In a recent webinar, DistillerSR customer Andrew Purchase, Director of Pharmacovigilance, UK QPPV, Site Head (Swansea, UK) at ICON was joined by Michael-Braun Boghos, Senior Director Safety Strategy at Oracle Life Science, and moderated by Peter O’Blenis, DistillerSR CEO. They discussed the challenges of ensuring the safety and efficacy of products from clinical development stage through post-marketing and the likelihood of high volumes of information from heterogeneous data sources.
Q: What should be the initial step when setting up a literature search at ICON?
A: It’s crucial to start by defining the research objective. Whether we’re dealing with a product in clinical trials or a commercial setting, the focus of our literature search can vary. We work closely with librarians to develop a well-defined search strategy that aligns with our objectives.
Q: How does the literature search strategy evolve as a product progresses through its lifecycle?
A: The strategy evolves significantly as a product moves through different lifecycle phases. In clinical trials, our focus might shift as we uncover new risks and safety information. When transitioning to a commercial setting, we face a higher data volume and less controlled conditions. Our strategy adapts to incorporate real-world factors like pregnancy exposure and misuse, which become more relevant.
Q: How does technology come into play when monitoring and adapting literature search strategies at ICON?
A: Technology plays a crucial role in continuously monitoring and adapting our literature search strategies. It allows us to assess the strategy’s effectiveness by running metrics and gauging how well it retrieves relevant information. Additionally, technology helps us manage the increasing volume of data efficiently, and it motivates our team by automating administrative tasks, enabling us to focus more on the scientific aspects of our work.
Q: Michael, what are some challenges associated with monitoring literature and pharmacovigilance, as per your discussion?
A: Several challenges were discussed regarding monitoring literature and pharmacovigilance:
- Adapting to changing regulations: Staying compliant with evolving regulatory requirements poses a constant challenge.
- Managing multiple data sources: Selecting and accessing the right data sources while ensuring sufficient recall can be daunting.
- Signal detection: Identifying safety signals within a vast amount of data is akin to searching for a needle in a haystack.
- Dealing with data heterogeneity: Integrating and analyzing data from diverse sources, often in different formats, can be complex.
- Navigating cost and resource constraints: Balancing budgets and resources, especially in post-marketing scenarios, presents significant challenges.
Q: How does the implementation of automation and AI address these challenges in pharmacovigilance and literature monitoring, in your view, Michael?
A: Automation and AI offer effective solutions to tackle these challenges in pharmacovigilance and literature monitoring:
- Handling data volume: Automation efficiently processes extensive literature and individual cases, reducing manual workload.
- Expedited ICSR processing: AI automates data entry into safety databases, ensuring faster submissions to health authorities while maintaining compliance.
- Enhanced signal detection: Machine learning and AI assist in recognizing safety signals and patterns across diverse data sets.
- Reducing manual data entry: Automation streamlines data transfer between systems, eliminating the need for manual input.
- Real-time monitoring: Automation enables real-time data analysis and monitoring, minimizing decision-making delays.
Q: How is ICON currently leveraging AI and technology to enhance pharmacovigilance and literature monitoring, Andrew?
A: ICON is actively exploring various avenues to leverage AI and technology for the improvement of pharmacovigilance and literature monitoring:
- Automating repetitive tasks: AI is being used to automate rule-based processes, reducing manual efforts and enhancing efficiency.
- System integration: ICON is working towards seamlessly integrating different systems to streamline data transfer and eliminate manual data entry.
- Real-time monitoring: Technology solutions are facilitating real-time monitoring and analysis of safety data.
- Continuous improvement: ICON is committed to exploring AI’s potential to enhance literature searches, ICSR processing, and signal detection, all with a primary focus on improving compliance and operational efficiency.
Q: Michael, could you outline the framework utilized at Oracle for managing data and insights?
A: This framework is structured into four layers. At the foundational level, we have the tech stack, encompassing elements like databases, networking, and data storage infrastructure such as data lakes. Just above that, we find the data layer, which entails the management of diverse data sets, including literature, clinical trial data, EHR data, and health authority data. This layer also entails the vital processes of data standardization and formatting. Moving up, we encounter the applications layer, housing the tools utilized for data analysis. Finally, at the pinnacle, there’s the people layer, representing the users who leverage these tools to derive insights from the data.
Q: How does this framework intend to enhance data management and signal detection in the context of pharmacovigilance?
A: The framework’s primary objective is to elevate data management and signal detection by offering a unified structure for data and insights management. It effectively eliminates the existence of silos and replaces scattered Excel spreadsheets. This unified framework simplifies communication with health authorities. Moreover, it introduces multimodal signaling, enabling the aggregation of signal scores from different data sets. This approach ensures a more comprehensive understanding, especially when signals emerge in multiple data sources with varying scores.
Q: Andrew, what advantages are associated with the integration of technology and automation in the field of pharmacovigilance?
A: The integration of technology and automation in pharmacovigilance brings forth several significant benefits. It enhances operational efficiency by automating repetitive tasks, mitigates the risk of missing cases or signals, and expedites data processing with heightened accuracy. Furthermore, it elevates the quality of signal detection and ensures timely reporting, thereby maintaining a favorable reputation with regulatory bodies. Automation can also contribute to employee motivation by relieving them of manual workloads and enabling a more concentrated focus on meaningful tasks.
Q: How do you go about quantifying the return on investment (ROI) for the implementation of automation within the realm of pharmacovigilance?
A: Measuring the ROI for automation in pharmacovigilance involves a systematic approach. It begins with the identification of specific metrics for each workflow step, such as time saved or cost reduction. Subsequently, the potential savings at each step after automation implementation are calculated. The overall impact on efficiency and costs is assessed, ensuring that the achieved savings offset the initial investment in automation tools. Demonstrating that automation leads to cost savings and process enhancements is pivotal in justifying the ROI.
Q: Could you elaborate on the continuous innovation approach when implementing automation for case processing?
A: The continuous innovation approach entails a meticulous breakdown of the case processing workflow into individual steps and substeps. Instead of attempting to automate the entire process in one go, organizations focus their efforts on automating specific steps or substeps that present the most significant bottlenecks or challenges. They commence with a targeted area, establish baseline metrics, proceed with automation implementation, and subsequently gauge the efficiency gains achieved. It’s a gradual journey where one success leads to the next, ultimately automating the entire workflow.
Q: What strategy does Oracle employ for the integration of AI and technology into the field of pharmacovigilance, Michael?
A: Oracle’s strategy for incorporating AI and technology into pharmacovigilance revolves around two primary objectives. Firstly, we endeavor to automate various facets of the pharmacovigilance process to enhance operational efficiency and ensure compliance with regulatory requirements. Secondly, we harness the power of AI and large language models to extract deeper insights from the ever-expanding pool of data sources, encompassing literature, clinical trial data, and more. Our strategy is twofold, focusing on both streamlining processes and bolstering our ability to promptly identify and assess risks.
