Statistical Methods of Signal Detection

DIMACS
Center for Discrete Mathematics & Theoretical Computer Science

Student: Chris LaVallee
School: University of Saint Thomas
E-mail: chrislavallee@gmail.com
Research Area: Biostatics
Project Name: Statistical Methods of Signal Detection
Mentor: Ivan Zorych, Rutgers University

Project Description

The World Health Organization (WHO) has defined a signal as "Reported information on a possible causal relationship between an adverse event and a drug, the relationship being unknown or incompletely documented previously".

We will explore different statistical models and methods used to deal with count data and contingency tables. Potential difficulties such as Simpson's paradox will be investigated. All methods will be tested on sample datasets.

Our ultimate goal is to look for adverse drug reaction signals in real data from the FDA Adverse Event Reporting System (AERS) or CADRMP Adverse Reaction Database maintained by Health Canada.