04 May 2008

Safety Informatics : never heard of it ?

Business Intelligence, Analysts, KPO etc., etc., all the buzz words that indicate a transition from back office services to high end knowledge and decision making processes. One area that is related to drug safety and decision support is Safety Informatics and data mining.

As we discussed in the past week safety signals are often complex and are reported to one central location at the regulatory side. So FDA faces a significant resource constraint to identify safety issues and signals from this huge database. Sponsors (biopharma and device companies) also have to constantly watch out for underlying safety issues and patterns. This process of signal detection until 2000 was a purely manual effort, but has now given way to more intelligent and advanced software tools. These tools can be classified into 1. Safety databases 2. Signal/Pattern detection tools for data mining. 

There are several software tools in this class but we will limit the discussion to 2-3 of them.

i3 Pharma's Aperio

i3Aperio is a signal detection and assessment tool and is built on top of their proprietary claims database, which contains  the health care experiences of more than 25 million individuals—so that users can monitor the safety of  new drugs quickly and effectively. The power of the safety database lies in integration of data from demographics, medical history, pharmacy and socioeconomic data.


WebVDME™ from Lincoln Technologies (or PhaseForward)

FDA initiated development of a drug safety data mining tool with Lincoln technologies in 2003, the software was released by 2005, and Lincoln technologies was bought over by PhaseForward. 

The WebVDME application supports the detection and quantification of safety signals through the use of advanced data mining techniques applied to a variety of spontaneous reporting databases, including a company’s internal safety database. 

WebVDME integrates with FDA’s Adverse Event Reporting System (AERS) and Vaccine Adverse Event Reporting System (VAERS) databases etc.,

No comments:

Post a Comment