Predictive algorithm for adverse drug events based on the Empirica Signal outputs from FDA’s AERS database
Project
Predictive algorithm for adverse drug events based on the Empirica Signal outputs from FDA’s AERS database
Principal Investigator
Oracle Principal Investigator
Bill DuMouchel
Summary
AZCERT proposes to work with Oracle’s Empirica Signal® and the FDA’s AERS database to determine whether a pattern of events that are potential surrogates for TdP can predict the subsequent development of a positive signal for drug-induced TdP. Analyzing data for drugs known to have a positive TdP signal, we will identify the pattern of signals that precede the first positive TdP signal. We will then compare the rate of development of a TdP signal in a group of drugs that have the surrogate pattern to the rate for a group of drugs that lack the pattern. The goal is to develop potential predictive algorithms that are based on the early presence of the surrogate pattern and predict the subsequent appearance of a true signal for TdP. The work proposed herein will bring a new dimension to the analysis of spontaneous report databases like AERS, i.e. prediction.