This paper reports on implementation of a novel, rapid, and systematic screening and alert system, based on data mining algorithms developed by DuMouchel (1,2), to assist in the monitoring of adverse drug reactions in FDA's postmarketing spontaneous reporting system database. The statistical algorithms produce signals that are coupled to interactive visualization tools (2,3) to place the drug safety signals in a contextual framework that facilitates interpretation.
In February 1998, we began to systematically analyze higher than expected drug-event combinations by drug, event code, gender, time interval, and age groups. Since March 2001, we have also been studying synergic interactions between multiple drugs (drug interactions) and event codes (syndromes).
We will illustrate the potential of this approach to improve the analysis of related safety issues such as drug interactions and of events coded as drug errors or maladministration. Public access to both the FDA's post-marketing surveillance database, and the programs for data mining, simplifies independent validation of these approaches and application to other areas of aberration detection.
References:
1. DuMouchel W. Bayesian Data Mining in Large Frequency Tables, With an Application to the FDA Spontaneous Reporting System. The American Statistician, 1999; 53:177-90
2. Szarfman A. The Application Of Bayesian Data Mining And Graphic Visualization Tools To Screen FDA's Spontaneous Reporting System Database. 2000 Proceedings of the Section on Bayesian Statistical Science. American Statistical Association, pages 67-71.
3. Levine J.G., Szarfman A. Standardized Data Structures and Visualization Tools: A way to accelerate the regulatory review of the integrated summary of safety of new drug Applications. Biopharmaceutical Report 4(3):12-.17,1996.
See www.research.att.com/~dumouchelLearning Objectives: To be added.
Keywords: Biostatistics, Health Care Quality
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.