Visualize Data in Population and PK/PD Analysis for Exposure Response

Presented: Thursday, February 15, 2007

Speaker: Simon Zhou, Wyeth Research

Listen to the archived web cast. Download instructions on how to view the web cast.

Download the web cast presentation.

Traditional hypothesis testing in exploratory drug development requires large clinical trials, offers limited learning, and is inefficient.  This fact was acknowledged by FDA in its Critical Path Initiative document to call for innovation in early drug development.  Exposure-response analysis based on PK/PD modeling is such a tool to integrate cumulative learning into clinical trials to meet this challenge.  It can help to focus on the right hypothesis to be tested in subsequent trials, with potential to demonstrate safety and effectiveness more rapidly, with more certainty, and at lower costs.

PK/PD modeling is a data and computational intensive approach to explore exposure-response.  It involves pooling, cleaning, and analyzing data across a large number of studies.  The data pooling and cleaning process is tedious and prone to error.  Model building requires experience and skill to drill down many possibilities to arrive at the most plausible and parsimonious solution.  Model validation may take various forms, and usually generates more numbers than the input observations.  Conveying the PK/PD modeling results to a non-technical audience is another challenge but critical to effect decision making.  The presentation will focus on how to use data visualization at each of abovementioned steps to expedite and guide PK/PD model building for better presentation and communication.