Get SMART to Stay in CONTROL and Avoid KAOS: Signal Management for Adverse Events in Real Time
Aug 12, 2015 1:49 PM
Peter Mroz, Statistical Programmer, Janssen Pharmaceutical Companies of Johnson & Johnson
Pharmacovigilance is defined as “the science and activities relating to the detection, assessment, understanding and prevention of adverse effects or any other drug-related problem” (World Health Organization). At Janssen we perform pharmacovigilance to comply with regulations and to protect patient safety. SMART is a decision-support tool that provides an efficient, comprehensive and timely system for conducting post-marketing safety surveillance. Its features include: monitoring product safety data for a range of potentially concerning reporting patterns; alerting appropriate staff when concerning reporting patterns are noted; analyzing and assessing/evaluating case data and basic trends for resultant alerts; documenting reviewer comments and signoff for resulting alerts; managing workload; and generating reports and trend graphs. SMART employs several advanced techniques for signal detection, including data mining, control chart techniques and case-level triage. SMART allows users to filter large amounts of data down to a few items of concern. SMART has been developed using JMP JSL for the user interface, reports and graphs, and Oracle SQL and PL/SQL for data storage and back-end logic. This presentation will provide a general overview and then focus on some interesting technical details of the system. These details include designing the user interfaces using the Application Builder in JMP, connecting JMP to Oracle, and using control chart techniques for signal detection.