Figure legend

Figure 1. Receiver operating characteristic (ROC) curves for predicting adverse events from voriconazole concentrati
Figure 2. Diagnostic goodness-of-fit plots for basic model (A1, A2) and final model (B1, B2). A1 and B1, Observed concentrations versus population-predicted concentrations; A2 and B2, Observed voriconazole plasma concentrations versus individual-predicted concentrations; the lines are the lines of unity y=x.
Figure 3. Diagnostic goodness-of-fit plots for basic model (C1, C2) and final model (D1, D2). C1 and D1, conditional weighted residualsversus population-predicted concentrations; C2 and D2, conditional weighted residuals versus time.
Figure 4. The median voriconazole Ctroughversus time profiles for 30 days based on the optimal intravenous (right) or oral (left) dosing regimen. The loading doses of TBIL-1, TBIL-2 and TBIL-3 patients were 400 mg q12h, 200 mg q12h and 200 mg q12h for first day, respectively. The maintenance doses of TBIL-1, TBIL-2 and TBIL-3 patients were 100 mg q12h (squares), 50 mg q12h (filled dots) or 100mg qd (triangles) and 50 mg qd (hollow dots), respectively.