
8
Confounding bias and causal inference in
exposure–response modeling
8.1 Introduction
In a randomized clinical tr ial the mean response difference b etween treatment
groups is an unbiased estimate for the causal treatment effect, since potential
confounding factors between the treatment and resp onse are eliminated by
randomization. However, it is more complex to determine exposure–response
relationships s inc e drug exposure, e.g., drug conce ntration, is g enerally not
controlled even in a randomized clinical trial and is often affected by con-
founding factors. The Food and Drug Administration (FDA) has issued a
technical document for exposure–resp o ...