Table 2 Parameter values of the model \({\mathrm{PK}}_{\mathrm{fix},\mathrm{bio},\mathrm{lin},\mathrm{peri}}^{\mathrm{MTX}}\) estimated with the whole and the cross-validation data set, relative standard error (RSE) of fixed effect parameters and residual unexplained variability in brackets.

From: Pharmacokinetic–pharmacodynamic modeling of maintenance therapy for childhood acute lymphoblastic leukemia

 

\({PK}_{fix,bio,lin,peri}^{MTX}\)

\({PK}_{fix,bio,lin,peri}^{MTX}\), cross-validation

Fixed effects parameters with RSE in %

\({K}_{in}^{MTX}\)[1/day]

0.031 (5)

0.030 (6)

\({K}_{eff}^{MTX}\)[1/day]

0.018 (5)

0.019 (5)

CV with 95% CI, η-Shrinkage in % and p-values

\({K}_{in}^{MTX}\)

34 [29, 38], 29, 0.22

37 [32, 42], 32, 0.12

\({K}_{eff}^{MTX}\)

31 [26,35], 32, 0.45

21 [14, 27], 53, 0.24

Residual unexplained variability

Additive residual error

0.000018 (7)

0.000021 (15)

Proportional residual error

0.024 (8)

0.0068 (56)

Errors of the individual predictions

Median of RMSEs

0.0042 (0.0028)

0.0054 (0.0045)

Mean of RMSEs

0.0047 (0.0028)

0.0064 (0.0045)

Median of MAEs

0.0033 (0.0022)

0.0041 (0.0035)

Mean of MAEs

0.0036 (0.0022)

0.0049 (0.0035)

  1. RMSEs and MAEs of the patient trajectories computed with the individual patient parameters are each calculated for the whole data set with the standard deviation in brackets. Coefficients of variation (CV) are calculated as \(\sqrt{\mathrm{exp}\left({\upomega }^{2}\right)-1}\) with \({\upomega }^{2}\) as the variance of the interindividual variability estimated by NONMEM, and the 95% CI being calculated accordingly.
Back to article page