Leveraging real-world data for safety signal detection and risk management in pre- and post-market settings

TL;DR

This paper discusses how real-world data can be used to improve drug safety monitoring before and after market approval while protecting patient privacy.

Contribution

The paper introduces privacy-preserving methods for integrating real-world data to enhance pharmacovigilance and regulatory compliance.

Findings

Privacy-preserving record linkage improves longitudinal safety monitoring with patient-level insights.

Linked real-world data enables detection of rare events and long-term risks not captured by traditional methods.

Regulatory best practices include early engagement and incorporating RWD into risk management plans.

Abstract

The evolving regulatory landscape has increasingly recognized the value of real-world data (RWD) in enhancing drug safety surveillance across the clinical development lifecycle. Enabled by frameworks such as the FDA’s Real-World Evidence (RWE) Programs and other international regulatory bodies, sponsors now have expanded opportunities to use RWD to detect, evaluate, and manage safety signals in both pre- and post-market settings. This paper examines how the integration of RWD, particularly through privacy-preserving record linkage (PPRL) methods like tokenization, can improve pharmacovigilance by enabling longitudinal safety monitoring while protecting patient privacy. Traditional safety surveillance methods, such as spontaneous adverse event reporting and aggregate signal detection, are limited by under-reporting and fragmented data sources. In contrast, linked RWD offers more comprehensive, patient-level insights into safety outcomes, including rare events, off-label use, and long-term risks. The paper outlines regulatory considerations for using de-identified, linked RWD in safety reporting, emphasizing the importance of clear protocols, IRB engagement, and legal compliance with HIPAA. It further highlights emerging best practices for integrating RWD into clinical development, such as early regulatory engagement and the incorporation of linked RWD-derived safety signals into risk management plans. Ultimately, we propose that leveraging linked RWD in a privacy-focused manner enables more proactive, scalable, and effective pharmacovigilance. This approach supports earlier detection of safety issues, enhances post-market follow-up, and promotes data continuity between trial and real-world settings, positioning RWD as a cornerstone of modern safety surveillance.

Figures