Post-Conference takeaways from DIA Pharmacovigilance and Risk Management Strategies Conference 2023, Bethesda, MD.
Status
Pharmacovigilance and drug safety have undergone significant changes in recent years due to technological advancements, increased regulatory scrutiny, and the growing demand for more efficient and effective monitoring of adverse events associated with pharmaceutical products. At the DIA Global Pharmacovigilance and Risk Management Strategies Conference (PVRMS) in Bethesda, Maryland, many topics were discussed, with several presentations touching on ways to collect more data and extract insights from that data. The need for integrating often siloed Pharmacovigilance (PV) operations is obvious. Platforms have seen an unprecedented caseload and as a result, regulatory authorities and pharmaceutical companies have implemented several measures to enhance the efficiency and effectiveness of pharmacovigilance systems. These measures include the use of digital tools for reporting and processing, data-driven decision-making, and the collaboration between different stakeholders around the world.
The benefits are clear, more data equals better decisions and better decision-making lends to improved public health and safety overall. The volume of data is only as good as the ability to collect, analyze, and act upon that information. Spontaneous reporting systems, clinics trials, and other data sources generate large volumes of data which are analyzed to identify previously unknown risks or patterns of adverse events. Once a medication is on the market, post-marketing surveillance and/or monitoring is conducted to identify any adverse events that may occur in larger and more diverse patient populations, that might not have otherwise been detected in clinical trials. From this monitoring and identification, the benefit-risk profile of a medicine is characterized, high-risk patient populations are identified, and risk management strategies are formed.
As mentioned, the outcome of larger scale data collection ultimately leads to innovations that can further drug development and enable data driven decision-making. This is done by:
- better reporting and oversight to health authorities (HAs) and key stakeholders,
- utilizing a scalable system to store, version, and act upon findings as they evolve, and
- proactively respond to risks with strategies, better implementation of Additional Risk Minimization Measures (aRMMs), and overall better risk management strategies.
The three above benefits are often difficult to track and distribute quickly or even globally without a systematic approach to manage these processes. Typically, this involves multiple teams, internally and externally, that contribute to the many moving parts.
New Tools and Methods
As the volume of data in drug safety continues to grow, new tools and methods are being developed to help alleviate this burden, make better decisions, and communicate and share that information globally.
A specific emphasis at this year’s DIA PVRMS Conference was on new tools and methods used to alleviate the volume of data associated with adverse drug reactions (ADRs) and adverse events (AEs), which enable quicker data-driven decision-making, data-sharing, and global distribution by the PV and Drug Safety Experts. The advancements in technology aid but will not replace the human expertise and judgment critical for many tasks in drug safety.
FMQ Method
The Frequentist/Multi-Criteria Quantitative (FMQ) method and analysis was reinforced as a powerful tool for evaluating the benefit-risk profile of medications. The FMQ method is used to analyze data from multiple sources, including spontaneous adverse event reports, electronic health records, and other clinical databases to provide an overall assessment of the drug, assessing its efficacy and safety. The goal of the FMQ method is to identify potential safety signals by evaluating the frequency of reported adverse events associated with a drug and comparing it to the expected background frequency of those events in the general population.
The FMQ method has several benefits in pharmacovigilance, including:
- Improved sensitivity: The method can identify potential safety signals that may be missed by traditional signal detection methods.
- Scalability: The FMQ method can be used to analyze large volumes of data, making it suitable for pharmacovigilance applications.
- Flexibility: The method can be adapted to different types of data sources and can be used in combination with other signal detection methods.
- Automation: The FMQ method can be automated, which reduces the burden on human analysts and improves the speed and accuracy of analysis.
This statistical technique combines frequentist and machine learning approaches to analyze large datasets of adverse event reports and identify potential safety signals. Where other traditional signal detection tools lack, the FMQ method can be adapted to different types of data sources, offering a more comprehensive approach that considers multiple criteria, such as efficacy, safety, and patient preferences.
AI and Machine Learning
The buzz around Machine Learning (ML) and Artificial Intelligence (AI) continues to grow. These tools are used to identify patterns or emerging safety signals in large datasets, such as adverse event reports. Both Artificial Intelligence and Machine Learning can automate tasks, reduce manual labor, improve productivity, yet the key differentiator is the level of human involvement required. AI refers to the general concept of creating machines or software to perform tasks that typically require human intelligence, whereas ML is a subfield of AI that involves statistical models or algorithms to enable computers to learn from data and improve their performance on specific tasks.
Social Media Monitoring
Another growing method is in social media monitoring. By analyzing social media posts, comments, and reviews, PV teams can gain insights into how patients are using their products, and whether there are any emerging safety concerns. Social media platforms, such as Twitter, Facebook, or Reddit, provide a rich source of data on consumer opinions, behaviors, and experiences related to health and wellness. Of all the benefits this tool provides, most important might be the engagement factor, where pharma companies can engage with consumers and respond to their concerns and questions related to pharmaceutical products. However social media monitoring also presents its challenges, such as, inaccurate, or biased data entry, privacy concerns, or data overload. Social media data may be incomplete, it may contain personal information which raises privacy concerns, or the data can be overwhelming and require analysis to extract meaningful insights. Yes, despite these challenges, there is still benefit in the early detection, real-time data, and large-scale consumer engagement with this platform.
The Benefit and Challenge
The benefits of these methods are vast, revolutionizing the way medical professionals diagnose, treat, and prevent diseases. AI can help to identify patterns in patient data or perform tasks that typically require human intelligence, ML enables computers to learn from data and improve their performance, and social media monitoring provides an engagement opportunity for companies to address questions and concerns on a publicly accessible platform.
As we know, one of the biggest burdens in pharmacovigilance is the effective management of large amounts of data. As pharmaceutical products continue to reach patients worldwide, the volume of data generated by these reports, trials, and other sources has and will become increasingly more challenging. And so, having proper systems in place to alleviate some of the manual workload, enable safety professionals to do more of what matters most, extracting insights and making decisions based on that data is where technology alleviates some of that challenge.
More Data, Better Decisions
Technology has enabled the collection, storage, analysis, and sharing of data in ways that were not possible before. For example, electronic health records have made it easier to identify adverse events and monitor the safety of medications in real-time. Machine Learning Algorithms can be used to identify patterns and associations in large datasets, which can help identify previously unknown or unexpected signals. Social media and other digital platforms have been used to collect patient-reported data and identify emerging safety issues. Technology has improved communication and collaboration among pharmaceutical companies, regulatory authorities, stakeholders, and teams. Real-world data, Pharmacovigilance databases, wearable devices, AI, etc., have all contributed dynamically in improving patient safety.
These advancements speed up the ability to get more insights and help filter through the “noise” inherent with ever increasing amounts of data that safety experts and scientists are inundated with. The benefits of more data in drug safety are clear, but the effective use of that data requires a combination of technical knowledge, analytical skills, clinical expertise, and effective communication and collaboration among different stakeholders. Technology is just one leg of a three-legged stool, which still requires People and Processes to be effective.
It is certainly exciting to see and hear about the enormous progress we are making in Pharmacovigilance with new and improving technologies. However, it’s important to keep in mind the role that technology plays – along with an optimized process – is to enable industry experts to do their jobs more effectively, not to replace them. By equipping drug safety experts, regulatory agencies, and healthcare professionals with quality data and insights, we can better safeguard patients and ensure that medications are used safely and appropriately. These developments will continue to lead to more accurate, personalized, and efficient drug safety monitoring and risk management in years to come.
