.39 ± 0.14|| 0.004 ± 0.0007 DOS2 0.31 ± 0.40 1.18 ± 0.17§,† 0.40 ± 0.11‡ 0.84 ± 0.31 WIN1 0.24 ± 0.03‡,† 1.27 ± 0.22‡,† 0.43 ± 0.06*,† 0.004 ± 0.001|| WIN2 0.22 ± 0.02‡,† 0.69 ± 0.47 WIN3 0.23 ± 0.03‡,† 1.25 ± 0.18‡,† 0.44 ± 0.06||,† 0.85 ± 0.33* WIN4 0.46 ± 0.20‡,* 0.72 ± 0.40 Tab. 3. Vancomycin serum levels (VCM), % prediction error (%PE), RMSE, Bland-Altman bias – comparison to measured values. Data presented as mean ± SD. ‡P <0.0001, **P <0.005, *P <0.05 compared to measured; †P <0.0001 compared to DOS1 VCM concentration [mg/L] %PE [%] RMSE [%] Bland-Altman bias ± SD (95% limits of agreement) measured 24.1 ± 10.8 DOS1 20.2 ± 5.3** −5.7 ± 34.5 35 3.96 ± 9.15 (−14.0–21.9) DOS2 21.2 ± 5.5* −3.2 ± 33.0 33 3.29 ± 8.74 (−13,8–20.4) WIN1 20.6 ± 6.2* −4.4 ± 36.4 36 3.55 ± 9.17 (−15.0–22.1) WIN2 20.0 ± 6.4** −7.4 ± 36.7 37 4.09 ± 9.10 (−13.7–21.9) WIN3 20.5 ± 5.7* −4.5 ± 36.2 36 3.64 ± 9.22 (−14.4–21.7) WIN4 16.7 ± 6.3‡† −20.8 ± 39.4† 44 7.41 ± 11.44 (−15.0–29.8) Tab. 4. Vancomycin serum levels (VCM), %prediction error (%PE), RMSE, Bland-Altman bias – comparison to DOS1 model. Data presented as mean ± SD %PE [%] RMSE [%] Bland-Altman bias ± SD (95% limits of agreement) DOS2 4.1 ± 13.9 14 −0.66 ± 2.48 (−5.5–4.2) WIN −16.8 ± 24.9 30 3.44 ± 5.06 (−6.5–13.4) WIN1 1.7 ± 15.2 15 −0.42 ± 2.75 (−5.8–5.0) WIN2 −1.4 ± 15.8 16 0.13 ± 3.03 (−5.8–6.1) WIN3 1.7 ± 12.7 13 −0.33 ± 2.31 (−4.8–4.2) ORIGINÁLNÍ PRÁCE COMPARISON OF MW\PHARM 3.30 (DOS) AND MW\PHARM ++ (WINDOWS) VERSIONS OF PHARMACOKINETIC SOFTWARE FOR PK/PD MODELLING OF VANCOMYCIN IN CONTINUOUS ADMINISTRATION
RkJQdWJsaXNoZXIy NDA4Mjc=