SC-11 - Pharmacovigilance and Signal Detection

1. Course Aim

This course will introduce participants to the principles of surveillance in pharmacovigilance with a focus on quantitative aspects and recent advances in the field. The course will cover classical pharmacovigilance signal detection using adverse event reports and also describe quantitative signal detection methods for electronic medical record and claims data. Real-world examples will be used across the lecture sessions to illustrate and connect different concepts. Participants will engage in interactive discussions about contemporary issues in and challenges facing pharmacovigilance.

2. Requisites Statement

Entry level.

3. Course Objectives

1. Explain the need for pharmacovigilance and its component activities with a focus on safety surveillance
2. Describe the potential data sources (e.g., individual case reports, claims, electronic medical records, social media data) for safety surveillance in pharmacovigilance
3. Understand the array of core analytic methods for pharmacovigilance, including those applied to adverse events reports for signal detection and active monitoring in electronic healthcare databases
4. Recognize the limitations of each data source and approach and how they complement each other
5. Appreciate some of the recent advances and controversies in quantitative signal detection

4. Syllabus Outline

The course will be live followed by a panel Q&A session

1. Signal detection and quantitative analysis of individual case reports and beyond (Greg Powell): This session will cover basic concepts of safety surveillance in pharmacovigilance, describe how adverse event reports are generated and collected, introduce quantitative tools for separating potential signals from noise among large collections of adverse event reports, and describe how the concepts and tools of signal detection can be applied to observational databases. It will also touch on some preliminary work using social media for signal detection.

2. Statistical methods for signal detection using individual case reports (Niklas Norén): Many large organizations rely on statistical pattern discovery as one component in their pharmacovigilance processes to direct and inform clinical review of individual case reports. This session will describe some of the core analytic methods, with an emphasis on their strengths and limitations. It will highlight some recent advances including methods to detect possible risk factors and to improve statistical signal detection with multivariate analyses.

3. Novel quantitative methods for identifying drug safety signals in electronic healthcare data (Judy Maro): With the expanding availability of longitudinal electronic healthcare data, quantitative methods for safety signal detection are expanding rapidly in sophistication (e.g., leveraging inverse probability of treatment weighting) and in various populations including during pregnancy and in children. This session will focus on technical aspects of these new methods in both self-controlled and cohort designs.

4. Pharmacovigilance in practice: Getting from signal to insight (Monica Muñoz): This session will provide an overview of how identified signals are managed from a regulatory and clinical point of view. Attendees will gain an understanding of signal management principles with an emphasis on case series development and information synthesis.
In addition to questions raised by participants, the live panel discussion will cover topics related to contemporary issues in and challenges facing pharmacovigilance and signal detection.