
60 Exposure-Response Modeling: Methods and Practical Implementation
not determine if a further exposure increase will further incre ase the risk,
nor quantify the r isk at a very low ex posure level. Therefore, the repeated
measure data did not provide much more information than the analysis of the
events and exposure at 2 hours in Chapter 2 . One rea son is that extr eme QT
prolongation occurs mostly around the peak of exposure, hence the majority
of subjects w ho had no event at 2 hours also had no event a t other time points.
As a final note for non- and semiparametric models, we would like to
draw the reader’s attention to advanced approaches to ...