Amendment_1.pdf

AMENDMENT TO FDA BROAD AGENCY ANNOUNCEMENT FOR THE ADVANCE

RESEARCH AND DEVELOPMENT OF REGULATORY SCIENCE AND

INNOVATION

Broad Agency Announcement Number: FDABAA-15-00121 Amendment Number: One (1)

Amendment Posting Date: April 28, 2015

Department of Health and Human Services, Food and Drug Administration, Office of Acquisitions & Grants Service, 5630 Fishers Lane, Rockville, MD 20857

Point of Contact: Sean Wybenga, Contracting Officer

Sean.Wybenga@fda.hhs.gov

SUMMARY OF CHANGES

PURPOSE: The purpose of this amendment is to add an additional area of research interest to “Part I:

Research Areas of Interest, Research Area 5. Strengthening Social and Behavioral Science at FDA by Enhancing Audience Understanding” of the announcement document.

The following research area is hereby added to Part I of FDA’s Broad Agency Announcement No.

FDABAA-15-00121.

5.4 Increase the accuracy and efficiency of FDA review through the use of novel software tools

Approaches would include exploratory research on and conceptual development of automated tools that apply text mining techniques using natural language processing technologies, analytical methods and machine learning to accurately classify unstructured data within MedWatch reports based on the World Health Organization - Uppsala Monitoring Centre causality assessment scale for drug relatedness. FDA safety evaluators review MedWatch reports and conduct causality assessments to determine whether a medication may have caused an adverse event. Clinical judgment as well as more formalized causality assessment methods are used for this task. However, conducting manual case report reviews for causality assessment is labor- and time-intensive. Therefore, FDA is seeking tools and methods that will increase efficiency and facilitate the triage, review and processing of these reports. Information supplied by such tools and methods would enable safety evaluators to filter reports based on varying degrees of likelihood that a medication may have caused an adverse event. Performance characteristics: This pilot effort would involve training the tool using MedWatch reports scored by safety evaluators, and then tested with additional reports to determine the tool’s ability to identify MedWatch reports that are most likely to be informative in the triage and review process of adverse event signal detection in pharmacovigilance.

mailto:Sean.Wybenga@fda.hhs.gov

PURPOSE: The purpose of this amendment is to add an additional area of research interest to “Part I: Research Areas of Interest, Research Area 5. Strengthening Social and Behavioral Science at FDA by Enhancing Audience Understanding” of the announceme...