Pharm Res (2020) 37: 171

https://doi.org/10.1007/s11095-020-02908-7

This article was updated to correct Figs. 1 and 4 as author corrections were overlooked during the production process.

Fig. 1
figure 1

Schematic diagram of methods of MAP estimation used in the study. Standard MAP executes estimation once using all historical TDM data (a); Adaptive MAP executes estimation iteratively using each segment of historical TDM data with updated prior mean by posterior mode from its preceding iteration and repeats until the last iteration, i.e. 0 to m1, …, mn-1 to mn (b). Weighted MAP executes estimation once using all historical TDM data with weighted importance (likelihood) of each segment of data during the estimation (c).

Fig. 4
figure 2

The percentage error of Bayesian forecasting using the weighted MAP method for all combinations of weighting factors ΔTRef and α. ΔTRef and α are both weighting factors. ΔTRef is the reference day, defined as the cutoff value of the time distance where w(Ys) is 1, i.e.where ΔTRef equals ΔT Ys . α is the unitless effect size of the weighting function w(Ys) on the likelihood f(η| Y).