Season 1 Episode 1: Building the Right Data and Technology Foundation for Safety

This video, from the Veeva podcast "Safety Revolution," features David Kološić (Veeva) and Aniket Agarwal (Director for Data Operations and Analytics in Patient Safety at Sandoz), discussing the critical role of a robust data and technology foundation in pharmacovigilance (PV). The conversation centers on how Sandoz, as a large generic and biosimilar company, is navigating its growth while enhancing operational efficiency and speed in patient safety through strategic technology adoption.

The discussion highlights Sandoz's deliberate shift from a historical landscape of "best-in-class" siloed systems to a platform-based approach. This transformation is driven by the need to overcome challenges associated with maintaining complex integrations between disparate systems and to ensure sustainable growth without a proportional increase in operational teams. Aniket explains that while other domains like clinical and regulatory have adopted platforms earlier, PV's slower pace is due to stringent regulations, frequent inspections, and the necessity of maintaining data compatibility for products with long market lifecycles (30-40 years). The core idea is to harmonize data and technology across global development (clinical, regulatory, safety) to establish a single source of truth, reducing manual reconciliation and improving data quality.

A significant portion of the conversation is dedicated to the strategic application of automation and AI in PV. Aniket advocates for a "grounded approach," emphasizing that organizations should first identify specific problems and leverage simpler automation for quick efficiencies before deploying more complex AI solutions. He identifies high-impact AI use cases, particularly in the Individual Case Safety Report (ICSR) space, such as ingesting unstructured data (e.g., non-E2B reports which constitute a significant portion of incoming data) and generating human-readable narratives. Beyond ICSRs, AI is seen as transformative for moving from traditional to predictive signal detection, enhancing the quality of detected signals. The speakers also touch upon the balance between making reporting easy for healthcare professionals and patients (e.g., supporting regional languages) and the need for technology to structure this diverse intake downstream. The ultimate vision for 2030 is "no-touch" end-to-end case processing, with AI solving the remaining 40% of complex scenarios, contingent on evolving regulatory frameworks and building confidence in AI-generated data.

Key Takeaways:

• Shift to Platform-Based PV: Sandoz is moving from siloed, "best-in-class" systems to a unified platform approach to achieve sustainable operations, reduce integration complexities, and ensure systems evolve at a consistent pace. This is crucial for long-term efficiency and growth in pharmacovigilance.
• Drivers for PV Platform Adoption: The need for harmonization of data and technology across global development (clinical, regulatory, safety) is a prime driver. A platform approach simplifies data flow, reduces maintenance, and supports cross-functional collaboration.
• Challenges in PV Technology Evolution: PV has been slower to adopt platform solutions due to strict regulations, frequent inspections, the need for data compatibility for products with decades-long market presence, and the inherent risk associated with system changes.
• Importance of Data Standardization: Standardizing data across regulatory, clinical, and safety domains is critical for establishing a "one source of truth," reducing manual reconciliation efforts, and improving the efficiency and quality of reporting (e.g., for PSURs/DSURs).
• Overcoming Data Silos and Mindset Shifts: Achieving data standardization requires breaking down historical departmental silos and fostering a mindset shift towards common organizational goals, even if it involves an iterative process and governance to align definitions.
• Grounded Approach to Automation and AI: Prioritize solving specific problems with simpler automation for quick wins and agility. Reserve AI for more complex, high-impact use cases where traditional automation is insufficient, adopting a phased approach if necessary.
• High-Impact AI Use Cases in PV: Key areas where AI can drive significant value include ingesting and structuring unstructured incoming data (e.g., non-E2B reports), generating advanced, human-readable narratives for ICSRs, and transitioning from traditional to predictive signal detection.
• Balancing Reporting Ease and Data Structure: To encourage higher reporting rates, it's essential to make the reporting process as simple as possible for users (e.g., supporting regional languages, flexible input formats). Technology, particularly AI, can then be leveraged downstream to decipher and structure this diverse information.
• Benefits of Cross-System Analytics: A platform approach enables efficient, real-time cross-system analytics for periodic reports, reducing manual data extraction and reconciliation, and ensuring consistent information for regulatory submissions and inspections.
• Regulatory Harmonization and Frameworks: Organizations like ICH, EMA, and FDA are actively working on frameworks and guidance to support the adoption of automation and AI in PV, indicating a growing openness and a shared goal with the industry towards safer and more efficient processes.
• Vision for "No-Touch" PV: The aspirational goal for 2030 is end-to-end automated (no-touch) case processing, where systems can consistently handle a vast majority of scenarios, driving both quality and efficiency, and freeing up resources for more complex tasks.
• Building Regulator Confidence in AI: As AI technologies advance, it's crucial to collaborate with regulators to understand their expectations, address concerns like hallucination, and ensure AI-generated data remains usable and fit for purpose within a compliant framework.

Tools/Resources Mentioned:

• Veeva: A leading platform provider in the pharmaceutical industry, hosting the podcast.
• E2B: An international standard for the electronic transmission of individual case safety reports (ICSRs).
• ICH (International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use): Mentioned as an organization striving for harmonization.
• EMA (European Medicines Agency): Referenced for its efforts in standardization (e.g., through UdraVigilance) and guidance on AI.
• FDA (U.S. Food and Drug Administration): Referenced for its guidance on AI.
• CIOMS (Council for International Organizations of Medical Sciences): Mentioned as producing material on AI.

Key Concepts:

• Pharmacovigilance (PV): The science and activities relating to the detection, assessment, understanding and prevention of adverse effects or any other drug-related problem.
• Individual Case Safety Report (ICSR): A report detailing a single suspected adverse drug reaction in a patient.
• Periodic Safety Update Report (PSUR): A periodic report providing an update on the worldwide safety experience of a medicinal product.
• Development Safety Update Report (DSUR): A periodic report providing an update on the worldwide safety experience of an investigational medicinal product.
• Signal Detection: The process of identifying and assessing potential safety signals from various data sources.
• Generics and Biosimilars: Types of pharmaceutical products Sandoz specializes in.
• GxP (Good x Practice): A collection of quality guidelines and regulations created to ensure that products are safe and meet their intended use.
• 21 CFR Part 11: Regulations issued by the FDA that set forth the criteria under which electronic records and electronic signatures are considered trustworthy, reliable, and equivalent to paper records.