药物安全创新的下一个前沿：人工智能支持的信号管理

The Next Frontier of DrugSafety Innovation: AI-SupportedSignal Management STEPHANIE SENN,Product Manager, Vigilance Signal, IQVIA Table of contents Introduction1Opportunities for innovation in the signal management space2Where could AI implementation benefit safety signal management?3How is IQVIA approaching AI in safety signal management?4What’s on the horizon?4References5About the authors5 Introduction Safety signal management is the process of detecting and assessing adverseevent data and relevant supporting pharmacologic, clinical, and epidemiologicevidence to determine if there is a new risk associated with a medicinal productor if an existing risk has evolved. For over 20 years, companies have utilizedquantitative and qualitative methods to detect potential signals that requirefurther investigation. Manual selection of relevant data and its subsequent reviewhave underpinned the signal analysis process and been a resource-intensiveexercise performed by a Team of Pharmacovigilance (PV) experts. The rapid growth of data volume and source complexity,coupled with advances in technology has led pharmacompanies, regulators, technology providers, andindustry bodies to explore opportunities for AI tostrengthen signal detection and analysis, enabling fasteridentification of patterns across treatment categoriesand patient populations. Agentic AI offers a promising approach for assistingsignal management. These systems can autonomouslyadapt to new data, prioritize tasks, and proactivelyidentify emerging risks while maintaining traceabilityand alerting for human oversight as needed. Thiscapability aligns perfectly with the dynamic nature of PV,where timely and context-aware decisions are critical. been the standard data sources, regulators and industryhave demonstrated successful use of Real-World Data(RWD) in specific scenarios, including in retrospectiveobservational studies utilizing electronic health recordsas a supplementary source. With advances in AI, it isnow more achievable to harness RWD for routine signalmanagement by utilizing Natural Language Processing(NLP) and Generative AI (GenAI) for the extraction ofrelevant data and to address identified challenges withstandardization and quality. This stands to benefit smallcompanies working in therapeutic areas with lowerspontaneous reporting rates and more specialized data,e.g., rare diseases and precision medicine. Opportunities for innovation inthe signal management space Technological advances and the desire to reducethe operational burden of signal management havestrengthened the drive to improve signal-relatedprocesses. Over the past few years, the following trendshave been driving the need and opportunity for change: 1. A growing regulatory emphasis on improving signaland risk management 2. Increased interest in utilizing supplementary datasources for evidence Across the industry and all stages of the drugdevelopment and manufacturing lifecycle, AI tools arebeing implemented to increase efficiency and reduceresource overload. In June 2025, the FDA launchedElsa, a GenAI tool to help FDA employees work moreefficiently; thus far, Elsa has been used to speed upclinical protocol reviews and evaluations.1From a drugcompany perspective, GSK has implemented PVLens,an automated system that extracts labeled safetyinformation from FDA Structured Product Labels (SPLs)to enhance PV with improved accuracy and insight.2These advancements indicate a changing tide as theindustry realizes the benefits of how these AI solutionscan augment product lifecycle management. 3. Increased industry adoption of AI 4. A desire to move toward widespread use ofpredictive analytics Data mining algorithms and manual data review havelong been the accepted approach of regulators todetect and validate signals. Regulators and MarketingAuthorization Holders (MAHs) have researched andrefined disproportionality algorithms over the years, buttraditional frequentist and Bayesian methods have notsignificantly improved the false positive rate for manyMAHs, resulting in a time-consuming analysis step.There is a desire and mindset shift across industry toidentify how technology can be best utilized to supportPV experts in signal management. In an IQVIA-commissioned survey and study conductedby IDC, real-time signal detection is forecasted tobecome the primary focus area for automation initiativesin the next two years.3We have already started toimplement AI solutions to support signal management,but within the near future we can see the potential torevolutionize how we manage patient safety as a whole.Predictive AI analytics can be used to identify long-termpatterns in the data to help companies predict signaltrends for their drugs and answer questions that haveyet to be asked. Agentic AI can be used to seek outpatterns, contextualize findings using multiple sources,and recommend next steps. As a result, they can takemore proactive approaches to safety, replacing theretrospective rou