Automating IDMP preparation with PhlexGlobal

Informa Connect Life Sciences

/@Ibclifesciences

Published: March 11, 2021

Open in YouTube
Insights

This video provides an in-depth exploration of how artificial intelligence and machine learning are being applied to automate regulatory processes within the pharmaceutical industry, with a particular focus on preparing for the Identification of Medicinal Products (IDMP) mandate. Jim Nichols, Chief Product Officer at Phlexglobal, discusses the company's "Flex Neuron" AI technology and its role in transforming document-driven regulatory operations into more efficient, data-driven processes. The discussion highlights the critical shift in regulatory compliance from unstructured content to structured data, emphasizing the strategic value this transformation brings to pharmaceutical companies.

The core of the conversation revolves around the "datification of regulatory," a trend driven by global health authorities like the EMA and FDA. Phlexglobal's Flex Neuron platform leverages AI and machine learning to extract, classify, and prepare data from various regulatory documents, such as SmPCs (Summary of Product Characteristics) and Module 3 documents. This automation is crucial for populating data management solutions like Flex IDMP, ensuring that pharmaceutical companies are ready for upcoming mandatory submissions. The video also details how AI assists in Trial Master File (TMF) management by automating document indexing and placement, significantly reducing manual effort and enhancing accuracy.

Nichols provides concrete examples, including Phlexglobal's work with AstraZeneca, where AI has been used to recursively extract IDMP data from regulatory documents and to automate the cataloging and metadata population of health authority communications into systems like Veeva Vault. He underscores that IDMP is currently the most significant regulatory update, with the recent release of Implementation Guide v2 initiating a two-year countdown to mandatory submissions. The ultimate goal of IDMP, as part of EMA's Article 57, is to enhance patient safety through a comprehensive, hierarchical collection of data that enables better impact analysis and signal detection for adverse events. For pharmaceutical companies, this transition means regulatory information evolves from a compliance burden into a valuable strategic asset, offering greater visibility into their product portfolios and facilitating proactive decision-making.

Key Takeaways: • The "datification of regulatory" is a transformative trend in the pharmaceutical industry, shifting focus from unstructured documents to structured data, driven by global health authorities like the EMA and FDA. This fundamental change is vital for modernizing compliance and leveraging regulatory information strategically. • IDMP (Identification of Medicinal Products) is identified as the most critical new regulation, with the recent release of Implementation Guide v2 starting a two-year countdown to mandatory data submissions, requiring urgent preparation from pharmaceutical companies. • Artificial intelligence and machine learning, exemplified by Phlexglobal's Flex Neuron, are essential for automating the extraction and encoding of structured data from vast volumes of unstructured regulatory documents (e.g., SmPCs, Module 3 documents) to meet IDMP requirements. • Automation significantly improves Trial Master File (TMF) management through AI-assisted document indexing and classification, which reduces labor-intensive manual processes, ensures proper attribution, and accurately places documents within the TMF reference model structure. • Transitioning from document-driven to data-driven processes enables pharmaceutical companies to perform more effective analytics, conduct impact analyses, and enhance signal detection for patient safety, ultimately leading to better product lifecycle and regulatory management. • Structured regulatory information transforms from a mere compliance overhead into a strategic asset, providing pharmaceutical companies with valuable internal visibility into their products and enabling proactive decision-making regarding product changes and regulatory impacts. • Companies should identify high-volume document areas where critical data is currently "trapped" to prioritize and target automation efforts effectively, focusing on problems that offer the biggest value return for AI implementation. • AI solutions can be integrated with existing content management systems, such as Veeva Vault, to automate the cataloging, classification, and metadata population of incoming health authority communications, streamlining regulatory operations. • Companies currently compliant with xEVMPD (Extended EudraVigilance Medicinal Product Dictionary) must understand the migration path to IDMP and be prepared to operate in both worlds simultaneously, requiring robust bridging mechanisms to maintain compliance during the transition. • The EMA's ultimate goal for IDMP, under Article 57, is to enhance patient safety by centralizing detailed, hierarchical product data, enabling more sophisticated impact analysis and signal detection for adverse events at a granular level. • Pharmaceutical companies benefit from IDMP by gaining enhanced internal visibility into their own products, allowing them to conduct more precise impact analyses on product changes and better understand how regulatory and product lifecycles affect their operations. • Regulatory Information Management (RIM) encompasses IDMP, emphasizing the need for robust connectivity across all regulatory activities, ensuring that actions like filing a Type 2 variation are seamlessly tied to corresponding IDMP submissions. • Phlexglobal offers comprehensive support for IDMP readiness, including data extraction services using Flex Neuron, intuitive data management applications, and analytical services to help companies map and aggregate data from various internal sources.

Tools/Resources Mentioned:

  • Flex Neuron: Phlexglobal's AI/ML technology for content extraction, document classification, and data preparation.
  • Flex TMF: Phlexglobal's Trial Master File software, which embeds Flex Neuron for AI-assisted document indexing.
  • Flex IDMP: Phlexglobal's data management solution for staging and managing IDMP data.
  • Flex xEVMPD: Phlexglobal's solution for xEVMPD submissions, serving as a precursor to IDMP.
  • Veeva Vault: A content management system mentioned as being integrated with AI for categorizing health authority communications.

Key Concepts:

  • IDMP (Identification of Medicinal Products): International standards for the unique identification and exchange of medicinal product information, mandated by the EMA.
  • xEVMPD (Extended EudraVigilance Medicinal Product Dictionary): A precursor to IDMP, used for submitting medicinal product data to the EMA.
  • TMF (Trial Master File): A collection of essential documents that individually and collectively permit the evaluation of the conduct of a clinical trial and the quality of the data produced.
  • RIM (Regulatory Information Management): The systematic management of all regulatory data, documents, and processes throughout the product lifecycle.
  • Datification of Regulatory: The industry trend of transforming unstructured regulatory content into structured, actionable data for improved compliance, analytics, and strategic decision-making.
  • AI-assisted Document Indexing: Using artificial intelligence to automatically suggest or apply metadata and classification to documents, reducing manual effort and improving accuracy.
  • Structured Product Labeling (SPL): An FDA-regulated standard for exchanging product and establishment information.
  • PQ CMC Initiative (Product Quality and Chemistry, Manufacturing, and Controls): An FDA initiative aimed at modernizing CMC data submission and review.
  • Structured Product Monograph: A Health Canada initiative for structured product information.
  • Electronic Prescribing Information (ePI): A European initiative to transition from paper-based prescribing information to structured electronic formats.

Examples/Case Studies:

  • AstraZeneca: Phlexglobal assisted AstraZeneca in preparing for IDMP by recursively extracting IDMP data from numerous regulatory documents, quality checking it, and providing an up-to-date data collection for their IDMP management system.
  • Health Authority Communications: Phlexglobal worked on a project where approximately 15,000 health authority communications per year needed to be cataloged into a content management system (Veeva Vault). Their AI system was trained to identify attributes like product, agency, and registration, pre-populating metadata for proper categorization and classification.