Pharmacovigilance plays a crucial role in drug safety and is defined here as the science behind the detection, assessment, understanding, and prevention of, adverse effects of medicines. The aim is to monitor and ensure the safety of medicinal products both during clinical trials and post-marketing, crucial for protecting patient health.
A vital component of pharmacovigilance is the collection and analysis of reports of adverse drug reactions (ADRs) from patients and healthcare professionals. This data is essential for safety analysis and is regulated by the FDA, MHRA (UK) and EU[i] among others. As such, pharmacovigilance is a global effort, with international organizations like the World Health Organization (WHO) playing a key role in facilitating information sharing and harmonizing standards[ii].
However, significant shortcomings, including under-reporting, reactive processes, and issues with population diversity, may be leading to delays and potential harm to patients.
Under-reporting and diversity
According to a 2024 World Health Organization (WHO) report[iii], patient safety indicators face numerous challenges, including issues with data quality and availability, as health care facilities often have incomplete or inconsistent records. This has led the WHO to publish a strategic objective (Strategy 6.2[iv]) to 鈥渃reate a patient safety information system based on all sources of data related to risks and harm inherent in the delivery of health integrated with existing health management information systems.鈥�
One major issue in the under-reporting of ADRs is that current pharmacovigilance processes have been described as reactive rather than proactive, often responding to adverse events after they occur rather than preventing them through proactive monitoring and risk management[v]. This slow response can hinder the timely identification of safety issues, which may be critical for patient health.
Another significant challenge is the lack of diversity in the populations from which data is collected. Pharmacovigilance systems may not fairly represent all demographic groups, leading to gaps in understanding how different populations may respond to medications. This can result in safety concerns being overlooked, as adverse effects may manifest differently across groups.
Real-world data (RWD)
Utilizing real-world data (RWD) can play a crucial role in enhancing the detection of ADRs through various mechanisms, largely driven by advancements in technology, including the use of AI and automated data tools, as well as the increasing availability of digitized medical records.
One of the primary advantages of using RWD is the shift from reactive to proactive signal detection. RWD allows for ongoing monitoring and analysis of patient experiences, enabling pharmacovigilance teams to identify potential signals before they escalate into more significant issues. This approach can significantly improve the accuracy and efficiency of signal detection, as it incorporates a broader range of data sources, gathered from diverse settings beyond controlled clinical trials. This includes data from patient-generated reports, electronic health records and wearable technologies that monitor patient health metrics in real-time. Inclusion of this data allows for a more comprehensive analysis, revealing correlations between drug usage and adverse effects that may not be evident in traditional clinical trial settings.
A significant challenge in pharmacovigilance is the underreporting of adverse events. Real-world data can help mitigate this issue by providing more extensive datasets that capture a wider array of patient experiences. Further, the use of advanced analytics, AI and machine learning techniques, can lower the incidence of false positives in signal detection, allowing pharmacovigilance teams to focus their resources on genuine safety concerns.
Regulatory bodies such as the FDA have recognized the importance of RWD in post market safety surveillance. The FDA’s Sentinel System[vi], for example, uses real-world data to monitor medication usage and medical diagnoses across a large population, which aids in identifying ADRs more effectively.
Data from the real world
Data-driven solutions employ AI, machine learning, natural language processing (NLP), and analysis of RWD 鈥� the data points that, once analyzed, become real-world evidence (RWE). In today鈥檚 highly digitized world, RWD collection can create a wealth of data points from which researchers can build RWE.
One company making its mark in this area is US-based Veradigm, an integrated data systems and services company that combines data-driven clinical insights with actionable tools. Veradigm works alongside organizations to achieve real-world data insights with the collection, preparation and analysis of expansive datasets. Veradigm can execute actionable retrospective analyses and prospective programs, all with the support of a dedicated research team and custom data visualization. Digital dashboards can be provided so that stakeholders can better understand the patient journey utilizing data captured directly at the point of care.
Veradigm Network EHR Data, captures both structured and unstructured data across diverse patient populations and geographies, and by pairing AI with clinical validation, ensures data accuracy. This method has already been seen to improve post-market drug safety surveillance.
For example, Veradigm recently announced an advancement in the use of AI to scale the generation of RWE for GLP-1 receptor agonists (GLP-1 RAs), including semaglutide and tirzepatide[vii]. Veradigm discovered that while GLP-1 therapies are transforming the management of type 2 diabetes and obesity, significant gaps remain in understanding real-world usage, especially in identifying reasons for discontinuation or capturing side effects hidden in physician notes.
By applying AI to EHR deidentified data within the Veradigm Network, researchers were able to analyze certain factors only found in unstructured fields that could influence adherence and health outcomes. AI-enabled analysis of unstructured notes, when combined with other clinical details, can enable researchers to gain a deeper understanding of the patient clinical experience, identifying potential patterns that could indicate emerging safety concerns. Human experts can then investigate further.
This combination of AI-assisted monitoring with human medical expertise allows companies to maintain comprehensive safety surveillance while managing a growing volume of data, assisting safety teams to process information more efficiently and potentially identify safety signals earlier in the drug development process.
For more on how Veradigm can help with your real-world data, download the free paper below.
[i] https://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2012:316:0038:0040:EN:PDF
[ii] The Importance of Pharmacovigilance, (Safety monitoring of medicinal products) ISBN 92 4 159015 7 ( https://iris.who.int/bitstream/handle/10665/42493/a75646.pdf
[iii] . Global patient safety report 2024. Geneva: World Health Organization; 2024. Licence: CC BY-NC-SA 3.0 IGO. https://iris.who.int/bitstream/handle/10665/376928/9789240095458-eng.pdf
[iv] Global patient safety report 2024. Geneva: World Health Organization; 2024. Licence: CC BY-NC-SA 3.0 IGO. Strategy 6.2, p241
[v] https://www.pharmaceutical-technology.com/features/harnessing-pharmacovigilance-to-turn-reactive-investigation-into-active-discovery/
[vi] https://www.fda.gov/safety/fdas-sentinel-initiative
[vii] https://www.ispor.org/heor-resources/presentations-database/presentation-cti/ispor-2025/poster-session-2/real-world-impact-of-semaglutide-liraglutide-and-tirzepatide-on-weight-loss-and-cardiometabolic-lab-measures-a-look-into-drug-persistence-and-reasons-for-discontinuation-r-dc
